Microeconomics, Global Edition • NEW: Math Review Exercises in MyEconLab—MyEconLab now offers an array of assignable and auto-graded exercises that cover fundamental math concepts. Geared specifically...

Microeconomics, Global Edition
• NEW: Math Review Exercises in MyEconLab—MyEconLab now offers an a
ay of
assignable and auto-graded exercises that cover fundamental math concepts. Geared
specifically toward principles and intermediate economics students, these exercises aim
to increase student confidence and success in these courses. Our new Math Review is
accessible from the assignment manager and contains over 150 graphing, alge
a, and
calculus exercises for homework, quiz, and test use.
• Digital Interactives—Engaging assessment activities that promote critical thinking and
application of key economic principles. Each Digital Interactive has progressive levels
where students can explore, apply, compare, and analyze economic principles. Many
Digital Interactives include real time data from FRED® that displays, in graph and table
form, up-to-the-minute data on key macro variables. Digital Interactives can be assigned
and graded within MyEconLab, or used as a lecture tool to encourage engagement,
classroom conversation, and group work.
• Pearson eText—The Pearson eText gives students access to their textbook
anytime, anywhere. In addition to note-taking, highlighting, and bookmarking,
the Pearson eText offers interactive and sharing features. Instructors can share
comments or highlights, and students can add their own, for a tight community
of learners in any class.
• Learning Resources—Personalized learning aids such as Help Me Solve This
problem walkthroughs and Figure Animations provide on-demand help when
students need it most.
• Personalized Study Plan—Assists students in monitoring their own progress
y offering them a customized study plan based on Homework, Quiz, and
Test results. Includes regenerated exercises with unlimited practice, as
well as the opportunity to earn mastery points by completing quizzes on
ecommended learning objectives.
• Practice—Algorithmically generated homework and study plan exercises with instant
feedback ensure varied and productive practice, helping students improve their
understanding and prepare for quizzes and tests. Draw-graph exercises encourage
students to practice the language of economics.
Practice, Engage, and Assess with MyLab Economics®
A L W A Y S L E A R N I N G
MicroeconoMics
ninth edition
GLoBAL edition
The Pearson series in economics
Abel/Bernanke/Croushore†
Macroeconomics*
Acemoglu/Laibson/List†
Economics*
Bade/Parkin†
Foundations of Economics*
Berck/Helfand
The Economics of the
Environment
Bierman/Fernandez
Game Theory with Economic
Applications
Blanchard†
Macroeconomics*
Boye
Principles of Transportation
Economics
Branson
Macroeconomic Theory
and Policy
Bruce
Public Finance and the
American Economy
Carlton/Perloff†
Modern Industrial Organization
Case/Fai
Oster†
Principles of Economics*
Chapman
Environmental Economics:
Theory, Application, and Policy
Daniels/VanHoose
International Monetary &
Financial Economics
Downs
An Economic Theory
of Democracy
Farnham†
Economics for Managers
Froyen
Macroeconomics: Theories
and Policies
Fusfeld
The Age of the Economist
Ge
er†
International Economics*
Gordon
Macroeconomics*
Greene†
Econometric Analysis
Gregory/Stuart
Russian and Soviet Economic
Performance and Structure
Hartwick/Olewile
The Economics of Natural
Resource Use
Heil
one
Milberg
The Making of the Economic
Society
Heyne/Boettke/Prychitko
The Economic Way of Thinking
Hu
ard/O’Brien†
Economics*
InEcon
Money, Banking, and the
Financial System*
Hu
ard/O’Brien/Rafferty
Macroeconomics*
Hughes/Cain
American Economic History
Husted/Melvin
International Economics
Jehle/Reny
Advanced Microeconomic
Theory
Keat/Young/Erfle
Managerial Economics
Klein
Mathematical Methods for
Economics
Krugman/Obstfeld/Melitz†
International Economics:
Theory & Policy*
Laidle
The Demand for Money
Lynn
Economic Development:
Theory and Practice for
a Divided World
Mille
Economics Today*
Mille
Benjamin
The Economics of Macro Issues
Mille
Benjamin/North
The Economics of Public Issues
Mishkin†
The Economics of Money,
Banking, and Financial
Markets*
The Economics of Money,
Banking, and Financial
Markets, Business School
Edition*
Macroeconomics: Policy
and Practice*
Mu
ay
Econometrics: A Modern
Introduction
O’Sullivan/Sheffrin/Perez
Economics: Principles,
Applications and Tools*
Parkin†
Economics*
Perloff†
Microeconomics*
Microeconomics: Theory and
Applications with Calculus*
Perloff/Brande
Managerial Economics
and Strategy*
Pindyck/Rubinfeld†
Microeconomics*
Riddell/Shackelford/Stamos
Schneide
Economics: A Tool for Critically
Understanding Society
Roberts
The Choice: A Fable of Free
Trade and Protection
Schere
Industry Structure, Strategy,
and Public Policy
Schille
The Economics of Poverty and
Discrimination
Sherman
Market Regulation
Stock/Watson†
Introduction to Econometrics
Studenmund†
Using Econometrics:
A Practical Guide
Todaro/Smith
Economic Development
Walters/Walters/Appel/
Callahan/Centanni/Maex
O’Neill
Econversations:
Today’s Students Discuss
Today’s Issues
Williamson†
Macroeconomics
*denotes Pearson MyLab Economics titles
†denotes availability of Global Edition titles Visit www.myeconlab.com to learn more.
http:
www.myeconlab.com
MicroeconoMics
Massachusetts Institute of Technology
RobeRt S. Pindyck
University of California, Berkeley
daniel l. Rubinfeld
ninth edition
GLoBAL edition
Harlow, England • London • New York • Boston • San Francisco • Toronto • Sydney • Dubai • Singapore • Hong Kong
Tokyo • Seoul • Taipei • New Delhi • Cape Town • Sao Paulo • Mexico City • Madrid • Amsterdam • Munich • Paris • Milan
Vice President, Business Publishing: Donna Battista
Director of Portfolio Management: Adrienne D’Am
osio
Portfolio Manager: Ashley Bryan
Associate Project Editor, Global Edition: Paromita Banerjee
Editorial Assistant: Nicole Nedwidek
Vice President, Product Marketing: Roxanne McCarley
Director of Strategic Marketing: Brad Parkins
Strategic Marketing Manager: Deborah Strickland
Product Marketer: Tricia Murphy
Field Marketing Manager: Ramona Elme
Field Marketing Assistant: Kristen Compton
Product Marketing Assistant: Jessica Quazza
Vice President, Production and Digital Studio, Arts
and Business: Etain O’Dea
Director of Production, Business: Jeff Holcom
Managing Producer, Business: Alison Kalil
Content Producer, Global Edition: Pooja Aggarwal
Manufacturing Controller,
Global Edition: Kay Holman
Content Producer: Mary Kate Mu
ay
Operations Specialist: Carol Melville
Creative Director: Blair Brown
Manager, Learning Tools: Brian Surette
Managing Producer, Digital Studio, Arts
and Business: Diane Lombardo
Digital Studio Producer: Melissa Honig
Digital Studio Producer: Alana Coles
Digital Content Team Lead: Noel Lotz
Digital Content Project Lead: Noel Lotz
Media Production Manager, Global Edition: Vikram Kuma
Full-Service Project Management and Composition:
Integra Software Services
Interior Design: Integra Software Services
Cover Art: LeksusTuss/Shutterstock
For information regarding permissions, request forms, and the appropriate contacts within the Pearson
Education Global Rights and Permissions department, please visit www.pearsoned.com/permissions/.
Acknowledgments of third-party content appear on page 768, which constitutes an extension of this
copyright page.
PEARSON, ALWAYS LEARNING, and Pearson MyLab Economics® are exclusive trademarks owned by
Pearson Education, Inc. or its affiliates in the U.S. and/or other countries.
Pearson Education Limited
Edinburgh Gate
Harlow
Essex CM20 2JE
England
and Associated Companies throughout the world
Visit us on the World Wide Web at: www.pearsonglobaleditions.com
© Pearson Education Limited 2018
The rights of Robert Pindyck and Daniel Rubinfeld to be identified as the authors of this work have been
asserted by them in accordance with the Copyright, Designs and Patents Act 1988.
Authorized adaptation from the United States edition, entitled Microeconomics, 9th Edition, ISBN 978-0-13-418424-1
y Robert Pindyck and Daniel Rubinfeld, published by Pearson Education © 2018.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted
in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without either
the prior written permission of the publisher or a license permitting restricted copying in the United King-
dom issued by the Copyright Licensing Agency Ltd, Saffron House, 6–10 Ki
y Street, London EC1N 8TS.
All trademarks used herein are the property of their respective owners. The use of any trademark in this text
does not vest in the author or publisher any trademark ownership rights in such trademarks, nor does the
use of such trademarks imply any affiliation with or endorsement of this book by such owners.
ISBN 10: 1-292-21331-0
ISBN 13: 978-1-292-21331-6
British Li
ary Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Li
ary
10 9 8 7 6 5 4 3 2 1
Typeset in Palatino LT Pro by Integra-PDY IN
Printed and bound by Vivar in Malaysia
http:
www.pearsoned.com/permissions
http:
www.pearsonglobaleditions.com
To our daughters,
Maya, Talia, and Shira
Sarah and Rachel
6
Revising a textbook every three or four years is hard work, and the last edition was well-liked by students. “So why is our publisher pushing for a new edition?” the authors wondered. “Were some of the examples
ecoming stale? Or might it have something to do with the used book market?”
Could be both. In any case, here they are again, with a new edition that has sub-
stantial improvements and lots of new examples.
Robert S. Pindyck is the Bank of Tokyo-Mitsubishi Ltd. Professor of Econom-
ics and Finance in the Sloan School of Management at M.I.T. Daniel L. Rubinfeld
is the Robert L. Bridges Professor of Law and Professor of Economics Emeritus
at the University of California, Berkeley, and Professor of Law at NYU. Both re-
ceived their Ph.D.s from M.I.T., Pindyck in 1971 and Rubinfeld in 1972. Professor
Pindyck’s research and writing have covered a variety of topics in microeconom-
ics, including the effects of uncertainty on firm behavior and market structure;
the behavior of natural resource, commodity, and financial markets; environmen-
tal economics; and criteria for investment decisions. Professor Rubinfeld, who
served as chief economist at the Department of Justice in 1997 and 1998, is the
author of a variety of articles relating to antitrust, competition policy, law and
economics, law and statistics, and public economics.
Pindyck and Rubinfeld are also co-authors of Econometric Models and Economic
Forecasts, another best-selling textbook that makes a perfect gift (birthdays, wed-
dings, bar mitzvahs, you name it) for the man or woman who has everything.
(Buy several—bulk pricing is available.) These two authors are always looking
for ways to earn some extra spending money, so they enrolled as human subjects
in a double-blind test of a new hair restoration medication. Rubinfeld strongly
suspects that he is being given the placebo.
This is probably more than you want to know about these authors, but for fur-
ther information, see their Web sites: http:
web.mit.edu
pindyck/www/ and
https:
www.law.berkeley.edu/our-faculty/faculty-profiles/daniel-rubinfeld
abouT The auThors
The authors, back again for a
new edition, reflect on their
years of successful textbook
collaboration. Pindyck is on the
ight and Rubinfeld on the left.
http:
web.mit.edu
pindyck/www
https:
www.law.berkeley.edu/our-faculty/faculty-profiles/daniel-rubinfeld
ief conTenTs
7
PArt one
introduction: Markets and Prices 23
1 Preliminaries 25
2 The Basics of Supply and Demand 43
PArt two
Producers, consumers, and competitive Markets 87
3 Consumer Behavior 89
4 Individual and Market Demand 131
5 Uncertainty and Consumer Behavior 179
6 Production 209
7 The Cost of Production 237
8 Profit Maximization and Competitive Supply 289
9 The Analysis of Competitive Markets 327
PArt three
Market Structure and competitive Strategy 367
10 Market Power: Monopoly and Monopsony 369
11 Pricing with Market Power 413
12 Monopolistic Competition and Oligopoly 465
13 Game Theory and Competitive Strategy 501
14 Markets for Factor Inputs 543
15 Investment, Time, and Capital Markets 573
PArt Four
information, Market failure, and the
Role of Government 607
16 General Equili
ium and Economic Efficiency 609
17 Markets with Asymmetric Information 645
18 Externalities and Public Goods 675
19 Behavioral Economics 713
Appendix: The Basics of Regression 735
Glossary 743
Answers to Selected Exercises 753
Photo Credits 768
Index 769
This page intentionally left blank
conTenTs
Preface 15
PArt one
introduction: Markets and Prices 23
1 Preliminaries 25
1.1 The Themes of Microeconomics 26
Trade-Offs 26
Prices and Markets 27
Theories and Models 27
Positive versus Normative Analysis 28
1.2 What Is a Market? 29
Competitive versus Noncompetitive Markets 30
Market Price 30
Market Definition—The Extent of a Market 31
1.3 Real versus Nominal Prices 34
1.4 Why Study Microeconomics? 39
Corporate Decision Making: The Toyota
Prius 39
Public Policy Design: Fuel Efficiency Standards for
the Twenty-First Century 40
Summary 41
Questions for Review 41
Exercises 42
2 The basics of supply and
Demand 43
2.1 Supply and Demand 44
The Supply Curve 44
The Demand Curve 45
2.2 The Market Mechanism 47
2.3 Changes in Market Equili
ium 48
2.4 Elasticities of Supply and Demand 55
Point versus Arc Elasticities 58
2.5 Short-Run versus Long-Run Elasticities 62
Demand 62
Supply 67
*2.6 Understanding and Predicting the Effects of
Changing Market Conditions 71
2.7 Effects of Government Intervention—Price
Controls 80
Summary 83
Questions for Review 83
Exercises 84
PArt two
Producers, consumers,
and competitive Markets 87
3 consumer behavior 89
Consumer Behavior 89
3.1 Consumer Preferences 91
Market Baskets 91
Some Basic Assumptions about Preferences 92
Indifference Curves 93
Indifference Maps 94
The Shape of Indifference Curves 95
The Marginal Rate of Substitution 96
Perfect Substitutes and Perfect Complements 98
3.2 Budget Constraints 104
The Budget Line 104
The Effects of Changes in Income and Prices 106
3.3 Consumer Choice 108
Corner Solutions 111
3.4 Revealed Preference 114
3.5 Marginal Utility and Consumer Choice 117
Rationing 120
*3.6 Cost-of-Living Indexes 122
Ideal Cost-of-Living Index 123
Laspeyres Index 124
Paasche Index 125
Price Indexes in the United States: Chain
Weighting 126
Summary 127
Questions for Review 128
Exercises 128
9
10 contents
4 individual and market
Demand 131
4.1 Individual Demand 132
Price Changes 132
The Individual Demand Curve 132
Income Changes 134
Normal versus Inferior Goods 135
Engel Curves 136
Substitutes and Complements 138
4.2 Income and Substitution Effects 139
Substitution Effect 140
Income Effect 141
A Special Case: The Giffen Good 142
4.3 Market Demand 144
From Individual to Market Demand 144
Elasticity of Demand 146
Speculative Demand 149
4.4 Consumer Surplus 152
Consumer Surplus and Demand 152
4.5 Network Externalities 155
Positive Network Externalities 155
Negative Network Externalities 157
*4.6 Empirical Estimation of Demand 159
The Statistical Approach to Demand Estimation 160
The Form of the Demand Relationship 161
Interview and Experimental Approaches to Demand
Determination 163
Summary 164
Questions for Review 164
Exercises 165
5 uncertainty and consumer
ehavior 179
5.1 Describing Risk 180
Probability 180
Expected Value 181
Variability 181
Decision Making 183
5.2 Preferences Toward Risk 185
Different Preferences Toward Risk 186
5.3 Reducing Risk 190
Diversification 190
Insurance 191
The Value of Information 194
*5.4 The Demand for Risky Assets 196
Assets 196
Risky and Riskless Assets 197
Asset Returns 197
The Trade-Off Between Risk and Return 199
The Investor’s Choice Problem 200
Summary 205
Questions for Review 205
Exercises 205
6 Production 209
The Production Decisions of a Firm 209
6.1 Firms and Their Production Decisions 210
Why Do Firms Exist? 211
The Technology of Production 212
The Production Function 212
The Short Run versus the Long Run 213
6.2 Production with One Variable Input (Labor) 214
Average and Marginal Products 214
The Slopes of the Product Curve 215
The Average Product of Labor Curve 217
The Marginal Product of Labor Curve 217
The Law of Diminishing Marginal Returns 218
Labor Productivity 222
6.3 Production with Two Variable Inputs 224
Isoquants 224
Input Flexibility 226
Diminishing Marginal Returns 226
Substitution Among Inputs 226
Production Functions—Two Special Cases 228
6.4 Returns to Scale 231
Describing Returns to Scale 232
Summary 234
Questions for Review 234
Exercises 235
7 The cost of Production 237
7.1 Measuring Cost: Which Costs Matter? 237
Economic Cost versus Accounting Cost 238
Opportunity Cost 238
Sunk Costs 239
Fixed Costs and Variable Costs 241
Fixed versus Sunk Costs 242
Marginal and Average Cost 244
7.2 Cost in the Short Run 245
The Determinants of Short-Run Cost 245
The Shapes of the Cost Curves 246
7.3 Cost in the Long Run 251
The User Cost of Capital 251
The Cost-Minimizing Input Choice 252
The Isocost Line 253
Choosing Inputs 253
contents 11
Cost Minimization with Varying Output Levels 257
The Expansion Path and Long-Run Costs 258
7.4 Long-Run versus Short-Run Cost Curves 261
The Inflexibility of Short-Run Production 261
Long-Run Average Cost 262
Economies and Diseconomies of Scale 263
The Relationship between Short-Run
and Long-Run Cost 266
7.5 Production with Two Outputs—Economies
of Scope 267
Product Transformation Curves 267
Economies and Diseconomies of Scope 268
The Degree of Economies of Scope 269
*7.6 Dynamic Changes in Costs— The Learning
Curve 270
Graphing the Learning Curve 270
Learning versus Economies of Scale 271
*7.7 Estimating and Predicting Cost 275
Cost Functions and the Measurement of Scale
Economies 276
Summary 278
Questions for Review 279
Exercises 280
8 Profit maximization and
competitive supply 289
8.1 Perfectly Competitive Markets 289
When Is a Market Highly Competitive? 291
8.2 Profit Maximization 292
Do Firms Maximize Profit? 292
Alternative Forms of Organization 293
8.3 Marginal Revenue, Marginal Cost, and Profit
Maximization 294
Demand and Marginal Revenue for a Competitive
Firm 295
Profit Maximization by a Competitive Firm 297
8.4 Choosing Output in the Short Run 297
Short-Run Profit Maximization by a Competitive
Firm 297
When Should the Firm Shut Down? 299
8.5 The Competitive Firm’s Short-Run Supply
Curve 302
The Firm’s Response to an Input Price Change 303
8.6 The Short-Run Market Supply Curve 305
Elasticity of Market Supply 306
Producer Surplus in the Short Run 308
8.7 Choosing Output in the Long Run 310
Long-Run Profit Maximization 310
Long-Run Competitive Equili
ium 311
Economic Rent 314
Producer Surplus in the Long Run 315
8.8 The Industry’s Long-Run Supply Curve 316
Constant-Cost Industry 317
Increasing-Cost Industry 318
Decreasing-Cost Industry 319
The Effects of a Tax 320
Long-Run Elasticity of Supply 321
Summary 324
Questions for Review 324
Exercises 325
9 The analysis of competitive
markets 327
9.1 Evaluating the Gains and Losses from
Government Policies—Consumer and Producer
Surplus 327
Review of Consumer and Producer Surplus 328
Application of Consumer and Producer
Surplus 329
9.2 The Efficiency of a Competitive Market 333
9.3 Minimum Prices 338
9.4 Price Supports and Production Quotas 342
Price Supports 342
Production Quotas 344
9.5 Import Quotas and Tariffs 351
9.6 The Impact of a Tax or Subsidy 355
The Effects of a Subsidy 359
Summary 362
Questions for Review 362
Exercises 363
PArt three
Market Structure and competitive
Strategy 367
10 market Power: monopoly and
monopsony 369
10.1 Monopoly 370
Average Revenue and Marginal Revenue 370
The Monopolist’s Output Decision 371
An Example 373
A Rule of Thumb for Pricing 375
Shifts in Demand 377
The Effect of a Tax 378
*The Multiplant Firm 379
12 contents
10.2 Monopoly Power 380
Production, Price, and Monopoly Power 383
Measuring Monopoly Power 383
The Rule of Thumb for Pricing 384
10.3 Sources of Monopoly Power 387
The Elasticity of Market Demand 388
The Number of Firms 388
The Interaction Among Firms 389
10.4 The Social Costs of Monopoly Power 389
Rent Seeking 390
Price Regulation 391
Natural Monopoly 392
Regulation in Practice 393
10.5 Monopsony 394
Monopsony and Monopoly Compared 397
10.6 Monopsony Power 398
Sources of Monopsony Power 398
The Social Costs of Monopsony Power 399
Bilateral Monopoly 400
10.7 Limiting Market Power: The Antitrust Laws 401
Restricting What Firms Can Do 402
Enforcement of the Antitrust Laws 404
Antitrust in Europe 404
Summary 408
Questions for Review 409
Exercises 409
11 Pricing with market Power 413
11.1 Capturing Consumer Surplus 414
11.2 Price Discrimination 415
First-Degree Price Discrimination 415
Second-Degree Price Discrimination 418
Third-Degree Price Discrimination 418
11.3 Intertemporal Price Discrimination
and Peak-Load Pricing 424
Intertemporal Price Discrimination 425
Peak-Load Pricing 426
11.4 The Two-Part Tariff 428
*11.5 Bundling 433
Relative Valuations 434
Mixed Bundling 436
Bundling in Practice 440
Tying 443
*11.6 Advertising 443
A Rule of Thumb for Advertising 445
Summary 448
Questions for Review 448
Exercises 449
12 monopolistic competition and
oligopoly 465
12.1 Monopolistic Competition 466
The Makings of Monopolistic Competition 466
Equili
ium in the Short Run and the Long Run 467
Monopolistic Competition and Economic
Efficiency 468
12.2 Oligopoly 470
Equili
ium in an Oligopolistic Market 471
The Cournot Model 472
The Linear Demand Curve—An Example 475
First Mover Advantage—The Stackelberg Model 477
12.3 Price Competition 478
Price Competition with Homogeneous Products—
The Bertrand Model 478
Price Competition with Differentiated Products 479
12.4 Competition versus Collusion: The Prisoners’
Dilemma 483
12.5 Implications of the Prisoners’ Dilemma for
Oligopolistic Pricing 486
Price Rigidity 486
Price Signaling and Price Leadership 487
The Dominant Firm Model 490
12.6 Cartels 491
Analysis of Cartel Pricing 492
Summary 496
Questions for Review 497
Exercises 497
13 Game Theory and competitive
strategy 501
13.1 Gaming and Strategic Decisions 501
Noncooperative versus Cooperative Games 502
13.2 Dominant Strategies 504
13.3 The Nash Equili
ium Revisited 506
Maximin Strategies 508
*Mixed Strategies 510
13.4 Repeated Games 512
13.5 Sequential Games 517
The Extensive Form of a Game 517
The Advantage of Moving First 518
13.6 Threats, Commitments, and Credibility 519
Empty Threats 520
Commitment and Credibility 520
Bargaining Strategy 522
13.7 Entry Dete
ence 524
Strategic Trade Policy and International
Competition 527
contents 13
*13.8 Auctions 530
Auction Formats 531
Valuation and Information 531
Private-Value Auctions 532
Common-Value Auctions 533
Maximizing Auction Revenue 535
Bidding and Collusion 535
Summary 538
Questions for Review 538
Exercises 539
14 markets for factor inputs 543
14.1 Competitive Factor Markets 543
Demand for a Factor Input When Only One Input Is
Variable 544
Demand for a Factor Input When Several Inputs Are
Variable 547
The Market Demand Curve 548
The Supply of Inputs to a Firm 551
The Market Supply of Inputs 553
14.2 Equili
ium in a Competitive Factor Market 556
Economic Rent 556
14.3 Factor Markets with Monopsony power 560
Monopsony Power: Marginal and Average
Expenditure 560
Purchasing Decisions with Monopsony Power 561
Bargaining Power 562
14.4 Factor Markets with Monopoly Power 564
Monopoly Power over the Wage Rate 564
Unionized and Nonunionized Workers 566
Summary 569
Questions for Review 569
Exercises 570
15 investment, Time, and capital
markets 573
15.1 Stocks versus Flows 574
15.2 Present Discounted Value 575
Valuing Payment Streams 576
15.3 The Value of a Bond 578
Perpetuities 579
The Effective Yield on a Bond 580
15.4 The Net Present Value Criterion for Capital
Investment Decisions 583
The Electric Motor Factory 584
Real versus Nominal Discount Rates 585
Negative Future Cash Flows 586
15.5 Adjustments for Risk 587
Diversifiable versus Nondiversifiable Risk 588
The Capital Asset Pricing Model 589
15.6 Investment Decisions by Consumers 592
15.7 Investments in Human Capital 594
*15.8 Intertemporal Production Decisions—Depletable
Resources 598
The Production Decision of an Individual Resource
Producer 598
The Behavior of Market Price 599
User Cost 599
Resource Production by a Monopolist 600
15.9 How Are Interest Rates Determined? 602
A Variety of Interest Rates 603
Summary 604
Questions for Review 605
Exercises 605
PArt Four
information, Market failure, and the
Role of Government 607
16 General equili
ium and
economic efficiency 609
16.1 General Equili
ium Analysis 609
Two Interdependent Markets—Moving to General
Equili
ium 610
Reaching General Equili
ium 611
Economic Efficiency 615
16.2 Efficiency in Exchange 616
The Advantages of Trade 617
The Edgeworth Box Diagram 617
Efficient Allocations 618
The Contract Curve 620
Consumer Equili
ium in a Competitive Market 621
The Economic Efficiency of Competitive Markets 623
16.3 Equity and Efficiency 624
The Utility Possibilities Frontier 624
Equity and Perfect Competition 626
16.4 Efficiency in Production 627
Input Efficiency 627
The Production Possibilities Frontier 628
Output Efficiency 629
Efficiency in Output Markets 631
16.5 The Gains from Free Trade 632
Comparative Advantage 632
An Expanded Production Possibilities Frontier 633
16.6 An Overview—The Efficiency of Competitive
Markets 637
14 contents
16.7 Why Markets Fail 638
Market Power 639
Incomplete Information 639
Externalities 639
Public Goods 640
Summary 641
Questions for Review 641
Exercises 642
17 markets with asymmetric
information 645
17.1 Quality Uncertainty and the Market for
Lemons 646
The Market for Used Cars 646
Implications of Asymmetric Information 648
The Importance of Reputation and
Standardization 649
17.2 Market Signaling 653
A Simple Model of Job Market Signaling 654
Guarantees and Wa
anties 656
17.3 Moral Hazard 658
17.4 The Principal–Agent Problem 660
The Principal–Agent Problem in Private
Enterprises 660
The Principal–Agent Problem in Public
Enterprises 663
Incentives in the Principal–Agent Framework 664
*17.5 Managerial Incentives in an Integrated
Firm 666
Asymmetric Information and Incentive Design in the
Integrated Firm 666
Applications 668
17.6 Asymmetric Information in Labor Markets:
Efficiency Wage Theory 669
Summary 671
Questions for Review 672
Exercises 672
18 externalities and Public
Goods 675
18.1 Externalities 675
Negative Externalities and Inefficiency 676
Positive Externalities and Inefficiency 678
18.2 Ways of Co
ecting Market Failure 681
An Emissions Standard 682
An Emissions Fee 682
Standards versus Fees 683
Tradeable Emissions Permits 686
Recycling 689
18.3 Stock Externalities 693
Stock Buildup and Its Impact 694
18.4 Externalities and Property Rights 699
Property Rights 699
Bargaining and Economic Efficiency 700
Costly Bargaining—The Role of Strategic
Behavior 701
A Legal Solution—Suing for Damages 701
18.5 Common Property Resources 703
18.6 Public Goods 705
Efficiency and Public Goods 706
Public Goods and Market Failure 708
Summary 709
Questions for Review 710
Exercises 711
19 behavioral economics 713
19.1 Reference Points and Consumer
Preferences 714
19.2 Fairness 718
19.3 Rules of Thumb and Biases in Decision
Making 719
19.4 Bu
les 726
Informational Cascades 728
19.5 Behavioral Economics and Public
Policy 731
Summing Up 733
Summary 733
Questions for Review 734
Exercises 734
Appendix: The Basics of Regression 735
Glossary 743
Answers to Selected Exercises 753
Photo Credits 768
Index 769
Preface
For students who care about how the world works, microeconomics is prob-ably the most relevant, interesting, and important subject they can study. (Macroeconomics is the second-most important subject.) A good grasp
of microeconomics is vital for managerial decision making, for designing and
understanding public policy, and, more generally, for appreciating how a mod-
ern economy functions. In fact, even understanding the news each day often
equires knowledge of microeconomics.
We wrote this book, Microeconomics, because we believe that students need to
e exposed to the new topics that have come to play a central role in microeco-
nomics over the years—topics such as game theory and competitive strategy, the
oles of uncertainty and information, and the analysis of pricing by firms with
market power. We also felt that students need to be shown how microeconomics
can help us to understand what goes on in the world and how it can be used as
a practical tool for decision making. Microeconomics is an exciting and dynamic
subject, but students need to be given an appreciation of its relevance and use-
fulness. They want and need a good understanding of how microeconomics can
actually be used outside the classroom.
To respond to these needs, the ninth edition of Microeconomics provides a
treatment of microeconomic theory that stresses its relevance and application
to both managerial and public policy decision making. This applied emphasis
is accomplished by including examples that cover such topics as the analysis of
demand, cost, and market efficiency; the design of pricing strategies; investment
and production decisions; and public policy analysis. Because of the importance
that we attach to these examples, they are included in the flow of the text. (A
complete list is included on the endpapers inside the front cover.)
The coverage in this edition of Microeconomics incorporates the dramatic chang-
es that have occu
ed in the field in recent years. There has been growing interest
in game theory and the strategic interactions of firms (Chapters 12 and 13), in
the role and implications of uncertainty and asymmetric information (Chapters 5
and 17), in the pricing strategies of firms with market power (Chapters 10 and 11),
in the design of policies to deal efficiently with externalities such as environmental
pollution (Chapter 18), and in behavioral economics (Chapter 19).
That the coverage in Microeconomics is comprehensive and up to date does
not mean that it is “advanced” or difficult. We have worked hard to make the
exposition clear and accessible as well as lively and engaging. We believe that
the study of microeconomics should be enjoyable and stimulating. We hope that
our book reflects this belief. Except for appendices and footnotes, Microeconom-
ics uses no calculus. As a result, it should be suitable for students with a
oad
ange of backgrounds. (Those sections that are more demanding are marked
with an asterisk and can be easily omitted.) 15
16 PreFAce
Changes in the Ninth Edition
Each new edition of this book is built on the success of prior editions by adding some new topics, by adding and updating examples, and by im-proving the exposition of existing materials. We continue that tradition in
this ninth edition. We have made a number of changes throughout the book, but
the most important are the following:
•    We added a new chapter (Chapter 19) on behavioral economics. Behav-
ioral economics goes beyond the simple paradigm of maximizing some-
thing (e.g., utility, output, profit) subject to a constraint (e.g., income, cost,
demand and cost). While this paradigm has been extremely powerful in
helping us understand how markets work, it does not accurately describe
how real-world consumers and firms behave. The new and flourishing field
of behavioral economics incorporates findings from psychology into our
descriptions of how consumers and firms make decisions. Although the
previous edition of this book had a section on behavioral economics (that
appeared in Chapter 5), we decided that this topic was sufficiently impor-
tant to deserve a chapter of its own.
We have updated many of the examples (as we do in every new edition), but
we also added several new ones.
•    We now have several examples of taxicab markets that include the entry of
“ride-share” services like Uber and Lyft (Chapters 9 and 13).
•    We added an example about Tesla’s new battery factory (its “Gigafactory”)
and how scale economies will reduce the cost of batteries for electric cars
(Chapter 7).
•    We added an example on merger policy (Chapter 10) and one on the Auto
Parts Cartel (Chapter 12).
•    We even have two examples (in Chapters 1 and 12) that deal with the pricing
of this textbook.
•    As part of the new Chapter 19, we added several examples that are “behav-
ioral” in nature, including consumers’ use of credit card debt (and apparent
willingness to pay extremely high interest rates) and decisions to join and
use health clubs.
•    With the exception of the new Chapter 19, the layout of this edition is simi-
lar to that of the prior edition. This has allowed us to continue to define key
terms in the margins (as well as in the Glossary at the end of the book) and to
use the margins to include Concept Links that relate newly developed ideas
to concepts introduced previously in the text.
Alternative Course Designs
This new edition of Microeconomics offers instructors considerable flexibil-ity in course design. For a one-quarter or one-semester course stressing the basic core material, we would suggest using the following chapters
and sections of chapters: 1 through 6, 7.1–7.4, 8 through 10, 11.1–11.3, 12, 14,
15.1–15.4, 18.1–18.2, and 18.5. A somewhat more ambitious course might also in-
clude parts of Chapters 5, 16, and 19 and additional sections in Chapters 7 and 9.
PreFAce 17
To emphasize uncertainty and market failure, an instructor should also include
substantial parts of Chapters 5 and 17.
Depending on one’s interests and the goals of the course, other sections could
e added or used to replace the materials listed above. A course emphasizing
modern pricing theory and business strategy would include all of Chapters 11,
12, and 13 and the remaining sections of Chapter 15. A course in managerial
economics might also include the appendices to Chapters 4, 7, and 11, as well as
the appendix on regression analysis at the end of the book. A course stressing
welfare economics and public policy should include Chapter 16 and additional
sections of Chapters 18 and 19.
Finally, we want to stress that those sections or subsections that are more
demanding and/or peripheral to the core material have been marked with an
asterisk. These sections can easily be omitted without detracting from the flow
of the book.
Supplementary Materials
Ancillaries of an exceptionally high quality are available to instructors and students using this book. The Instructor’s Manual, prepared by Duncan M. Holthausen of North Carolina State University, provides detailed so-
lutions to all end-of-chapter Questions for Review and Exercises. The ninth edi-
tion contains many entirely new review questions and exercises, and a number
of exercises have been revised and updated. The new instructor’s manual has
een revised accordingly. Each chapter also contains Teaching Tips to summa-
ize key points.
The Test Item File contains approximately 2,000 multiple-choice and short-
answer questions with solutions. All of this material has been thoroughly
eviewed, accuracy checked, and revised for this edition. TestGen is a com-
puterized test generation program, available exclusively from Pearson, that
allows instructors to easily create and administer tests on paper, electronically,
or online. Instructors can select test items from the publisher-supplied test
ank, which is organized by chapter and based on the associated textbook ma-
terial, or create their own questions from scratch. With both quick and simple
test creation and flexible and robust editing tools, TestGen is a complete test
generator system for today’s educators.
The PowerPoint Presentation has been revised for this edition by Fernando
Quijano. Instructors can edit the detailed outlines to create their own full- color,
professional-looking presentations and customized handouts for students.
The PowerPoint Presentation also contains lecture notes and a complete set of
animated textbook figures.
For your convenience, all instructor resources are available online via our
centralized supplements Web site, the Instructor Resource Center (www
.pearsonglobaleditions.com/Pindyck). For access or more information, con-
tact your local Pearson representative or request access online at the Instruc-
tor Resource Center.
http:
www.pearsonglobaleditions.com/Pindyck
http:
www.pearsonglobaleditions.com/Pindyck
18 PreFAce
Pearson MyLab Economics is a content-rich Web site with homework, quiz,
test, and tutorial options related to the ninth edition of Microeconomics. Pearson
MyLab Economics offers students an opportunity to sharpen their problem-
solving skills and to assess their understanding of text material in one program.
Similarly, instructors can manage all assessment needs in one program.
Pearson MyLab Economics contains:
•    End-of-chapter exercises available for practice or auto-graded assignment.
These exercises include algorithmic, numerical, and draw-graph exercises.
•    Additional exercises for assignment that draws upon material in the text.
•    Instant tutorial feedback on a student’s problem and graphing responses.
•    Interactive Learning Aids including Help Me Solve This step-by-step tutorials
and graph animations.
•    Auto Graded Problems and Graphs for all assignments.
•    Digital Interactives are engaging assessment activities that promote critical
thinking and application of key economic principles.
•    Test Item File questions for homework assignment.
•    A Custom Exercise Builder that allows instructors to create their own problems.
•    A Gradebook that records student performance and generates reports by
student or chapter.
•    Experiments in two versions, Single Player (for easy, asynchronous, interac-
tive homework assignments) and Multiplayer (for a fast paced, instructor-
led, synchronous, interactive experience).
•    The Pearson eText gives students access to their textbook anytime, any-
where. Students actively read, with access to note-taking, highlighting, and
ookmarking. Instructors can share comments or highlights, and students
can add their own, for a tight community of learners in any class.
•    Communication tools that enable students and instructors to communicate
through email, discussion board, chat, and ClassLive.
•    Customization options that provide additional ways to share documents
and add content.
•    Prebuilt courses offer a turn-key way for instructors to create a course that
includes pre-built assignments distributed by chapter.
•    A fourteen-day grace period that offers students temporary access as they
wait for financial aid.
The Pearson MyLab Economics exercises for Microeconomics were created by
Duncan M. Holthausen at North Carolina State University. For additional infor-
mation and a demonstration, visit www.myeconlab.com.
Acknowledgments
As the saying goes, it takes a village to revise a textbook. Because the ninth edition of Microeconomics has been the outgrowth of years of experience in the classroom, we owe a debt of gratitude to our students and to the
colleagues with whom we often discuss microeconomics and its presentation.
http:
www.myeconlab.com
PreFAce 19
We  have also had the help of capable research assistants. For the first eight
editions of the book, these included Peter Adams, Walter Athier, Smita Brun-
nerneier, Corola Conces, Phillip Gi
s, Matt Hartman, Salar Jahedi, Jamie Jue,
Rashmi Khare, Jay Kim, Maciej Kotowski, Catherine Martin, Tammy McGav-
ock, Masaya Okoshi, Kathy O’Regan, Shira Pindyck, Karen Randig, Subi Ran-
gan, Deborah Senior, Ashesh Shah, Nicola Stafford, and Wilson Tai. Kathy Hill
helped with the art, while Assunta Kent, Mary Knott, and Dawn Elliott Linahan
provided secretarial assistance with the first edition. We especially want to thank
Lynn Steele and Jay Tharp, who provided considerable editorial support for the
second edition. Mark Glickman and Steve Wiggins assisted with the examples in
the third edition, while Andrew Guest, Jeanette Sayre, and Lynn Steele provided
valuable editorial support for the third, fourth, and fifth editions, as did Brandi
Henson and Jeanette Sayre for the sixth edition, Ida Ng for the seventh edition,
and Ida Ng and Dagmar Trantinova for the eighth and ninth editions. In addi-
tion, Caterina Castellano and Sarah Tang provided supe
research assistance
on this ninth edition.
Writing this book has been both a painstaking and enjoyable process. At each
stage we received exceptionally fine guidance from teachers of microeconomics
throughout the country. After the first draft of the first edition of the book had
een edited and reviewed, it was discussed at a two-day focus group meeting
in New York. This provided an opportunity to get ideas from instructors with
a variety of backgrounds and perspectives. We would like to thank the follow-
ing focus group members for advice and criticism: Carl Davidson of Michigan
State University; Richard Eastin of the University of Southern California; Judith
Roberts of California State University, Long Beach; and Charles Strein of the
University of Northern Iowa.
We would like to thank the reviewers who provided comments and ideas that
have contributed significantly to the ninth edition of Microeconomics:
Bahram Adrangi, University of Portland
Richard Anderson, Texas A&M University
Bryan D. Buckley, University of Illinois
at U
ana-Champaign
Michael Enz, Framingham State University
Da
in Gulla, University of Kentucky
John Horn, Washington University in St. Louis
Robert Horn, James Madison University
Muhammad Husain, Georgia State University
Siew Hoon Lim, North Dakota State University
Frank Limehouse, DePaul University
Edward Scahill, The University of Scranton
Kimberly Sims, University of Tennessee Knoxville
Ralph Sonenshine, American University
Tom Vukina, North Carolina State University
Roger E. Wehr, The University of Texas at Arlington
Nii Adote A
ahams, Missouri Southern State
College
Jack Adams, University of Arkansas, Little Rock
Sheri Aggarwal, Dartmouth College
Anca Alecsandru, Louisiana State University
Anita Alves Pena, Colorado State University
Ted Amato, University of North Carolina, Charlotte
John J. Antel, University of Houston
Albert Assibey-Mensah, Kentucky State University
Ke
y Back, Northwestern University
Dale Ballou, University of Massachusetts, Amherst
William Baxter, Stanford University
Charles A. Bennett, Gannon University
Gregory Besharov, Duke University
Maharukh Bhiladwalla, Rutgers University
Victor Brajer, California State University, Fullerton
James A. Brander, University of British Columbia
David S. Bullock, University of Illinois
Jeremy Bulow, Stanford University
Donald L. Bumpass, Sam Houston State University
Raymonda Burgman, DePauw University
H. Stuart Burness, University of New Mexico
We would also like to thank all those who reviewed the first eight editions at
various stages of their evolution:
20 PreFAce
Peter Calcagno, College of Charleston
Winston Chang, State University of New York,
Buffalo
Henry Chappel, University of South Carolina
Joni Charles, Texas State University–San Marcos
La
y A. Chenault, Miami University
Ha
ison Cheng, University of Southern California
Eric Chiang, Florida Atlantic University
Kwan Choi, Iowa State University
Charles Clotfelter, Duke University
Ben Collier, Northwest Missouri State University
Kathryn Combs, California State University,
Los Angeles
Tom Cooper, Georgetown College
Richard Corwall, Middlebury College
John Coupe, University of Maine at Orono
Robert Crawford, Ma
iott School, Brigham Young
University
Jacques Cremer, Virginia Polytechnic Institute and
State University
Julie Cullen, University of California, San Diego
Carl Davidson, Michigan State University
Gilbert Davis, University of Michigan
Arthur T. Denzau, Washington University
Tran Dung, Wright State University
Richard V. Eastin, University of Southern California
Lee Endress, University of Hawaii
Maxim Engers, University of Virginia
Carl E. Enomoto, New Mexico State University
Ray Fa
ow, Seattle University
Tammy R. Feldman, University of Michigan
Gary Fe
ier, Southern Methodist University
Todd Matthew Fitch, University of San Francisco
John Francis, Auburn University, Montgomery
Roger Frantz, San Diego State University
Delia Furtado, University of Connecticut
Craig Gallet, California State University, Sacramento
Patricia Gladden, University of Missouri
Michele Glower, Lehigh University
Otis Gilley, Louisiana Tech University
Tiffani Gottschall, Washington & Jefferson College
William H. Greene, New York University
Thomas J. Grennes, North Carolina State University
Thomas A. Gresik, Notre Dame University
John Gross, University of Wisconsin at Milwaukee
Adam Grossberg, Trinity College
Philip Grossman, Saint Cloud State University
Nader Habibi, Brandeis University
Jonathan Hamilton, University of Florida
Claire Hammond, Wake Forest University
Robert G. Hansen, Dartmouth College
Bruce Hartman, California State University,
The California Maritime Academy
James Hartigan, University of Oklahoma
Daniel Henderson, Binghamton University
George Heitman, Pennsylvania State University
Wayne Hickenbottom, University of Texas at Austin
George E. Hoffer, Virginia Commonwealth
University
Stella Hofrenning, Augsburg College
Donald Holley, Boise State University
Duncan M. Holthausen, North Carolina State
University
Robert Inman, The Wharton School, University of
Pennsylvania
Brian Jacobsen, Wisconsin Lutheran College
Joyce Jacobsen, Rhodes College
Jonatan Jelen, New York University
Changik Jo, Anderson University
B. Patrick Joyce, Michigan Technological University
Mahbubul Kabir, Lyon College
Folke Kafka, University of Pittsburgh
David Kaserman, Auburn University
Brian Kench, University of Tampa
Michael Kende, INSEAD, France
Philip G. King, San Francisco State University
Paul Koch, Olivet Nazarene University
Tetteh A. Kofi, University of San Francisco
Dennis Kovach, Community College of Allegheny
County
Anthony Krautman, DePaul University
Leonard Lardaro, University of Rhode Island
Sang Lee, Southeastern Louisiana University
Robert Lemke, Florida International University
Peter Linneman, University of Pennsylvania
Leonard Loyd, University of Houston
R. Ashley Lyman, University of Idaho
James MacDonald, Rensselaer Polytechnical
Institute
Wesley A. Magat, Duke University
Peter Marks, Rhode Island College
Anthony M. Marino, University of Southern
California
Lawrence Martin, Michigan State University
John Makum Mbaku, Weber State University
Richard D. McGrath, College of William and Mary
Douglas J. Miller, University of Missouri–Columbia
David Mills, University of Virginia, Charlottesville
Richard Mills, University of New Hampshire
Jennifer Moll, Fairfield University
Michael J. Moore, Duke University
W. D. Morgan, University of California at Santa
Ba
ara
Julianne Nelson, Stern School of Business, New York
University
George Norman, Tufts University
PreFAce 21
Laudo Ogura, Grand Valley State University
June Ellenoff O’Neill, Baruch College
Daniel O
, Virginia Polytechnic Institute and State
University
Ozge Ozay, University of Utah
Christos Paphristodoulou, Mälardalen University
Lourenço Paz, Syracuse University
Sharon J. Pearson, University of Alberta, Edmonton
Ivan P’ng, University of California, Los Angeles
Michael Podgursky, University of Massachusetts,
Amherst
Jonathan Powers, Knox College
Lucia Quesada, Universidad Torcuato Di Telia
Benjamin Rashford, Oregon State University
Charles Ratliff, Davidson College
Judith Roberts, California State University,
Long Beach
Fred Rodgers, Medaille College
William Rogers, University of Missouri–Saint Louis
Geoffrey Rothwell, Stanford University
Nestor Ruiz, University of California, Davis
Edward L. Sattler, Bradley University
Roger Sherman, University of Virginia
Nachum Sicherman, Columbia University
Sigbjørn Sødal, Agder University College
Menahem Spiegel, Rutgers University
Houston H. Stokes, University of Illinois, Chicago
Richard W. Stratton, University of Akron
Houston Stokes, University of Illinois at Chicago
Charles T. Strein, University of Northern Iowa
Charles Stuart, University of California, Santa
Ba
ara
Valerie Suslow, University of Michigan
Theofanis Tsoulouhas, North Carolina State
Mira Tsymuk, Hunter College, CUNY
Abdul Turay, Radford University
Sevin Ugural, Eastern Medite
anean University
Nora A. Underwood, University of California, Davis
Nikolaos Vettas, Duke University
David Vrooman, St. Lawrence University
Michael Wasylenko, Syracuse University
Thomas Watkins, Eastern Kentucky University
Robert Whaples, Wake Forest University
David Wharton, Washington College
Lawrence J. White, New York University
Michael F. Williams, University of St. Thomas
Beth Wilson, Humboldt State University
Arthur Woolf, University of Vermont
Chiou-nan Yeh, Alabama State University
Philip Young, University of Maryland
Peter Zaleski, Villanova University
Joseph Ziegler, University of Arkansas, Fayetteville
Apart from the formal review process, we are especially grateful to Jean
Andrews, Paul Anglin, J. C. K. Ash, Ernst Berndt, George Bittlingmayer, Severin
Borenstein, Paul Carlin, Whewon Cho, Setio Anga
o Dewo, Avinash Dixit, Frank
Fabozzi, Joseph Fa
ell, Frank Fisher, Jonathan Hamilton, Robert Inman, Joyce
Jacobsen, Paul Joskow, Stacey Kole, Preston McAfee, Jeannette Mortensen, John
Mullahy, Krishna Pendakur, Jeffrey Perloff, Ivan P’ng, A. Mitchell Polinsky, Judith
Roberts, Geoffrey Rothwell, Garth Saloner, Joel Schrag, Daniel Siegel, Thomas
Stoker, David Storey, James Walker, and Michael Williams, who were kind enough
to provide comments, criticisms, and suggestions as the various editions of this
ook developed.
Chapter 19 of this edition contains new and updated material on behavioral
economics, whose genesis owes much to the thoughtful comments of George
Akerlof. We also want to thank Caterina Castellano, who helped update the
examples, created new examples and end-of-chapter questions and exercises,
provided editorial assistance at all stages of the book’s production, and carefully
eviewed the page proofs of this edition.
We also wish to express our sincere thanks for the extraordinary effort those
at Macmillan, Prentice Hall, and Pearson made in the development of the vari-
ous editions of our book. Throughout the writing of the first edition, Bonnie
Lieberman provided invaluable guidance and encouragement; Ken MacLeod
kept the progress of the book on an even keel; Gerald Lombardi provided mas-
terful editorial assistance and advice; and John Molyneux ably oversaw the
ook’s production.
In the development of the second edition, we were fortunate to have the
encouragement and support of David Boelio, and the organizational and
22 PreFAce
editorial help of two Macmillan editors, Caroline Carney and Jill Lectka. The
second edition also benefited greatly from the supe
development editing of
Gerald Lombardi and from John Travis, who managed the book’s production.
Jill Lectka and Denise A
ott were our editors for the third edition, and we ben-
efited greatly from their input. Leah Jewell was our editor for the fourth edition;
her patience, thoughtfulness, and perseverance were greatly appreciated. Chris
Rogers provided continual and loyal guidance through editions five through
seven. With respect to this ninth edition, we are grateful to our Portfolio Manager
Ashley Bryan who has worked diligently through this major revision. We also ap-
preciate the efforts of our Content Producer, Mary Kate Mu
ay; Project Manager
with Integra, Gina Linko; Product Marketer, Tricia Murphy; Field Marketing
Manager, Ramona Elmer; Digital Content Project Lead, Noel Lotz; and Digital
Studio Producer, Melissa Honig.
We owe a special debt of thanks to Catherine Lynn Steele, whose supe
edi-
torial work ca
ied us through five editions of this book. Lynn passed away
on December 10, 2002. We miss her very much.
R.S.P.
D.L.R.
Part 1 surveys the scope of microeconomics and intro-
duces some basic concepts and tools.
Chapter 1 discusses the range of problems that microeconomics
addresses, and the kinds of answers it can provide. It also explains
what a market is, how we determine the boundaries of a market,
and how we measure market price.
Chapter 2 covers one of the most important tools of microeco-
nomics: supply-demand analysis. We explain how a competitive
market works and how supply and demand determine the prices
and quantities of goods and services. We also show how supply-
demand analysis can be used to determine the effects of changing
market conditions, including government intervention.
1 Preliminaries
25
2 The Basics of Supply
and Demand
43
ChaPTerS
Introduction: Markets
and Prices
ParT One
This page intentionally left blank
25
economics is divided into two main
anches: microeconomics and macroeconomics. Microeconomics deals with the behavior of individual economic units. These units include consumers,
workers, investors, owners of land, business firms—in fact, any indi-
vidual or entity that plays a role in the functioning of our economy.1
Microeconomics explains how and why these units make economic
decisions. For example, it explains how consumers make purchasing
decisions and how their choices are affected by changing prices and
incomes. It also explains how firms decide how many workers to hire
and how workers decide where to work and how much work to do.
Another important concern of microeconomics is how eco-
nomic units interact to form larger units—markets and industries.
Microeconomics helps us to understand, for example, why the
American automobile industry developed the way it did and how
producers and consumers interact in the market for automobiles. It ex-
plains how automobile prices are determined, how much automobile
companies invest in new factories, and how many cars are produced
each year. By studying the behavior and interaction of individual firms
and consumers, microeconomics reveals how industries and markets
operate and evolve, why they differ from one another, and how they
are affected by government policies and global economic conditions.
By contrast, macroeconomics deals with aggregate economic
quantities, such as the level and growth rate of national output, inter-
est rates, unemployment, and inflation. But the boundary between
macroeconomics and microeconomics has become less and less
distinct in recent years. The reason is that macroeconomics also in-
volves the analysis of markets—for example, the aggregate markets
for goods and services, labor, and corporate bonds. To understand
how these aggregate markets operate, we must first understand
the behavior of the firms, consumers, workers, and investors who
constitute them. Thus macroeconomists have become increasingly
concerned with the microeconomic foundations of aggregate eco-
nomic phenomena, and much of macroeconomics is actually an ex-
tension of microeconomic analysis.
1.1 The Market for Sweeteners 32
1.2 a Bicycle Is a Bicycle.
Or Is It? 33
1.3 The Price of eggs and
the Price of a College
education 35
1.4 The authors Debate the
Minimum Wage 36
1.5 health Care and
College Textbooks 37
LIST Of exaMPLeS
1.1 The Themes of
Microeconomics 26
1.2 What Is a Market? 29
1.3 real versus nominal
Prices 34
1.4 Why Study
Microeconomics? 39
ChaPTer OuTLIne
Preliminaries
ChaPTer 1
1The prefix micro- is derived from the Greek word meaning “small.” However, many of
the individual economic units that we will study are small only in relation to the U.S.
economy as a whole. For example, the annual sales of General Motors, IBM, or Microsoft
are larger than the gross national products of many countries.
26 ParT 1 Introduction: Markets and Prices
1.1 The Themes of Microeconomics
The Rolling Stones once said: “You can’t always get what you want.” This is
true. For most people (even Mick Jagger), that there are limits to what you can
have or do is a simple fact of life learned in early childhood. For economists,
however, it can be an obsession.
Much of microeconomics is about limits—the limited incomes that consum-
ers can spend on goods and services, the limited budgets and technical know-
how that firms can use to produce things, and the limited number of hours in
a week that workers can allocate to labor or leisure. But microeconomics is also
about ways to make the most of these limits. More precisely, it is about the alloca-
tion of scarce resources. For example, microeconomics explains how consumers
can best allocate their limited incomes to the various goods and services avail-
able for purchase. It explains how workers can best allocate their time to labor
instead of leisure, or to one job instead of another. And it explains how firms
can best allocate limited financial resources to hiring additional workers versus
uying new machinery, and to producing one set of products versus another.
In a planned economy such as that of Cuba, North Korea, or the former
Soviet Union, these allocation decisions are made mostly by the government.
Firms are told what and how much to produce, and how to produce it; workers
have little flexibility in choice of jobs, hours worked, or even where they live;
and consumers typically have a very limited set of goods to choose from. As
a result, many of the tools and concepts of microeconomics are of limited rel-
evance in those countries.
Trade-Offs
In modern market economies, consumers, workers, and firms have much
more flexibility and choice when it comes to allocating scarce resources.
Microeconomics describes the trade-offs that consumers, workers, and firms
face, and shows how these trade-offs are best made.
The idea of making optimal trade-offs is an important theme in micro-
economics—one that you will encounter throughout this book. Let’s look at it in
more detail.
Consumers Consumers have limited incomes, which can be spent on a wide
variety of goods and services, or saved for the future. Consumer theory, the sub-
ject matter of Chapters 3, 4, and 5 of this book, describes how consumers, based
on their preferences, maximize their well-being by trading off the purchase of
more of some goods for the purchase of less of others. We will also see how con-
sumers decide how much of their incomes to save, thereby trading off cu
ent
consumption for future consumption.
Workers Workers also face constraints and make trade-offs. First, people
must decide whether and when to enter the workforce. Because the kinds of
jobs—and co
esponding pay scales—available to a worker depend in part on
educational attainment and accumulated skills, one must trade off working
now (and earning an immediate income) for continued education (and the hope
of earning a higher future income). Second, workers face trade-offs in their
choice of employment. For example, while some people choose to work for
large corporations that offer job security but limited potential for advancement,
others prefer to work for small companies where there is more opportunity for
microeconomics Branch of
economics that deals with the
ehavior of individual economic
units—consumers, firms, workers,
and investors—as well as the
markets that these units comprise.
macroeconomics Branch
of economics that deals with
aggregate economic variables,
such as the level and growth rate
of national output, interest rates,
unemployment, and inflation.
ChaPTer 1 PreLIMInarIeS 27
advancement but less security. Finally, workers must sometimes decide how
many hours per week they wish to work, thereby trading off labor for leisure.
Firms Firms also face limits in terms of the kinds of products that they can
produce, and the resources available to produce them. General Motors, for ex-
ample, is very good at producing cars and trucks, but it does not have the abil-
ity to produce airplanes, computers, or pharmaceuticals. It is also constrained in
terms of financial resources and the cu
ent production capacity of its factories.
Given these constraints, GM must decide how many of each type of vehicle to
produce. If it wants to produce a larger total number of cars and trucks next
year or the year after, it must decide whether to hire more workers, build new
factories, or do both. The theory of the firm, the subject matter of Chapters 6
and 7, describes how these trade-offs can best be made.
Prices and Markets
A second important theme of microeconomics is the role of prices. All of the
trade-offs described above are based on the prices faced by consumers, work-
ers, or firms. For example, a consumer trades off beef for chicken based partly
on his or her preferences for each one, but also on their prices. Likewise, work-
ers trade off labor for leisure based in part on the “price” that they can get for
their labor—i.e., the wage. And firms decide whether to hire more workers or
purchase more machines based in part on wage rates and machine prices.
Microeconomics also describes how prices are determined. In a centrally
planned economy, prices are set by the government. In a market economy,
prices are determined by the interactions of consumers, workers, and firms.
These interactions occur in markets—collections of buyers and sellers that to-
gether determine the price of a good. In the automobile market, for example,
car prices are affected by competition among Ford, General Motors, Toyota, and
other manufacturers, and also by the demands of consumers. The central role
of markets is the third important theme of microeconomics. We will say more
about the nature and operation of markets shortly.
Theories and Models
Like any science, economics is concerned with the explanations of observed phe-
nomena. Why, for example, do firms tend to hire or lay off workers when the prices
of their raw materials change? How many workers are likely to be hired or laid off
y a firm or an industry if the price of raw materials increases by, say, 10 percent?
In economics, as in other sciences, explanation and prediction are based on
theories. Theories are developed to explain observed phenomena in terms of a set
of basic rules and assumptions. The theory of the firm, for example, begins with
a simple assumption—firms try to maximize their profits. The theory uses this
assumption to explain how firms choose the amounts of labor, capital, and raw
materials that they use for production and the amount of output they produce.
It also explains how these choices depend on the prices of inputs, such as labor,
capital, and raw materials, and the prices that firms can receive for their outputs.
Economic theories are also the basis for making predictions. Thus the theory
of the firm tells us whether a firm’s output level will increase or decrease in
esponse to an increase in wage rates or a decrease in the price of raw materi-
als. With the application of statistical and econometric techniques, theories can
e used to construct models from which quantitative predictions can be made.
A model is a mathematical representation, based on economic theory, of a firm, a
28 ParT 1 Introduction: Markets and Prices
market, or some other entity. For example, we might develop a model of a par-
ticular firm and use it to predict by how much the firm’s output level will change
as a result of, say, a 10-percent drop in the price of raw materials.
Statistics and econometrics also let us measure the accuracy of our predictions.
For example, suppose we predict that a 10-percent drop in the price of raw materi-
als will lead to a 5-percent increase in output. Are we sure that the increase in out-
put will be exactly 5 percent, or might it be somewhere between 3 and 7 percent?
Quantifying the accuracy of a prediction can be as important as the prediction itself.
No theory, whether in economics, physics, or any other science, is perfectly
co
ect. The usefulness and validity of a theory depend on whether it succeeds
in explaining and predicting the set of phenomena that it is intended to explain
and predict. Theories, therefore, are continually tested against observation. As
a result of this testing, they are often modified or refined and occasionally even
discarded. The process of testing and refining theories is central to the develop-
ment of economics as a science.
When evaluating a theory, it is important to keep in mind that it is invariably
imperfect. This is the case in every
anch of science. In physics, for example,
Boyle’s law relates the volume, temperature, and pressure of a gas.2 The law is
ased on the assumption that individual molecules of a gas behave as though
they were tiny, elastic billiard balls. Physicists today know that gas molecules
do not, in fact, always behave like billiard balls, which is why Boyle’s law
eaks down under extremes of pressure and temperature. Under most condi-
tions, however, it does an excellent job of predicting how the temperature of a
gas will change when the pressure and volume change, and it is therefore an
essential tool for engineers and scientists.
The situation is much the same in economics. For example, because firms do
not maximize their profits all the time, the theory of the firm has had only limited
success in explaining certain aspects of firms’ behavior, such as the timing of
capital investment decisions. Nonetheless, the theory does explain a
oad range
of phenomena regarding the behavior, growth, and evolution of firms and indus-
tries, and has thus become an important tool for managers and policymakers.
Positive versus Normative Analysis
Microeconomics is concerned with both positive and normative questions.
Positive questions deal with explanation and prediction, normative questions
with what ought to be. Suppose the U.S. government imposes a quota on the
import of foreign cars. What will happen to the price, production, and sales of
cars? What impact will this policy change have on American consumers? On
workers in the automobile industry? These questions belong to the realm of
positive analysis: statements that describe relationships of cause and effect.
Positive analysis is central to microeconomics. As we explained above, theo-
ies are developed to explain phenomena, tested against observations, and used
to construct models from which predictions are made. The use of economic
theory for prediction is important both for the managers of firms and for pub-
lic policy. Suppose the federal government is considering raising the tax on
gasoline. The change would affect the price of gasoline, consumers’ purchasing
choices for small or large cars, the amount of driving that people do, and so on.
positive analysis analysis
describing relationships of cause
and effect.
2Robert Boyle (1627–1691) was a British chemist and physicist who discovered experimentally that
pressure (P), volume (V), and temperature (T) were related in the following way: PV = RT, where
R is a constant. Later, physicists derived this relationship as a consequence of the kinetic theory of
gases, which describes the movement of gas molecules in statistical terms.
ChaPTer 1 PreLIMInarIeS 29
To plan sensibly, oil companies, automobile companies, producers of auto-
mobile parts, and firms in the tourist industry would all need to estimate the
impact of the change. Government policymakers would also need quantitative
estimates of the effects. They would want to determine the costs imposed on
consumers (perhaps
oken down by income categories); the effects on profits
and employment in the oil, automobile, and tourist industries; and the amount
of tax revenue likely to be collected each year.
Sometimes we want to go beyond explanation and prediction to ask such
questions as “What is best?” This involves normative analysis, which is also
important for both managers of firms and those making public policy. Again,
consider a new tax on gasoline. Automobile companies would want to deter-
mine the best (profit-maximizing) mix of large and small cars to produce once
the tax is in place. Specifically, how much money should be invested to make
cars more fuel-efficient? For policymakers, the primary issue is likely to be
whether the tax is in the public interest. The same policy objectives (say, an in-
crease in tax revenues and a decrease in dependence on imported oil) might be
met more cheaply with a different kind of tax, such as a tariff on imported oil.
Normative analysis is not only concerned with alternative policy options; it
also involves the design of particular policy choices. For example, suppose it
has been decided that a gasoline tax is desirable. Balancing costs and benefits,
we then ask what is the optimal size of the tax.
Normative analysis is often supplemented by value judgments. For example,
a  comparison between a gasoline tax and an oil import tariff might conclude
that the gasoline tax will be easier to administer but will have a greater impact
on lower-income consumers. At that point, society must make a value judgment,
weighing equity against economic efficiency. When value judgments are involved,
microeconomics cannot tell us what the best policy is. However, it can clarify the
trade-offs and thereby help to illuminate the issues and sharpen the debate.
1.2 What Is a Market?
Business people, journalists, politicians, and ordinary consumers talk about
markets all the time—for example, oil markets, housing markets, bond markets,
labor markets, and markets for all kinds of goods and services. But often what
they mean by the word “market” is vague or misleading. In economics, markets
are a central focus of analysis, so economists try to be as clear as possible about
what they mean when they refer to a market.
It is easiest to understand what a market is and how it works by dividing
individual economic units into two
oad groups according to function—
uyers and sellers. Buyers include consumers, who purchase goods and ser-
vices, and firms, which buy labor, capital, and raw materials that they use
to produce goods and services. Sellers include firms, which sell their goods
and services; workers, who sell their labor services; and resource owners,
who rent land or sell mineral resources to firms. Clearly, most people and
most firms act as both buyers and sellers, but we will find it helpful to think
of them as simply buyers when they are buying something and sellers when
they are selling something.
Together, buyers and sellers interact to form markets. A market is the collec-
tion of buyers and sellers that, through their actual or potential interactions, determine
the price of a product or set of products. In the market for personal computers, for
example, the buyers are business firms, households, and students; the sellers are
normative analysis analysis
examining questions of what
ought to be.
market Collection of buyers
and sellers that, through their
actual or potential interactions,
determine the price of a product
or set of products.
30 ParT 1 Introduction: Markets and Prices
Hewlett-Packard, Lenovo, Dell, Apple, and a number of other firms. Note that a
market includes more than an industry. An industry is a collection of firms that sell the
same or closely related products. In effect, an industry is the supply side of the market.
Economists are often concerned with market definition—with determining
which buyers and sellers should be included in a particular market. When de-
fining a market, potential interactions of buyers and sellers can be just as impor-
tant as actual ones. An example of this is the market for gold. A New Yorker who
wants to buy gold is unlikely to travel to Zurich to do so. Most buyers of gold
in New York will interact only with sellers in New York. But because the cost of
transporting gold is small relative to its value, buyers of gold in New York could
purchase their gold in Zurich if the prices there were significantly lower.
Significant differences in the price of a commodity create a potential for
a
itrage: buying at a low price in one location and selling at a higher price some-
where else. The possibility of a
itrage prevents the prices of gold in New York
and Zurich from differing significantly and creates a world market for gold.
Markets are at the center of economic activity, and many of the most interest-
ing issues in economics concern the functioning of markets. For example, why
do only a few firms compete with one another in some markets, while in others
a great many firms compete? Are consumers necessarily better off if there are
many firms? If so, should the government intervene in markets with only a few
firms? Why have prices in some markets risen or fallen rapidly, while in other
markets prices have hardly changed at all? And which markets offer the best
opportunities for an entrepreneur thinking of going into business?
Competitive versus Noncompetitive Markets
In this book, we study the behavior of both competitive and noncompetitive
markets. A perfectly competitive market has many buyers and sellers, so that
no single buyer or seller has any impact on price. Most agricultural markets are
close to being perfectly competitive. For example, thousands of farmers pro-
duce wheat, which thousands of buyers purchase to produce flour and other
products. As a result, no single farmer and no single buyer can significantly af-
fect the price of wheat.
Many other markets are competitive enough to be treated as if they were
perfectly competitive. The world market for copper, for example, contains a few
dozen major producers. That number is enough for the impact on price to be
small if any one producer goes out of business. The same is true for many other
natural resource markets, such as those for coal, iron, tin, or lumber.
Other markets containing a small number of producers may still be treated
as competitive for purposes of analysis. For example, the U.S. airline industry
contains several dozen firms, but most routes are served by only a few firms.
Nonetheless, because competition among those firms is often fierce, for some
purposes airline markets can be treated as competitive. Finally, some markets
contain many producers but are noncompetitive; that is, individual firms can
jointly affect the price. The world oil market is one example. Since the early
1970s, that market has been dominated by the OPEC cartel. (A cartel is a group
of producers that acts collectively.)
Market Price
Markets make possible transactions between buyers and sellers. Quantities of
a good are sold at specific prices. In a perfectly competitive market, a single
price—the market price—will usually prevail. The price of wheat in Kansas
market definition
Determination of the buyers,
sellers, and range of products that
should be included in a particular
market.
a
itrage Practice of buying
at a low price at one location
and selling at a higher price in
another.
perfectly competitive
market Market with many
uyers and sellers, so that no
single buyer or seller has a
significant impact on price.
market price Price prevailing
in a competitive market.
ChaPTer 1 PreLIMInarIeS 31
City and the price of gold in New York are two examples. These prices are usu-
ally easy to measure. For example, you can find the price of corn, wheat, or gold
each day in the business section of a newspaper.
In markets that are not perfectly competitive, different firms might charge
different prices for the same product. This might happen because one firm is
trying to win customers from its competitors, or because customers have
and
loyalties that allow some firms to charge higher prices than others. For example,
two
ands of laundry detergent might be sold in the same supermarket at differ-
ent prices. Or two supermarkets in the same town might sell the same
and of
laundry detergent at different prices. In cases such as this, when we refer to the
market price, we will mean the price averaged across
ands or supermarkets.
The market prices of most goods will fluctuate over time, and for many
goods the fluctuations can be rapid. This is particularly true for goods sold in
competitive markets. The stock market, for example, is highly competitive be-
cause there are typically many buyers and sellers for any one stock. As anyone
who has invested in the stock market knows, the price of any particular stock
fluctuates from minute to minute and can rise or fall substantially during a
single day. Likewise, the prices of commodities such as wheat, soybeans, coffee,
oil, gold, silver, and lumber can rise or fall dramatically in a day or a week.
Market Definition—The Extent of a Market
As we saw, market definition identifies which buyers and sellers should be in-
cluded in a given market. However, to determine which buyers and sellers to
include, we must first determine the extent of a market—its boundaries, both
geographically and in terms of the range of products to be included in it.
When we refer to the market for gasoline, for example, we must be clear about
its geographic boundaries. Are we refe
ing to downtown Los Angeles, southern
California, or the entire United States? We must also be clear about the range of
products to which we are refe
ing. Should regular-octane and high-octane pre-
mium gasoline be included in the same market? Gasoline and diesel fuel?
For some goods, it makes sense to talk about a market only in terms of very
estrictive geographic boundaries. Housing is a good example. Most people
who work in downtown Chicago will look for housing within commuting
distance. They will not look at homes 200 or 300 miles away, even though
those homes might be much cheaper. And homes (together with the land
they are sitting on) 200 miles away cannot be easily moved closer to Chicago.
Thus the housing market in Chicago is separate and distinct from, say, that in
Cleveland, Houston, Atlanta, or Philadelphia. Likewise, retail gasoline mar-
kets, though less limited geographically, are still regional because of the ex-
pense of shipping gasoline over long distances. Thus the market for gasoline
in southern California is distinct from that in northern Illinois. On the other
hand, as we mentioned earlier, gold is bought and sold in a world market; the
possibility of a
itrage prevents the price from differing significantly from
one location to another.
We must also think carefully about the range of products to include in
a market. For example, there is a market for single-lens reflex (SLR) digital
cameras, and many
ands compete in that market. But what about compact
“point-and-shoot” digital cameras? Should they be considered part of the
same market? Probably not, because they are typically used for different pur-
poses and so do not compete with SLR cameras. Gasoline is another example.
Regular- and premium-octane gasolines might be considered part of the same
extent of a market
Boundaries of a market, both
geographical and in terms of
ange of products produced and
sold within it.
32 ParT 1 Introduction: Markets and Prices
market because most consumers can use either. Diesel fuel, however, is not part
of this market because cars that use regular gasoline cannot use diesel fuel, and
vice versa.3
Market definition is important for two reasons:
•     A company must understand who its actual and potential competitors are for
the various products that it sells or might sell in the future. It must also know
the product boundaries and geographical boundaries of its market in order to
set price, determine advertising budgets, and make capital investment decisions.
•     Market definition can be important for public policy decisions. Should the
government allow a merger or acquisition involving companies that produce
similar products, or should it challenge it? The answer depends on the im-
pact of that merger or acquisition on future competition and prices; often this
can be evaluated only by defining a market.
3How can we determine the extent of a market? Since the market is where the price of a good is
established, one approach focuses on market prices. We ask whether product prices in different
geographic regions (or for different product types) are approximately the same, or whether they
tend to move together. If either is the case, we place them in the same market. For a more detailed
discussion, see George J. Stigler and Robert A. Sherwin, “The Extent of the Market,” Journal of Law
and Economics 27 (October 1985): 555–85.
4This example is based on F. M. Scherer, “Archer-Daniels-Midland Corn Processing,” Case C16-92-
1126, John F. Kennedy School of Government, Harvard University, 1992.
ExAMPlE 1.1 The markeT For sWeeTeners
In 1990, the archer-Daniels-Midland Company (aDM)
acquired the Clinton Corn Processing Company (CCP).4
aDM was a large company that produced many agri-
cultural products, one of which was high-fructose corn
syrup (hfCS). CCP was another major u.S. corn syrup
producer. The u.S. Department of Justice (DOJ) chal-
lenged the acquisition on the grounds that it would
lead to a dominant producer of corn syrup with the
power to push prices above competitive levels. Indeed,
aDM and CCP together accounted for over 70 percent
of u.S. corn syrup production.
aDM fought the DOJ decision, and the case went
to court. The basic issue was whether corn syrup
epresented a distinct market. If it did, the combined
market share of aDM and CCP would have been about
40  percent, and the DOJ’s concern might have been
wa
anted. aDM, however, argued that the co
ect
market definition was much
oader—a market for
sweeteners which included sugar as well as corn syrup.
Because the aDM–CCP combined share of a sweetener
market would have been quite small, there would be
no concern about the company’s power to raise prices.
aDM argued that sugar and corn syrup should be
considered part of the same market because they are
used interchangeably to sweeten a vast a
ay of food
products, such as soft drinks, spaghetti sauce, and
pancake syrup. aDM also showed that as the level of
prices for corn syrup and sugar fluctuated, industrial
food producers would change the proportions of
each sweetener that they used in their products. In
October 1990, a federal judge agreed with aDM’s
argument that sugar and corn syrup were both part of
a
oad market for sweeteners. The acquisition was
allowed to go through.
Sugar and corn syrup continue to be used almost
interchangeably to satisfy americans’ strong taste
for sweetened foods. The use of all sweeteners rose
steadily through the 1990s, reaching 150 pounds per
person in 1999. But starting in 2000, sweetener use
egan to decline as health concerns led people to
find substitute snacks with less added sugar. By 2014,
american per-capita consumption of sweeteners had
dropped to 131 pounds per person. In addition,
people consumed more sugar (68 pounds per person)
than corn syrup (46 pounds per person). Part of the
shift from corn syrup to sugar was due to a growing
elief that sugar is somehow more “natural”—and
therefore healthier—than corn syrup.
ChaPTer 1 PreLIMInarIeS 33
ExAMPlE 1.2 a BiCyCle is a BiCyCle. or is iT?
Where did you buy your last bicycle?
You might have bought a used bike from
a friend or from a posting on Craigslist.
But if it was new, you probably bought
it from either of two types of stores.
If you were looking for something in-
expensive, just a functional bicycle to get
you from a to B, you would have done
well by going to a mass merchandiser
such as Target, Wal-Mart, or Sears. There
you could easily find a decent bike cost-
ing around $100 to $200. On the other
hand, if you are a serious cyclist (or at
least like to think of yourself as one), you
would probably go to a bicycle dealer—
a store that specializes in bicycles and
icycle equipment. There it would be difficult to find a
ike costing less than $400, and you could easily spend
far more. But of course you would have been happy to
spend more, because you are serious cyclist.
What does a $1000 Trek bike give you that a $120
huffy bike doesn’t? Both might have 21-speed gear
shifts (3 in front and 7 in back), but the shifting mecha-
nisms on the Trek will be higher quality and probably
shift more smoothly and evenly. Both bikes will have
front and rear hand
akes, but the
akes on the Trek
will likely be stronger and more durable. and the Trek
is likely to have a lighter frame than the huffy, which
could be important if you are a competitive cyclist.
So there are actually two different markets for bi-
cycles, markets that can be identified by the type of
store in which the bicycle is sold. This is illustrated in
Table 1.1. “Mass market” bicycles, the
ones that are sold in Target and Wal-
Mart, are made by companies such as
huffy, Schwinn, and Mantis, are priced
as low as $90 and rarely cost more
than $250. These companies are fo-
cused on producing functional bicycles
as cheaply as possible, and typically do
their manufacturing in China. “Dealer”
icycles, the ones sold in your local
icycle store, include such
ands as
Trek, Cannondale, Giant, Gary fisher,
and ridley, and are priced from $400
and up—way up. for these companies
the emphasis is on performance, as
measured by weight and the quality of
the
akes, gears, tires, and other hardware.
Companies like huffy and Schwinn would never
try to produce a $1000 bicycle, because that is
simply not their forte (or competitive advantage, as
economists like to say). Likewise, Trek and ridley
have developed a reputation for quality, and they
have neither the skills nor the factories to produce
$100 bicycles. Mongoose, on the other hand, strad-
dles both markets. They produce mass market bi-
cycles costing as little as $120, but also high-quality
dealer bicycles costing $700 to $2000.
after you buy your bike, you will need to lock
it up carefully due to the unfortunate reality of yet
another market—the black market for used bikes and
their parts. We hope that you—and your bike—stay
out of that market!
Table 1.1 markeTs For BiCyCles
Type of Bicycle companies and prices (2011)
Mass Market Bicycles: sold by mass
merchandisers such as Target, Wal-mart,
Kmart, and sears.
Huffy: $90–$140
schwinn: $140–$240
mantis: $129–$140
mongoose: $120–$280
Dealer Bicycles: sold by bicycle dealers—
stores that sell only (or mostly) bicycles and
icycle equipment.
Trek: $400–$2500
cannondale: $500–$2000
Giant: $500–$2500
Gary fisher: $600–$2000
mongoose: $700–$2000
idley: $1300–$2500
scott: $1000–$3000
ibis: $2000 and up
34 ParT 1 Introduction: Markets and Prices
1.3 Real versus Nominal Prices
We often want to compare the price of a good today with what it was in the past
or is likely to be in the future. To make such a comparison meaningful, we need
to measure prices relative to an overall price level. In absolute terms, the price of
a dozen eggs is many times higher today than it was 50 years ago. Relative to
prices overall, however, it is actually lower. Therefore, we must be careful to
co
ect for inflation when comparing prices across time. This means measuring
prices in real rather than nominal terms.
The nominal price of a good (sometimes called its “cu
ent-dollar” price) is
its absolute price. For example, the nominal price of a pound of butter was about
$0.87 in 1970, $1.88 in 1980, about $1.99 in 1990, and about $3.48 in 2015. These
are the prices you would have seen in supermarkets in those years. The real price
of a good (sometimes called its “constant-dollar” price) is the price relative to an
aggregate measure of prices. In other words, it is the price adjusted for inflation.
For consumer goods, the aggregate measure of prices most often used is the
Consumer Price Index (CPI). The CPI is calculated by the U.S. Bureau of Labor
Statistics by surveying retail prices, and is published monthly. It records how the cost
of a large market basket of goods purchased by a “typical” consumer changes over
time. Percentage changes in the CPI measure the rate of inflation in the economy.
Sometimes we are interested in the prices of raw materials and other in-
termediate products bought by firms, as well as in finished products sold at
wholesale to retail stores. In this case, the aggregate measure of prices often
used is the Producer Price Index (PPI). The PPI is also calculated by the U.S.
Bureau of Labor Statistics and published monthly, and records how, on aver-
age, prices at the wholesale level change over time. Percentage changes in the
PPI measure cost inflation and predict future changes in the CPI.
So which price index should you use to convert nominal prices to real prices?
It depends on the type of product you are examining. If it is a product or service
normally purchased by consumers, use the CPI. If instead it is a product nor-
mally purchased by businesses, use the PPI.
Because we are examining the price of butter in supermarkets, the relevant
price index is the CPI. After co
ecting for inflation, do we find that the price of
utter was more expensive in 2015 than in 1970? To find out, let’s calculate the
2015 price of butter in terms of 1970 dollars. The CPI was 38.8 in 1970 and rose
to about 237.0 in 2015. (There was considerable inflation in the United States
during the 1970s and early 1980s.) In 1970 dollars, the price of butter was

38.8
237.0
* $3.48 = $0.57
In real terms, therefore, the price of butter was lower in 2015 than it was in 1970.5 Put
another way, the nominal price of butter went up by about 300 percent, while the
CPI went up 511 percent. Relative to the aggregate price level, butter prices fell.
In this book, we will usually be concerned with real rather than nominal
prices because consumer choices involve analyses of price comparisons. These
elative prices can most easily be evaluated if there is a common basis of com-
parison. Stating all prices in real terms achieves this objective. Thus, even
though we will often measure prices in dollars, we will be thinking in terms of
the real purchasing power of those dollars.
nominal price absolute
price of a good, unadjusted for
inflation.
eal price Price of a good
elative to an aggregate measure
of prices; price adjusted for
inflation.
Consumer Price Index
Measure of the aggregate price
level.
Producer Price Index
Measure of the aggregate price
level for intermediate products
and wholesale goods.
5Two good sources of data on the national economy are the Economic Report of the President and the
Statistical Abstract of the United States. Both are published annually and are available from the U.S.
Government Printing Office.
ChaPTer 1 PreLIMInarIeS 35
6You can get data on the cost of a college education by visiting the National Center for Education
Statistics and download the Digest of Education Statistics at http:
nces.ed.gov. Historical and cur-
ent data on the average retail price of eggs can be obtained from the Bureau of Labor Statistics
(BLS) at http:
www.bls.gov, by selecting CPI—Average Price Data.
ExAMPlE 1.3 The PriCe oF eggs and The PriCe
oF a College eduCaTion
In 1970, Grade a large eggs cost about 61 cents a dozen. In the same year,
the average annual cost of a college education at a private four-year college,
including room and board, was about $2112. By 2016, the price of eggs
had risen to $2.47 a dozen, and the average cost of a college education was
$25,694. In real terms, were eggs more expensive in 2016 than in 1970? had a
college education become more expensive?
Table 1.2 shows the nominal price of eggs, the nominal cost of a college
education, and the CPI for 1970–2016. (The CPI is based on 1983 = 100.)
Table 1.2 The real PriCes oF eggs and oF a College
eduCaTion6
1970 1980 1990 2000 2016
consumer price
index
38.8 82.4 130.7 172.2 241.7
Nominal Prices
Grade a large eggs $0.61 $0.84 $1.01 $0.91 $2.47
college education $1,784 $3,499 $7,602 $12,922 $25,694
Real Prices ($1970)
Grade a large eggs $0.61 $0.40 $0.30 $0.21 $0.40
college education $1,784 $1,624 $2,239 $2,912 $4,125
also shown are the real prices of eggs and college education in 1970 dollars,
calculated as follows:
eal price of eggs in 1980 =
CPI1970
CPI1980
* nominal price in 1980
eal price of eggs in 1990 =
CPI1970
CPI1990
* nominal price in 1990
and so forth.
The table shows clearly that the real cost of a college education rose (by
231 percent) during this period, while the real cost of eggs fell (by 34 percent).
It is these relative changes in prices that are important for the choices that
consumers make, not the fact that both eggs and college cost more in nominal
dollars today than they did in 1970.
In the table, we calculated real prices in terms of 1970 dollars, but we
could just as easily have calculated them in terms of dollars of some other base
http:
nces.ed.gov
http:
www.bls.gov
36 ParT 1 Introduction: Markets and Prices
year. for example, suppose we want to calculate the real price of eggs in 1990
dollars. Then:
real price of eggs in 1970 =
CPI1990
CPI1970
* nominal price in 1970
=
130.7
38.8
* 0.61 = 2.05
real price of eggs in 2016 =
CPI1990
CPI2016
* nominal price in 2016
=
130.7
241.7
* 2.47 = 1.34
Percentage change in real price =
eal price in 2016 - real price in 1970
eal price in 1970

=
1.34 - 2.05
2.05
= -0.34
notice that the percentage decline in real price is the same whether we use
1970 dollars or 1990 dollars as the base year.
ExAMPlE 1.4 The auThors deBaTe The minimum Wage
Many workers in the united States are dissatisfied.
They feel that their wages have not grown in the past
two decades, and are finding it hard to make ends
meet. This is especially the case for unskilled work-
ers, many of whom earn the minimum wage. as a re-
sult, some economists and politicians have proposed
aising the minimum wage, while others argue that
doing so would lead to fewer jobs for teenagers and
other new entrants into the labor market. So what to
do? Should the minimum wage be increased?
“Leave it where it is,” says Pindyck. “It has been
aised regularly since it was first introduced in 1938
(at $0.25 per hour), and is much higher than it was
just a decade or two ago.” “Wait,” shouts rubinfeld.
You are ignoring inflation, and confusing the nomi-
nal minimum wage and the real (inflation-adjusted)
minimum wage. Just look at figure 1.1, which shows
the minimum wage in both nominal and real terms.
In real terms, the minimum wage is actually much
lower than it was during the 1970s.”
“Good point,” replies Pindyck. “as we tell our
students, one should never ignore inflation. I hate
to say it, but you are right on this one. It is the
eal minimum wage that matters, and that has in-
deed been declining.” “But,” adds Pindyck, “there
are two points you shouldn’t forget. first, many
states have a minimum wage that is significantly
higher than the federal minimum. for example,
California’s 2016 minimum wage of $10 and new
York’s $9 minimum are substantially higher than
the federal minimum (in 2016) of $7.25. Second
and probably more important, the minimum wage
can reduce the incentive of employers to hire
entry-level workers, so it might still be best not to
aise it.”
“You are certainly co
ect about some states hav-
ing a higher minimum wage,” responds rubinfeld.
“as for the impact of a higher minimum wage
on employment, that’s a tough call, and many
economists disagree about the likely impact. Let’s
come back to this problem when we discuss labor
markets in Chapter 14. In the meantime, read-
ers can learn more about the minimum wage at
http:
www.dol.gov.”
http:
www.dol.gov.%E2%80%9D
ChaPTer 1 PreLIMInarIeS 37
Real Wage (2000$)
Nominal Wage
0
2
4
6
8
1938 1948 1958 1968 1978 1988 1998 2008 2018
D
ol
la
s
p
e
H
ou
Yea
FIguRE 1.1
The minimum Wage
In nominal terms the minimum wage has increased steadily over the past 80 years. However, in real
terms its 2016 level is below that of the 1970s.
ExAMPlE 1.5 healTh Care and College TexTBooks
health care costs have been
ising in the united States,
and some argue that the cost
increases are the result of
inefficiencies in the health
care system. The claim has
een made that textbook
prices have also been rising,
and students who must buy
the textbooks often com-
plain bitterly about the high
prices. first, is it true that health care and college
textbooks have been becoming increasingly expen-
sive? remember that we have to consider prices in
the context of overall inflation. So the question is
whether the prices of health care and college text-
ooks have been rising faster than inflation.
figure 1.2 provides the an-
swers. It shows a price index
for textbooks and a price index
for health care, both in nominal
terms, along with the CPI, and all
scaled to equal 100 in 1980. Let’s
first look at health care. Clearly
the cost has been rising in real
terms; the nominal price has risen
about twice as fast as the CPI. The
CPI during the period tripled, go-
ing from 100 in 1980 to about 300 in 2016, but the
cost of health care increased six-fold, going from 100
to 600. Why have health care costs risen so much?
first, as consumers have become wealthier, they have
shifted their purchases away from other goods and
towards health care, straining the system. Second, as
38 ParT 1 Introduction: Markets and Prices
life expectancy has increased, we have reached the
point of diminishing returns—it becomes more and
more costly to achieve an extra few months of life
expectancy. Sound confusing? Stay tuned – we will
explain the increase in health care costs in more de-
tail in Chapters 3 and 6.
What about the prices of college textbooks?
They have risen at a phenomenal rate, increasing
nine-fold between 1980 and 2016, compared to
the three-fold increase in the CPI. The rate of in-
crease has been especially rapid since 1995. So
it is no surprise that students are often outraged
when they go to the college bookstore to buy the
ooks they need for their classes. how can the
publishing companies get away with charging
higher and higher prices? first, the choice of text-
ook for a course is usually made by the instruc-
tor, not the students, and the instructor is often
unaware of the price and in some cases may not
care about it. Second, as a result of mergers and
acquisitions, the textbook industry has become
highly concentrated. There are now only three
major firms that publish a large fraction of all text-
ooks, and they have found it in their own self-
interest to avoid any aggressive price competition.
Sound strange? Wait until Chapter 12 where we
will explain what has happened to textbook prices
in more detail.
1980
100
200
300
400
500
600
700
800
900
1000
1985 1990 1995 2000 2005 2010 2015 2020
CPI All
Textbooks
Health Care
FIguRE 1.2
PriCes oF healTh Care and College TexTBooks
The prices of both health care and college textbooks have been rising much faster than overall
inflation. This is especially true of college textbook prices, which have increased about three times
as fast as the CPI.
ChaPTer 1 PreLIMInarIeS 39
1.4 Why Study Microeconomics?
We think that after reading this book you will have no doubt about the im-
portance and
oad applicability of microeconomics. In fact, one of our major
goals is to show you how to apply microeconomic principles to actual decision-
making problems. Nonetheless, some extra motivation early on never hurts.
Here are two examples that not only show the use of microeconomics in prac-
tice, but also provide a preview of this book.
Corporate Decision Making: The Toyota Prius
In 1997, Toyota Motor Corporation introduced the Prius in Japan, and started
selling it worldwide in 2001. The Prius, the first hy
id car to be sold in the
United States, can run off both a gasoline engine and a battery, and the momen-
tum of the car charges the battery. Hy
id cars are more energy efficient than
cars with just a gasoline engine; the Prius, for example, can get 45 to 55 miles
per gallon. The Prius was a big success, and within a few years other manufac-
turers began introducing hy
id versions of some of their cars.
The design and efficient production of the Prius involved not only some im-
pressive engineering, but a lot of economics as well. First, Toyota had to think care-
fully about how the public would react to the design and performance of this new
product. How strong would demand be initially, and how fast would it grow?
How would demand depend on the prices that Toyota charged? Understanding
consumer preferences and trade-offs and predicting demand and its responsive-
ness to price are essential to Toyota and every other automobile manufacturer. (We
discuss consumer preferences and demand in Chapters 3, 4, and 5.)
Next, Toyota had to be concerned with the cost of manufacturing these cars —
whether produced in Japan or, starting in 2010, in the United States. How high
would production costs be? How would the cost of each car depend on the total
number of cars produced each year? How would the cost of labor and the prices
of steel and other raw materials affect costs? How much and how fast would
costs decline as managers and workers gained experience with the production
process? And to maximize profits, how many of these cars should Toyota plan to
produce each year? (We discuss production and cost in Chapters 6 and 7, and the
profit-maximizing choice of output in Chapters 8 and 10.)
Toyota also had to design a pricing strategy and consider how competitors
would react to it. Although the Prius was the first hy
id car, Toyota knew
that it would compete with other small fuel-efficient cars, and that soon other
manufacturers would introduce their own hy
id cars. Should Toyota charge
a relatively low price for a basic stripped-down version of the Prius and high
prices for individual options like leather seats? Or would it be more profitable
to make these options “standard” items and charge a higher price for the whole
package? Whatever pricing strategy Toyota chose, how were competitors likely
to react? Would Ford or Nissan try to undercut by lowering the prices of its
smaller cars, or rush to
ing out their own hy
id cars at lower prices? Might
Toyota be able to deter Ford and Nissan from lowering prices by threatening to
espond with its own price cuts? (We discuss pricing in Chapters 10 and 11, and
competitive strategy in Chapters 12 and 13.)
Manufacturing the Prius required large investments in new capital equip-
ment, so Toyota had to consider both the risks and possible outcomes of its
decisions. Some of this risk was due to uncertainty over the future price of
oil and thus the price of gasoline (lower gasoline prices would reduce the
40 ParT 1 Introduction: Markets and Prices
demand for small fuel-efficient cars). Some of the risk was due to uncertainty
over the wages that Toyota would have to pay its workers at its plants in Japan
and in the United States. (Oil and other commodity markets are discussed in
Chapters  2 and 9. Labor markets and the impact of unions are discussed in
Chapter 14. Investment decisions and the implications of uncertainty are dis-
cussed in Chapters 5 and 15.)
Toyota also had to wo
y about organizational problems. Toyota is an inte-
grated firm in which separate divisions produce engines and parts and then
assemble finished cars. How should the managers of different divisions be re-
warded? What price should the assembly division be charged for the engines it
eceives from another division? (We discuss internal pricing and organizational
incentives for the integrated firm in Chapters 11 and 17.)
Finally, Toyota had to think about its relationship to the government and the
effects of regulatory policies. For example, all of its cars sold in the United States
must meet federal emissions standards, and U.S. production-line operations must
comply with health and safety regulations. How might those regulations and
standards change over time? How would they affect costs and profits? (We discuss
the role of government in limiting pollution and promoting health and safety in
Chapter 18.)
Public Policy Design: Fuel Efficiency Standards
for the Twenty-First Century
In 1975, the U.S. government imposed regulations designed to improve the av-
erage fuel economy of domestically-sold cars and light trucks (including vans
and sport utility vehicles). The CAFE (Corporate Average Fuel Economy) stan-
dards have become increasingly stringent over the years. In 2007, President
George W. Bush signed into law the Energy Independence and Security Act,
which required automakers to boost fleet wide gas mileage to 35 miles per
gallon (mpg) by 2020. In 2011, the Obama administration pushed the 35 mpg
target forward to 2016, and (with the agreement of 13 auto companies) set a
standard of 55 mpg for 2020. While the program’s primary goal is to increase
energy security by reducing the U.S. dependence on imported oil, it would
also generate substantial environmental benefits, such as a reduction in green-
house gas emissions.
A number of important decisions have to be made when designing a fuel
efficiency program, and most of those decisions involve economics. First, the
government must evaluate the monetary impact of the program on consum-
ers. Higher fuel economy standards will increase the cost of purchasing a car
(the cost of achieving higher fuel economy will be borne in part by consumers),
ut will lower the cost of operating it (gas mileage will be higher). Analyzing
the ultimate impact on consumers means analyzing consumer preferences and
demand. For example, would consumers drive less and spend more of their in-
come on other goods? If so, would they be nearly as well off? (Consumer prefer-
ences and demand are discussed in Chapters 3 and 4).
Before imposing CAFE standards, it is important to estimate the likely im-
pact those standards will have on the cost of producing cars and light trucks.
Might automobile companies minimize cost increases by using new lightweight
materials or by changing the footprint of new model cars? (Production and cost
are discussed in Chapters 6 and 7.) Then the government needs to know how
changes in production costs will affect the production levels and prices of new
automobiles and light trucks. Are the additional costs likely to be abso
ed by
ChaPTer 1 PreLIMInarIeS 41
manufacturers or passed on to consumers in the form of higher prices? (Output
determination is discussed in Chapter 8 and pricing in Chapters 10 through 13.)
The government must also ask why problems related to oil consumption
are not solved by our market-oriented economy. One answer is that oil prices
are determined in part by a cartel (OPEC) that is able to push the price of oil
above competitive levels. (Pricing in markets in which firms have the power to
control prices are discussed in Chapters 10 through 12.) Finally, the high U.S.
demand for oil has led to a substantial outflow of dollars to the oil-producing
countries, which in turn has created political and security issues that go beyond
the confines of economics. What economics can do, however, is help us evaluate
how best to reduce our dependence on foreign oil. Are standards like those of
the CAFE program prefe
ed to fees on oil consumption? What are the environ-
mental implications of increasingly stringent standards? (These problems are
discussed in Chapter 18.)
These are just two examples of how microeconomics can be applied in the
arenas of private and public-policy decision making. You will discover many
more applications as you read this book.
Summary
1. Microeconomics is concerned with the decisions made
y individual economic units—consumers, workers,
investors, owners of resources, and business firms. It
is also concerned with the interaction of consumers
and firms to form markets and industries.
2. Microeconomics relies heavily on the use of theory,
which can (by simplification) help to explain how
economic units behave and to predict what behavior
will occur in the future. Models are mathematical rep-
esentations of theories that can help in this explana-
tion and prediction process.
3. Microeconomics is concerned with positive questions
that have to do with the explanation and prediction of
phenomena. But microeconomics is also important for
normative analysis, in which we ask what choices are
est—for a firm or for society as a whole. Normative
analyses must often be combined with individual
value judgments because issues of equity and fairness
as well as of economic efficiency may be involved.
4. A market refers to a collection of buyers and sellers who
interact, and to the possibility for sales and purchases
that result from that interaction. Microeconomics in-
volves the study of both perfectly competitive mar-
kets, in which no single buyer or seller has an impact
on price, and noncompetitive markets, in which indi-
vidual entities can affect price.
5. The market price is established by the interaction of
uyers and sellers. In a perfectly competitive market,
a single price will usually prevail. In markets that
are not perfectly competitive, different sellers might
charge different prices. In this case, the market price
efers to the average prevailing price.
6. When discussing a market, we must be clear about
its extent in terms of both its geographic boundaries
and the range of products to be included in it. Some
markets (e.g., housing) are highly localized, whereas
others (e.g., gold) are global in nature.
7. To account for the effects of inflation, we measure
eal (or constant-dollar) prices, rather than nominal
(or cu
ent-dollar) prices. Real prices use an ag-
gregate price index, such as the CPI, to co
ect for
inflation.
QueStIonS for revIew
1. It is often said that a good theory is one that can be
efuted by an empirical, data-oriented study. Explain
why a theory that cannot be evaluated empirically is
not a good theory.
2. Which of the following two statements involves posi-
tive economic analysis and which normative? How do
the two kinds of analysis differ?
a. Gasoline rationing (allocating to each individual
a maximum amount of gasoline that can be pur-
chased each year) is poor social policy because it
interferes with the workings of the competitive
market system.
. Gasoline rationing is a policy under which more
people are made worse off than are made better off.
42 ParT 1 Introduction: Markets and Prices
exerCISeS
1. Decide whether each of the following statements is
true or false and explain why:
a. Fast-food chains like McDonald’s, Burger King,
and Wendy’s operate all over the United States.
Therefore, the market for fast food is a national
market.
. People generally buy clothing in the city in which
they live. Therefore, there is a clothing market in,
say, Atlanta that is distinct from the clothing mar-
ket in Los Angeles.
c. Some consumers strongly prefer Pepsi and some
strongly prefer Coke. Therefore, there is no single
market for colas.
2. The following table shows the average retail price of
utter and the Consumer Price Index from 1980 to
2010, scaled so that the CPI = 100 in 1980.
1980 1990 2000 2010
cpi 100 158.56 208.98 218.06
etail price of butter
(salted, Grade aa, per lb.)
$1.88 $1.99 $2.52 $2.88
a. Calculate the real price of butter in 1980 dollars.
Has the real price increased/decreased/stayed the
same from 1980 to 2000? From 1980 to 2010?
. What is the percentage change in the real price
(1980 dollars) from 1980 to 2000? From 1980 to 2010?
c. Convert the CPI into 1990 = 100 and determine the
eal price of butter in 1990 dollars.
d. What is the percentage change in real price (1990
dollars) from 1980 to 2000? Compare this with your
answer in (b). What do you notice? Explain.
3. At the time this book went to print, the minimum
wage was $7.25. To find the cu
ent value of the
CPI, go to http:
www.bls.gov/cpi/home.htm. Click
on “CPI Tables,” which is found on the left side of
the website. Then, click on “Table Containing History
of CPI-U U.S. All Items Indexes and Annual Percent
Changes from 1913 to Present.” This will give you the
CPI from 1913 to the present.
a. With these values, calculate the cu
ent real mini-
mum wage in 1990 dollars.
. Stated in real 1990 dollars, what is the percentage
change in the real minimum wage from 1985 to the
present?
3. Suppose the price of regular-octane gasoline was 20
cents per gallon higher in New Jersey than in Oklahoma.
Do you think there would be an opportunity for a
i-
trage (i.e., that firms could buy gas in Oklahoma and
then sell it at a profit in New Jersey)? Why or why not?
4. In Example 1.3, what economic forces explain why
the real price of eggs has fallen while the real price of
a college education has increased? How have these
changes affected consumer choices?
5. Suppose that the Japanese yen rises against the U.S.
dollar—that is, it will take more dollars to buy a
given amount of Japanese yen. Explain why this
increase simultaneously increases the real price
of Japanese cars for U.S. consumers and lowers
the real price of U.S. automobiles for Japanese
consumers.
6. The price of long-distance telephone service fell from
40 cents per minute in 1996 to 22 cents per minute in
1999, a 45-percent (18 cents/40 cents) decrease. The
Consumer Price Index increased by 10 percent over
this period. What happened to the real price of tele-
phone service?
http:
www.bls.gov/cpi/home.htm
43
One of the best ways to appreciate the relevance of economics is to begin with the basics of supply and demand. Supply-demand analysis is a fundamental and powerful tool that can
e applied to a wide variety of interesting and important problems. To
name a few:
•     Understanding and predicting how changing world economic con-
ditions affect market price and production
•     Evaluating the impact of government price controls, minimum
wages, price supports, and production incentives
•     Determining how taxes, subsidies, tariffs, and import quotas affect
consumers and producers
We begin with a review of how supply and demand curves are used
to describe the market mechanism. Without government intervention
(e.g., through the imposition of price controls or some other regulatory
policy), supply and demand will come into equili
ium to determine
oth the market price of a good and the total quantity produced. What
that price and quantity will be depends on the particular characteris-
tics of supply and demand. Variations of price and quantity over time
depend on the ways in which supply and demand respond to other
economic variables, such as aggregate economic activity and labor
costs, which are themselves changing.
We will, therefore, discuss the characteristics of supply and de-
mand and show how those characteristics may differ from one market
to another. Then we can begin to use supply and demand curves to
understand a variety of phenomena—for example, why the prices of
some basic commodities have fallen steadily over a long period while
the prices of others have experienced sharp fluctuations; why short-
ages occur in certain markets; and why announcements about plans
for future government policies or predictions about future economic
conditions can affect markets well before those policies or conditions
ecome reality.
Besides understanding qualitatively how market price and quantity
are determined and how they can vary over time, it is also important
to learn how they can be analyzed quantitatively. We will see how
simple “back of the envelope” calculations can be used to analyze and
predict evolving market conditions. We will also show how markets
2.1 The Price of Eggs and the
Price of a College
Education Revisited 50
2.2 Wage Inequality in the
United States 51
2.3 The Long-Run Behavior of
Natural Resource Prices 51
2.4 The Effects of 9/11 on the
Supply and Demand for
New York City Office Space 53
2.5 The Market for Wheat 59
2.6 The Demand for Gasoline
and Automobiles 66
2.7 The Weather in Brazil
and the Price of Coffee
in New York 69
2.8 The Behavior of Copper
Prices 74
2.9 Upheaval in the World
Oil Market 76
2.10 Price Controls and Natural
Gas Shortages 81
LIST Of ExAMPLES
2.1 Supply and Demand 44
2.2 The Market Mechanism 47
2.3 Changes in Market
Equili
ium 48
2.4 Elasticities of Supply and
Demand 55
2.5 Short-Run versus Long-Run
Elasticities 62
*2.6 Understanding and Predicting
the Effects of Changing
Market Conditions 71
2.7 Effects of Government
Intervention—Price
Controls 80
ChAPTER OUTLINE
The Basics of Supply
and Demand
ChAPTER 2
44 PART 1 Introduction: Markets and Prices
espond both to domestic and international macroeconomic fluctuations and to
the effects of government interventions. We will try to convey this understand-
ing through simple examples and by urging you to work through some exer-
cises at the end of the chapter.
2.1 Supply and Demand
The basic model of supply and demand is the workhorse of microeconomics. It
helps us understand why and how prices change, and what happens when the
government intervenes in a market. The supply-demand model combines two
important concepts: a supply curve and a demand curve. It is important to under-
stand precisely what these curves represent.
The Supply Curve
The supply    curve shows the quantity of a good that producers are willing to
sell at a given price, holding constant any other factors that might affect the
quantity supplied. The curve labeled S in Figure 2.1 illustrates this. The verti-
cal axis of the graph shows the price of a good, P, measured in dollars per unit.
This is the price that sellers receive for a given quantity supplied. The horizon-
tal axis shows the total quantity supplied, Q, measured in the number of units
per period.
The supply curve is thus a relationship between the quantity supplied and
the price. We can write this relationship as an equation:
QS = QS(P)
Or we can draw it graphically, as we have done in Figure 2.1.
Note that the supply curve in Figure 2.1 slopes upward. In other words, the
higher the price, the more that firms are able and willing to produce and sell. For
example, a higher price may enable cu
ent firms to expand production by
hiring extra workers or by having existing workers work overtime (at greater
cost to the firm). Likewise, they may expand production over a longer period
of time by increasing the size of their plants. A higher price may also attract
supply curve Relationship
etween the quantity of a good
that producers are willing to sell
and the price of the good.
Quantity
Price
Q1 Q2
S S′
P1
P2
Figure 2.1
The Supply Curve
The supply curve, labeled S in the figure, shows how the quan-
tity of a good offered for sale changes as the price of the good
changes. The supply curve is upward sloping: The higher the
price, the more firms are able and willing to produce and sell.
If production costs fall, firms can produce the same quantity at
a lower price or a larger quantity at the same price. The supply
curve then shifts to the right (from S to S′).
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 45
new firms to the market. These newcomers face higher costs because of their
inexperience in the market and would therefore have found entry uneconomi-
cal at a lower price.
OTher variableS ThaT affeCT Supply The quantity supplied can de-
pend on other variables besides price. For example, the quantity that produc-
ers are willing to sell depends not only on the price they receive but also on
their production costs, including wages, interest charges, and the costs of raw
materials. The supply curve labeled S in Figure 2.1 was drawn for particular
values of these other variables. A change in the values of one or more of these
variables translates into a shift in the supply curve. Let’s see how this might
happen.
The supply curve S in Figure 2.1 says that at a price P1, the quantity pro-
duced and sold would be Q1. Now suppose that the cost of raw materials falls.
How does this affect the supply curve?
Lower raw material costs—indeed, lower costs of any kind—make production
more profitable, encouraging existing firms to expand production and enabling
new firms to enter the market. If at the same time the market price stayed con-
stant at P1, we would expect to observe a greater quantity supplied. Figure 2.1
shows this as an increase from Q1 to Q2. When production costs decrease, output
increases no matter what the market price happens to be. The entire supply curve
thus shifts to the right, which is shown in the figure as a shift from S to S′.
Another way of looking at the effect of lower raw material costs is to imag-
ine that the quantity produced stays fixed at Q1 and then ask what price firms
would require to produce this quantity. Because their costs are lower, they
would accept a lower price—P2. This would be the case no matter what quan-
tity was produced. Again, we see in Figure 2.1 that the supply curve must shift
to the right.
We have seen that the response of quantity supplied to changes in price can
e represented by movements along the supply curve. However, the response of
supply to changes in other supply-determining variables is shown graphically
as a shift of the supply curve itself. To distinguish between these two graphical
depictions of supply changes, economists often use the phrase change in supply
to refer to shifts in the supply curve, while reserving the phrase change in the
quantity supplied to apply to movements along the supply curve.
The Demand Curve
The demand    curve shows how much of a good consumers are willing to buy
as the price per unit changes. We can write this relationship between quantity
demanded and price as an equation:
QD = QD(P)
or we can draw it graphically, as in Figure 2.2. Note that the demand curve in
that figure, labeled D, slopes downward: Consumers are usually ready to buy
more if the price is lower. For example, a lower price may encourage consumers
who have already been buying the good to consume larger quantities. Likewise,
it may allow other consumers who were previously unable to afford the good to
egin buying it.
Of course the quantity of a good that consumers are willing to buy can
depend on other things besides its price. Income is especially important. With
greater incomes, consumers can spend more money on any good, and some
consumers will do so for most goods.
demand curve Relationship
etween the quantity of a good
that consumers are willing to buy
and the price of the good.
46 PART 1 Introduction: Markets and Prices
ShifTing The DemanD Curve Let’s see what happens to the demand
curve if income levels increase. As you can see in Figure 2.2, if the market price
were held constant at P1, we would expect to see an increase in the quantity
demanded—say, from Q1 to Q2, as a result of consumers’ higher incomes.
Because this increase would occur no matter what the market price, the result
would be a shift to the right of the entire demand curve. In the figure, this is shown
as a shift from D to D’. Alternatively, we can ask what price consumers would
pay to purchase a given quantity Q1. With greater income, they should be will-
ing to pay a higher price—say, P2 instead of P1 in Figure 2.2. Again, the demand
curve will shift to the right. As we did with supply, we will use the phrase change
in demand to refer to shifts in the demand curve, and reserve the phrase change in
the quantity demanded to apply to movements along the demand curve.1
SubSTiTuTe anD COmplemenTary gOODS Changes in the prices of
elated goods also affect demand. Goods are substitutes when an increase in
the price of one leads to an increase in the quantity demanded of the other. For
example, copper and aluminum are substitute goods. Because one can often be
substituted for the other in industrial use, the quantity of copper demanded will in-
crease if the price of aluminum increases. Likewise, beef and chicken are substitute
goods because most consumers are willing to shift their purchases from one to
the other when prices change.
Goods are complements when an increase in the price of one leads to a decrease
in the quantity demanded of the other. For example, automobiles and gasoline are
complementary goods. Because they tend to be used together, a decrease in the
price of gasoline increases the quantity demanded for automobiles. Likewise, com-
puters and computer software are complementary goods. The price of computers
has dropped dramatically over the past decade, fueling an increase not only in
purchases of computers, but also purchases of software packages.
We attributed the shift to the right of the demand curve in Figure 2.2 to an
increase in income. However, this shift could also have resulted from either
an increase in the price of a substitute good or a decrease in the price of a
substitutes Two goods for
which an increase in the price of
one leads to an increase in the
quantity demanded of the other.
complements Two goods for
which an increase in the price
of one leads to a decrease in the
quantity demanded of the other.
P2
D
D ′
Price
Quantity
P1
Q2Q1
Figure 2.2
The DemanD Curve
The demand curve, labeled D, shows how the quantity of a
good demanded by consumers depends on its price. The de-
mand curve is downward sloping; holding other things equal,
consumers will want to purchase more of a good as its price
goes down. The quantity demanded may also depend on other
variables, such as income, the weather, and the prices of other
goods. For most products, the quantity demanded increases
when income rises. A higher income level shifts the demand
curve to the right (from D to D′).
1Mathematically, we can write the demand curve as
QD = D(P, I )
where I is disposable income. When we draw a demand curve, we are keeping I fixed.
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 47
complementary good. Or it might have resulted from a change in some other
variable, such as the weather. For example, demand curves for skis and snow-
oards will shift to the right when there are heavy snowfalls.
2.2 The Market Mechanism
The next step is to put the supply curve and the demand curve together. We
have done this in Figure 2.3. The vertical axis shows the price of a good, P, again
measured in dollars per unit. This is now the price that sellers receive for a
given quantity supplied, and the price that buyers will pay for a given quantity
demanded. The horizontal axis shows the total quantity demanded and sup-
plied, Q, measured in number of units per period.
equili
ium The two curves intersect at the equili
ium,     or     market-    
clearing,    price and quantity. At this price (P0 in Figure 2.3), the quantity sup-
plied and the quantity demanded are just equal (to Q0). The market    mechanism
is the tendency in a free market for the price to change until the market clears—
i.e., until the quantity supplied and the quantity demanded are equal. At this
point, because there is neither excess demand nor excess supply, there is no
pressure for the price to change further. Supply and demand might not always
e in equili
ium, and some markets might not clear quickly when conditions
change suddenly. The tendency, however, is for markets to clear.
To understand why markets tend to clear, suppose the price were initially
above the market-clearing level—say, P1 in Figure 2.3. Producers will try to pro-
duce and sell more than consumers are willing to buy. A surplus—a situation
in which the quantity supplied exceeds the quantity demanded—will result.
To sell this surplus—or at least to prevent it from growing—producers would
egin to lower prices. Eventually, as price fell, quantity demanded would in-
crease, and quantity supplied would decrease until the equili
ium price P0
was reached.
The opposite would happen if the price were initially below P0—say, at P2.
A shortage—a situation in which the quantity demanded exceeds the quan-
tity supplied—would develop, and consumers would be unable to purchase
equili
ium (or market-
clearing) price Price that
equates the quantity supplied to
the quantity demanded.
market mechanism
Tendency in a free market for
price to change until the market
clears.
surplus Situation in which the
quantity supplied exceeds the
quantity demanded.
shortage Situation in which
the quantity demanded exceeds
the quantity supplied.
S
Price
P1
(dollars per unit)
P0
P2
Quantity
Shortage
Surplus
D
Q0
Figure 2.3
Supply anD DemanD
The market clears at price P0 and quantity Q0. At the
higher price P1, a surplus develops, so price falls. At the
lower price P2, there is a shortage, so price is bid up.
48 PART 1 Introduction: Markets and Prices
all they would like. This would put upward pressure on price as consumers
tried to outbid one another for existing supplies and producers reacted by
increasing price and expanding output. Again, the price would eventually
each P0.
When Can We uSe The Supply-DemanD mODel? When we draw and
use supply and demand curves, we are assuming that at any given price, a
given quantity will be produced and sold. This assumption makes sense only
if a market is at least roughly competitive. By this we mean that both sellers and
uyers should have little market power—i.e., little ability individually to affect the
market price.
Suppose instead that supply were controlled by a single producer—a
mono polist. In this case, there will no longer be a simple one-to-one relation-
ship between price and the quantity supplied. Why? Because a monopolist’s
ehavior depends on the shape and position of the demand curve. If the
demand curve shifts in a particular way, it may be in the monopolist’s interest
to keep the quantity fixed but change the price, or to keep the price fixed and
change the quantity. (How this could occur is explained in Chapter 10.) Thus
when we work with supply and demand curves, we implicitly assume that we
are refe
ing to a competitive market.
2.3 Changes in Market equili
ium
We have seen how supply and demand curves shift in response to changes in
such variables as wage rates, capital costs, and income. We have also seen how
the market mechanism results in an equili
ium in which the quantity supplied
equals the quantity demanded. Now we will see how that equili
ium changes
in response to shifts in the supply and demand curves.
Let’s begin with a shift in the supply curve. In Figure 2.4, the supply curve
has shifted from S to S′ (as it did in Figure 2.1), perhaps as a result of a de-
crease in the price of raw materials. As a result, the market price drops (from
P1 to P3), and the total quantity produced increases (from Q1 to Q3). This is
what we would expect: Lower costs result in lower prices and increased sales.
Q1
P1
S
D
Price
Quantity
P3
Q3
S ′
Figure 2.4
neW equili
ium fOllOWing ShifT
in Supply
When the supply curve shifts to the right, the market clears at
a lower price P3 and a larger quantity Q3.
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 49
(Indeed, gradual decreases in costs resulting from technological progress and
etter management are an important driving force behind economic growth.)
Figure 2.5 shows what happens following a rightward shift in the demand
curve resulting from, say, an increase in income. A new price and quantity re-
sult after demand comes into equili
ium with supply. As shown in Figure 2.5,
we would expect to see consumers pay a higher price, P3, and firms produce a
greater quantity, Q3, as a result of an increase in income.
In most markets, both the demand and supply curves shift from time to time.
Consumers’ disposable incomes change as the economy grows (or contracts,
during economic recessions). The demands for some goods shift with the sea-
sons (e.g., fuels, bathing suits, um
ellas), with changes in the prices of related
goods (an increase in oil prices increases the demand for natural gas), or simply
with changing tastes. Similarly, wage rates, capital costs, and the prices of raw
materials also change from time to time, and these changes shift the supply
curve.
Supply and demand curves can be used to trace the effects of these changes. In
Figure 2.6, for example, shifts to the right of both supply and demand result in a
slightly higher price (from P1 to P2) and a much larger quantity (from Q1 to Q2). In
general, price and quantity will change depending both on how much the supply
and demand curves shift and on the shapes of those curves. To predict the sizes
and directions of such changes, we must be able to characterize quantitatively the
dependence of supply and demand on price and other variables. We will turn to
this task in the next section.
Q1
S
D
D ′
Price
Quantity
P1
P3
Q3
Figure 2.5
neW equili
ium fOllOWing
ShifT in DemanD
When the demand curve shifts to the right, the market
clears at a higher price P3 and a larger quantity Q3.
Q1
P2
S S ′
D
D ′
Price
QuantityQ2
P1
Figure 2.6
neW equili
ium fOllOWing
ShifTS in Supply anD DemanD
Supply and demand curves shift over time as market condi-
tions change. In this example, rightward shifts of the supply
and demand curves lead to a slightly higher price and a
much larger quantity. In general, changes in price and quan-
tity depend on the amount by which each curve shifts and
the shape of each curve.
50 PART 1 Introduction: Markets and Prices
exaMple 2.1 The priCe Of eggS anD The priCe Of a COllege
eDuCaTiOn reviSiTeD
In Example 1.3 (page 35), we saw
that from 1970 to 2016, the real
(constant-dollar) price of eggs
fell by 34 percent, while the real
price of a college education rose
y 131  percent. What caused this
large decline in egg prices and large
increase in the price of college?
We can understand these price
changes by examining the behav-
ior of supply and demand for each good, as shown
in figure 2.7. for eggs, the mechanization of poultry
farms sharply reduced the cost of producing eggs,
shifting the supply curve downward. At the same
time, the demand curve for eggs shifted to the left as
a more health-conscious population changed its eat-
ing habits and tended to avoid eggs. As a result, the
eal price of eggs declined sharply
while total annual consumption in-
creased (from 5300 million dozen
to about 6400 million dozen).
As for college, supply and de-
mand shifted in the opposite di-
ections. Increases in the costs of
equipping and maintaining modern
classrooms, laboratories, and li
ar-
ies, along with increases in faculty
salaries, pushed the supply curve up. At the same time,
the demand curve shifted to the right as a larger per-
centage of a growing number of high school graduates
decided that a college education was essential. Thus,
despite the increase in price, 2016 found 12.5 million
students enrolled in four-year undergraduate college
degree programs, compared with 6.9 million in 1970.
D1970
S1970
S2016
D2016
P
(1970
dollars
pe
dozen)
$0.61
Q
(million dozens)
63925300
$0.27
D1970
S1970
S2016
D2016
P
(annual
cost in
1970
dollars)
Q
(millions of students enrolled)
12.56.9
$2112
(a) (b)
$3835
Figure 2.7
(a) markeT fOr eggS (b) markeT fOr COllege eDuCaTiOn
(a) The supply curve for eggs shifted downward as production costs fell; the demand curve shifted to the left as consumer pref-
erences changed. As a result, the real price of eggs fell sharply and egg consumption rose. (b) The supply curve for a college
education shifted up as the costs of equipment, maintenance, and staffing rose. The demand curve shifted to the right as a
growing number of high school graduates desired a college education. As a result, both price and enrollments rose sharply.
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 51
exaMple 2.2 Wage inequaliTy in The uniTeD STaTeS
Although the U.S. economy has grown vigorously
over the past two decades, the gains from this growth
have not been shared equally by all. The wages
of skilled high-income workers have grown sub-
stantially, while the wages of unskilled low-income
workers have, in real terms, actually fallen slightly.
Overall, there has been growing inequality in the
distribution of earnings, a phenomenon which began
around 1980 and has accelerated in recent years.
for example, from 1978 to 2009, people in the top
20 percent of the income distribution experienced an
increase in their average real (inflation-adjusted) pre-
tax household income of 45 percent, while those in
the bottom 20 percent saw their average real pretax
income increase by only 4 percent.2
Why has income distribution become so much more
unequal during the past two decades? The answer is in
the supply and demand for workers. While the supply
of unskilled workers—people with limited educations—
has grown substantially, the demand for them has risen
only slightly. This shift of the supply curve to the right,
combined with little movement of the demand curve,
has caused wages of unskilled workers to fall. On the
other hand, while the supply of skilled workers—e.g.,
engineers, scientists, managers, and economists—has
grown slowly, the demand has risen dramatically, push-
ing wages up. (We leave it to you as an exercise to draw
supply and demand curves and show how they have
shifted, as was done in Example 2.1.)
These trends are evident in the behavior of wages
for different categories of employment. from 1980
to 2009, for example, the real (inflation-adjusted)
weekly earnings of skilled workers (such as finance,
insurance, and real estate workers) rose by more than
20 percent. Over the same period, the weekly real
incomes of relatively unskilled workers (such as retail
trade workers) rose by only 5 percent.3
Most projections point to a continuation of this
phenomenon during the coming decade. As the high-
tech sectors of the American economy grow, the de-
mand for highly skilled workers is likely to increase
further. At the same time, the computerization of
offices and factories will further reduce the demand
for unskilled workers. (This trend is discussed further
in Example 14.7.) These changes can only exace
ate
wage inequality.
2In after-tax terms, the growth of inequality has been even greater; the average real after-tax income
of the bottom 20 percent of the distribution fell over this period. For historical data on income in-
equality in the United States, see the Historical Income Inequality Tables at the U.S. Census Bureau
website: http:
www.census.gov/.
3For detailed earnings data, visit the Detailed Statistics section of the website of the Bureau of Labor
Statistics (BLS): http:
www.bls.gov/. Select Employment, Hours, and Earnings from the Cu
ent
Employment Statistics survey (National).
exaMple 2.3 The lOng-run behaviOr Of naTural reSOurCe priCeS
Many people are concerned about
the earth’s natural resources. At
issue is whether our energy and
mineral resources are likely to be
depleted in the near future, leading
to sharp price increases that could
ing an end to economic growth.
An analysis of supply and demand
can give us some perspective.
The earth does indeed have only
a finite amount of mineral resources,
such as copper, iron, coal, and oil.
During the past century, however,
the prices of these and most other
natural resources have declined or
emained roughly constant relative
to overall prices. figure 2.8, for ex-
ample, shows the price of copper
in real terms (adjusted for inflation),
together with the quantity con-
sumed from 1880 to 2016. (Both are
shown as an index, with 1880 = 1.)
http:
www.census.gov
http:
www.bls.gov
52 PART 1 Introduction: Markets and Prices
0
15
30
45
60
75
90
105
1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
In
d
ex
(1
88
0=
1)
Consumption
Price
Figure 2.8
COnSumpTiOn anD priCe Of COppe
Although annual consumption of copper has increased about a hundredfold, the real (inflation-
adjusted) price has not changed much.
Despite short-term variations in price, no significant
long-term increase has occu
ed, even though annual
consumption is now about 600 times greater than in
1880. Similar patterns hold for other mineral resources,
such as iron, oil, and coal.4
how can we explain this huge increase in copper
consumption but very little change in price? The an-
swer is shown graphically in figure 2.9. As you can
see from that figure, the demands for these resources
grew along with the world economy. But as demand
grew, production costs fell. The decline in costs was
due, first, to the discovery of new and bigger depos-
its that were cheaper to mine, and then to technical
progress and the economic advantage of mining and
efining on a large scale. As a result, the supply curve
shifted over time to the right. Over the long term, be-
cause increases in supply were greater than increases
in demand, price often fell, as shown in figure 2.9.
4The index of U.S. copper consumption was around 102 in 1999 and 2000 but then dropped off
significantly due to falling demand from 2001 to 2006. After continuing to decline through 2009,
copper consumption saw modest increases from 2010 through 2016, but consumption remained
significantly lower than its 1999–2000 levels. Consumption data (1880–1899) and price data (1880–
1969) in Figure 2.8 are from Robert S. Manthy, Natural Resource Commodities—A Century of Statistics
(Baltimore: Johns Hopkins University Press, 1978). More recent price (1970–2016) and consumption
data (1970–2016) are from the U.S. Geological Survey—Minerals Information, Copper Statistics and
Information (http:
minerals.usgs.gov/minerals/pubs/commodity/copper).
http:
minerals.usgs.gov/minerals/pubs/commodity/coppe
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 53
This is not to say that the prices of copper,
iron, and coal will decline or remain constant
forever. After all, these resources are finite. But as
prices begin to rise, consumption will likely shift,
at least in part, to substitute materials. Copper,
for example, has already been replaced in many
applications by aluminum and, more recently,
in electronic applications by fiber optics. (See
Example 2.8 for a more detailed discussion of
copper prices.)
Price
Quantity
Long-Run Path of
Price and Consumption
S1900
S1950 S2000
D1900 D1950 D2000
Figure 2.9
lOng-run mOve-
menTS Of Supply
anD DemanD fOr
mineral reSOurCeS
Although demand for most re-
sources has increased dramatically
over the past century, prices have
fallen or risen only slightly in real
(inflation-adjusted) terms because
cost reductions have shifted the
supply curve to the right just as
dramatically.
exaMple 2.4 The effeCTS Of 9/11 On The Supply anD DemanD
fOr neW yOrk CiTy OffiCe SpaCe
The September 11, 2001, te
orist attack on the World
Trade Center (WTC) complex damaged or destroyed
21 buildings, accounting for 31.2 million square feet
(msf) of Manhattan office space—nearly 10 percent
of the city’s entire inventory. Just prior to the attack,
the Manhattan office vacancy rate was 8.0 percent,
and the average asking rent was $52.50 per square
foot (psf). Given the huge unexpected reduction in
the quantity of office space supplied, we might expect
the equili
ium rental price of office space to increase
and, as a result, the equili
ium quantity of rented
office space to decrease. And because it takes time
to construct new office buildings and restore dam-
aged ones, we might also expect the vacancy rate to
decline sharply.
Surprisingly, however, the vacancy rate in
Manhattan increased from 8.0 percent in August
2001 to 9.3 percent in November 2001. Moreover,
the average rental price fell from $52.50 to $50.75
per square foot. In downtown Manhattan, the loca-
tion of the Trade Center, the changes were even
more dramatic: The vacancy rate rose from 7.5 per-
cent to 10.6 percent, and the average rental price
fell nearly 8 percent, to $41.81. What happened?
Rental prices fell because the demand for office
space fell.
figure 2.10 describes the market for office
space in downtown Manhattan. The supply and
demand curves before 9/11 appear as SAug and
DAug. The equili
ium price and quantity of down-
town Manhattan office space were $45.34 psf and
76.4 msf, respectively. The reduction in supply from
August until November is indicated by a leftward
shift in the supply curve (from SAug to S′Nov); the
esult is a higher equili
ium price P’ and a lower
equili
ium quantity, Q′. This is the outcome that
54 PART 1 Introduction: Markets and Prices
most forecasters predicted for the months following
September 11.
Many forecasters, however, failed to predict the
significant decrease in demand for office space
complementing the loss in supply. first, many firms,
oth displaced and non-displaced, chose not to
elocate downtown because of quality-of-life con-
cerns (i.e., the WTC ruins, pollution, disabled trans-
portation, and aging inventory). firms displaced by
the attack were also forced to reevaluate their office-
space needs, and they ultimately repurchased a little
more than 50 percent of their original office space
in Manhattan. Others left Manhattan but stayed in
New York City; still others moved to New Jersey.5
furthermore, in late 2001, the U.S. economy was
experiencing an economic slowdown (exace
ated
y the events of September 11) that further reduced
the demand for office space. Therefore, the cumula-
tive decrease in demand (a shift from DAug to D′Nov)
actually caused the average rental price of down-
town Manhattan office space to decrease rather than
increase in the months following September 11.
By November, even though the price had fallen to
$41.81, there were 57.2 msf on the market.
There is evidence that office real estate mar-
kets in other major U.S. cities experienced similar
surges in vacancy rates following 9/11. for in-
stance, in Chicago, not only did vacancy rates in-
crease in downtown office buildings, this increase
was significantly more pronounced in properties
in or near landmark buildings that are considered
prefe
ed targets for te
orist attacks.6
The Manhattan commercial real estate market
ounced back strongly after 2001. In 2007 the office
vacancy rate was only 5.8 percent, its lowest figure
since 9/11, and the average asking rent was over
$74  psf. But by May 2009, following the financial
crisis, the vacancy rate had risen above 13 percent.
financial services firms occupy more than a quarter
of Manhattan office space, and with the financial cri-
sis came a slump in commercial real estate. But the
eal estate market recovered along with the overall
economy, and by 2016 the vacancy rate dropped
elow 9 percent.
Price
($/psf)
Quantity (msf)57.20
41.81
45.34
P′
76.4Q′
SAugS′Nov
DAug
D′Nov
Figure 2.10
Supply anD DemanD
fOr neW yOrk CiTy
OffiCe SpaCe
Following 9/11 the supply curve shifted
to the left, but the demand curve also
shifted to the left, so that the average
ental price fell.
5See Jason Bram, James O
, and Carol Rapaport, “Measuring the Effects of the September 11 Attack
on New York City,” Federal Reserve Bank of New York, Economic Policy Review, November, 2002.
6See Alberto Abadie and Sofia Dermisi, “Is Te
orism Eroding Agglomeration Economies in Central
Business Districts? Lessons from the Office Real Estate Market in Downtown Chicago,” Journal of
U
an Economics, Volume 64, Issue 2, September 2008, 451–463.
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 55
2.4 elasticities of Supply and Demand
We have seen that the demand for a good depends not only on its price, but also
on consumer income and on the prices of other goods. Likewise, supply de-
pends both on price and on variables that affect production cost. For example,
if the price of coffee increases, the quantity demanded will fall and the quantity
supplied will rise. Often, however, we want to know how much the quantity
supplied or demanded will rise or fall. How sensitive is the demand for coffee
to its price? If price increases by 10 percent, how much will the quantity de-
manded change? How much will it change if income rises by 5 percent? We use
elasticities to answer questions like these.
An elasticity measures the sensitivity of one variable to another. Specifically,
it is a number that tells us the percentage change that will occur in one variable in
esponse to a 1-percent increase in another variable. For example, the price elasticity of
demand measures the sensitivity of quantity demanded to price changes. It tells
us what the percentage change in the quantity demanded for a good will be fol-
lowing a 1-percent increase in the price of that good.
priCe elaSTiCiTy Of DemanD Let’s look at this in more detail. We write the
price    elasticity    of    demand, Ep, as
Ep = (%∆Q)/(%∆P)
where %∆Q means “percentage change in quantity demanded” and %∆P means
“percentage change in price.” (The symbol ∆ is the Greek capital letter delta; it
means “the change in.” So ∆X means “the change in the variable X,” say, from
one year to the next.) The percentage change in a variable is just the absolute change
in the variable divided by the original level of the variable. (If the Consumer Price Index
were 200 at the beginning of the year and increased to 204 by the end of the year,
the percentage change—or annual rate of inflation—would be 4/200 = .02, or 2
percent.) Thus we can also write the price elasticity of demand as follows:7
Ep =
∆Q/Q
∆P/P
=
P∆Q
Q∆P
     (2.1)
The price elasticity of demand is usually a negative number. When the price
of a good increases, the quantity demanded usually falls. Thus ∆Q/∆P (the
change in quantity for a change in price) is negative, as is Ep. Sometimes we
efer to the magnitude of the price elasticity—i.e., its absolute size. For example,
if Ep = -2, we say that the elasticity is 2 in magnitude.
When the price elasticity is greater than 1 in magnitude, we say that demand
is price elastic because the percentage decline in quantity demanded is greater
than the percentage increase in price. If the price elasticity is less than 1 in
magnitude, demand is said to be price inelastic. In general, the price elasticity of
demand for a good depends on the availability of other goods that can be sub-
stituted for it. When there are close substitutes, a price increase will cause the
consumer to buy less of the good and more of the substitute. Demand will then
e highly price elastic. When there are no close substitutes, demand will tend to
e price inelastic.
elasticity Percentage change
in one variable resulting from a
1-percent increase in another.
price elasticity of
demand Percentage change
in quantity demanded of a
good resulting from a 1-percent
increase in its price.
7In terms of infinitesimal changes (letting the ∆P become very small), Ep = (P/Q)(dQ/dP).
56 PART 1 Introduction: Markets and Prices
linear DemanD Curve Equation (2.1) says that the price elasticity of de-
mand is the change in quantity associated with a change in price (∆Q/∆P)
times the ratio of price to quantity (P/Q). But as we move down the demand
curve, ∆Q/∆P may change, and the price and quantity will always change.
Therefore, the price elasticity of demand must be measured at a particular point
on the demand curve and will generally change as we move along the curve.
This principle is easiest to see for a linear    demand    curve—that is, a demand
curve of the form
Q = a - bP
As an example, consider the demand curve
Q = 8 - 2P
For this curve, ∆Q/∆P is constant and equal to -2 (a ∆P of 1 results in a ∆Q
of -2). However, the curve does not have a constant elasticity. Observe from
Figure 2.11 that as we move down the curve, the ratio P/Q falls; the elastic-
ity therefore decreases in magnitude. Near the intersection of the curve with
the price axis, Q is very small, so Ep = -2(P/Q) is large in magnitude. When
P = 2 and Q = 4, Ep = -1. At the intersection with the quantity axis, P = 0 so
Ep = 0.
Because we draw demand (and supply) curves with price on the vertical
axis and quantity on the horizontal axis, ∆Q/∆P = (1/slope of curve). As a re-
sult, for any price and quantity combination, the steeper the slope of the curve,
the less elastic is demand. Figure 2.12 shows two special cases. Figure 2.12(a)
shows a demand curve reflecting infinitely     elastic     demand: Consumers will
uy as much as they can at a single price P*. For even the smallest increase
in price above this level, quantity demanded drops to zero, and for any de-
crease in price, quantity demanded increases without limit. The demand curve
in Figure  2.12(b), on the other hand, reflects completely     inelastic     demand:
Consumers will buy a fixed quantity Q*, no matter what the price.
OTher DemanD elaSTiCiTieS We will also be interested in elasticities of
demand with respect to other variables besides price. For example, demand for
most goods usually rises when aggregate income rises. The income    elasticity    of    
linear demand curve
Demand curve that is a
straight line.
infinitely elastic
demand Principle that
consumers will buy as much of a
good as they can get at a single
price, but for any higher price
the quantity demanded drops to
zero, while for any lower price
the quantity demanded increases
without limit.
completely inelastic
demand Principle that
consumers will buy a fixed
quantity of a good regardless of
its price.
Ep = 0
Ep = –1
Price
Quantity
Ep = –
Q = 8 – 2P
2
4
4
8
�
Figure 2.11
linear DemanD Curve
The price elasticity of demand depends not
only on the slope of the demand curve but
also on the price and quantity. The elastic-
ity, therefore, varies along the curve as price
and quantity change. Slope is constant for
this linear demand curve. Near the top, be-
cause price is high and quantity is small, the
elasticity is large in magnitude. The elasticity
ecomes smaller as we move down the curve.
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 57
demand is the percentage change in the quantity demanded, Q, resulting from a
1-percent increase in income I:
EI =
∆Q/Q
∆I/I
=
I
Q

∆Q
∆I
     (2.2)
The demand for some goods is also affected by the prices of other goods. For
example, because butter and margarine can easily be substituted for each other,
the demand for each depends on the price of the other. A cross-price    elastic-
ity    of    demand refers to the percentage change in the quantity demanded for a
good that results from a 1-percent increase in the price of another good. So the
elasticity of demand for butter with respect to the price of margarine would be
written as
EQbPm =
∆Q
Q
∆Pm/Pm
=
Pm
Q

∆Q
∆Pm
     (2.3)
where Qb is the quantity of butter and Pm is the price of margarine.
In this example, the cross-price elasticities will be positive because the goods
are substitutes: Because they compete in the market, a rise in the price of mar-
garine, which makes butter cheaper relative to margarine, leads to an increase
in the quantity of butter demanded. (Because the demand curve for butter will
shift to the right, the price of butter will rise.) But this is not always the case.
Some goods are complements: Because they tend to be used together, an increase
in the price of one tends to push down the consumption of the other. Take
gasoline and motor oil. If the price of gasoline goes up, the quantity of gasoline
income elasticity of
demand Percentage change in
the quantity demanded resulting
from a 1-percent increase in
income.
cross-price elasticity of
demand Percentage change
in the quantity demanded of one
good resulting from a 1-percent
increase in the price of another.
Price
P*
Q*
DPrice
Quantity
D
Quantity
(b)(a)
Figure 2.12
(a) infiniTely elaSTiC DemanD (b) COmpleTely inelaSTiC DemanD
(a) For a horizontal demand curve, ∆Q /∆P is infinite. Because a tiny change in price leads to an enormous change
in demand, the elasticity of demand is infinite. (b) For a vertical demand curve, ∆Q /∆P is zero. Because the quantity
demanded is the same no matter what the price, the elasticity of demand is zero.
58 PART 1 Introduction: Markets and Prices
demanded falls—motorists will drive less. And because people are driving
less, the demand for motor oil also falls. (The entire demand curve for motor
oil shifts to the left.) Thus, the cross-price elasticity of motor oil with respect to
gasoline is negative.
elaSTiCiTieS Of Supply Elasticities of supply are defined in a similar man-
ner. The price     elasticity     of     supply is the percentage change in the quantity
supplied resulting from a 1-percent increase in price. This elasticity is usually
positive because a higher price gives producers an incentive to increase output.
We can also refer to elasticities of supply with respect to such variables as
interest rates, wage rates, and the prices of raw materials and other intermedi-
ate goods used to manufacture the product in question. For example, for most
manufactured goods, the elasticities of supply with respect to the prices of raw
materials are negative. An increase in the price of a raw material input means
higher costs for the firm; other things being equal, therefore, the quantity sup-
plied will fall.
point versus arc elasticities
So far, we have considered elasticities at a particular point on the demand curve
or the supply curve. These are called point elasticities. The point     elasticity     of    
demand, for example, is the price elasticity of demand at a particular point on the de-
mand curve and is defined by Equation (2.1). As we demonstrated in Figure 2.11
using a linear demand curve, the point elasticity of demand can vary depend-
ing on where it is measured along the demand curve.
There are times, however, when we want to calculate a price elasticity over
some portion of the demand curve (or supply curve) rather than at a single
point. Suppose, for example, that we are contemplating an increase in the
price of a product from $8.00 to $10.00 and expect the quantity demanded
to fall from 6 units to 4. How should we calculate the price elasticity of de-
mand? Is the price increase 25 percent (a $2 increase divided by the original
price of $8), or is it 20 percent (a $2 increase divided by the new price of $10)?
Is the percentage decrease in quantity demanded 33 1/3 percent (2/6) or
50 percent (2/4)?
There is no co
ect answer to such questions. We could calculate the price
elasticity using the original price and quantity. If so, we would find that
Ep = (-3 3 1 /3 percent/2 5 percent) = -1 .3 3 . Or we could use the new price
and quantity, in which case we would find that Ep = (-5 0 percent/2 0 percent)
= -2 .5 . The difference between these two calculated elasticities is large, and
neither seems preferable to the other.
arC elaSTiCiTy Of DemanD We can resolve this problem by using the
arc    elasticity    of    demand: the elasticity calculated over a range of prices. Rather
than choose either the initial or the final price, we use an average of the two,
P; for the quantity demanded, we use Q. Thus the arc elasticity of demand
is given by
Arc elasticity: Ep = (∆Q/∆P)(P/Q)     (2.4)
In our example, the average price is $9 and the average quantity 5 units. Thus
the arc elasticity is
Ep = (-2/$2)($9/5) = -1.8
price elasticity of
supply Percentage change in
quantity supplied resulting from a
1-percent increase in price.
point elasticity of demand
Price elasticity at a particular
point on the demand curve.
arc elasticity of
demand Price elasticity
calculated over a range of prices.
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 59
The arc elasticity will always lie somewhere (but not necessarily halfway) be-
tween the point elasticities calculated at the lower and the higher prices.
Although the arc elasticity of demand is sometimes useful, economists gener-
ally use the word “elasticity” to refer to a point elasticity. Throughout the rest of
this book, we will do the same, unless noted otherwise.
8For a survey of statistical studies of the demand and supply of wheat and an analysis of
evolving market conditions, see La
y Salathe and Sudchada Langley, “An Empirical Analysis
of Alternative Export Subsidy Programs for U.S. Wheat,” Agricultural Economics Research 38:1
(Winter 1986). The supply and demand curves in this example are based on the studies they
surveyed.
exaMple 2.5 The markeT fOr WheaT
Wheat is an important agricultural com-
modity, and the wheat market has been
studied extensively by agricultural econ-
omists. During recent decades, changes
in the wheat market had major implica-
tions for both American farmers and
U.S. agricultural policy. To understand
what happened, let’s examine the be-
havior of supply and demand beginning
in 1981.
from statistical studies, we know that for 1981 the supply curve for wheat
was approximately as follows:8
Supply: QS = 1800 + 240P
where price is measured in nominal dollars per bushel and quantities in mil-
lions of bushels per year. These studies also indicate that in 1981, the demand
curve for wheat was
Demand: QD = 3550 - 266P
By setting the quantity supplied equal to the quantity demanded, we can deter-
mine the market-clearing price of wheat for 1981:
QS = QD
1800 + 240P = 3550 - 266P
506P = 1750
P = $3.46 per bushel
To find the market-clearing quantity, substitute this price of $3.46 into either
the supply curve equation or the demand curve equation. Substituting into the
supply curve equation, we get
Q = 1800 + (240)(3.46) = 2630 million bushels
60 PART 1 Introduction: Markets and Prices
What are the price elasticities of demand and supply at this price and quan-
tity? We use the demand curve to find the price elasticity of demand:
EPD =
P
Q

∆QD
∆P
=
3.46
2630
(-266) = -0.35
Thus demand is inelastic. We can likewise calculate the price elasticity of supply:
EPS =
P
Q

∆QS
∆P

=
3.46
2630
(240) = 0.32
Because these supply and demand curves are linear, the price elasticities will
vary as we move along the curves. for example, suppose that a drought caused the
supply curve to shift far enough to the left to push the price up to $4.00 per bushel.
In this case, the quantity demanded would fall to 3550 - (266)(4.00) = 2486
million bushels. At this price and quantity, the elasticity of demand would be
EPD =
4.00
2486
(-266) = -0.43
The wheat market has evolved over the years, in part because of
changes in demand. The demand for wheat has two components: domestic
(demand by U.S. consumers) and export (demand by foreign consumers).
During the past 30  years, domestic demand for wheat rose only slightly
(due to modest increases in population and income). Export demand,
however, fell sharply. There were several reasons. first and foremost was
the success of the Green Revolution in agriculture: Developing countries
like India, which had been large importers of wheat, became increas-
ingly self-sufficient. In addition, European countries adopted protectionist
policies that subsidized their own production and imposed tariff ba
iers
against imported wheat.
In 2007, demand and supply were
Demand: QD = 2900 - 125P
Supply: QS = 1460 + 115P
Once again, equating quantity supplied and quantity demanded yields the
market-clearing (nominal) price and quantity:
1460 + 115P = 2900 - 125P
P = $6.00 per bushel
Q = 1460 + (115)(6) = 2150 million bushels
Thus the price of wheat (in nominal terms) rose considerably since 1981. In fact,
nearly all of this increase occu
ed during 2005 to 2007. (In 2002, for example,
the price of wheat was only $2.78 per bushel.) The causes? Dry weather in
2005, even dryer weather in 2006, and heavy rains in 2007 combined with in-
creased export demand. You can check to see that, at the 2007 price and quan-
tity, the price elasticity of demand was - 0.35 and the price elasticity of supply
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 61
0.32. Given these low elasticities, it is not surprising that the price of wheat rose
so sharply.9
International demand for U.S. wheat fluctuates with the weather and political
conditions in other major wheat producing countries, such as China, India and
Russia. Between 2008 and 2010, U.S. wheat exports fell by 30% in the face of ro-
ust international production, so the price of wheat reached a low of $4.87 in 2010,
down from $6.48 two years earlier. Inclement weather led to shortfalls in 2011,
however, and U.S. exports shot up by 33%, driving the price up to $5.70 in 2011.
Although export demand for U.S. wheat has followed a downward trend,
the weather’s influence on wheat production is more ambiguous due to the
interaction of climate patterns with different wheat classes. for instance, in
2015 the unusually cold winter season led to declines in the crop conditions
for winter wheat with negligible effects on its production. Durum wheat pro-
duction, on the other hand, surged 53 percent from the previous year due to
favorable weather conditions during its spring growing season. figure 2.13
shows how the price of wheat has changed from 1980.
9These are short-run elasticity estimates from the Economics Research Service (ERS) of the U.S. Department
of Agriculture (USDA). For more information, consult the following publications: William Lin, Paul C.
Westcott, Robert Skinner, Scott Sanford, and Daniel G. De La To
e Ugarte, Supply Response Under the 1996
Farm Act and Implications for the U.S. Field Crops Sector (Technical Bulletin No. 1888, ERS, USDA, July 2000,
http:
www.ers.usda.gov/); and James Barnes and Dennis Shields, The Growth in U.S. Wheat Food Demand
(Wheat Situation and Outlook Yea
ook, WHS-1998, http:
www.ers.usda.gov/). For background informa-
tion on U.S. wheat production, see http:
www.ers.usda.gov/topics/crops/wheat
ackground.aspx.
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
1980
D
ol
la
s
p
e

u
sh
el
Nominal Price   
Real Price (2000$)
1985 1990 1995 2000
Yea
2005 2010 2015 2020
Figure 2.13
The priCe Of WheaT in The u.S.
The price of wheat fluctuates in response to the weather and changes in export demand.
http:
www.ers.usda.gov
http:
www.ers.usda.gov
http:
www.ers.usda.gov/topics/crops/wheat
ackground.aspx
62 PART 1 Introduction: Markets and Prices
2.5 Short-run versus long-run elasticities
When analyzing demand and supply, we must distinguish between the short run
and the long run. In other words, if we ask how much demand or supply changes
in response to a change in price, we must be clear about how much time is allowed
to pass before we measure the changes in the quantity demanded or supplied. If we allow
only a short time to pass—say, one year or less—then we are dealing with the
short run. When we refer to the long run we mean that enough time is allowed for
consumers or producers to adjust fully to the price change. In general, short-run
demand and supply curves look very different from their long-run counterparts.
Demand
For many goods, demand is much more price elastic in the long run than in the short
un. For one thing, it takes time for people to change their consumption habits. For
example, even if the price of coffee rises sharply, the quantity demanded will fall
only gradually as consumers begin to drink less. In addition, the demand for a good
might be linked to the stock of another good that changes only slowly. For example,
the demand for gasoline is much more elastic in the long run than in the short run. A
sharply higher price of gasoline reduces the quantity demanded in the short run by
causing motorists to drive less, but it has its greatest impact on demand by inducing
consumers to buy smaller and more fuel-efficient cars. But because the stock of cars
changes only slowly, the quantity of gasoline demanded falls only slowly. Figure
2.14(a) shows short-run and long-run demand curves for goods such as these.
DemanD anD DurabiliTy On the other hand, for some goods just the op-
posite is true—demand is more elastic in the short run than in the long run.
Because these goods (automobiles, refrigerators, televisions, or the capital
equipment purchased by industry) are durable, the total stock of each good
owned by consumers is large relative to annual production. As a result, a small
change in the total stock that consumers want to hold can result in a large per-
centage change in the level of purchases.
Suppose, for example, that the price of refrigerators goes up 10 percent, caus-
ing the total stock of refrigerators that consumers want to hold to drop 5 per-
cent. Initially, this will cause purchases of new refrigerators to drop much more
than 5 percent. But eventually, as consumers’ refrigerators depreciate (and units
We found that the market-clearing price of wheat was $3.46 in 1981, but
in fact the price was greater than this. Why? Because the U.S. government
ought wheat through its price support program. In addition, farmers have been
eceiving direct subsidies for the wheat they produce. This aid to farmers (at
the expense of taxpayers) has increased in magnitude. In 2002—and again in
2008—Congress passed legislation continuing (and in some cases expanding)
subsidies to farmers. The food, Conservation, and Energy Act of 2008 autho-
ized farm aid through 2012, at a projected cost of $284 billion over five years.
Recent U.S. budget crises, however, have given support to those in Congress
who feel these subsidies should end.
Agricultural policies that support farmers exist in the United States, Europe,
Japan, and many other countries. We discuss how these policies work, and
evaluate the costs and benefits for consumers, farmers, and the government
udget in Chapter 9.
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 63
must be replaced), the quantity demanded will increase again. In the long run,
the total stock of refrigerators owned by consumers will be about 5 percent less
than before the price increase. In this case, while the long-run price elasticity
of demand for refrigerators would be - .05/.10 = -0.5, the short-run elasticity
would be much larger in magnitude.
Or consider automobiles. Although annual U.S. demand—new car purchases—
is about 10 to 12 million, the stock of cars that people own is around 130 million.
If automobile prices rise, many people will delay buying new cars. The quantity
demanded will fall sharply, even though the total stock of cars that consumers
might want to own at these higher prices falls only a small amount. Eventually,
however, because old cars wear out and must be replaced, the quantity of new
cars demanded picks up again. As a result, the long-run change in the quantity
demanded is much smaller than the short-run change. Figure 2.14(b) shows de-
mand curves for a durable good like automobiles.
inCOme elaSTiCiTieS Income elasticities also differ from the short run to
the long run. For most goods and services—foods, beverages, fuel, entertain-
ment, etc.—the income elasticity of demand is larger in the long run than in the
short run. Consider the behavior of gasoline consumption during a period of
strong economic growth during which aggregate income rises by 10 percent.
Eventually people will increase gasoline consumption because they can afford
to take more trips and perhaps own larger cars. But this change in consumption
Price
Quantity
Price
Quantity
DSR
(a) (b)
DLR
DLR
DSR
Figure 2.14
(a) gaSOline: ShOrT-run anD lOng-run DemanD CurveS
(b) auTOmObileS: ShOrT-run anD lOng-run DemanD CurveS
(a) In the short run, an increase in price has only a small effect on the quantity of gasoline demanded. Motorists may
drive less, but they will not change the kinds of cars they are driving overnight. In the longer run, however, because
they will shift to smaller and more fuel-efficient cars, the effect of the price increase will be larger. Demand, there-
fore, is more elastic in the long run than in the short run. (b) The opposite is true for automobile demand. If price
increases, consumers initially defer buying new cars; thus annual quantity demanded falls sharply. In the longer run,
however, old cars wear out and must be replaced; thus annual quantity demanded picks up. Demand, therefore, is
less elastic in the long run than in the short run.
64 PART 1 Introduction: Markets and Prices
takes time, and demand initially increases only by a small amount. Thus, the
long-run elasticity will be larger than the short-run elasticity.
For a durable good, the opposite is true. Again, consider automobiles. If
aggregate income rises by 10 percent, the total stock of cars that consumers
will want to own will also rise—say, by 5 percent. But this change means a
much larger increase in cu
ent purchases of cars. (If the stock is 130 million, a
5- percent increase is 6.5 million, which might be about 60 to 70 percent of nor-
mal demand in a single year.) Eventually consumers succeed in increasing the
total number of cars owned; after the stock has been rebuilt, new purchases
are made largely to replace old cars. (These new purchases will still be greater
than before because a larger stock of cars outstanding means that more cars
need to be replaced each year.) Clearly, the short-run income elasticity of de-
mand will be much larger than the long-run elasticity.
CyCliCal inDuSTrieS Because the demands for durable goods fluctuate so
sharply in response to short-run changes in income, the industries that produce
these goods are quite vulnerable to changing macroeconomic conditions, and in
particular to the business cycle—recessions and booms. Thus, these industries
are often called cyclical    industries—their sales patterns tend to magnify cycli-
cal changes in gross domestic product (GDP) and national income.
Figures 2.15 and 2.16 illustrate this principle. Figure 2.15 plots two variables
over time: the annual real (inflation-adjusted) rate of growth of GDP and the
annual real rate of growth of investment in producers’ durable equipment
cyclical industries Industries
in which sales tend to magnify
cyclical changes in gross domestic
product and national income.
220
215
210
25
0
5
10
15
20
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020
G
o
w
th
R
at
e
(a
nn
ua
l p
e
ce
nt
ag
e)
Yea
GDP
Equipment
Investment
Figure 2.15
gDp anD inveSTmenT in Durable equipmenT
Annual growth rates are compared for GDP and investment in durable equipment. Because the short-run
GDP elasticity of demand is larger than the long-run elasticity for long-lived capital equipment, changes
in investment in equipment magnify changes in GDP. Thus capital goods industries are considered
“cyclical.”
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 65
(i.e., machinery and other equipment purchased by firms). Note that although
the durable equipment series follows the same pattern as the GDP series, the
changes in GDP are magnified. For example, in 1961–1966 GDP grew by at least
4 percent each year. Purchases of durable equipment also grew, but by much
more (over 10 percent in 1963–1966). Equipment investment likewise grew
much more quickly than GDP during 1993–1998. On the other hand, during the
ecessions of 1974–1975, 1982, 1991, 2001, and 2008, equipment purchases fell by
much more than GDP.
Figure 2.16 also shows the real rate of growth of GDP, along with the
annual real rates of growth of spending by consumers on durable goods
(automobiles, appliances, etc.) and nondurable goods (food, fuel, clothing,
etc.). Note that while both consumption series follow GDP, only the durable
goods series tends to magnify changes in GDP. Changes in consumption of
nondurables are roughly the same as changes in GDP, but changes in con-
sumption of durables are usually several times larger. This is why companies
such as General Motors and General Electric are considered “cyclical”: Sales
of cars and electrical appliances are strongly affected by changing macroeco-
nomic conditions.
210
25
0
5
10
15
20
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020
G
o
w
th
R
at
e
(a
nn
ua
l p
e
ce
nt
ag
e)
Yea
Durables
Nondurables
GDP
Figure 2.16
COnSumpTiOn Of DurableS verSuS nOnDurableS
Annual growth rates are compared for GDP, consumer expenditures on durable goods (automobiles, appliances, fur-
niture, etc.), and consumer expenditures on nondurable goods (food, clothing, services, etc.). Because the stock of
durables is large compared with annual demand, short-run demand elasticities are larger than long-run elasticities.
Like capital equipment, industries that produce consumer durables are “cyclical” (i.e., changes in GDP are magni-
fied). This is not true for producers of nondurables.
66 PART 1 Introduction: Markets and Prices
exaMple 2.6 The DemanD fOr gaSOline
anD auTOmObileS
Gasoline and automobiles exemplify some of the different characteristics of
demand discussed above. They are complementary goods—an increase in the
price of one tends to reduce the demand for the other. In addition, their respec-
tive dynamic behaviors (long-run versus short-run elasticities) are just the oppo-
site from each other. for gasoline, the long-run price and income elasticities are
larger than the short-run elasticities; for automobiles, the reverse is true.
There have been a number of statistical studies of the demands for gasoline
and automobiles. here we report elasticity estimates based on several that em-
phasize the dynamic response of demand.10 Table 2.1 shows price and income
elasticities of demand for gasoline in the United States for the short run, the
long run, and just about everything in between.
10For gasoline and automobile demand studies and elasticity estimates, see R. S. Pindyck,
The Structure of World Energy Demand (Cam
idge, MA: MIT Press, 1979); Carol Dahl and
Thomas Sterner, “Analyzing Gasoline Demand Elasticities: A Survey,” Energy Economics (July 1991);
Molly Espey, “Gasoline Demand Revised: An International Meta-Analysis of Elasticities,” Energy
Economics (July 1998); David L. Greene, James R. Kahn, and Robert C. Gibson, “Fuel Economy
Rebound Effects for U.S. Household Vehicles,” The Energy Journal 20 (1999); Daniel Graham and
Stephen Glaister, “The Demand for Automobile Fuel: A Survey of Elasticities,” Journal of Transport
Economics and Policy 36 (January 2002); and Ian Pa
y and Kenneth Small, “Does Britain or the
United States Have the Right Gasoline Tax?” American Economic Review 95 (2005).
Note the large differences between the long-run and the short-run elastici-
ties. following the sharp increases that occu
ed in the price of gasoline with
the rise of the OPEC oil cartel in 1974, many people (including executives in
the automobile and oil industries) claimed that the quantity of gasoline de-
manded would not change much—that demand was not very elastic. Indeed,
for the first year after the price rise, they were right. But demand did eventually
change. It just took time for people to alter their driving habits and to replace
large cars with smaller and more fuel-efficient ones. This response continued
after the second sharp increase in oil prices that occu
ed in 1979–1980. It
was partly because of this response that OPEC could not maintain oil prices
above $30 per ba
el, and prices fell. The oil and gasoline price increases that
occu
ed in 2005–2011 likewise led to a gradual demand response. In 2014,
oil prices began to decrease at an accelerated pace, leading to an increased
demand for light trucks and a decreased demand for cars.
Table 2.2 shows price and income elasticities of demand for automobiles.
Note that the short-run elasticities are much larger than the long-run elasticities.
Table 2.1 DemanD fOr gaSOline
Number of Years allowed to Pass followiNg
a Price or iNcome chaNge
Elasticity 1 2 3 5 10
Price -0.2 -0.3 -0.4 -0.5 -0.8
income 0.2 0.4 0.5 0.6 1.0
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 67
It should be clear from the income elasticities why the automobile industry is
so highly cyclical. for example, GDP fell 2 percent in real (inflation-adjusted)
terms during the 1991 recession, but automobile sales fell by about 8 percent.
Auto sales began to recover in 1993, and rose sharply between 1995 and 1999.
During the 2008 recession, GDP fell by nearly 3 percent, and car and truck
sales decreased by 21%. Sales began to recover in 2010, when they increased
y nearly 10%. By 2015, sales had risen above pre-recession levels.
Table 2.2 DemanD fOr auTOmObileS
Number of Years allowed to Pass followiNg
a Price or iNcome chaNge
Elasticity 1 2 3 5 10
Price -1.2 -0.9 -0.8 -0.6 -0.4
income 3.0 2.3 1.9 1.4 1.0
Supply
Elasticities of supply also differ from the long run to the short run. For most
products, long-run supply is much more price elastic than short-run supply:
Firms face capacity constraints in the short run and need time to expand capacity
y building new production facilities and hiring workers to staff them. This is
not to say that the quantity supplied will not increase in the short run if price
goes up sharply. Even in the short run, firms can increase output by using their
existing facilities for more hours per week, paying workers to work overtime,
and hiring some new workers immediately. But firms will be able to expand
output much more when they have the time to expand their facilities and hire
larger permanent workforces.
For some goods and services, short-run supply is completely inelastic. Rental
housing in most cities is an example. In the very short run, there is only a fixed
number of rental units. Thus an increase in demand only pushes rents up. In
the longer run, and without rent controls, higher rents provide an incentive to
enovate existing buildings and construct new ones. As a result, the quantity
supplied increases.
For most goods, however, firms can find ways to increase output even in the
short run—if the price incentive is strong enough. However, because various
constraints make it costly to increase output rapidly, it may require large price
increases to elicit small short-run increases in the quantity supplied. We discuss
these characteristics of supply in more detail in Chapter 8.
Supply anD DurabiliTy For some goods, supply is more elastic in the
short run than in the long run. Such goods are durable and can be recycled
as part of supply if price goes up. An example is the secondary supply of met-
als: the supply from scrap metal, which is often melted down and refa
i-
cated. When the price of copper goes up, it increases the incentive to convert
scrap copper into new supply, so that, initially, secondary supply increases
sharply. Eventually, however, the stock of good-quality scrap falls, making the
68 PART 1 Introduction: Markets and Prices
melting, purifying, and refa
icating more costly. Secondary supply then con-
tracts. Thus the long-run price elasticity of secondary supply is smaller than
the short-run elasticity.
Figures 2.17(a) and 2.17(b) show short-run and long-run supply curves for
primary (production from the mining and smelting of ore) and secondary cop-
per production. Table 2.3 shows estimates of the elasticities for each component
of supply and for total supply, based on a weighted average of the component
elasticities.11 Because secondary supply is only about 20 percent of total supply,
the price elasticity of total supply is larger in the long run than in the short run.
11These estimates were obtained by aggregating the regional estimates reported in Franklin
M. Fisher, Paul H. Cootner, and Martin N. Baily, “An Econometric Model of the World Copper
Industry,” Bell Journal of Economics 3 (Autumn 1972): 568–609.
Price
Quantity
Price
Quantity
SSR
(a) (b)
SLR
SSR
SLR
Figure 2.17
COpper: ShOrT-run anD lOng-run Supply CurveS
Like that of most goods, the supply of primary copper, shown in part (a), is more elastic in the long run.
If price increases, firms would like to produce more but are limited by capacity constraints in the short
un. In the longer run, they can add to capacity and produce more. Part (b) shows supply curves for
secondary copper. If the price increases, there is a greater incentive to convert scrap copper into new
supply. Initially, therefore, secondary supply (i.e., supply from scrap) increases sharply. But later, as the
stock of scrap falls, secondary supply contracts. Secondary supply is therefore less elastic in the long
un than in the short run.
Table 2.3 Supply Of COppe
PricE Elasticity of: short-run long-run
Primary supply 0.20 1.60
secondary supply 0.43 0.31
total supply 0.25 1.50
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 69
exaMple 2.7 The WeaTher in
azil anD The priCe Of COffee in neW yOrk
Droughts or subfreezing weather
occasionally destroy or dam-
age many of Brazil’s coffee trees.
Because Brazil is by far the world’s
largest coffee producer the result is
a decrease in the supply of coffee
and a sharp run-up in its price.
In July 1975, for example, a frost
destroyed most of Brazil’s 1976–
1977 coffee crop. (Remember that it is winter in
Brazil when it is summer in the northern hemi-
sphere.) As figure 2.18 shows, the price of a pound
of coffee in New York went from 68 cents in 1975
to $1.23 in 1976 and $2.70 in 1977. Prices fell but
then jumped again in 1986, after a seven-month
drought in 1985 ruined much of Brazil’s crop. finally,
starting in June 1994, freezing weather followed
y a drought destroyed nearly half of Brazil’s crop.
As a result, the price of coffee in
1994–1995 was about double its
1993 level. By 2002, however, the
price had dropped to its lowest
level in 30 years. (Researchers pre-
dict that over the next 50 years,
global warming may eliminate as
much as 60 percent of Brazil’s cof-
fee-growing areas, resulting in a
major decline in coffee production and an increase in
prices. Should that happen, we will discuss it in the
twentieth edition of this book.)
The important point in figure 2.18 is that any run-
up in price following a freeze or drought is usually
short-lived. Within a year, price begins to fall; within
three or four years, it returns to its earlier levels. In
1978, for example, the price of coffee in New York
fell to $1.48 per pound, and by 1983, it had fallen in
$0.00
$0.50
$1.00
$1.50
$2.00
$2.50
$3.00
$3.50
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020
N
om
in
al
P
i
ce
(d
ol
la
s
p
e
p
ou
nd
)
Yea
Figure 2.18
priCe Of
azilian COffee
When droughts or freezes damage Brazil’s coffee trees, the price of coffee can soar. The
price usually falls again after a few years, as demand and supply adjust.
70 PART 1 Introduction: Markets and Prices
12During 1980, however, prices temporarily went just above $2.00 per pound as a result of export
quotas imposed under the International Coffee Agreement (ICA). The ICA is essentially a cartel
agreement implemented by the coffee-producing countries in 1968. It has been largely ineffective
and has seldom had an effect on price. We discuss cartel pricing in detail in Chapter 12.
13You can learn more about the world coffee market from the Foreign Agriculture Service of the U.S.
Department of Agriculture by visiting their website at http:
www.fas.usda.gov/htp/coffee.asp.
Another good source of information is http:
www.nationalgeographic.com/coffee.
eal (inflation-adjusted) terms to within a few cents
of its prefreeze 1975 price.12 Likewise, in 1987 the
price of coffee fell to below its predrought 1984 level,
and then continued declining until the 1994 freeze.
After hitting a low of 45 cents per pound in 2002,
coffee prices increased at an average rate of 17% per
year, reaching $1.46—equal to the 1995 peak—in
2010. Brazilian coffee growers have worked to in-
crease their production in the past decade, but bad
weather has led to inconsistent crop yields.
Coffee prices behave this way because both
demand and supply (especially supply) are much
more elastic in the long run than in the short run.
figure 2.19 illustrates this fact. Note from part (a) of
the figure that in the very short run (within one or two
months after a freeze), supply is completely inelastic:
There are simply a fixed number of coffee beans,
some of which have been damaged by the frost.
Demand is also relatively inelastic. As a result of the
frost, the supply curve shifts to the left, and price in-
creases sharply, from P0 to P1.
In the intermediate run—say, one year after the
freeze—both supply and demand are more elastic,
supply because existing trees can be harvested more
intensively (with some decrease in quality), and de-
mand because consumers have had time to change
their buying habits. As part (b) shows, although the in-
termediate-run supply curve also shifts to the left, price
has come down from P1 to P2. The quantity supplied
has also increased somewhat from the short run, from
Q1 to Q2. In the long run shown in part (c), price returns
to its normal level because growers have had time to re-
place trees damaged by the freeze. The long-run supply
curve, then, simply reflects the cost of producing coffee,
including the costs of land, of planting and caring for
the trees, and of a competitive rate of profit.13
Q2
P2
Price
QuantityQ0
S′ S
D
P0
Q0
P0 S
D
Price
QuantityQ1
P0
S′
D
Price
Quantity
S
P1
Q0
(c)(a) (b)
Figure 2.19
Supply anD DemanD fOr COffee
(a) A freeze or drought in Brazil causes the supply curve to shift to the left. In the short run, supply is completely inelastic;
only a fixed number of coffee beans can be harvested. Demand is also relatively inelastic; consumers change their habits only
slowly. As a result, the initial effect of the freeze is a sharp increase in price, from P0 to P1. (b) In the intermediate run, supply
and demand are both more elastic; thus price falls part of the way back, to P2. (c) In the long run, supply is extremely elastic;
ecause new coffee trees will have had time to mature, the effect of the freeze will have disappeared. Price returns to P0.
http:
www.fas.usda.gov/htp/coffee.asp
http:
www.nationalgeographic.com/coffee
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 71
*2.6 understanding and predicting
the effects of Changing Market
Conditions
So far, our discussion of supply and demand has been largely qualitative. To
use supply and demand curves to analyze and predict the effects of changing
market conditions, we must begin attaching numbers to them. For example, to
see how a 50-percent reduction in the supply of Brazilian coffee may affect the
world price of coffee, we must determine actual supply and demand curves and
then calculate the shifts in those curves and the resulting changes in price.
In this section, we will see how to do simple “back of the envelope” cal-
culations with linear supply and demand curves. Although they are often
approximations of more complex curves, we use linear curves because they
are easier to work with. It may come as a surprise, but one can do some infor-
mative economic analyses on the back of a small envelope with a pencil and a
pocket calculator.
First, we must learn how to “fit” linear demand and supply curves to
market data. (By this we do not mean statistical fitting in the sense of linear
egression or other statistical techniques, which we will discuss later in the
ook.) Suppose we have two sets of numbers for a particular market: The
first set consists of the price and quantity that generally prevail in the mar-
ket (i.e., the price and quantity that prevail “on average,” when the market
is in equili
ium or when market conditions are “normal”). We call these
numbers the equili
ium price and quantity and denote them by P* and Q*.
The second set consists of the price elasticities of supply and demand for
the market (at or near the equili
ium), which we denote by ES and ED, as
efore.
These numbers may come from a statistical study done by someone else;
they may be numbers that we simply think are reasonable; or they may be
numbers that we want to try out on a “what if” basis. Our goal is to write down
the supply and demand curves that fit (i.e., are consistent with) these numbers. We can
then determine numerically how a change in a variable such as GDP, the price
of another good, or some cost of production will cause supply or demand to
shift and thereby affect market price and quantity.
Let’s begin with the linear curves shown in Figure 2.20. We can write these
curves alge
aically as follows:
Demand: Q = a - bP     (2.5a)
Supply: Q = c + dP     (2.5b)
Our problem is to choose numbers for the constants a, b, c, and d. This is done,
for supply and for demand, in a two-step procedure:
•     Step 1: Recall that each price elasticity, whether of supply or demand, can be
written as
E = (P/Q)(∆Q/∆P)
where ∆Q/∆P is the change in quantity demanded or supplied result-
ing from a small change in price. For linear curves, ∆Q/∆P is constant.
From equations (2.5a) and (2.5b), we see that ∆Q/∆P = d for supply and
72 PART 1 Introduction: Markets and Prices
∆Q/∆P = -b for demand. Now, let’s substitute these values for ∆Q/∆P
into the elasticity formula:
Demand: ED = -b(P∗/Q ∗)     (2.6a)
Supply: ES = d(P∗/Q ∗)     (2.6b)
where P* and Q* are the equili
ium price and quantity for which we have
data and to which we want to fit the curves. Because we have numbers for ES,
ED, P*, and Q*, we can substitute these numbers in equations (2.6a) and (2.6b)
and solve for b and d.
•     Step 2: Since we now know b and d, we can substitute these numbers, as
well as P* and Q*, into equations (2.5a) and (2.5b) and solve for the remain-
ing constants a and c. For example, we can rewrite equation (2.5a) as
a = Q ∗ + bP∗
and then use our data for Q* and P*, together with the number we calculated
in Step 1 for b, to obtain a.
Let’s apply this procedure to a specific example: long–run supply and demand
for the world copper market. The relevant numbers for this market are as follows:
Quantity Q ∗ = 18 million metric tons per year (mmt/yr)
Price P∗ = $3.00 per pound
Elasticity of suppy ES = 1.5
Elasticity of demand ED = -0.5.
(The price of copper has fluctuated during the past few decades between
$0.60 and more than $4.00, but $3.00 is a reasonable average price for 2008–2011).
We begin with the supply curve equation (2.5b) and use our two-step proce-
dure to calculate numbers for c and d. The long-run price elasticity of supply is
1.5, P∗ = $3.00, and Q ∗ = 18.
Q*
P*
Price
Quantitya
–c/d
a
Demand: Q = a – bP
Supply: Q = c + dP
ED = –b(P*/Q*)
ES = d(P*/Q*)
Figure 2.20
fiTTing linear
Supply anD
DemanD CurveS TO
DaTa
Linear supply and demand
curves provide a convenient tool
for analysis. Given data for the
equili
ium price and quantity P*
and Q*, as well as estimates of
the elasticities of demand and
supply ED and ES, we can cal-
culate the parameters c and d
for the supply curve and a and
for the demand curve. (In the
case drawn here, c 6 0.) The
curves can then be used to ana-
lyze the behavior of the market
quantitatively.
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 73
•     Step 1: Substitute these numbers in equation (2.6b) to determine d:
1.5 = d(3/18) = d/6
so that d = (1.5)(6) = 9.
•     Step 2: Substitute this number for d, together with the numbers for P* and
Q*, into equation (2.5b) to determine c:
18 = c + (9)(3.00) = c + 27
so that c = 18 - 27 = -9. We now know c and d, so we can write our supply
curve:
Supply: Q = -9 + 9P
We can now follow the same steps for the demand curve equation (2.5a).
An estimate for the long-run elasticity of demand is -0.5.14 First, substitute this
number, as well as the values for P* and Q*, into equation (2.6a) to determine b:
-0.5 = -b(3/18) = -
6
so that b = (0.5)(6) = 3. Second, substitute this value for b and the values for P*
and Q* in equation (2.5a) to determine a:
18 = a = (3)(3) = a - 9
so that a = 18 + 9 = 27. Thus, our demand curve is:
Demand: Q = 27 - 3P
To check that we have not made a mistake, let’s set the quantity supplied
equal to the quantity demanded and calculate the resulting equili
ium price:
Supply = -9 + 9P = 27 - 3P = Demand
9P + 3P = 27 + 9
or P = 36/12 = 3.00, which is indeed the equili
ium price with which we began.
Although we have written supply and demand so that they depend only on
price, they could easily depend on other variables as well. Demand, for exam-
ple, might depend on income as well as price. We would then write demand as
Q = a - bP + f I     (2.7)
where I is an index of the aggregate income or GDP. For example, I might equal
1.0 in a base year and then rise or fall to reflect percentage increases or decreases
in aggregate income.
For our copper market example, a reasonable estimate for the long-run
income elasticity of demand is 1.3. For the linear demand curve (2.7), we can
then calculate f by using the formula for the income elasticity of demand:
E = (I/Q)(∆Q/∆I). Taking the base value of I as 1.0, we have
1.3 = (1.0/18)( f ).
Thus f = (1.3)(18)/(1.0) = 23.4. Finally, substituting the values b = 3, f = 23.4,
P∗ = 3.00, and Q ∗ = 18 into equation (2.7), we can calculate that a must equal 3.6.
We have seen how to fit linear supply and demand curves to data. Now, to
see how these curves can be used to analyze markets, let’s look at Example 2.8,
which deals with the behavior of copper prices, and Example 2.9, which con-
cerns the world oil market.
14 See Claudio Agostini, “Estimating Market Power in the U.S. Copper Industry,” Review of Industrial
Organization 28 (2006), 17–39.
74 PART 1 Introduction: Markets and Prices
exaMple 2.8 The behaviOr Of COpper priCeS
After reaching a level of about $1.00 per pound in
1980, the price of copper fell to about 60 cents per
pound in 1986. In real (inflation-adjusted) terms,
this price was even lower than during the Great
Depression 50 years earlier. Prices increased in
1988–1989 and in 1995, largely as a result of strikes
y miners in Peru and Canada that disrupted sup-
plies, but then fell again from 1996 through 2003.
Prices increased sharply, however, between 2003
and 2007, and while copper fell along with many
other commodities during the 2008–2009 recession,
the price of copper had recovered by early 2010.
Yet the recovery was short lived. After 2011, copper
prices declined through 2016. figure 2.21 shows the
ehavior of copper prices from 1965 to 2016 in both
eal and nominal terms.
Worldwide recessions in 1980 and 1982 contrib-
uted to the decline of copper prices; as mentioned
above, the income elasticity of copper demand is
about 1.3. But copper demand did not pick up as
the industrial economies recovered during the mid-
1980s. Instead, the 1980s saw a steep decline in
demand.
The price decline through 2003 occu
ed for two
easons. first, a large part of copper consumption is
for the construction of equipment for electric power
generation and transmission. But by the late 1970s,
the growth rate of electric power generation had
fallen dramatically in most industrialized countries.
In the United States, for example, the growth rate
fell from over 6 percent per annum in the 1960s and
early 1970s to less than 2 percent in the late 1970s
0
20
40
60
80
100
120
140
160
180
200
220
240
260
280
300
320
340
360
380
400
420
440
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
P
ic
e
(c
en
ts
p
e
p
ou
nd
)
Nominal Price
Real Price (2000$)
Figure 2.21
COpper priCeS, 1965–2016
Copper prices are shown in both nominal (no adjustment for inflation) and real (inflation-adjusted) terms.
In real terms, copper prices declined steeply from the early 1970s through the mid-1980s as demand fell.
In 1988–1990, copper prices rose in response to supply disruptions caused by strikes in Peru and Canada
ut later fell after the strikes ended. Prices declined during the 1996–2002 period but then increased
sharply starting in 2005. The increase lasted through 2011, with prices declining again between 2012 and
2016.
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 75
15 Our thanks to Patricia Foley, Executive Director of the American Bureau of Metal Statistics, for
supplying the data on China. Other data are from the Monthly and Annual Reports of the U.S.
Geological Survey Mineral Resources Program—http:
minerals.usgs.gov/minerals/pubs/copper.
16 The boom created hundreds of new jobs, which in turn led to increases in housing prices: “Copper
Boom Creates Housing Crunch,” The Arizona Republic, July 12, 2007.
and 1980s. This decline meant a big drop in what
had been a major source of copper demand. Second,
in the 1980s, other materials, such as aluminum and
fiber optics, were increasingly substituted for copper.
Why did the price increase so sharply after 2003?
first, the demand for copper from China and other
Asian countries began increasing dramatically, re-
placing the demand from Europe and the U.S. Chinese
copper consumption, for example, has nearly tripled
since 2001. Second, because prices had dropped
so much from 1996 through 2003, producers in the
U.S., Canada, and Chile closed unprofitable mines
and cut production. Between 2000 and 2003, for
example, U.S. mine production of copper declined by
23 percent.15
One might expect increasing prices to stimulate
investments in new mines and increases in produc-
tion, and that is indeed what has happened. Arizona,
for example, experienced a copper boom as Phelps
Dodge opened a major new mine in 2007.16 By
2007, producers began to wo
y that prices would
decline again, either as a result of these new invest-
ments or because demand from Asia would level off
or even drop.
The producers’ wo
ies were certainly wa
anted. By
2011, China had become the world’s largest importer
of copper, representing some 40 percent of world
demand. But China’s economy began to weaken that
year, and its consumption of copper and other metals
egan to fall. That decline in Chinese copper demand
0
1
2
3
4
5
0 5 10 15 20 25 30
Quantity (million metric tons/yr)
P
ic
e
(d
ol
la
s
p
e
p
ou
nd
) P* 5 3.00
P9 5 2.04
Q9 5 9.4
Q* 5 18
S
D
D9
Figure 2.22
COpper Supply anD DemanD
The shift in the demand curve leads to a nearly 55-percent decline in price.
http:
minerals.usgs.gov/minerals/pubs/coppe
76 PART 1 Introduction: Markets and Prices
17Note that because we have multiplied the demand function by 0.45—i.e., reduced the quantity de-
manded at every price by 55 percent—the new demand curve is not parallel to the old one. Instead,
the curve rotates downward at its intersection with the price axis.
continued for the next five years, and was accompa-
nied by declining demand in Russia, Brazil, India, and
other countries. The total drop in demand was some-
thing on the order of 50 percent! Not surprisingly, cop-
per prices responded; they declined substantially from
2012 to 2016, falling to around $2 per pound.
What would a decline in demand do to the price
of copper? To find out, we can use the linear supply
and demand curves that we just derived. Let’s cal-
culate the effect on price of a 55-percent decline in
demand. Because we are not concerned here with
the effects of GDP growth, we can leave the income
term, fI, out of the demand equation.
We want to shift the demand curve to the left by
55 percent. In other words, we want the quantity
demanded to be 45 percent of what it would be
otherwise for every value of price. for our linear de-
mand curve, we simply multiply the right–hand side
y 0.45:
Q = (0.45)(27 - 3P) = 12.15 - 1.35P
Supply is again Q = -9 + 9P. Now we can equate
the quantity supplied and the quantity demanded
and solve for price:
-9 + 9P = 12.15 - 1.35P
or P = 21.15/10.35 = $2.04 per pound. A decline
in demand of 55 percent therefore results in a drop
in price from $3.00 per pound to just over $2.00 per
pound as shown in figure 2.22.17
exaMple 2.9 upheaval in The WOrlD Oil markeT
Since the early 1970s, the world oil mar-
ket has been buffeted by the OPEC cartel
and by political turmoil in the Persian
Gulf. In 1974, by collectively restrain-
ing output, OPEC (the Organization of
Petroleum Exporting Countries) pushed
world oil prices well above what they
would have been in a competitive mar-
ket. OPEC could do this because it ac-
counted for much of world oil produc-
tion. During 1979–1980, oil prices shot up again, as the Iranian revolution
and the out
eak of the Iran-Iraq war sharply reduced Iranian and Iraqi
production. During the 1980s, the price gradually declined, as demand fell
and competitive (i.e., non-OPEC) supply rose in response to price. Prices
emained relatively stable during 1988–2001, except for a temporary spike
in 1990 following the Iraqi invasion of Kuwait. Prices increased again in
2002–2003 as a result of a strike in Venezuela and then the war with Iraq
that began in the spring of 2003. Oil prices continued to increase through
the summer of 2008 as a result of rising demand in Asia and reductions
in OPEC output. By the end of 2008, the recession had reduced demand
around the world, leading prices to plummet 127% in six months. Between
2009 and 2012, oil prices partially recovered, partially buoyed by contin-
ued growth in China, Brazil, and India. But after 2012, oil prices dropped
dramatically for two reasons. first, they were driven down by a sharp
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 77
18For a nice overview of the factors that have affected world oil prices, see James D. Hamilton,
“Understanding Crude Oil Prices,” The Energy Journal, 2009, Vol. 30, pp. 179–206.
economic slowdown in China, Brazil, and Russia, and stagnant growth
in much of Europe. Second, they responded to a large increase in supply,
mostly from new U.S oil sources. By 2015, the price of oil had dropped be-
low $40 per ba
el. figure 2.23 shows the world price of oil from 1970 to
2015, in both nominal and real terms.18
The Persian Gulf is one of the less stable regions of the world—a fact that
has led to concern over the possibility of new oil supply disruptions and sharp
increases in oil prices. What would happen to oil prices—in both the short run
and longer run—if a war or revolution in the Persian Gulf caused a sharp cut-
ack in oil production? Let’s see how simple supply and demand curves can be
used to predict the outcome of such an event.
Because this example is set in 2015–2016, all prices are measured in 2015
dollars. here are some rough figures:
•     2015–2016 world price = $50 per ba
el
•     World demand and total supply = 35 billion ba
els per year (
yr)
•     OPEC supply = 12
y
•     Competitive (non-OPEC) supply = 23
y
•     Saudi production = 3.6
yr (part of OPEC)
0
20
40
60
80
100
120
140
160
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
P
ic
e
(d
ol
la
s
p
e

a
e
l)

Real Price (2015$)
Nominal
Figure 2.23
priCe Of CruDe Oil
The OPEC cartel and political events caused the price of oil to rise sharply at times. It later
fell as supply and demand adjusted.
78 PART 1 Introduction: Markets and Prices
19For the sources of these numbers and a more detailed discussion of OPEC oil pricing, see Robert S.
Pindyck, “Gains to Producers from the Cartelization of Exhaustible Resources,” Review of Economics
and Statistics 60 (May 1978): 238–51; James M. Griffin and David J. Teece, OPEC Behavior and World Oil
Prices (London: Allen and Unwin, 1982); and John C. B. Cooper, “Price Elasticity of Demand for Crude
Oil: Estimates for 23 Countries,” Organization of the Petroleum Exporting Countries Review (March 2003).
The following table gives price elasticity estimates for oil supply and demand:19
short run long run
World demand: -0.05 -0.30
competitive supply: 0.05 0.30
You should verify that these numbers imply the following for demand and
competitive supply in the short run:
Short-run demand: D = 36.75 - 0.035P
Short@run competitive supply: SC = 21.85 + 0.023P
Of course, total supply is competitive supply plus OPEC supply, which we take
as constant at 12
yr. Adding this 12
yr to the competitive supply curve
above, we obtain the following for the total short-run supply:
Short@run total supply: ST = 33.85 + 0.023P
You should verify that the quantity demanded and the total quantity supplied
are equal at an equili
ium price of $50 per ba
el.
You should also verify that the co
esponding demand and supply curves for
the long run are as follows:
Long@run demand: D = 45.5 - 0.210P
Long@run competitive supply: SC = 16.1 + 0.138P
Long@run total supply: ST = 28.1 + 0.138P
Again, you can check that the quantities supplied and demanded equate at a
price of $50 per ba
el.
Saudi Arabia is one of the world’s largest oil producers, accounting for roughly
3.6
yr, which is nearly 10 percent of total world production. What would
happen to the price of oil if, because of war or political upheaval, Saudi Arabia
stopped producing oil? We can use our supply and demand curves to find out.
for the short run, simply subtract 3.6 from short-run total supply:
Short@run demand: D = 36.75 - 0.035P
Short@run total supply: ST = 30.25 + 0.023P
By equating this total quantity supplied with the quantity demanded, we can
see that in the short run, the price will more than double to $112.07 per ba
el.
figure 2.24 shows this supply shift and the resulting short-run increase in price.
The initial equili
ium is at the intersection of ST and D. After the drop in Saudi
production, the equili
ium occurs where ST′ and D cross.
In the long run, however, things will be different. Because both demand and
competitive supply are more elastic in the long run, the 3.6
yr cut in oil
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 79
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
0 5 10 15 20 25 30 35 40
Quantity (billion ba
els/year)
P
ic
e
(d
ol
la
s
p
e

a
e
l)
P* 5 50.00
P9 5 112.07
Q* 5 35
S9TSC ST
D
(a)
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
0 10 20 30 40 50 60
P
ic
e
(d
ol
la
s
p
e

a
e
l)
Quantity (billion ba
els/year)
D
S9TSC ST
P* 5 50.00
P9 5 60.34
Q* 5 35
(b)
Figure 2.24
impaCT Of SauDi prODuCTiOn CuT
The total supply is the sum of competitive (non-OPEC) supply and the 12
yr of OPEC supply. Part (a)
shows the short-run supply and demand curves. If Saudi Arabia stops producing, the supply curve will shift to
the left by 3.6
yr. In the short-run, price will increase sharply. Part (b) shows long-run curves. In the long run,
ecause demand and competitive supply are much more elastic, the impact on price will be much smaller.
80 PART 1 Introduction: Markets and Prices
20You can obtain recent data and learn more about the world oil market by accessing the websites of
the American Petroleum Institute at www.api.org or the U.S. Energy Information Administration at
www.eia.doe.gov.
production will no longer support such a high price. Subtracting 3.6 from long-
un total supply and equating with long-run demand, we can see that the price
will fall to $60.34, only $10.34 above the initial $50 price.
Thus, if Saudi Arabia suddenly stops producing oil, we should expect to see
about a doubling in price. however, we should also expect to see the price
gradually decline afterward, as demand falls and competitive supply rises.
This is indeed what happened following the sharp decline in Iranian and
Iraqi production in 1979–1980. history may or may not repeat itself, but if it
does, we can at least predict the impact on oil prices.20
2.7 effects of government
intervention—price Controls
In the United States and most other industrial countries, markets are rarely free
of government intervention. Besides imposing taxes and granting subsidies,
governments often regulate markets (even competitive markets) in a variety of
ways. In this section, we will see how to use supply and demand curves to ana-
lyze the effects of one common form of government intervention: price controls.
Later, in Chapter 9, we will examine the effects of price controls and other forms
of government intervention and regulation in more detail.
Figure 2.25 illustrates the effects of price controls. Here, P0 and Q0 are the equi-
li
ium price and quantity that would prevail without government regulation.
P0
S
D
Price
QuantityQ2Q0Q1
Excess Demand
Pmax
Figure 2.25
effeCTS Of priCe COnTrOlS
Without price controls, the market clears at the
equili
ium price and quantity P0 and Q0. If price
is regulated to be no higher than Pmax, the quan-
tity supplied falls to Q1, the quantity demanded
increases to Q2, and a shortage develops.
http:
www.api.org
http:
www.eia.doe.gov
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 81
The government, however, has decided that P0 is too high and mandated that the
price can be no higher than a maximum allowable ceiling price, denoted by Pmax.
What is the result? At this lower price, producers (particularly those with higher
costs) will produce less, and the quantity supplied will drop to Q1. Consumers, on
the other hand, will demand more at this low price; they would like to purchase
the quantity Q2. Demand therefore exceeds supply, and a shortage develops—i.e.,
there is excess demand. The amount of excess demand is Q2 - Q1.
This excess demand sometimes takes the form of queues, as when drivers lined
up to buy gasoline during the winter of 1974 and the summer of 1979. In both
instances, the lines were the result of price controls; the government prevented do-
mestic oil and gasoline prices from rising along with world oil prices. Sometimes
excess demand results in curtailments and supply rationing, as with natural gas
price controls and the resulting gas shortages of the mid-1970s, when industrial
consumers closed factories because gas supplies were cut off. Sometimes it spills
over into other markets, where it artificially increases demand. For example, natu-
al gas price controls caused potential buyers of gas to use oil instead.
Some people gain and some lose from price controls. As Figure 2.25 suggests,
producers lose: They receive lower prices, and some leave the industry. Some but
not all consumers gain. While those who can purchase the good at a lower price
are better off, those who have been “rationed out” and cannot buy the good at
all are worse off. How large are the gains to the winners and how large are the
losses to the losers? Do total gains exceed total losses? To answer these questions,
we need a method to measure the gains and losses from price controls and other
forms of government intervention. We discuss such a method in Chapter 9.
21This regulation began with the Supreme Court’s 1954 decision requiring the then Federal Power
Commission to regulate wellhead prices on natural gas sold to interstate pipeline companies. These
price controls were largely removed during the 1980s, under the mandate of the Natural Gas Policy
Act of 1978. For a detailed discussion of natural gas regulation and its effects, see Paul W. MacAvoy
and Robert S. Pindyck, The Economics of the Natural Gas Shortage (Amsterdam: North-Holland, 1975);
R. S. Pindyck, “Higher Energy Prices and the Supply of Natural Gas,” Energy Systems and Policy
2(1978): 177–209; and Arlon R. Tussing and Connie C. Barlow, The Natural Gas Industry (Cam
idge,
MA: Ballinger, 1984).
exaMple 2.10 priCe COnTrOlS anD naTural gaS ShOrTageS
In 1954, the federal government began regulating the
wellhead price of natural gas. Initially the controls
were not binding; the ceiling prices were above those
that cleared the market. But in about 1962, when these
ceiling prices did become binding, excess demand for
natural gas developed and slowly began to grow. In
the 1970s, this excess demand, spu
ed by higher oil
prices, became severe and led to widespread curtail-
ments. Soon ceiling prices were far below prices that
would have prevailed in a free market.21
Today, producers and industrial consumers of nat-
ural gas, oil, and other commodities are concerned
that the government might respond, once again, with
price controls if prices rise sharply. Let’s calculate the
likely impact of price controls on natural gas, based
on market conditions in 2007.
figure 2.26 shows the wholesale price of natural
gas, in both nominal and real (2000 dollars) terms,
from 1950 through 2007. The following numbers de-
scribe the U.S. market in 2007:
•     The (free-market) wholesale price of natural gas
was $6.40 per mcf (thousand cubic feet);
•     Production and consumption of gas were 23
Tcf (trillion cubic feet);
•     the average price of crude oil (which affects
the supply and demand for natural gas) was
about $50 per ba
el.
82 PART 1 Introduction: Markets and Prices
A reasonable estimate for the price elasticity of
supply is 0.2. higher oil prices also lead to more
natural gas production because oil and gas are often
discovered and produced together; an estimate of the
cross-price elasticity of supply is 0.1. As for demand,
the price elasticity is about -0.5, and the cross-price
elasticity with respect to oil price is about 1.5. You
can verify that the following linear supply and de-
mand curves fit these numbers:
Supply: Q = 15.90 + 0.72PG + 0.05PO
Demand: Q = 0.02 - 1.8PG + 0.69PO
where Q is the quantity of natural gas (in Tcf), PG
is the price of natural gas (in dollars per mcf), and
PO is the price of oil (in dollars per ba
el). You can
also verify, by equating the quantities supplied and
demanded and substituting $50 for PO, that these
supply and demand curves imply an equili
ium
free-market price of $6.40 for natural gas.
Suppose the government determines that the free-
market price of $6.40 per mcf is too high, decides to
impose price controls, and sets a maximum price of
$3.00 per mcf. What impact would this have on the
quantity of gas supplied and the quantity demanded?
Substitute $3.00 for PG in both the supply and de-
mand equations (keeping the price of oil, PO, fixed at
$50). You should find that the supply equation gives a
quantity supplied of 20.6 Tcf and the demand equation
a quantity demanded of 29.1 Tcf. Therefore, these price
controls would create an excess demand (i.e., short-
age) of 29.1 - 20.6 = 8.5 Tcf. In Example 9.1 we’ll
show how to measure the resulting gains and losses to
producers and consumers.
1950 1960 1970 1980
Yea
1990 2000 2010 2020
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
D
ol
la
s
p
e
th
ou
sa
nd
c
u
ic
fe
et
Real Price (2000$)
Nominal Price
Figure 2.26
priCe Of naTural gaS
The price of natural gas started rising in 1976 following deregulation of the market. Then, like
other fuels, the price rose sharply starting around 2000, but then plummeted during 2009 to 2011
as new sources of gas were discovered (namely “fracking”).
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 83
Summary
QueStionS for review
1.     Supply-demand analysis is a basic tool of microeco-
nomics. In competitive markets, supply and demand
curves tell us how much will be produced by firms
and how much will be demanded by consumers as a
function of price.
2.     The market mechanism is the tendency for supply and
demand to equili
ate (i.e., for price to move to the
market-clearing level), so that there is neither excess
demand nor excess supply. The equili
ium price is
the price that equates the quantity demanded with the
quantity supplied.
3.     Elasticities describe the responsiveness of supply and
demand to changes in price, income, or other vari-
ables. For example, the price elasticity of demand
measures the percentage change in the quantity de-
manded resulting from a 1-percent increase in price.
4.     Elasticities pertain to a time frame, and for most goods
it is important to distinguish between short-run and
long-run elasticities.
5.     We can use supply-demand diagrams to see how
shifts in the supply curve and/or demand curve can
explain changes in the market price and quantity.
1.     Suppose that unusually hot weather causes the demand
curve for ice cream to shift to the right. Why will the
price of ice cream rise to a new market-clearing level?
2.     Use supply and demand curves to illustrate how each
of the following events would affect the price of butter
and the quantity of butter bought and sold: (a) an in-
crease in the price of margarine; (b) an increase in the
price of milk; (c) a decrease in average income levels.
3.     If a 3-percent increase in the price of corn flakes causes
a 6-percent decline in the quantity demanded, what is
the elasticity of demand?
4.     Explain the difference between a shift in the supply
curve and a movement along the supply curve.
5.     Explain why for many goods, the long-run price elas-
ticity of supply is larger than the short-run elasticity.
6.     Why do long-run elasticities of demand differ from
short-run elasticities? Consider two goods: paper towels
and televisions. Which is a durable good? Would you
expect the price elasticity of demand for paper towels to
e larger in the short run or in the long run? Why? What
about the price elasticity of demand for televisions?
7.     Are the following statements true or false? Explain
your answers.
a.     The elasticity of demand is the same as the slope
of the demand curve.
.     The cross-price elasticity will always be positive.
c.     The supply of apartments is more inelastic in the
short run than the long run.
6.     If we can estimate, at least roughly, the supply and
demand curves for a particular market, we can calcu-
late the market-clearing price by equating the quan-
tity supplied with the quantity demanded. Also, if
we know how supply and demand depend on other
economic variables, such as income or the prices of
other goods, we can calculate how the market-clearing
price and quantity will change as these other variables
change. This is a means of explaining or predicting
market behavior.
7.     Simple numerical analyses can often be done by fit-
ting linear supply and demand curves to data on price
and quantity and to estimates of elasticities. For many
markets, such data and estimates are available, and
simple “back of the envelope” calculations can help
us understand the characteristics and behavior of the
market.
8.     When a government imposes price controls, it keeps
the price below the level that equates supply and de-
mand. A shortage develops; the quantity demanded
exceeds the quantity supplied.
8.     Suppose the government regulates the prices of beef
and chicken and sets them below their market-clear-
ing levels. Explain why shortages of these goods will
develop and what factors will determine the sizes of
the shortages. What will happen to the price of pork?
Explain
iefly.
9.     The city council of a small college town decides to reg-
ulate rents in order to reduce student living expenses.
Suppose the average annual market-clearing rent for
a two-bedroom apartment had been $700 per month
and that rents were expected to increase to $900 within
a year. The city council limits rents to their cu
ent
$700-per-month level.
a.     Draw a supply and demand graph to illustrate
what will happen to the rental price of an apart-
ment after the imposition of rent controls.
.     Do you think this policy will benefit all students?
Why or why not?
10.     In a discussion of tuition rates, a university official
argues that the demand for admission is completely
price inelastic. As evidence, she notes that while the
university has doubled its tuition (in real terms) over
the past 15 years, neither the number nor quality
of students applying has decreased. Would you ac-
cept this argument? Explain
iefly. (Hint: The official
makes an assertion about the demand for admission,
ut does she actually observe a demand curve? What
else could be going on?)
84 PART 1 Introduction: Markets and Prices
11.     Suppose the demand curve for a product is given by
Q = 10 - 2P + PS
where P is the price of the product and PS is the price
of a substitute good. The price of the substitute good
is $2.00.
a.     Suppose P = $1.00. What is the price elasticity
of demand? What is the cross-price elasticity of
demand?
.     Suppose the price of the good, P, goes to $2.00.
Now what is the price elasticity of demand? What
is the cross-price elasticity of demand?
12.     Suppose that rather than the declining demand as-
sumed in Example 2.8, a decrease in the cost of copper
production causes the supply curve to shift to the right
y 40 percent. How will the price of copper change?
13.     Suppose the demand for natural gas is perfectly in-
elastic. What would be the effect, if any, of natural gas
price controls?
exerciSeS
1.     Suppose the demand curve for a product is given by
Q = 300 - 2P + 4I, where I is average income mea-
sured in thousands of dollars. The supply curve is
Q = 3P - 50.
a.     If I = 25, find the market-clearing price and quan-
tity for the product.
.     If I = 50, find the market-clearing price and quan-
tity for the product.
c.     Draw a graph to illustrate your answers.
2.     Consider a competitive market for which the quanti-
ties demanded and supplied (per year) at various
prices are given as follows:
PricE
(Dollars)
DEmanD
(millions)
suPPly
(millions)
60 22 14
80 20 16
100 18 18
120 16 20
a.     Calculate the price elasticity of demand when the
price is $80 and when the price is $100.
.     Calculate the price elasticity of supply when the
price is $80 and when the price is $100.
c.     What are the equili
ium price and quantity?
d.     Suppose the government sets a price ceiling of $80.
Will there be a shortage, and if so, how large will
it be?
3.     Refer to Example 2.5 (page 59) on the market for
wheat. In 1998, the total demand for U.S. wheat was
Q = 3244 - 283P and the domestic supply was
QS = 1944 + 207P. At the end of 1998, both Brazil
and Indonesia opened their wheat markets to U.S.
farmers. Suppose that these new markets add 200 mil-
lion bushels to U.S. wheat demand. What will be the
free-market price of wheat and what quantity will be
produced and sold by U.S. farmers?
4.     A vegetable fiber is traded in a competitive world
market, and the world price is $9 per pound.
Unlimited quantities are available for import into the
United States at this price. The U.S. domestic supply
and demand for various price levels are shown as
follows:
PricE
u.s. suPPly
(million lBs)
u.s. DEmanD
(million lBs)
3 2 34
6 4 28
9 6 22
12 8 16
15 10 10
18 12 4
a.     What is the equation for demand? What is the
equation for supply?
.     At a price of $9, what is the price elasticity of de-
mand? What is it at a price of $12?
c.     What is the price elasticity of supply at $9? At $12?
d.     In a free market, what will be the U.S. price and
level of fiber imports?
    *5.     Much of the demand for U.S. agricultural output has
come from other countries. In 1998, the total demand
for wheat was Q = 3244 - 283P. Of this, total domes-
tic demand was QD = 1700 - 107P, and domestic
supply was QS = 1944 + 207P. Suppose the export
demand for wheat falls by 40 percent.
a.     U.S. farmers are concerned about this drop in ex-
port demand. What happens to the free-market
price of wheat in the United States? Do farmers
have much reason to wo
y?
.     Now suppose the U.S. government wants to buy
enough wheat to raise the price to $3.50 per bushel.
With the drop in export demand, how much wheat
ChAPTER 2 ThE BASICS Of SUPPLY AND DEMAND 85
would the government have to buy? How much
would this cost the government?
6.     The rent control agency of New York City has found
that aggregate demand is QD = 160 - 8P. Quantity
is measured in tens of thousands of apartments. Price,
the average monthly rental rate, is measured in hun-
dreds of dollars. The agency also noted that the in-
crease in Q at lower P results from more three-person
families coming into the city from Long Island and
demanding apartments. The city’s board of realtors
acknowledges that this is a good demand estimate and
has shown that supply is QS = 70 + 7P.
a.     If both the agency and the board are right about
demand and supply, what is the free-market price?
What is the change in city population if the agency
sets a maximum average monthly rent of $300 and
all those who cannot find an apartment leave the
city?
.     Suppose the agency bows to the wishes of the
oard and sets a rental of $900 per month on all
apartments to allow landlords a “fair” rate of re-
turn. If 50 percent of any long-run increases in
apartment offerings comes from new construction,
how many apartments are constructed?
7.     In 2010, Americans smoked 315 billion cigarettes, or
15.75 billion packs of cigarettes. The average retail price
(including taxes) was about $5.00 per pack. Statistical
studies have shown that the price elasticity of demand
is -0.4, and the price elasticity of supply is 0.5.
a.     Using this information, derive linear demand and
supply curves for the cigarette market.
.     In 1998, Americans smoked 23.5 billion packs of
cigarettes, and the retail price was about $2.00 per
pack. The decline in cigarette consumption from
1998 to 2010 was due in part to greater public
awareness of the health hazards from smoking,
ut was also due in part to the increase in price.
Suppose that the entire decline was due to the in-
crease in price. What could you deduce from that
about the price elasticity of demand?
8.     In Example 2.8 we examined the effect of a 20-percent
decline in copper demand on the price of copper, us-
ing the linear supply and demand curves developed
in Section 2.6. Suppose the long-run price elasticity of
copper demand were -0.75 instead of -0.5.
a.     Assuming, as before, that the equili
ium price and
quantity are P∗ = $3 per pound and Q∗ = 18 mil-
lion metric tons per year, derive the linear demand
curve consistent with the smaller elasticity.
.     Using this demand curve, recalculate the effect of a
55-percent decline in copper demand on the price
of copper.
9.     In Example 2.8 (page 74), we discussed the recent
decline in world demand for copper, due in part to
China’s decreasing consumption. What would hap-
pen, however, if China’s demand were increasing?
a.     Using the original elasticities of demand and sup-
ply (i.e., ES = 1.5 and ED = -0.5), calculate the
effect of a 20-percent increase in copper demand on
the price of copper.
.     Now calculate the effect of this increase in demand
on the equili
ium quantity, Q*.
c.     As we discussed in Example 2.8, the U.S. produc-
tion of copper declined between 2000 and 2003.
Calculate the effect on the equili
ium price and
quantity of both a 20-percent increase in copper de-
mand (as you just did in part a) and of a 20-percent
decline in copper supply.
10.     Example 2.9 (page 76) analyzes the world oil market.
Using the data given in that example:
a.     Show that the short-run demand and competitive
supply curves are indeed given by
D = 36.75 - 0.035P
SC = 21.85 + 0.023P
.     Show that the long-run demand and competitive
supply curves are indeed given by
D = 45.5 - 0.210P
SC = 16.1 + 0.138P
c.     In Example 2.9 we examined the impact on price
of a disruption of oil from Saudi Arabia. Suppose
that instead of a decline in supply, OPEC pro-
duction increases by 2 billion ba
els per year
(
yr) because the Saudis open large new oil
fields. Calculate the effect of this increase in pro-
duction on the price of oil in both the short run
and the long run.
11.     Refer to Example 2.10 (page 81), which analyzes the
effects of price controls on natural gas.
a.     Using the data in the example, show that the fol-
lowing supply and demand curves describe the
market for natural gas in 2005–2007:
Supply: Q = 15.90 + 0.72PG + 0.05PO
Demand: Q = 0.02 - 1.8PG + 0.69PO
Also, verify that if the price of oil is $50, these
curves imply a free-market price of $6.40 for natu-
al gas.
.     Suppose the regulated price of gas were $4.50 per
thousand cubic feet instead of $3.00. How much
excess demand would there have been?
86 PART 1 Introduction: Markets and Prices
c.     Suppose that the market for natural gas remained
unregulated. If the price of oil had increased from
$50 to $100, what would have happened to the free-
market price of natural gas?
    12.     The table below shows the retail price and sales for
instant coffee and roasted coffee for two years.
a.     Using these data alone, estimate the short-run price
elasticity of demand for roasted coffee. Derive a
linear demand curve for roasted coffee.
.     Now estimate the short-run price elasticity of de-
mand for instant coffee. Derive a linear demand
curve for instant coffee.
*
c.     Which coffee has the higher short-run price elas-
ticity of demand? Why do you think this is the
case?
yEa
Etail
PricE of
instant
coffEE
($/lB)
salEs of
instant
coffEE
(million
lBs)
Etail
PricE of
oastED
coffEE
($/lB)
salEs of
oastED
coffEE
(million
lBs)
year 1 10.35 75 4.11 820
year 2 10.48 70 3.76 850
Part 2 presents the theoretical core of microeconomics.
Chapters 3 and 4 explain the principles underlying consumer de-
mand. We see how consumers make consumption decisions, how
their preferences and budget constraints determine their demands
for various goods, and why different goods have different demand
characteristics. Chapter 5 contains more advanced material that
shows how to analyze consumer choice under uncertainty. We
explain why people usually dislike risky situations and show how
they can reduce risk and choose among risky alternatives.
Chapters 6 and 7 develop the theory of the firm. We see how
firms combine inputs, such as capital, labor, and raw materi-
als, to produce goods and services in a way that minimizes the
costs of production. We also see how a firm’s costs depend on
its rate of production and production experience. Chapter 8 then
shows how firms choose profit-maximizing rates of production.
We also see how the production decisions of individual firms
combine to determine the competitive market supply curve and its
characteristics.
Chapter 9 applies supply and demand curves to the analysis
of competitive markets. We show how government policies, such
as price controls, quotas, taxes, and subsidies, can have wide-
anging effects on consumers and producers, and we explain how
supply-demand analysis can be used to evaluate these effects.
3 Consumer Behavior
89
4 Individual and
Market Demand
131
5 Uncertainty and
Consumer Behavior
179
6 Production
209
7 The Cost of Production
237
8 Profit Maximization and
Competitive Supply
289
9 The Analysis of Competitive
Markets
327
ChAPTerS
Producers, Consumers,
and Competitive
Markets
PArT TWo
This page intentionally left blank
89
Some time ago, General Mills introduced a new
eakfast cereal. The new
and, Apple-Cinnamon Cheerios, was a sweetened and more flavorful variant on General Mills’ classic Cheerios product.
But before Apple-Cinnamon Cheerios could be extensively marketed,
the company had to resolve an important problem: How high a price
should it charge? No matter how good the cereal was, its profitability
would depend on the company’s pricing decision. Knowing that
consumers would pay more for a new product was not enough. The
question was how much more. General Mills, therefore, had to conduct a
careful analysis of consumer preferences to determine the demand for
Apple-Cinnamon Cheerios.
General Mills’ problem in determining consumer preferences mir-
ors the more complex problem faced by the U.S. Congress in evaluat-
ing the federal Food Stamps program. The goal of the program is to
give low-income households coupons that can be exchanged for food.
But there has always been a problem in the program’s design that
complicates its assessment: To what extent do food stamps provide
people with more food, as opposed to simply subsidizing the purchase
of food that they would have bought anyway? In other words, has the
program turned out to be little more than an income supplement that
people spend largely on nonfood items instead of a solution to the
nutritional problems of the poor? As in the cereal example, we need
an analysis of consumer behavior. In this case, the federal government
must determine how spending on food, as opposed to spending on
other goods, is affected by changing income levels and prices.
Solving these two problems—one involving corporate policy and
the other public policy—requires an understanding of the theory
of consumer behavior: the explanation of how consumers allocate
incomes to the purchase of different goods and services.
Consumer Behavio
How can a consumer with a limited income decide which goods and
services to buy? This is a fundamental issue in microeconomics—one
that we address in this chapter and the next. We will see how consum-
ers allocate their incomes across goods and explain how these alloca-
tion decisions determine the demands for various goods and services.
In turn, understanding consumer purchasing decisions will help us to
understand how changes in income and prices affect the demand for
3.1 Designing New
Automobiles (I) 99
3.2 Can Money Buy
happiness? 103
3.3 Designing New
Automobiles (II) 110
3.4 The Authors Argue about
health Care 112
3.5 A College Trust Fund 114
3.6 revealed Preference for
ecreation 116
3.7 Marginal Utility and
happiness 119
LIST oF exAMPLeS
3.1 Consumer Preferences 91
3.2 Budget Constraints 104
3.3 Consumer Choice 108
3.4 revealed Preference 112
3.5 Marginal Utility
and Consumer Choice 117
3.6 Cost-of-Living Indexes 122*
ChAPTer oUTLINe
Consumer Behavio
ChAPTer 3
90 PArT 2 Producers, Consumers, and Competitive Markets
goods and services and why the demand for some products is more sensitive
than others to changes in prices and income.
Consumer behavior is best understood in three distinct steps:
1. Consumer Preferences: The first step is to find a practical way to describe
the reasons people might prefer one good to another. We will see how a
consumer’s preferences for various goods can be described graphically and
alge
aically.
2. Budget Constraints: Of course, consumers also consider prices. In Step
2, therefore, we take into account the fact that consumers have limited
incomes which restrict the quantities of goods they can buy. What does
a consumer do in this situation? We find the answer to this question by
putting consumer preferences and budget constraints together in the
third step.
3. Consumer Choices: Given their preferences and limited incomes,
consumers choose to buy combinations of goods that maximize their
satisfaction. These combinations will depend on the prices of various
goods. Thus, understanding consumer choice will help us understand
demand—i.e., how the quantity of a good that consumers choose to pur-
chase depends on its price.
These three steps are the basics of consumer theory, and we will go through
them in detail in the first three sections of this chapter. Afterward, we will ex-
plore a number of other interesting aspects of consumer behavior. For example,
we will see how one can determine the nature of consumer preferences from ac-
tual observations of consumer behavior. Thus, if a consumer chooses one good
over a similarly priced alternative, we can infer that he or she prefers the first
good. Similar kinds of conclusions can be drawn from the actual decisions that
consumers make in response to changes in the prices of the various goods and
services that are available for purchase.
At the end of this chapter, we will return to the discussion of real and nomi-
nal prices that we began in Chapter 1. We saw that the Consumer Price Index
can provide one measure of how the well-being of consumers changes over
time. In this chapter, we delve more deeply into the subject of purchasing
power by describing a range of indexes that measure changes in purchasing
power over time. Because they affect the benefits and costs of numerous social-
welfare programs, these indexes are significant tools in setting government
policy in the United States.
What do Consumers do? Before proceeding, we need to be clear
about our assumptions regarding consumer behavior, and whether those as-
sumptions are realistic. It is hard to argue with the proposition that consum-
ers have preferences among the various goods and services available to them,
and that they face budget constraints which put limits on what they can buy.
But we might take issue with the proposition that consumers decide which
combinations of goods and services to buy so as to maximize their satisfac-
tion. Are consumers as rational and informed as economists often make them
out to be?
We know that consumers do not always make purchasing decisions ra-
tionally. Sometimes, for example, they buy on impulse, ignoring or not fully
accounting for their budget constraints (and going into debt as a result).
Sometimes consumers are unsure about their preferences or are swayed by
the consumption decisions of friends and neighbors, or even by changes in
theory of consumer
ehavior Description of how
consumers allocate incomes
among different goods and
services to maximize their well-
eing.
ChAPTer 3 CoNSUMer BehAvIor 91
mood. And even if consumers do behave rationally, it may not always be
feasible for them to account fully for the multitude of prices and choices that
they face daily.
Economists have recently been developing models of consumer behavior
that incorporate more realistic assumptions about rationality and decision mak-
ing. This area of research, called behavioral economics, has drawn heavily from
findings in psychology and related fields. We will discuss some key results from
ehavioral economics in Chapter 5. At this point we simply want to make it
clear that our basic model of consumer behavior necessarily makes some sim-
plifying assumptions. But we also want to emphasize that this model has been
extremely successful in explaining much of what we actually observe regarding
consumer choice and the characteristics of consumer demand. As a result, this
model is a basic “workhorse” of economics. It is used widely, not only in eco-
nomics, but also in related fields such as finance and marketing.
3.1 Consumer Preferences
Given both the vast number of goods and services that our industrial economy
provides for purchase and the diversity of personal tastes, how can we describe
consumer preferences in a coherent way? Let’s begin by thinking about how a
consumer might compare different groups of items available for purchase. Will
one group of items be prefe
ed to another group, or will the consumer be indif-
ferent between the two groups?
Market Baskets
We use the term market basket to refer to such a group of items. Specifically, a
market basket is a list with specific quantities of one or more goods. A market
asket might contain the various food items in a grocery cart. It might also re-
fer to the quantities of food, clothing, and housing that a consumer buys each
month. Many economists also use the word bundle to mean the same thing as
market basket.
How do consumers select market baskets? How do they decide, for example,
how much food versus clothing to buy each month? Although selections may
occasionally be a
itrary, as we will soon see, consumers usually select market
askets that make them as well off as possible.
Table 3.1 shows several market baskets consisting of various amounts of
food and clothing purchased on a monthly basis. The number of food items
can be measured in any number of ways: by total number of containers, by
number of packages of each item (e.g., milk, meat, etc.), or by number of
pounds or grams. Likewise, clothing can be counted as total number of pieces,
as number of pieces of each type of clothing, or as total weight or volume.
Because the method of measurement is largely a
itrary, we will simply de-
scribe the items in a market basket in terms of the total number of units of
each commodity. Market basket A, for example, consists of 20 units of food
and 30 units of clothing, basket B consists of 10 units of food and 50 units of
clothing, and so on.
To explain the theory of consumer behavior, we will ask whether consumers
prefer one market basket to another. Note that the theory assumes that consum-
ers’ preferences are consistent and make sense. We explain what we mean by
these assumptions in the next subsection.
market basket (or bundle) List
with specific quantities of one or
more goods.
92 PArT 2 Producers, Consumers, and Competitive Markets
Some Basic Assumptions about Preferences
The theory of consumer behavior begins with three basic assumptions about
people’s preferences for one market basket versus another. We believe that
these assumptions hold for most people in most situations.
1. Completeness: Preferences are assumed to be complete. In other words,
consumers can compare and rank all possible baskets. Thus, for any two
market baskets A and B, a consumer will prefer A to B, will prefer B to A,
or will be indifferent between the two. By indifferent we mean that a per-
son will be equally satisfied with either basket. Note that these preferences
ignore costs. A consumer might prefer steak to hamburger but buy ham-
urger because it is cheaper.
2. Transitivity: Preferences are transitive. Transitivity means that if a con-
sumer prefers basket A to basket B and basket B to basket C, then the
consumer also prefers A to C. For example, if a Porsche is prefe
ed to a
Cadillac and a Cadillac to a Chevrolet, then a Porsche is also prefe
ed to
a Chevrolet. Transitivity is normally regarded as necessary for consumer
consistency.
3. More is better than less: Goods are assumed to be desirable—i.e., to
e good. Consequently, consumers always prefer more of any good to less. In
addition, consumers are never satisfied or satiated; more is always better,
even if just a little better.1 This assumption is made for pedagogic reasons;
namely, it simplifies the graphical analysis. Of course, some goods, such
as air pollution, may be undesirable, and consumers will always prefer
less. We ignore these “bads” in the context of our immediate discussion of
consumer choice because most consumers would not choose to purchase
them. We will, however, discuss them later in the chapter.
These three assumptions form the basis of consumer theory. They do not ex-
plain consumer preferences, but they do impose a degree of rationality and
easonableness on them. Building on these assumptions, we will now explore
consumer behavior in greater detail.
Table 3.1 alternative market Baskets
1Thus some economists use the term nonsatiation to refer to this third assumption.
Note: We will avoid the use of the letters C and F to represent market baskets, whenever market baskets
might be confused with the number of units of food and clothing.
Market Basket UNits oF Food UNits oF ClothiNg
A 20 30
B 10 50
D 40 20
E 30 40
G 10 20
H 10 40
ChAPTer 3 CoNSUMer BehAvIor 93
Indifference Curves
We can show a consumer’s preferences graphically with the use of indifference
curves. An indifference curve represents all combinations of market baskets that pro-
vide a consumer with the same level of satisfaction. That person is therefore indiffer-
ent among the market baskets represented by the points graphed on the curve.
Given our three assumptions about preferences, we know that a consumer
can always indicate either a preference for one market basket over another
or indifference between the two. We can then use this information to rank all
possible consumption choices. In order to appreciate this principle in graphic
form, let’s assume that there are only two goods available for consumption:
food F and clothing C. In this case, all market baskets describe combinations
of food and clothing that a person might wish to consume. As we have already
seen, Table 3.1 provides some examples of baskets containing various amounts
of food and clothing.
In order to graph a consumer’s indifference curve, it helps first to graph his
or her individual preferences. Figure 3.1 shows the same baskets listed in Table
3.1. The horizontal axis measures the number of units of food purchased each
week; the vertical axis measures the number of units of clothing. Market basket
A, with 20 units of food and 30 units of clothing, is prefe
ed to basket G be-
cause A contains more food and more clothing (recall our third assumption that
more is better than less). Similarly, market basket E, which contains even more
food and even more clothing, is prefe
ed to A. In fact, we can easily compare
all market baskets in the two shaded areas (such as E and G) to A because they
contain either more or less of both food and clothing. Note, however, that B
contains more clothing but less food than A. Similarly, D contains more food but
less clothing than A. Therefore, comparisons of market basket A with baskets
B, D, and H are not possible without more information about the consumer’s
anking.
This additional information is provided in Figure 3.2, which shows an indif-
ference curve, labeled U1, that passes through points A, B, and D. This curve
indicates that the consumer is indifferent among these three market baskets. It
indifference curve Curve
epresenting all combinations
of market baskets that provide a
consumer with the same level of
satisfaction.
10
10 20 30 40
Clothing
(units per week)
Food
(units per week)
B
A
•
H•
D•
E•
G•
•
50
40
30
20
FIgure 3.1
desCriBing individual
PreferenCes
Because more of each good is prefe
ed to
less, we can compare market baskets in the
shaded areas. Basket A is clearly prefe
ed
to basket G, while E is clearly prefe
ed to A.
However, A cannot be compared with B, D, or
H without additional information.
94 PArT 2 Producers, Consumers, and Competitive Markets
tells us that in moving from market basket A to market basket B, the consumer
feels neither better nor worse off in giving up 10 units of food to obtain 20 ad-
ditional units of clothing. Likewise, the consumer is indifferent between points
A and D: He or she will give up 10 units of clothing to obtain 20 more units of
food. On the other hand, the consumer prefers A to H, which lies below U1.
Note that the indifference curve in Figure 3.2 slopes downward from left to
ight. To understand why this must be the case, suppose instead that it sloped
upward from A to E. This would violate the assumption that more of any com-
modity is prefe
ed to less. Because market basket E has more of both food and
clothing than market basket A, it must be prefe
ed to A and therefore cannot be
on the same indifference curve as A. In fact, any market basket lying above and to
the right of indifference curve U1 in Figure 3.2 is prefe
ed to any market basket
on U1.
Indifference Maps
To describe a person’s preferences for all combinations of food and clothing,
we can graph a set of indifference curves called an indifference map. Each
indifference curve in the map shows the market baskets among which the
person is indifferent. Figure 3.3 shows three indifference curves that form part
of an indifference map (the entire map includes an infinite number of such
curves). Indifference curve U3 generates the highest level of satisfaction, fol-
lowed by indifference curves U2 and U1.
Indifference curves cannot intersect. To see why, we will assume the contrary
and see how the resulting graph violates our assumptions about consumer be-
havior. Figure 3.4 shows two indifference curves, U1 and U2, that intersect at A.
indifference map Graph
containing a set of indifference
curves showing the market
askets among which a consumer
is indifferent.
Clothing
(units per week)
U1
Food
(units per week)
B
H
G
A
D
E
10 20 30 40
50
40
30
20
10
FIgure 3.2
an indifferenCe
Curve
The indifference curve U1 that passes
through market basket A shows all
askets that give the consumer the
same level of satisfaction as does mar-
ket basket A; these include baskets B
and D. Our consumer prefers basket
E, which lies above U1, to A, but pre-
fers A to H or G, which lie below U1.
ChAPTer 3 CoNSUMer BehAvIor 95
Because A and B are both on indifference curve U1, the consumer must be
indifferent between these two market baskets. Because both A and D lie on
indifference curve U2, the consumer is also indifferent between these market
askets. Consequently, using the assumption of transitivity, the consumer is also
indifferent between B and D. But this conclusion can’t be true: Market basket
B must be prefe
ed to D because it contains more of both food and clothing.
Thus, intersecting indifference curves contradicts our assumption that more is
prefe
ed to less.
Of course, there are an infinite number of nonintersecting indifference
curves, one for every possible level of satisfaction. In fact, every possible market
asket (each co
esponding to a point on the graph) has an indifference curve
passing through it.
The Shape of Indifference Curves
Recall that indifference curves are all downward sloping. In our example of
food and clothing, when the amount of food increases along an indifference
curve, the amount of clothing decreases. The fact that indifference curves slope
downward follows directly from our assumption that more of a good is better
than less. If an indifference curve sloped upward, a consumer would be indif-
ferent between two market baskets even though one of them had more of both
food and clothing.
U1
U2
Clothing
(units pe
week)
Food
(units per week)
A
D
B
FIgure 3.4
indifferenCe Curves Cannot
interseCt
If indifference curves U1 and U2 intersect, one of the
assumptions of consumer theory is violated. According to
this diagram, the consumer should be indifferent among
market baskets A, B, and D. Yet B should be prefe
ed to
D because B has more of both goods.
U1
U2
U3
Clothing
(units pe
week)
Food
(units per week)
A
B
D
FIgure 3.3
an indifferenCe maP
An indifference map is a set of indifference curves that
describes a person’s preferences. Any market basket on
indifference curve U3, such as basket A, is prefe
ed to
any basket on curve U2 (e.g., basket B), which in turn is
prefe
ed to any basket on U1, such as D.
96 PArT 2 Producers, Consumers, and Competitive Markets
As we saw in Chapter 1, people face trade-offs. The shape of an indifference
curve describes how a consumer is willing to substitute one good for another.
Look, for example, at the indifference curve in Figure 3.5. Starting at market
asket A and moving to basket B, we see that the consumer is willing to give up
6 units of clothing to obtain 1 extra unit of food. However, in moving from B to
D, he is willing to give up only 4 units of clothing to obtain an additional unit of
food; in moving from D to E, he will give up only 2 units of clothing for 1 unit
of food. The more clothing and the less food a person consumes, the more cloth-
ing he will give up in order to obtain more food. Similarly, the more food that a
person possesses, the less clothing he will give up for more food.
The Marginal rate of Substitution
To quantify the amount of one good that a consumer will give up to obtain more
of another, we use a measure called the marginal rate of substitution (MRS).
The MRS of food F for clothing C is the maximum amount of clothing that a person is
willing to give up to obtain one additional unit of food. Suppose, for example, the
MRS is 3. This means that the consumer will give up 3 units of clothing to ob-
tain 1 additional unit of food. If the MRS is 1/2, the consumer is willing to give
up only 1/2 unit of clothing. Thus, the MRS measures the value that the individual
places on 1 extra unit of a good in terms of another.
Look again at Figure 3.5. Note that clothing appears on the vertical axis and
food on the horizontal axis. When we describe the MRS, we must be clear about
which good we are giving up and which we are getting more of. To be consis-
tent throughout the book, we will define the MRS in terms of the amount of the
good on the vertical axis that the consumer is willing to give up in order to obtain 1
extra unit of the good on the horizontal axis. Thus, in Figure 3.5 the MRS refers to
marginal rate of
substitution (MRS)
Maximum amount of a good that
a consumer is willing to give up
in order to obtain one additional
unit of another good.
Clothing
(units pe
week)
Food
(units per week)
1 2 3 4 5
16
14
12
10
8
6
4
2
A
B
D
E
G
–6
–4
–2
–1
1
1
1
1
FIgure 3.5
the marginal rate
of suBstitution
The magnitude of the slope of an indiffer-
ence curve measures the consumer’s mar-
ginal rate of substitution (MRS) between
two goods. In this figure, the MRS between
clothing (C) and food (F ) falls from 6 (be-
tween A and B) to 4 (between B and D) to 2
(between D and E) to 1 (between E and G).
When the MRS diminishes along an indiffer-
ence curve, the curve is convex.
ChAPTer 3 CoNSUMer BehAvIor 97
the amount of clothing that the consumer is willing to give up to obtain an ad-
ditional unit of food. If we denote the change in clothing by ∆C and the change
in food by ∆F, the MRS can be written as - ∆C/∆F. We add the negative sign to
make the marginal rate of substitution a positive number. (Remember that ∆C
is always negative; the consumer gives up clothing to obtain additional food.)
Thus the MRS at any point is equal in magnitude to the slope of the indiffer-
ence curve. In Figure 3.5, for example, the MRS between points A and B is 6: The
consumer is willing to give up 6 units of clothing to obtain 1 additional unit of
food. Between points B and D, however, the MRS is 4: With these quantities of
food and clothing, the consumer is willing to give up only 4 units of clothing to
obtain 1 additional unit of food.
Convexity Also observe in Figure 3.5 that the MRS falls as we move down
the indifference curve. This is not a coincidence. This decline in the MRS reflects
an important characteristic of consumer preferences. To understand this, we
will add an additional assumption regarding consumer preferences to the three
that we discussed earlier in this chapter (see page 92):
4. Diminishing marginal rate of substitution: Indifference curves are usu-
ally convex, or bowed inward. The term convex means that the slope of the
indifference curve increases (i.e., becomes less negative) as we move down
along the curve. In other words, an indifference curve is convex if the MRS
diminishes along the curve. The indifference curve in Figure 3.5 is convex.
As we have seen, starting with market basket A in Figure 3.5 and moving
to basket B, the MRS of food F for clothing C is - ∆C/∆F = -(-6)/1 = 6.
However, when we start at basket B and move from B to D, the MRS
falls to 4. If we start at basket D and move to E, the MRS is 2. Starting at
E and moving to G, we get an MRS of 1. As food consumption increases,
the slope of the indifference curve falls in magnitude. Thus the MRS
also falls.2
Is it reasonable to expect indifference curves to be convex? Yes. As more and
more of one good is consumed, we can expect that a consumer will prefer to
give up fewer and fewer units of a second good to get additional units of the
first one. As we move down the indifference curve in Figure 3.5 and consump-
tion of food increases, the additional satisfaction that a consumer gets from still
more food will diminish. Thus, he will give up less and less clothing to obtain
additional food.
Another way of describing this principle is to say that consumers generally
prefer balanced market baskets to market baskets that contain all of one good
and none of another. Note from Figure 3.5 that a relatively balanced market
asket containing 3 units of food and 6 units of clothing (basket D) generates
as much satisfaction as another market basket containing 1 unit of food and 16
units of clothing (basket A). It follows that a balanced market basket containing,
for example, 6 units of food and 8 units of clothing will generate a higher level
of satisfaction.
2With nonconvex preferences, the MRS increases as the amount of the good measured on the
horizontal axis increases along any indifference curve. This unlikely possibility might arise if one
or both goods are addictive. For example, the willingness to substitute an addictive drug for other
goods might increase as the use of the addictive drug increased.
98 PArT 2 Producers, Consumers, and Competitive Markets
Perfect Substitutes and Perfect Complements
The shape of an indifference curve describes the willingness of a consumer to
substitute one good for another. An indifference curve with a different shape
implies a different willingness to substitute. To see this principle, look at the
two somewhat extreme cases illustrated in Figure 3.6.
Figure 3.6 (a) shows Bob’s preferences for apple juice and orange juice. These
two goods are perfect substitutes for Bob because he is entirely indifferent be-
tween having a glass of one or the other. In this case, the MRS of apple juice for
orange juice is 1: Bob is always willing to trade 1 glass of one for 1 glass of the
other. In general, we say that two goods are perfect substitutes when the mar-
ginal rate of substitution of one for the other is a constant. Indifference curves
describing the trade-off between the consumption of the goods are straight
lines. The slope of the indifference curves need not be -1 in the case of perfect
substitutes. Suppose, for example, that Dan believes that one 16-megabyte
memory chip is equivalent to two 8-megabyte chips because both combinations
have the same memory capacity. In that case, the slope of Dan’s indifference
curve will be -2 (with the number of 8-megabyte chips on the vertical axis).
Figure 3.6 (b) illustrates Jane’s preferences for left shoes and right shoes.
For Jane, the two goods are perfect complements because a left shoe will not
increase her satisfaction unless she can obtain the matching right shoe. In this
case, the MRS of left shoes for right shoes is zero whenever there are more right
shoes than left shoes; Jane will not give up any left shoes to get additional right
shoes. Co
espondingly, the MRS is infinite whenever there are more left shoes
than right because Jane will give up all but one of her excess left shoes in order
to obtain an additional right shoe. Two goods are perfect complements when
the indifference curves for both are shaped as right angles.
in §2.1, we explain that two
goods are substitutes when
an increase in the price of
one leads to an increase in
the quantity demanded of
the other.
perfect substitutes Two
goods for which the marginal rate
of substitution of one for the other
is a constant.
perfect complements Two
goods for which the MrS is zero
or infinite; the indifference curves
are shaped as right angles.
in §2.1 we explain that
goods are complements
when an increase in the
price of one leads to a
decrease in the quantity
demanded of the other.
Apple
juice
(glasses)
Orange juice (glasses)
4
3
2
1
0
1 2 3 4
(a) Perfect Substitutes
Left
shoes
Right shoes
4
3
2
1
0
1 2 3 4
(b) Perfect Complements
FIgure 3.6
PerfeCt suBstitutes and PerfeCt ComPlements
In (a), Bob views orange juice and apple juice as perfect substitutes: He is always indifferent
etween a glass of one and a glass of the other. In (b), Jane views left shoes and right shoes
as perfect complements: An additional left shoe gives her no extra satisfaction unless she
also obtains the matching right shoe.
ChAPTer 3 CoNSUMer BehAvIor 99
Bads So far, all of our examples have involved products that are “goods”—i.e.,
cases in which more of a product is prefe
ed to less. However, some things are
ads: Less of them is prefe
ed to more. Air pollution is a bad; asbestos in housing
insulation is another. How do we account for bads in the analysis of consumer
preferences?
The answer is simple: We redefine the product under study so that consumer
tastes are represented as a preference for less of the bad. This reversal turns the
ad into a good. Thus, for example, instead of a preference for air pollution, we
will discuss the preference for clean air, which we can measure as the degree of
eduction in air pollution. Likewise, instead of refe
ing to asbestos as a bad, we
will refer to the co
esponding good, the removal of asbestos.
With this simple adaptation, all four of the basic assumptions of consumer
theory continue to hold, and we are ready to move on to an analysis of con-
sumer budget constraints.
ad Good for which less is
prefe
ed rather than more.
exAMPle 3.1 designing neW automoBiles (i)
Suppose you worked for the Ford
Motor Company and had to help plan
new models to introduce. Should the
new models emphasize interior space
or handling? horsepower or gas mile-
age? To decide, you would want to
know how people value the various
attributes of a car, such as power,
size, handling, gas mileage, interior
features, and so on. The more desirable the attributes,
the more people would be willing to pay for a car.
however, the better the attributes, the more the car
will cost to manufacture. A car with a more powerful
engine and more interior space, for example, will cost
more to produce than a car with a smaller engine and
less space. how should Ford trade off these different
attributes and decide which ones to emphasize?
The answer depends in part on the cost of produc-
tion, but it also depends on consumer preferences.
To find out how much people are willing to pay for
various attributes, economists and marketing experts
look at the prices that people actually do pay for a
wide range of models with a range of attributes. For
example, if the only difference between two cars is
interior space, and if the car with 2 additional cubic
feet sells for $1000 more than its smaller counter-
part, then interior space will be valued at $500 per
cubic foot. By evaluating car purchases over a range
of buyers and a range of models, one can estimate
the values associated with various
attributes, while accounting for
the fact that these valuations may
diminish as more and more of
each attribute is included in a car.
one way to obtain such infor-
mation is by conducting surveys
in which individuals are asked
about their preferences for vari-
ous automobiles with different combinations of at-
tributes. Another way is to statistically analyze past
consumer purchases of cars whose attributes varied.
one recent statistical study looked at a wide range
of Ford models with varying attributes.3 Figure 3.7 de-
scribes two sets of indifference curves, derived from
an analysis that varies two attributes: interior size
(measured in cubic feet) and acceleration (measured
in horsepower) for typical consumers of Ford auto-
mobiles. Figure 3.7 (a) describes the preferences of
typical owners of Ford Mustang coupes. Because they
tend to place greater value on acceleration than size,
Mustang owners have a high marginal rate of substitu-
tion for size versus acceleration; in other words, they
are willing to give up quite a bit of size to get better
acceleration. Compare these preferences to those
of Ford explorer owners, shown in Figure 3.7 (b).
They have a lower MrS and will consequently give
up a considerable amount of acceleration to get a car
with a roomier interior.
3Amil Petrin, “Quantifying the Benefits of New Products: The Case of the Minivan,” Journal of
Political Economy 110 (2002): 705–729. We wish to thank Amil Petrin for providing some of the em-
pirical information in this example.
100 PArT 2 Producers, Consumers, and Competitive Markets
utility You may have noticed a convenient feature of the theory of consumer
ehavior as we have described it so far: It has not been necessary to associate a
numerical level of satisfaction with each market basket consumed. For example, with
espect to the three indifference curves in Figure 3.3 (page 95), we know that
market basket A (or any other basket on indifference curve U3) gives more
satisfaction than any market basket on U2, such as B. Likewise, we know that
the market baskets on U2 are prefe
ed to those on U1. The indifference curves
simply allow us to describe consumer preferences graphically, building on the
assumption that consumers can rank alternatives.
We will see that consumer theory relies only on the assumption that
consumers can provide relative rankings of market baskets. Nonetheless,
it is often useful to assign numerical values to individual baskets. Using this
numerical approach, we can describe consumer preferences by assigning
scores to the levels of satisfaction associated with each indifference curve.
The concept is known as utility. In everyday language, the word utility has
ather
oad connotations, meaning, roughly, “benefit” or “well-being.”
Indeed, people obtain “utility” by getting things that give them pleasure
and by avoiding things that give them pain. In the language of economics,
the concept of utility refers to the numerical score representing the satisfaction
that a consumer gets from a market basket. In other words, utility is a device
used to simplify the ranking of market baskets. If buying three copies of
this textbook makes you happier than buying one shirt, then we say that the
three books give you more utility than the shirt.
utility Numerical score
epresenting the satisfaction that
a consumer gets from a given
market basket.
Space
(cubic feet)
(b)
Space
(cubic feet)
Acceleration
(horsepower)
(a)
50 100 150 200 250 Acceleration
(horsepower)
50 100 150 200 250
120
100
80
60
40
20
120
100
80
60
40
20
FIgure 3.7
PreferenCes for automoBile attriButes
Preferences for automobile attributes can be described by indifference curves. Each curve shows the combination
of acceleration and interior space that give the same satisfaction. Owners of Ford Mustang coupes (a) are willing to
give up considerable interior space for additional acceleration. The opposite is true for owners of Ford Explorers (b).
ChAPTer 3 CoNSUMer BehAvIor 101
utility funCtions A utility function is a formula that assigns a level of
utility to each market basket. Suppose, for example, that Phil’s utility func-
tion for food (F) and clothing (C) is u(F,C) = F + 2C. In that case, a market
asket consisting of 8 units of food and 3 units of clothing generates a util-
ity of 8 + (2)(3) = 14. Phil is therefore indifferent between this market bas-
ket and a market basket containing 6 units of food and 4 units of clothing
[6 + (2)(4) = 14]. On the other hand, either market basket is prefe
ed to a third
containing 4 units of food and 4 units of clothing. Why? Because this last market
asket has a utility level of only 4 + (4)(2) = 12.
We assign utility levels to market baskets so that if market basket A is pre-
fe
ed to basket B, the number will be higher for A than for B. For example,
market basket A on the highest of three indifference curves U3 might have a
utility level of 3, while market basket B on the second-highest indifference curve
U2 might have a utility level of 2; on the lowest indifference curve U1, basket D
has a utility level of 1. Thus the utility function provides the same information
about preferences that an indifference map does: Both order consumer choices
in terms of levels of satisfaction.
Let’s examine one particular utility function in some detail. The utility func-
tion u(F,C) = FC tells us that the level of satisfaction obtained from consuming F
units of food and C units of clothing is the product of F and C. Figure 3.8 shows
indifference curves associated with this function. The graph was drawn by ini-
tially choosing one particular market basket—say, F = 5 and C = 5 at point A.
This market basket generates a utility level U1 of 25. Then the indifference curve
(also called an isoutility curve) was drawn by finding all market baskets for which
FC = 25 (e.g., F = 10, C = 2.5 at point B; F = 2.5, C = 10 at point D). The sec-
ond indifference curve, U2, contains all market baskets for which FC = 50 and
the third, U3, all market baskets for which FC = 100.
It is important to note that the numbers attached to the indifference curves
are for convenience only. Suppose the utility function were changed to
u(F,C) = 4FC. Consider any market basket that previously generated a utility
level of 25—say, F = 5 and C = 5. Now the level of utility has increased, by
a factor of 4, to 100. Thus the indifference curve labeled 25 looks the same, al-
though it should now be labeled 100 rather than 25. In fact, the only difference
etween the indifference curves associated with the utility function 4FC and
utility function Formula
that assigns a level of utility to
individual market baskets.
U1 5 25
Clothing
(units pe
week)
Food
(units per week)
A
D
B
5
5 10 15
10
15
U3 5 100
U2 5 50
FIgure 3.8
utility funCtions and
indifferenCe Curves
A utility function can be represented by a set of indifference
curves, each with a numerical indicator. This figure shows
three indifference curves (with utility levels of 25, 50, and
100, respectively) associated with the utility function FC.
102 PArT 2 Producers, Consumers, and Competitive Markets
the utility function FC is that the curves are numbered 100, 200, and 400, rather
than 25, 50, and 100. It is important to stress that the utility function is simply
a way of ranking different market baskets; the magnitude of the utility difference
etween any two market baskets does not really tell us anything. The fact that
U3 has a level of utility of 100 and U2 has a level of 50 does not mean that mar-
ket baskets on U3 generate twice as much satisfaction as those on U2. This is so
ecause we have no means of objectively measuring a person’s satisfaction or
level of well-being from the consumption of a market basket. Thus whether we
use indifference curves or a measure of utility, we know only that U3 is better
than U2 and that U2 is better than U1. We do not, however, know by how much
one is prefe
ed to the other.
ordinal versus Cardinal utility The three indifference curves in
Figure 3.3 (page 95) provide a ranking of market baskets that is ordered, or ordi-
nal. For this reason, a utility function that generates a ranking of market baskets
is called an ordinal utility function. The ranking associated with the ordinal
utility function places market baskets in the order of most to least prefe
ed.
However, as explained above, it does not indicate by how much one is prefe
ed
to another. We know, for example, that any market basket on U3, such as A, is
prefe
ed to any on U2, such as B. However, the amount by which A is prefe
ed
to B (and B to D) is not revealed by the indifference map or by the ordinal utility
function that generates it.
When working with ordinal utility functions, we must be careful to avoid a
trap. Suppose that Juan’s ordinal utility function attaches a utility level of 5 to
a copy of this textbook; meanwhile Maria’s utility function attaches a level of
10. Will Maria be happier than Juan if each of them gets a copy of this book? We
don’t know. Because these numerical values are a
itrary, interpersonal com-
parisons of utility are impossible.
When economists first studied utility and utility functions, they hoped that
individual preferences could be quantified or measured in terms of basic units
and could therefore provide a ranking that allowed for interpersonal compari-
sons. Using this approach, we could say that Maria gets twice as much satisfac-
tion as Juan from a copy of this book. Or if we found that having a second copy
increased Juan’s utility level to 10, we could say that his happiness has doubled.
If the numerical values assigned to market baskets did have meaning in this
way, we would say that the numbers provided a cardinal ranking of alternatives.
A utility function that describes by how much one market basket is prefe
ed to
another is called a cardinal utility function. Unlike ordinal utility functions, a
cardinal utility function attaches to market baskets numerical values that can-
not a
itrarily be doubled or tripled without altering the differences between
the values of various market baskets.
Unfortunately, we have no way of telling whether a person gets twice
as much satisfaction from one market basket as from another. Nor do we
know whether one person gets twice as much satisfaction as another from
consuming the same basket. (Could you tell whether you get twice as much
satisfaction from consuming one thing versus another?) Fortunately, this
constraint is unimportant. Because our objective is to understand consumer
ehavior, all that matters is knowing how consumers rank different baskets.
Therefore, we will work only with ordinal utility functions. This approach
is sufficient for understanding both how individual consumer decisions are
made and what this knowledge implies about the characteristics of con-
sumer demand.
ordinal utility
function Utility function that
generates a ranking of market
askets in order of most to least
prefe
ed.
cardinal utility
function Utility function
describing by how much one
market basket is prefe
ed to
another.
ChAPTer 3 CoNSUMer BehAvIor 103
exAMPle 3.2 Can money Buy haPPiness?
economists use the term utility to represent a mea-
sure of the satisfaction or happiness that individuals
get from the consumption of goods and services.
Because a higher income allows one to consume
more goods and services, we say that utility increases
with income. But does greater income and consump-
tion really translate into greater happiness? research
comparing various measures of happiness suggests
that the answer is a qualified yes.4
In one study, an ordinal scale for happiness was
derived from the answer to the following question.
“how satisfied are you at present with your life, all
things considered?”5 Possible responses ran on a
scale from 0 (completely dissatisfied) to 10 (com-
pletely satisfied). Income was found to be a very
strong predictor of happiness (another strong predic-
tor was whether a person was employed or not). on
average, as income increased by one percent, the sat-
isfaction score increased one half a point. Knowing
that there is a positive relationship between utility
or satisfaction and income, it is reasonable to assign
utility values to the baskets of goods and services that
consumers buy. Whether that relationship is cardinal
or ordinal remains an ongoing debate.
Let’s take this inquiry one step further. Can one
compare levels of happiness across as well as within
countries? once again, the evidence says yes. In a
separate survey of individuals in 67 countries, a team
of researchers asked: “All things considered, how sat-
isfied are you with your life as a whole these days?”
esponses were given on a ten-point scale, with 1
epresenting the most dissatisfied and 10 the most
satisfied.6 Income was measured by each country’s
per-capita gross domestic product in U.S. dollars.
Figure 3.9 shows the results, with each data point
epresenting a different country. You can see that as
we move from poor countries with incomes below
$5000 per capita to those with incomes closer to
$10,000 per capita, satisfaction increases substan-
tially. once we move past the $10,000 level, the
index scale of satisfaction increases at a lower rate.
Comparisons across countries are difficult be-
cause there are likely to be many other factors that
explain satisfaction besides income (e.g., health,
climate, political environment, human rights, etc.).
Interestingly, a recent survey of 136,000 individ-
uals over 132 countries shows that the United
States, which had the highest GDP per capita, was
anked 16th overall in happiness. The number 1
ated country was Denmark. Generally, countries
in Northern europe and english-speaking countries
did well overall, as did a number of Latin American
countries. however, South Korea and russia were
not rated as high as their incomes would predict.
Does location affect feelings of well-being within
the United States? The answer is apparently yes, with
the top-ranked states (in order) being Utah, hawaii,
Wyoming, and Colorado, all west of the Mississippi
iver. (The lowest four, in reverse order, were West
virginia, Kentucky, Mississippi, and ohio, all east of
the Mississippi.) Moreover, it is possible that the rela-
tionship between income and satisfaction goes two
ways: Although higher incomes generate more satis-
faction, greater satisfaction offers greater motivation
for individuals to work hard and generate higher
incomes. Interestingly, even when studies account
for other factors, the positive relationship between
income and satisfaction remains.
4For a review of the relevant literature which underlies this example, see Raphael DiTella and
Robert MacCulloch, “Some Uses of Happiness Data in Economics,” Journal of Economic Perspectives
20 (Winter 2006): 25–46. See also the article by Nobel prize winner Angus Deaton, “Income, Health
and Well-Being around the World: Evidence from the Gallup World Poll,” Journal of Economic
Perspectives, 22 (Spring 2008): 53–72.
5Paul Frijters, John P. Haisken-Denew, and Michael A. Shields, “Money Does Matter! Evidence from
Increasing Real Income and Life Satisfaction in East Germany Following Reunification,” American
Economic Review 94 (June 2004): 730–40.
6Ronald Inglehart et al., European and World Values Surveys Four-Wave Integrated Data File, 1981–2004
(2006). Available online: http:
www.worldvaluessurvey.org.
http:
www.worldvaluessurvey.org
104 PArT 2 Producers, Consumers, and Competitive Markets
3.2 Budget Constraints
So far, we have focused only on the first element of consumer theory—
consumer preferences. We have seen how indifference curves (or, alternatively,
utility functions) can be used to describe how consumers value various baskets
of goods. Now we turn to the second element of consumer theory: the budget
constraints that consumers face as a result of their limited incomes.
The Budget line
To see how a budget constraint limits a consumer’s choices, let’s consider a situ-
ation in which a woman has a fixed amount of income, I, that can be spent on
food and clothing. Let F be the amount of food purchased and C be the amount
of clothing. We will denote the prices of the two goods PF and PC. In that case,
PFF (i.e., price of food times the quantity) is the amount of money spent on food
and PCC the amount of money spent on clothing.
The budget line indicates all combinations of F and C for which the total amount
of money spent is equal to income. Because we are considering only two goods
(and ignoring the possibility of saving), our hypothetical consumer will spend
her entire income on food and clothing. As a result, the combinations of food
and clothing that she can buy will all lie on this line:
PFF + PCC = I (3.1)
udget constraints
Constraints that consumers face as
a result of limited incomes.
udget line All combinations
of goods for which the total
amount of money spent is equal
to income.
Sa
ti
sf
ac
ti
on
w
it
h
lif
e
GDP per capita in 1996 U.S. $
0 5000 10,000 15,000 20,000 25,000 30,000 35,000 40,000
9
8
7
6
5
3
4
FIgure 3.9
inCome and haPPiness
A cross-country comparison shows that individuals living in countries with higher GDP
per capita are on average happier than those living in countries with lower per-capita
GDP.
ChAPTer 3 CoNSUMer BehAvIor 105
Suppose, for example, that our consumer has a weekly income of $80, the
price of food is $1 per unit, and the price of clothing is $2 per unit. Table 3.2
shows various combinations of food and clothing that she can purchase each
week with her $80. If her entire budget were allocated to clothing, the most that
she could buy would be 40 units (at a price of $2 per unit), as represented by
market basket A. If she spent her entire budget on food, she could buy 80 units
(at $1 per unit), as given by market basket G. Market baskets B, D, and E show
three additional ways in which her $80 could be spent on food and clothing.
Figure 3.10 shows the budget line associated with the market baskets given in
Table 3.2. Because giving up a unit of clothing saves $2 and buying a unit of food
costs $1, the amount of clothing given up for food along the budget line must be the
same everywhere. As a result, the budget line is a straight line from point A to point
G. In this particular case, the budget line is given by the equation F + 2C = $80.
The intercept of the budget line is represented by basket A. As our consumer
moves along the line from basket A to basket G, she spends less on clothing and
more on food. It is easy to see that the extra clothing which must be given up to
consume an additional unit of food is given by the ratio of the price of food to
the price of clothing ($1/$2 = 1/2). Because clothing costs $2 per unit and food
only $1 per unit, 1/2 unit of clothing must be given up to get 1 unit of food. In
Figure 3.10, the slope of the line, ∆C/∆F = -1/2, measures the relative cost of
food and clothing.
Table 3.2 market Baskets and the Budget line
Market Basket Food (F) ClothiNg (C)
total
speNdiNg
A 0 40 $80
B 20 30 $80
D 40 20 $80
E 60 10 $80
G 80 0 $80
Clothing
(units
per week)
Food
(units per week)
Budget Line F + 2C = $80
30
20
10
0 20 40 60
A
B
D
E
G
10
20
(I/PC) = 40
Slope ∆C/∆F = – 1 = –PF/PC

2
80 = (I/PF)
FIgure 3.10
a Budget line
A budget line describes the combinations of
goods that can be purchased given the consum-
er’s income and the prices of the goods. Line AG
(which passes through points B, D, and E) shows
the budget associated with an income of $80, a
price of food of PF = $1 per unit, and a price of
clothing of PC = $2 per unit. The slope of the
udget line (measured between points B and D) is
-PF/PC = -10/20 = -1/2.
106 PArT 2 Producers, Consumers, and Competitive Markets
Using equation (3.1), we can see how much of C must be given up to consume
more of F. We divide both sides of the equation by PC and then solve for C:
C = (I/PC) - (PF/PC)F (3.2)
Equation (3.2) is the equation for a straight line; it has a vertical intercept of
I/PC and a slope of - (PF/PC).
The slope of the budget line, - (PF/PC), is the negative of the ratio of the prices
of the two goods. The magnitude of the slope tells us the rate at which the two
goods can be substituted for each other without changing the total amount of
money spent. The vertical intercept (I/PC) represents the maximum amount of
C that can be purchased with income I. Finally, the horizontal intercept (I/PF)
tells us how many units of F can be purchased if all income were spent on F.
The effects of Changes in Income and Prices
We have seen that the budget line depends both on income and on the prices of
the goods, PF and PC. But of course prices and income often change. Let’s see
how such changes affect the budget line.
inCome Changes What happens to the budget line when income changes?
From the equation for the straight line (3.2), we can see that a change in income
alters the vertical intercept of the budget line but does not change the slope
(because the price of neither good changed). Figure 3.11 shows that if income
is doubled (from $80 to $160), the budget line shifts outward, from budget
line L1 to budget line L2. Note, however, that L2 remains parallel to L1. If she
desires, our consumer can now double her purchases of both food and cloth-
ing. Likewise, if her income is cut in half (from $80 to $40), the budget line shifts
inward, from L1 to L3.
Food
(units per week)
(I = $40) (I = $80) (I = $160)
L3 L1 L2
40
20
80
60
Clothing
(units per week)
1608040 120
FIgure 3.11
effeCts of a Change
in inCome on the
Budget line
A change in income (with prices un-
changed) causes the budget line to shift
parallel to the original line (L1). When the
income of $80 (on L1) is increased to $160,
the budget line shifts outward to L2. If the
income falls to $40, the line shifts inward
to L3.
ChAPTer 3 CoNSUMer BehAvIor 107
PriCe Changes What happens to the budget line if the price of one
good changes but the price of the other does not? We can use the equation
C = (I/PC) - (PF/PC)F to describe the effects of a change in the price of food
on the budget line. Suppose the price of food falls by half, from $1 to $0.50. In
that case, the vertical intercept of the budget line remains unchanged, although
the slope changes from -PF/PC = - $1/$2 = - $1/2 to - $0.50/$2 = - $1/4.
In Figure 3.12, we obtain the new budget line L2 by rotating the original bud-
get line L1 outward, pivoting from the C-intercept. This rotation makes sense
ecause a person who consumes only clothing and no food is unaffected by the
price change. However, someone who consumes a large amount of food will
experience an increase in his purchasing power. Because of the decline in the
price of food, the maximum amount of food that can be purchased has doubled.
On the other hand, when the price of food doubles from $1 to $2, the budget
line rotates inward to line L3 because the person’s purchasing power has dimin-
ished. Again, a person who consumed only clothing would be unaffected by the
food price increase.
What happens if the prices of both food and clothing change, but in a way
that leaves the ratio of the two prices unchanged? Because the slope of the bud-
get line is equal to the ratio of the two prices, the slope will remain the same.
The intercept of the budget line must shift so that the new line is parallel to the
old one. For example, if the prices of both goods fall by half, then the slope of
the budget line does not change. However, both intercepts double, and the bud-
get line is shifted outward.
This exercise tells us something about the determinants of a consumer’s
purchasing power—her ability to generate utility through the purchase of goods
and services. Purchasing power is determined not only by income, but also by
prices. For example, our consumer’s purchasing power can double either be-
cause her income doubles or because the prices of all the goods that she buys
fall by half.
Finally, consider what happens if everything doubles—the prices of both
food and clothing and the consumer’s income. (This can happen in an inflation-
ary economy.) Because both prices have doubled, the ratio of the prices has not
changed; neither, therefore, has the slope of the budget line. Because the price
of clothing has doubled along with income, the maximum amount of clothing
that can be purchased (represented by the vertical intercept of the budget line)
Clothing
(units pe
week)
Food
(units per week)
40
40 80 160120
(PF = 2)
L3 L1 L2
(PF = )
1
2(PF = 1)
FIgure 3.12
effeCts of a Change in
PriCe on the Budget line
A change in the price of one good (with income
unchanged) causes the budget line to rotate
about one intercept. When the price of food
falls from $1.00 to $0.50, the budget line rotates
outward from L1 to L2. However, when the price
increases from $1.00 to $2.00, the line rotates
inward from L1 to L3.
108 PArT 2 Producers, Consumers, and Competitive Markets
is unchanged. The same is true for food. Therefore, inflationary conditions in
which all prices and income levels rise proportionately will not affect the con-
sumer’s budget line or purchasing power.
3.3 Consumer Choice
Given preferences and budget constraints, we can now determine how individ-
ual consumers choose how much of each good to buy. We assume that consum-
ers make this choice in a rational way—that they choose goods to maximize the
satisfaction they can achieve, given the limited budget available to them. The maximiz-
ing market basket must satisfy two conditions:
1. It must be located on the budget line. To see why, note that any mar-
ket basket to the left of and below the budget line leaves some income
unallocated—income which, if spent, could increase the consumer’s sat-
isfaction. Of course, consumers can—and often do—save some of their
incomes for future consumption. In that case, the choice is not just
between food and clothing, but between consuming food or clothing now
and consuming food or clothing in the future. At this point, however, we
will keep things simple by assuming that all income is spent now. Note
also that any market basket to the right of and above the budget line
cannot be purchased with available income. Thus, the only rational and
feasible choice is a basket on the budget line.
2. It must give the consumer the most prefe
ed combination of goods and
services.
These two conditions reduce the problem of maximizing consumer satisfaction
to one of picking an appropriate point on the budget line.
In our food and clothing example, as with any two goods, we can graphically
illustrate the solution to the consumer’s choice problem. Figure 3.13 shows how
U1
U2
U3
Clothing
(units pe
week)
Food
(units per week)
Budget Line
D
A
B
20 40 80
20
30
40
–10C
+10F
FIgure 3.13
maximizing
Consumer
satisfaCtion
A consumer maximizes satisfaction
y choosing market basket A. At
this point, the budget line and indif-
ference curve U2 are tangent, and
no higher level of satisfaction (e.g.,
market basket D) can be attained.
At A, the point of maximization, the
MRS between the two goods equals
the price ratio. At B, however, be-
cause the MRS [- (-10/10) = 1] is
greater than the price ratio (1/2),
satisfaction is not maximized.
ChAPTer 3 CoNSUMer BehAvIor 109
the problem is solved. Here, three indifference curves describe a consumer’s
preferences for food and clothing. Remember that of the three curves, the outer-
most curve, U3, yields the greatest amount of satisfaction, curve U2 the next
greatest amount, and curve U1 the least.
Note that point B on indifference curve U1 is not the most prefe
ed choice,
ecause a reallocation of income in which more is spent on food and less on
clothing can increase the consumer’s satisfaction. In particular, by moving to
point A, the consumer spends the same amount of money and achieves the
increased level of satisfaction associated with indifference curve U2. In addi-
tion, note that baskets located to the right and above indifference curve U2, like
the basket associated with D on indifference curve U3, achieve a higher level of
satisfaction but cannot be purchased with the available income. Therefore, A
maximizes the consumer’s satisfaction.
We see from this analysis that the basket which maximizes satisfaction must
lie on the highest indifference curve that touches the budget line. Point A is the
point of tangency between indifference curve U2 and the budget line. At A, the
slope of the budget line is exactly equal to the slope of the indifference curve.
Because the MRS (- ∆C/∆F) is the negative of the slope of the indifference
curve, we can say that satisfaction is maximized (given the budget constraint) at
the point where
MRS = PF/PC (3.3)
This is an important result: Satisfaction is maximized when the marginal rate of
substitution (of F for C) is equal to the ratio of the prices (of F to C). Thus the con-
sumer can obtain maximum satisfaction by adjusting his consumption of goods
F and C so that the MRS equals the price ratio.
The condition given in equation (3.3) illustrates the kinds of optimization
conditions that arise in economics. In this instance, satisfaction is maximized
when the marginal benefit—the benefit associated with the consumption of
one additional unit of food—is equal to the marginal cost—the cost of the ad-
ditional unit of food. The marginal benefit is measured by the MRS. At point A,
it equals 1/2 (the magnitude of the slope of the indifference curve), which
implies that the consumer is willing to give up 1/2 unit of clothing to obtain
1 unit of food. At the same point, the marginal cost is measured by the magni-
tude of the slope of the budget line; it too equals 1/2 because the cost of getting
one unit of food is giving up 1/2 unit of clothing (PF = 1 and PC = 2 on the
udget line).
If the MRS is less or greater than the price ratio, the consumer’s satisfaction
has not been maximized. For example, compare point B in Figure 3.13 to point A.
At point B, the consumer is purchasing 20 units of food and 30 units of clothing.
The price ratio (or marginal cost) is equal to 1/2 because food costs $1 and cloth-
ing $2. However, the MRS (or marginal benefit) is greater than 1/2; it is approxi-
mately 1. As a result, the consumer is able to substitute 1 unit of food for 1 unit
of clothing without loss of satisfaction. Because food is cheaper than clothing, it
is in her interest to buy more food and less clothing. If our consumer purchases
1 less unit of clothing, for example, the $2 saved can be allocated to two units of
food, even though only one unit is needed to maintain her level of satisfaction.
The reallocation of the budget continues in this manner (moving along the
udget line), until we reach point A, where the price ratio of 1/2 just equals the
MRS of 1/2. This point implies that our consumer is willing to trade one unit of
clothing for two units of food. Only when the condition MRS = 1/2 = PF/PC
holds is she maximizing her satisfaction.
marginal benefit Benefit
from the consumption of one
additional unit of a good.
marginal cost Cost of one
additional unit of a good.
110 PArT 2 Producers, Consumers, and Competitive Markets
The result that the MRS equals the price ratio is deceptively powerful.
Imagine two consumers who have just purchased various quantities of food
and clothing. If both are maximizing, you can tell the value of each person’s
MRS by looking at the prices of the two goods. What you cannot tell, however,
is the quantity of each good purchased, because that decision is determined by
their individual preferences. If the two consumers have different tastes, they
will consume different quantities of food and clothing, even though each MRS
is the same.
exAMPle 3.3 designing neW automoBiles (ii)
our analysis of consumer choice allows us to see
how different preferences of consumer groups for
automobiles can affect their purchasing decisions.
Following up on example 3.1 (page 99), we con-
sider two groups of consumers planning to buy new
cars. Suppose that each consumer has an overall
car budget of $20,000, but has decided to allo-
cate $10,000 to interior size and acceleration and
$10,000 to all the other attributes of a new car. each
group, however, has different preferences for size
and acceleration.
Figure 3.14 shows the car-buying budget constraint
faced by individuals in each group. Those in the first
group, who are typical of Ford Mustang coupe owners
with preferences similar to those in Figure 3.7 (page
100), prefer acceleration to size. By finding the point
of tangency between a typical individual’s indiffer-
ence curve and the budget constraint, we see that con-
sumers in this group would prefer to buy a car whose
acceleration was worth $7000 and whose size was
worth $3000. Individuals in the second group, who
are typical of Ford explorer users, would prefer cars
$3000
$10,000
$7000 $10,000
Acceleration (horsepower)
$10,000
Acceleration (horsepower)
$2500
$10,000
Size
(cubic feet)
$7500
Size
(cubic feet)
(a) (b)
FIgure 3.14
Consumer ChoiCe of automoBile attriButes
The consumers in (a) are willing to trade off a considerable amount of interior space for some addi-
tional acceleration. Given a budget constraint, they will choose a car that emphasizes acceleration. The
opposite is true for consumers in (b).
ChAPTer 3 CoNSUMer BehAvIor 111
Corner Solutions
Sometimes consumers buy in extremes, at least within categories of goods.
Some people, for example, spend no money on travel and entertainment.
Indifference curve analysis can be used to show conditions under which con-
sumers choose not to consume a particular good.
In Figure 3.15, a man faced with budget line AB for snacks chooses to pur-
chase only ice cream (IC) and no frozen yogurt (Y). This decision reflects what
is called a corner solution. When one of the goods is not consumed, the con-
sumption bundle appears at the corner of the graph. At B, which is the point of
maximum satisfaction, the MRS of ice cream for frozen yogurt is greater than
the slope of the budget line. This inequality suggests that if the consumer had
more frozen yogurt to give up, he would gladly trade it for additional ice cream.
At this point, how ever, our consumer is already consuming all ice cream and
no frozen yogurt, and it is impossible to consume negative amounts of frozen
yogurt.
When a corner solution arises, the consumer’s MRS does not necessarily equal the
price ratio. Unlike the condition expressed in equation (3.3), the necessary con-
dition for satisfaction to be maximized when choosing between ice cream and
frozen yogurt in a corner solution is given by the following inequality.9
MRS Ú PIC/PY (3.4)
corner solution Situation
in which the marginal rate of
substitution of one good for
another in a chosen market basket
is not equal to the slope of the
udget line.
with $2500 worth of acceleration and $7500 worth
of size.7
We have simplified matters for this example by
considering only two attributes. In practice, an au-
tomobile company will use marketing and statistical
studies to learn how different groups of consumers
value a
oad set of attributes. Combined with infor-
mation about how these attributes will affect manu-
facturing costs, the company can design a production
and marketing plan.
In the context of our example, one potentially
profitable option is to appeal to both groups of
consumers by manufacturing a model emphasizing
acceleration to a slightly lesser degree than pre-
fe
ed by those in Figure 3.14 (a). A second option
is to produce a relatively large number of cars that
emphasize size and a smaller number emphasizing
acceleration.
Knowledge about the preferences of each group
(i.e., the actual indifference curves), along with in-
formation about the number of consumers in each,
would help the firm make a sensible business deci-
sion. In fact, an exercise similar to the one we’ve
described here was ca
ied out by General Motors
in a survey of a large number of automobile buyers.8
Some of the results were expected. For example,
households with children tended to prefer functional-
ity over style and so tended to buy minivans rather
than sedans and sporty cars. rural households, on the
other hand, tended to purchase pickups and all-wheel
drives. More interesting was the strong co
elation be-
tween age and attribute preferences. older consum-
ers tended to prefer larger and heavier cars with more
safety features and accessories (e.g., power windows
and steering). Further, younger consumers prefe
ed
greater horsepower and more stylish cars.
7The first set of indifference curves for the Ford Mustang coupe will be of the following form: U
(level of utility) = b0 (constant) + b1*S (space in cubic feet)* b2*S2 1 b3*H (horsepower) 1 b4*H2 1
5*O (a list of other attributes). Each indifference curve represents the combinations of S and H that
generate the same level of utility. The comparable relationship for the Ford Explorer will have the
same form, but different b’s.
8The survey design and the results are described in Steven Be
y, James Levinsohn, and Ariel Pakes,
“Differentiated Products Demand Systems from a Combination of Micro and Macro Data: The New
Car Market,” Journal of Political Economy, 112 (Fe
uary 2004): 68–105.
9Strict equality could hold if the slope of the budget constraint happened to equal the slope of the
indifference curve—a condition that is unlikely.
112 PArT 2 Producers, Consumers, and Competitive Markets
This inequality would, of course, be reversed if the corner solution were
at point A rather than B. In either case, we can see that the marginal benefit–
marginal cost equality that we described in the previous section holds only
when positive quantities of all goods are consumed.
An important lesson here is that predictions about how much of a product
consumers will purchase when faced with changing economic conditions de-
pend on the nature of consumer preferences for that product and related prod-
ucts and on the slope of the consumer’s budget line. If the MRS of ice cream
for frozen yogurt is substantially greater than the price ratio, as in Figure 3.15,
then a small decrease in the price of frozen yogurt will not alter the consumer’s
choice; he will still choose to consume only ice cream. But if the price of frozen
yogurt falls far enough, the consumer could quickly choose to consume a lot of
frozen yogurt.
U1
A
U2 U3
Frozen yogurt
(cups pe
month)
B Ice cream
(cups per month)
FIgure 3.15
a Corner solution
When the consumer’s marginal rate of
substitution is not equal to the price ra-
tio for all levels of consumption, a corner
solution arises. The consumer maximizes
satisfaction by consuming only one of the
two goods. Given budget line AB, the
highest level of satisfaction is achieved at
B on indifference curve U1, where the MRS
(of ice cream for frozen yogurt) is greater
than the ratio of the price of ice cream to
the price of frozen yogurt.
exAMPle 3.4 the authors argue aBout health Care
expenditures on health care in the
United States have risen dramati-
cally over the past few decades, a
phenomenon that some people—
including author rubinfeld—find
alarming. rubinfeld and many oth-
ers claim that these expenditures
have increased so much because
our health care system is very inefficient. “That may
well be,” argues Pindyck, “but there could also be
another reason. As consumers become better off
economically, their preferences shift to-
ward health care and away from other
goods. After all, if you already own a
nice home and two cars, what would
give you more satisfaction—a third car,
or additional medical care that might
extend your life by a year? I don’t know
about you, Dan, but I would choose the
extra health care.”
“I guess I would too,” replies rubinfeld. “But to
esolve this question, we need to look at indifference
ChAPTer 3 CoNSUMer BehAvIor 113
curves that describe consumer preferences for health
care versus other goods. If the story you are telling
is co
ect, the indifference curves should look like
the ones in Figure 3.16. That figure shows a series
of indifference curves and budget lines that charac-
terize the trade-off between consumption of health
care (h) versus other goods (o). Indifference curve
U1 applies to a consumer with low income; the
consumer’s budget line is tangent at point A, so that
the consumption of health care and consumption of
other goods that maximize the consumer’s satisfac-
tion are H1 and O1. Indifference curve U2 yields a
greater amount of satisfaction, but is only feasible
for a consumer with higher income. In this case util-
ity is maximized at point B. Curve U3 applies to a
high-income consumer, and implies less willingness
to give up health care for other goods. Moving from
point B to point C, the consumer’s consumption of
health care increases considerably (from H2 to H3),
while her consumption of other goods increases
only modestly (from O2 to O3).
“The question,” asks rubinfeld, “is whether
Figure 3.16 co
ectly characterizes the preferences of
consumers. After all, anyone can draw some curves.
But is there any evidence that Figure 3.16 describes
the real world?”
“Glad you asked,” replies Pindyck. “At least one
ecent statistical study provides strong evidence that
it does describe the real world.10 And common
sense provides further support. If your income were
high enough so that you could have most of the
things you wanted, would you prefer to spend ad-
ditional income on life-extending health care or on
another car?”
10See the interesting article by Robert E. Hall and Charles I. Jones, “The Value of Life and the Rise in
Health Spending,” Quarterly Journal of Economics, Fe
uary 2007, pp. 39–72. The authors explain that
the optimal composition of total spending shifts toward health as income increases. They predict
that the optimal heath share of spending is likely to exceed 30 percent by 2050.
U1
U2
U3
Health
Care
Othe
Goods
A
B
CH3
H2
H1
O1 O2 O3
FIgure 3.16
Consumer
PreferenCes for
health Care versus
other goods
These indifference curves show the
trade-off between consumption of
health care (H) versus other goods
(O). Curve U1 applies to a consumer
with low income; given the consum-
er’s budget constraint, satisfaction is
maximized at point A. As income
increases the budget line shifts to
the right, and curve U2 becomes fea-
sible. The consumer moves to point
B, with greater consumption of both
health care and other goods. Curve
U3 applies to a high-income con-
sumer, and implies less willingness to
give up health care for other goods.
Moving from point B to point C, the
consumer’s consumption of health
care increases considerably (from
H2 to H3), while her consumption of
other goods increases only modestly
(from O2 to O3).
114 PArT 2 Producers, Consumers, and Competitive Markets
3.4 revealed Preference
In Section 3.1, we saw how an individual’s preferences could be represented
y a series of indifference curves. Then in Section 3.3, we saw how preferences,
given budget constraints, determine choices. Can this process be reversed? If
we know the choices that a consumer has made, can we determine his or her
preferences?
exAMPle 3.5 a College trust fund
Jane Doe’s parents have provided a trust fund for her
college education. Jane, who is 18, can receive the
entire trust fund on the condition that she spend it
only on education. The fund is a welcome gift but
perhaps not as welcome as an unrestricted trust. To
see why Jane feels this way, consider Figure 3.17, in
which dollars per year spent on education are shown
on the horizontal axis and dollars spent on other
forms of consumption on the vertical.
The budget line that Jane faces before being awarded
the trust is given by line PQ. The trust fund expands the
udget line outward as long as the full amount of the
fund, shown by distance PB, is spent on education. By
accepting the trust fund and going to college, Jane in-
creases her satisfaction, moving from A on indifference
curve U1 to B on indifference curve U2.
Note that B represents a corner solution because
Jane’s marginal rate of substitution of education for
other consumption is lower than the relative price
of other consumption. Jane would prefer to spend a
portion of the trust fund on other goods in addition to
education. Without restriction on the trust fund, she
would move to C on indifference curve U3, decreas-
ing her spending on education (perhaps going to a
junior college rather than a four-year college) but in-
creasing her spending on items that she enjoys more
than education.
ecipients usually prefer unrestricted to restricted
trusts. restricted trusts are popular, however, because
they allow parents to control children’s expenditures
in ways that they believe are in the children’s long-
un best interests.
U1
U2
U3
A
B
C
P
Other
consumption
($)
Education ($)Q
FIgure 3.17
a College trust fund
When given a college trust fund that must be
spent on education, the student moves from
A to B, a corner solution. If, however, the trust
fund could be spent on other consumption as
well as education, the student would be better
off at C.
ChAPTer 3 CoNSUMer BehAvIor 115
We can if we have information about a sufficient number of choices that have
een made when prices and income levels varied. The basic idea is simple. If
a consumer chooses one market basket over another, and if the chosen market basket is
more expensive than the alternative, then the consumer must prefer the chosen market
asket.
Suppose that an individual, facing the budget constraint given by line l1
in Figure 3.18, chooses market basket A. Let’s compare A to baskets B and D.
Because the individual could have purchased basket B (and all baskets below
line l1) and did not, we say that A is prefe
ed to B.
It might seem at first glance that we cannot make a direct comparison be-
tween baskets A and D because D is not on l1. But suppose the relative prices of
food and clothing change, so that the new budget line is l2 and the individual
then chooses market basket B. Because D lies on budget line l2 and was not cho-
sen, B is prefe
ed to D (and to all baskets below line l2). Because A is prefe
ed
to B and B is prefe
ed to D, we conclude that A is prefe
ed to D. Furthermore,
note in Figure 3.18 that basket A is prefe
ed to all of the baskets that appear in
the green-shaded areas. However, because food and clothing are “goods” rather
than “bads,” all baskets that lie in the pink-shaded area in the rectangle above
and to the right of A are prefe
ed to A. Thus, the indifference curve passing
through A must lie in the unshaded area.
Given more information about choices when prices and income levels
vary, we can get a better fix on the shape of the indifference curve. Consider
Figure 3.19. Suppose that facing line l3 (which was chosen to pass through A),
the individual chooses market basket E. Because E was chosen even though A
was equally expensive (it lies on the same budget line), E is prefe
ed to A, as
are all points in the rectangle above and to the right of E. Now suppose that fac-
ing line l4 (which passes through A), the individual chooses market basket G.
Because G was chosen and A was not, G is prefe
ed to A, as are all market bas-
kets above and to the right of G.
We can go further by making use of the assumption that indifference curves
are convex. In that case, because E is prefe
ed to A, all market baskets above
and to the right of line AE in Figure 3.19 must be prefe
ed to A. Otherwise,
the indifference curve passing through A would have to pass through a point
above and to the right of AE and then fall below the line at E—in which case the
Clothing
(units pe
month)
Food
(units per month)
A
B
D
l1
l2
FIgure 3.18
evealed PreferenCe: tWo
Budget lines
If an individual facing budget line l1 chose market basket
A rather than market basket B, A is revealed to be pre-
fe
ed to B. Likewise, the individual facing budget line l2
chooses market basket B, which is then revealed to be
prefe
ed to market basket D. Whereas A is prefe
ed to
all market baskets in the green-shaded area, all baskets
in the pink-shaded area are prefe
ed to A.
116 PArT 2 Producers, Consumers, and Competitive Markets
indifference curve would not be convex. By a similar argument, all points on
AG or above are also prefe
ed to A. Therefore, the indifference curve must lie
within the unshaded area.
The revealed preference approach is valuable as a means of checking whether
individual choices are consistent with the assumptions of consumer theory. As
Example 3.6 shows, revealed preference analysis can help us understand the
implications of choices that consumers must make in particular circumstances.
Clothing
(units per month)
Food
(units per month)
l1
l4
l2
l3
E
A
B G
FIgure 3.19
evealed PreferenCe: four
Budget lines
Facing budget line l3 the individual chooses
E, which is revealed to be prefe
ed to A (be-
cause A could have been chosen). Likewise,
facing line l4, the individual chooses G which is
also revealed to be prefe
ed to A. Whereas A
is prefe
ed to all market baskets in the green-
shaded area, all market baskets in the pink-
shaded area are prefe
ed to A.
exAMPle 3.6 revealed PreferenCe for reCreation
A health club has been offering the
use of its facilities to anyone who is
willing to pay an hourly fee. Now
the club decides to alter its pricing
policy by charging both an an-
nual membership fee and a lower
hourly fee. Does this new financial
a
angement make individuals bet-
ter off or worse off than they were
under the old a
angement? The answer depends on
people’s preferences.
Suppose that roberta has $100 of income avail-
able each week for recreational activities, including
exercise, movies, restaurant meals, and so on. When
the health club charged a fee of $4 per hour, roberta
used the facility 10 hours per week. Under the new
a
angement, she is required to pay $30 per week but
can use the club for only $1 per hour.
Is this change beneficial for
oberta? revealed preference anal-
ysis provides the answer. In Figure
3.20, line l1 represents the budget
constraint that roberta faced under
the original pricing a
angement. In
this case, she maximized her satis-
faction by choosing market basket
A, with 10 hours of exercise and
$60 of other recreational activities. Under the new
a
angement, which shifts the budget line to l2, she
could still choose market basket A. But because U1
is clearly not tangent to l2, roberta will be better off
choosing another basket, such as B, with 25 hours of
exercise and $45 worth of other recreational activi-
ties. Because she would choose B when she could
still choose A, she prefers B to A. The new pricing
a
angement therefore makes roberta better off.
ChAPTer 3 CoNSUMer BehAvIor 117
3.5 Marginal utility
and Consumer Choice
In Section 3.3, we showed graphically how a consumer can maximize his or
her satisfaction, given a budget constraint. We do this by finding the highest
indifference curve that can be reached, given that budget constraint. Because
the highest indifference curve also has the highest attainable level of utility, it is
natural to recast the consumer’s problem as one of maximizing utility subject to
a budget constraint.
The concept of utility can also be used to recast our analysis in a way that
provides additional insight. To begin, let’s distinguish between the total utility
obtained by consumption and the satisfaction obtained from the last item con-
sumed. Marginal utility (MU) measures the additional satisfaction obtained from
consuming one additional unit of a good. For example, the marginal utility associ-
ated with a consumption increase from 0 to 1 unit of food might be 9; from 1 to
2, it might be 7; from 2 to 3, it might be 5.
These numbers imply that the consumer has diminishing marginal
utility: As more and more of a good is consumed, consuming additional
amounts will yield smaller and smaller additions to utility. Imagine, for ex-
ample, the consumption of television: Marginal utility might fall after the
marginal utility
(MU) Additional satisfaction
obtained from consuming one
additional unit of a good.
diminishing marginal
utility Principle that as
more of a good is consumed,
the consumption of additional
amounts will yield smaller
additions to utility.
(Note that B is also prefe
ed to C, which represents
the option of not using the health club at all.)
We could also ask whether this new pricing
system—called a two-part tariff—will increase the
club’s profits. If all members are like roberta and
more use generates more profit, then the answer is
yes. In general, however, the answer depends on
two factors: the preferences of all members and
the costs of operating the facility. We discuss the
two-part tariff in detail in Chapter 11, where we
study ways in which firms with market power set
prices.
FIgure 3.20
evealed PreferenCe for
eCreation
When facing budget line l1, an individual chooses
to use a health club for 10 hours per week at
point A. When the fees are altered, she faces
udget line l2. She is then made better off be-
cause market basket A can still be purchased, as
can market basket B, which lies on a higher indif-
ference curve.
Othe
ecreational
activities
($)
Amount of exercise (hours)
100
80
60
40
20
0
25 50 75
l1
l2
A
B
U1
U2
C
118 PArT 2 Producers, Consumers, and Competitive Markets
second or third hour and could become very small after the fourth or fifth
hour of viewing.
We can relate the concept of marginal utility to the consumer’s utility-
maximization problem in the following way. Consider a small movement
down an indifference curve in Figure 3.8 (page 101). The additional consump-
tion of food, ∆F, will generate marginal utility MUF. This shift results in a
total increase in utility of MUF∆F. At the same time, the reduced consumption
of clothing, ∆C, will lower utility per unit by MUC, resulting in a total loss of
MUC∆C.
Because all points on an indifference curve generate the same level of utility,
the total gain in utility associated with the increase in F must balance the loss
due to the lower consumption of C. Formally,
0 = MUF(∆F) + MUC(∆C)
Now we can rea
ange this equation so that
-(∆C/∆F) = MUF/MUC
But because -(∆C/∆F) is the MRS of F for C, it follows that
MRS = MUF/MUC (3.5)
Equation (3.5) tells us that the MRS is the ratio of the marginal utility of F to
the marginal utility of C. As the consumer gives up more and more of C to
obtain more of F, the marginal utility of F falls and that of C increases, so MRS
decreases.
We saw earlier in this chapter that when consumers maximize their satisfac-
tion, the MRS of F for C is equal to the ratio of the prices of the two goods:
MRS = PF/PC (3.6)
Because the MRS is also equal to the ratio of the marginal utilities of consuming
F and C (from equation 3.5), it follows that
MUF/MUC = PF/PC
o
MUF/PF = MUC/PC (3.7)
Equation (3.7) is an important result. It tells us that utility maximization is
achieved when the budget is allocated so that the marginal utility per dollar of
expenditure is the same for each good. To see why this principle must hold, sup-
pose that a person gets more utility from spending an additional dollar on
food than on clothing. In this case, her utility will be increased by spending
more on food. As long as the marginal utility of spending an extra dollar on
food exceeds the marginal utility of spending an extra dollar on clothing, she
can increase her utility by shifting her budget toward food and away from
clothing. Eventually, the marginal utility of food will decrease (because there
is diminishing marginal utility in its consumption) and the marginal utility
of clothing will increase (for the same reason). Only when the consumer has
satisfied the equal marginal principle—i.e., has equalized the marginal utility
per dollar of expenditure across all goods—will she have maximized utility. The
equal marginal principle is an important concept in microeconomics. It will
eappear in different forms throughout our analysis of consumer and pro-
ducer behavior.
equal marginal principle
Principle that utility is maximized
when the consumer has equalized
the marginal utility per dollar of
expenditure across all goods.
ChAPTer 3 CoNSUMer BehAvIor 119
exAMPle 3.7 marginal utility and haPPiness
In example 3.2 (page 103), we saw that
money (i.e., a higher income) can buy
happiness, at least to a degree. But what,
if anything, does research on consumer
satisfaction tell us about the relationship
etween happiness and the concepts of
utility and marginal utility? Interestingly,
that research is consistent with a pattern
of diminishing marginal utility of income,
oth in the U.S. and across countries.
To see why, let’s re-examine Figure 3.9
(page 104) in example 3.2. The data sug-
gest that as incomes increase from one country to the
next, satisfaction, happiness, or utility (we are using
the three words interchangeably) all increase as per-
capita income increases. The incremental increase in
satisfaction, however, declines as income increases.
If one is willing to accept that the satis-
faction index resulting from the survey
is a cardinal index, then the results are
consistent with a diminishing marginal
utility of income.
The results for the U.S. are quali-
tatively very similar to those for the
67 countries that make up the data for
Figure 3.9. Figure 3.21 calculates the
mean level of life satisfaction for nine
separate income groups in the popula-
tion; the lowest has a mean income of
$6,250, the next a mean income of $16,250, and so
on until the highest group, whose mean income is
$87,500. The solid curve is the one that best fits the
data. once again, we can see that reported happi-
ness increases with income, but at a diminishing rate.
Sa
ti
sf
ac
ti
on
w
it
h
Li
fe
6.8
7
7.2
7.4
7.6
7.8
8
8.2
Income in 1999 U.S. $
0 10000 20000 30000 40000 50000 60000 70000 80000 90000 100000
FIgure 3.21
marginal utility and haPPiness
A comparison of mean levels of satisfaction with life across income classes in the United States shows
that happiness increases with income, but at a diminishing rate.
120 PArT 2 Producers, Consumers, and Competitive Markets
ationing
In times of war and other crises, governments sometimes ration food, gasoline,
and other products, rather than allow prices to increase to competitive levels.
During World War II, for example, individual households in the United States
were limited to twelve ounces of sugar per week, one pound of coffee every five
weeks, and three gallons of gasoline per week. Rationing has often been used
with respect to water in periods of drought. Within the United States, California
has often faced water rationing for both household consumption and agricul-
tural production. Outside the United States, countries such as Rwanda, India,
Pakistan, and Egypt have imposed water rationing as recently as 2010.
Nonprice rationing is an alternative that some consider more equitable
than relying on uncontested market forces. Under a market system, those with
higher incomes can outbid those with lower incomes to obtain goods that are
in scarce supply. However, if products are rationed through a mechanism such
as the allocation of coupons to households or businesses, everyone will have an
equal chance to purchase a rationed good.
To understand how we can analyze rationing using the basic consumer
model, let’s use the gasoline rationing that occu
ed during 1979 as an example.
Following the 1979 Iranian Revolution, oil prices jumped, but the United States
imposed price controls that prevented increases in the price of gasoline, result-
ing in shortages. Gasoline was allocated by long lines at the gas pump: While
those who were willing to give up their time waiting got the gas they wanted,
others did not. By guaranteeing every eligible person a minimum amount of
gasoline, rationing can provide some people with access to a product that they
could not otherwise afford. But rationing hurts others by limiting the amount of
gasoline that they can buy.12
For those students concerned about future income
prospects, a recent survey by psychologist Daniel
Kahneman and economist Angus Deaton shows that
for this relatively high income group, making ad-
ditional money does not improve a person’s ability
to enjoy leisure time and good health—all of which
factor into one’s overall well-being.11
These results offer strong support for the modern
theory of economic decision making that underlies this
text, but they are still being carefully scrutinized. For
example, they do not account for the fact that satisfac-
tion tends to vary with age, with younger people often
expressing less satisfaction than older folks. or we can
look at this a different way. Students have something
positive to look forward to as they get older and wiser.
A second issue arises when we compare the
esults of happiness studies over time. Per-capita
incomes in the U.S., U.K., Belgium, and Japan
have all risen substantially over the past 20 years.
Average happiness, however, has remained rela-
tively unchanged. (Denmark, Germany, and Italy
did show some increased satisfaction.) one plau-
sible interpretation is that happiness is a rela-
tive, not absolute, measure of well-being. As a
country’s income increases over time, its citizens
increase their expectations; in other words, they
aspire to having higher incomes. To the extent that
satisfaction is tied to whether those aspirations
are met, satisfaction may not increase as income
grows over time.
11Daniel Kahneman and Angus Deaton, “High Income Improves Evaluation of Life but Not
Emotional Well-Being,” PNAS, Vol. 107 (September 21, 2010): 16489–16493.
12For a more extensive discussion of gasoline rationing, see H. E. Frech III and William C. Lee,
“The Welfare Cost of Rationing-by-Queuing Across Markets: Theory and Estimates from the U.S.
Gasoline Crises,” Quarterly Journal of Economics (1987): 97–108.
ChAPTer 3 CoNSUMer BehAvIor 121
We can see this principle clearly in Figure 3.22, which applies to a woman
with an annual income of $20,000. The horizontal axis shows her annual con-
sumption of gasoline, the vertical axis her remaining income after purchasing
gasoline. Suppose the controlled gasoline price is $1 per gallon. Because her
income is $20,000, she is limited to the points on budget line AB, which has a
slope of -1. Point A represents her total income of $20,000. (If no gasoline were
purchased, she would have $20,000 to spend on other goods.) At point B she
would be spending her entire income on gasoline. At $1 per gallon, she might
wish to buy 5000 gallons of gasoline per year and spend $15,000 on other goods,
epresented by C. At this point, she would have maximized her utility (by being
on the highest possible indifference curve U2), given her budget constraint of
$20,000.
Let’s assume that with rationing, our consumer can purchase up to a maxi-
mum of 2000 gallons of gasoline. Thus, she now faces budget line ADE, which is
not a straight line because purchases above 2000 gallons are not possible. Point
D represents the point of consumption of 2000 gallons per year. At that point,
the budget line become vertical, declining to point E, since rationing has lim-
ited gasoline consumption. The figure shows that her choice to consume at D
involves a lower level of utility, U1, than would be achieved without rationing,
U2, because she is consuming less gasoline and more of other goods than she
would otherwise prefer.
It is clear that at the rationed price the woman would be better off if her con-
sumption were not constrained. But is she better off under a rationing system
than she would be if there were no rationing at all? The answer, not surprisingly,
depends on what the competitive market price of gasoline would have been
without rationing. Figure 3.23 illustrates this point. Recall that had the price
of gasoline been determined by the market to be $1 per gallon, our consumer
would have been able to buy up to 20,000 gallons of gasoline per year—hence
the original budget line. With rationing, she chooses to buy the maximum allow-
able 2000 gallons per year, putting her on indifference curve U1. Now suppose
that the competitive market price had been $2.00 per gallon rather than $1.00.
U2
Spending
on othe
goods ($)
20,00050002000
C
D
A
0
15,000
BE
Gasoline (gallons per year)
18,000
20,000
U1
FIgure 3.22
ineffiCienCy of
gasoline rationing
When a good is rationed, less is
available than consumers would
like to buy. Consumers may be
worse off. Without gasoline ration-
ing, up to 20,000 gallons of gas-
oline are available for consump-
tion (at point B). The consumer
chooses point C on indifference
curve U2, consuming 5000 gallons
of gasoline. However, with a limit
of 2000 gallons of gasoline un-
der rationing (at point E), the con-
sumer moves to D on the lower
indifference curve U1.
122 PArT 2 Producers, Consumers, and Competitive Markets
Now the relevant budget line would be the line that was associated with a maxi-
mum gasoline consumption of only 10,000 gallons per year, and with no rationing
she would choose point F, which lies below indifference curve U1. (At point F, she
purchases 3,000 gallons of gasoline and has $14,000 to spend on other goods.)
But, consider what would happen if the price of gasoline were only $1.33 per
gallon. Then the relevant budget line would be the line associated with a maxi-
mum gasoline consumption of about 15,000 gallons per year ($20,000/$1.33). She
would choose a point such as G, where she purchases more than 3,000 gallons
of gasoline and has more than $14,000 to spend on other goods. In this case, she
would be better off without rationing, since point G lies above indifference curve
U1. We can conclude, therefore, that while rationing is a less efficient means of
allocating goods and services, under any particular rationing scheme some indi-
viduals may well be better off, even though others will necessarily be worse off.
*3.6 Cost-of-living Indexes
The Social Security system has been the subject of heated debate for some time
now. Under the present system, a retired person receives an annual benefit
that is initially determined at the time of retirement and is based on his or her
work history. The benefit then increases from year to year at a rate equal to the
ate of increase of the Consumer Price Index (CPI). Does the CPI accurately
eflect the cost of living for retirees? Is it appropriate to use the CPI as we now
do—as a cost-of-living index for other government programs, for private union
pensions, and for private wage agreements? On a similar note, we might ask
whether the Producer Price Index (PPI) accurately measures the change over
time in the cost of production. The answers to these questions lie in the eco-
nomic theory of consumer behavior. In this section, we describe the theoretical
underpinnings of cost indexes such as the CPI, using an example that describes
the hypothetical price changes that students and their parents might face.
in §1.3, we introduced the
Consumer Price Index as
a measure of the cost of a
“typical” consumer’s entire
market basket. as such,
changes in the Cpi also mea-
sure the rate of inflation.
cost-of-living index ratio
of the present cost of a typical
undle of consumer goods and
services compared with the cost
during a base period.
in §1.3, we explained that
the Producer Price Index
provides a measure of the
aggregate price level for
intermediate products and
wholesale goods.
Spending
on othe
goods ($)
D
0 10,0003000
20,000
14,000
Gasoline (gallons per year)
15,000 20,000
U1
G
F
FIgure 3.23
ComParing gasoline
ationing to the free
market
Some consumers will be worse off, but
others may be better off with rationing.
With rationing and a gasoline price of
$1.00 she buys the maximum allowable
2000 gallons per year, putting her on
indifference curve U1. Had the competi-
tive market price been $2.00 per gallon
with no rationing, she would have cho-
sen point F, which lies below indiffer-
ence curve U1. However, had the price
of gasoline been only $1.33 per gallon,
she would have chosen point G, which
lies above indifference curve U1.
ChAPTer 3 CoNSUMer BehAvIor 123
Ideal Cost-of-living Index
Let’s look at two sisters, Rachel and Sarah, whose preferences are identical.
When Sarah began her college education in 2000, her parents gave her a “dis-
cretionary” budget of $500 per quarter. Sarah could spend the money on food,
which was available at a price of $2.00 per pound, and on books, which were
available at a price of $20 each. Sarah bought 100 pounds of food (at a cost of
$200) and 15 books (at a cost of $300). Ten years later, in 2010, when Rachel (who
had worked during the interim) is about to start college, her parents promise
her a budget that is equivalent in buying power to the budget given to her older
sister. Unfortunately, prices in the college town have increased, with food now
$2.20 per pound and books $100 each. By how much should the discretionary
udget be increased to make Rachel as well off in 2010 as her sister Sarah was
in 2000? Table 3.3 summarizes the relevant data and Figure 3.24 provides the
answer.
The initial budget constraint facing Sarah in 2000 is given by line l1 in
Figure 3.24; her utility-maximizing combination of food and books is at point
A on indifference curve U1. We can check that the cost of achieving this level of
utility is $500, as stated in the table:
$500 = 100 lbs. of food * $2.00/lb. + 15 books * $20
ook
As Figure 3.24 shows, to achieve the same level of utility as Sarah while fac-
ing the new higher prices, Rachel requires a budget sufficient to purchase the
Table 3.3 ideal Cost-of-living index
2000 (SARAH) 2010 (RACHEL)
price of books $20
ook $100
ook
Number of books 15 6
price of food $2.00/lb. $2.20/lb.
pounds of food 100 300
expenditure $500 $1260
50 100 150 200 250 300 350 400 450 500 550 600
5
10
15
20
25
Books
(per quarter)
U1
A
B
l1
l3
l2
Food (lb. per quarter)
0
FIgure 3.24
Cost-of-living indexes
A price index, which represents the cost
of buying bundle A at cu
ent prices rela-
tive to the cost of bundle A at base-year
prices, overstates the ideal cost-of-living
index.
124 PArT 2 Producers, Consumers, and Competitive Markets
food-book consumption bundle given by point B on line l2 (and tangent to in-
difference curve U1), where she chooses 300 lbs. of food and 6 books. Note that
in doing so, Rachel has taken into account the fact that the price of books has
increased relative to food. Therefore, she has substituted toward food and away
from books.
The cost to Rachel of attaining the same level of utility as Sarah is given by
$1260 = 300 lbs. of food * $2.20/lb. + 6 books * $100
ook
The ideal cost-of-living adjustment for Rachel is therefore $760 (which is $1260
minus the $500 that was given to Sarah). The ideal cost-of-living index is
$1260/$500 = 2.52
Our index needs a base year, which we will set at 2000 = 100, so that the value
of the index in 2010 is 252. A value of 252 implies a 152 percent increase in the
cost of living, whereas a value of 100 would imply that the cost of living has not
changed. This ideal cost-of-living index represents the cost of attaining a given
level of utility at cu
ent (2010) prices relative to the cost of attaining the same utility
at base (2000) prices.
laspeyres Index
Unfortunately, such an ideal cost-of-living index would entail large amounts of in-
formation. We would need to know individual preferences (which vary across the
population) as well as prices and expenditures. Actual price indexes are therefore
ased on consumer purchases, not preferences. A price index that uses a fixed con-
sumption bundle in the base period is called a Laspeyres price index. The Laspeyres
price index answers the question: What is the amount of money at cu
ent-year prices
that an individual requires to purchase the bundle of goods and services that was chosen in
the base year divided by the cost of purchasing the same bundle at base-year prices?
The Laspeyres price index was illustrated in Figure 3.24. Calculating a
Laspeyres cost-of-living index for Rachel is a straightforward process. Buying
100 pounds of food and 15 books in 2010 would require an expenditure of
$1720 (100 * $2.20 + 15 * $100). This expenditure allows Rachel to choose
undle A on budget line l3 (or any other bundle on that line). Line l3 was con-
structed by shifting line l2 outward until it intersected point A. Note that l3 is
the budget line that allows Rachel to purchase, at cu
ent 2010 prices, the same
consumption bundle that her sister purchased in 2000. To compensate Rachel
for the increased cost of living, we must increase her discretionary budget by
$1220. Using 100 as the base in 2000, the Laspeyres index is therefore
100 * $1720/$500 = 344
ComParing ideal Cost-of-living and lasPeyres indexes In our
example, the Laspeyres price index is clearly much higher than the ideal price
index. Does a Laspeyres index always overstate the true cost-of-living index? The
answer is yes, as you can see from Figure 3.24. Suppose that Rachel was given
the budget associated with line l3 during the base year of 2000. She could choose
undle A, but clearly she could achieve a higher level of utility if she purchased
more food and fewer books (by moving to the right on line l3). Because A and B
generate equal utility, it follows that Rachel is better off receiving a Laspeyres
cost-of-living adjustment rather than an ideal adjustment. The Laspeyres index
overcompensates Rachel for the higher cost of living, and the Laspeyres cost-of-
living index is, therefore, greater than the ideal cost-of-living index.
ideal cost-of-living
index Cost of attaining a given
level of utility at cu
ent prices
elative to the cost of attaining the
same utility at base-year prices.
Laspeyres price
index Amount of money
at cu
ent year prices that an
individual requires to purchase
a bundle of goods and services
chosen in a base year divided by
the cost of purchasing the same
undle at base-year prices.
ChAPTer 3 CoNSUMer BehAvIor 125
This result holds generally. Why? Because the Laspeyres price index assumes
that consumers do not alter their consumption patterns as prices change. By changing
consumption, however—increasing purchases of items that have become rela-
tively cheaper and decreasing purchases of relatively more expensive items—
consumers can achieve the same level of utility without having to consume the
same bundle of goods that they did before the price change.
Paasche Index
Another commonly used cost-of-living index is the Paasche index. Unlike the
Laspeyres index, which focuses on the cost of buying a base-year bundle, the
Paasche index focuses on the cost of buying the cu
ent year’s bundle. In particu-
lar, the Paasche index answers another question: What is the amount of money at
cu
ent-year prices that an individual requires to purchase the cu
ent bundle of goods
and services divided by the cost of purchasing the same bundle in the base year?
ComParing the lasPeyres and PaasChe indexes It is helpful to com-
pare the Laspeyres and the Paasche cost-of-living indexes.
•     Laspeyres index: The amount of money at cu
ent-year prices that an indi-
vidual requires to purchase the bundle of goods and services that was chosen
in the base year divided by the cost of purchasing the same bundle at base-
year prices.
•     Paasche index: The amount of money at cu
ent-year prices that an indi-
vidual requires to purchase the bundle of goods and services chosen in the
cu
ent year divided by the cost of purchasing the same bundle in the base
year.
Both the Laspeyres (LI) and Paasche (PI) indexes are fixed-weight indexes:
The quantities of the various goods and services in each index remain un-
changed. For the Laspeyres index, however, the quantities remain unchanged
at base-year levels; for the Paasche they remain unchanged at cu
ent-year levels.
Suppose generally that there are two goods, food (F) and clothing (C). Let:
PFt and PCt be cu
ent-year prices
PFb and PCb be base-year prices
Ft and Ct be cu
ent-year quantities
Fb and Cb be base-year quantities
We can write the two indexes as:
LI =
PFtFb + PCtC
PFbFb + PCbC

PI =
PFtFt + PCtCt
PFbFt + PCbCt

Just as the Laspeyres index will overstate the ideal cost of living, the Paasche
will understate it because it assumes that the individual will buy the cu
ent-
year bundle in the base period. In actuality, facing base-year prices, consumers
would have been able to achieve the same level of utility at a lower cost by
changing their consumption bundles. Because the Paasche index is a ratio of
the cost of buying the cu
ent bundle divided by the cost of buying the cu
ent
Paasche index Amount of
money at cu
ent-year prices that
an individual requires to purchase
a cu
ent bundle of goods and
services divided by the cost of
purchasing the same bundle in a
ase year.
fixed-weight index Cost-
of-living index in which the
quantities of goods and services
emain unchanged.
126 PArT 2 Producers, Consumers, and Competitive Markets
undle at base-year prices, overstating the cost of the base-year bundle (the
denominator in the ratio) will cause the Paasche index itself to be understated.
To illustrate the Laspeyres-Paasche comparison, let’s return to our earlier
example and focus on Sarah’s choices of books and food. For Sarah (who went
to college in 2000), the cost of buying the base-year bundle of books and food
at cu
ent-year prices is $1720 (100 lbs. * $2.20/lb. + 15 books * $100
ook).
The cost of buying the same bundle at base-year prices is $500 (100 lbs *
$2/lb. + 15 books * $20
ook). The Laspeyres price index, LI, is therefore
100 * $1720/$500 = 344, as reported previously. In contrast, the cost of
uying the cu
ent-year bundle at cu
ent-year prices is $1260 (300 lbs. *
$2.20/lb. + 6 books * $100
ook). The cost of buying the same bundle at base-
year prices is $720 (300 lbs * $2/lb. + 6 books * $20
ook). Consequently, the
Paasche price index, PI, is 100 * $1260/$720 = 175. As expected, the Paasche
index is lower than the Laspeyres index and lower than the ideal index of 252.
Price Indexes in the united States: Chain Weighting
Historically, both the CPI and the PPI were measured as Laspeyres price in-
dexes. The overall CPI was calculated each month by the U.S. Bureau of Labor
Statistics as the ratio of the cost of a typical bundle of consumer goods and ser-
vices to the cost during a base period. A CPI for a particular category of goods
and services (e.g., housing) would utilize a bundle of goods and services from
that category. Similar calculations were done for the PPI using bundles of inter-
mediate and wholesale goods.
We have seen that the Laspeyres index overstates the amount needed to
compensate individuals for price increases. With respect to Social Security and
other government programs, this means that using the CPI with base weights
to adjust retirement benefits would tend to overcompensate most recipients and
would thus require greater government expenditure.
While economists have known of this problem for years, it was not until the
energy-price shocks of the 1970s, more recent fluctuations in food prices, and
concerns su
ounding federal deficits that dissatisfaction with the Laspeyres
index grew. It was estimated, for example, that a failure to account for changes
in computer-buying patterns in response to a sharp decrease in computer prices
had caused the CPI to overstate the cost of living substantially.
For this reason, the U.S. government changed the construction of the CPI and
the PPI, switching from a simple Laspeyres index to an index in which the base
weights are updated every few years. A chain-weighted price index is a cost-
of-living index that accounts for changes in quantities of goods and services
over time. Chain weighting was not new to the U.S. It had been adopted in 1995
as an improvement to the GDP deflator, a Paasche price index used to deflate
measures of gross domestic product (GDP) in order to obtain an estimate of real
GDP (GDP adjusted for inflation).13 Using chain-weighted versions of the CPI,
PPI, and GDP deflator has reduced the biases associated with the use of simple
Laspeyres and Paasche indexes, but because the weights are changed only in-
frequently, the biases have not been eliminated.14
chain-weighted price
index Cost-of-living index that
accounts for changes in quantities
of goods and services.
13For the latest changes in the CPI and PPI, see http:
www.bls.gov/cpi and http:
www.bls.gov
ppi. For information about the calculation of real GDP, see http:
www.bea.gov.
14Failures to account adequately for the appearance of new goods and improvements in the quality
of exisiting goods are additional sources of bias with respect to the CPI and PPI.
http:
www.bls.gov/cpi
http:
www.bls.gov/ppi
http:
www.bea.gov
http:
www.bls.gov/ppi
ChAPTer 3 CoNSUMer BehAvIor 127
The Bureau of Labor Statistics has continued to improve the measurement
of the CPI, which continues to rely on chain weighting. For example, in its 2015
elease of CPI statistics the BLS announced a na
ower and more focused meth-
odology for imputing prices for goods and services for which pricing informa-
tion is not available.
SUMMaRy
1. The theory of consumer choice rests on the assump-
tion that people behave rationally in an attempt to
maximize the satisfaction that they can obtain by
purchasing a particular combination of goods and
services.
2. Consumer choice has two related parts: the study of
the consumer’s preferences and the analysis of the
udget line that constrains consumer choices.
3. Consumers make choices by comparing market bas-
kets or bundles of commodities. Preferences are as-
sumed to be complete (consumers can compare all
possible market baskets) and transitive (if they prefer
asket A to B, and B to C, then they prefer A to C). In
addition, economists assume that more of each good is
always prefe
ed to less.
4. Indifference curves, which represent all combinations
of goods and services that give the same level of sat-
isfaction, are downward-sloping and cannot intersect
one another.
5. Consumer preferences can be completely described
y a set of indifference curves known as an indiffer-
ence map. An indifference map provides an ordinal
anking of all choices that the consumer might make.
6. The marginal rate of substitution (MRS) of F for C is
the maximum amount of C that a person is willing
to give up to obtain 1 additional unit of F. The MRS
diminishes as we move down along an indifference
curve. When there is a diminishing MRS, indifference
curves are convex.
7. Budget lines represent all combinations of goods for
which consumers expend all their income. Budget
lines shift outward in response to an increase in con-
sumer income. When the price of one good (on the
horizontal axis) changes while income and the price
of the other good do not, budget lines pivot and rotate
about a fixed point (on the vertical axis).
8. Consumers maximize satisfaction subject to budget
constraints. When a consumer maximizes satisfac-
tion by consuming some of each of two goods, the
marginal rate of substitution is equal to the ratio of the
prices of the two goods being purchased.
9. Maximization is sometimes achieved at a corner solu-
tion in which one good is not consumed. In such cases,
the marginal rate of substitution need not equal the
atio of the prices.
10. The theory of revealed preference shows how the
choices that individuals make when prices and in-
come vary can be used to determine their prefer-
ences. When an individual chooses basket A even
though he or she could afford B, we know that A is
prefe
ed to B.
11. The theory of the consumer can be presented by two
different approaches. The indifference curve approach
uses the ordinal properties of utility (that is, it allows
for the ranking of alternatives). The utility function
approach obtains a utility function by attaching a
number to each market basket; if basket A is prefe
ed
to basket B, A generates more utility than B.
12. When risky choices are analyzed or when compari-
sons must be made among individuals, the cardinal
properties of the utility function can be important.
Usually the utility function will show diminishing
marginal utility: As more and more of a good is con-
sumed, the consumer obtains smaller and smaller
increments of utility.
13. When the utility function approach is used and both
goods are consumed, utility maximization occurs
when the ratio of the marginal utilities of the two
goods (which is the marginal rate of substitution) is
equal to the ratio of the prices.
14. In times of war and other crises, governments some-
times ration food, gasoline, and other products,
ather than allow prices to increase to competi-
tive levels. Some consider nonprice rationing to be
more equitable than relying on uncontested market
forces.
15. An ideal cost-of-living index measures the cost of buy-
ing, at cu
ent prices, a bundle of goods that generates
the same level of utility as was provided by the bundle
of goods consumed at base-year prices. The Laspeyres
price index, however, represents the cost of buying
the bundle of goods chosen in the base year at cu
ent
prices relative to the cost of buying the same bundle at
ase-year prices. The CPI, even with chain weighting,
overstates the ideal cost-of-living index. By contrast,
the Paasche index measures the cost at cu
ent-year
prices of buying a bundle of goods chosen in the
cu
ent year divided by the cost of buying the same
undle at base-year prices. It thus understates the
ideal cost-of-living index.
128 PArT 2 Producers, Consumers, and Competitive Markets
QUeStionS foR Review
1. What are the four basic assumptions about individual
preferences? Explain the significance or meaning of
each.
2. Can a set of indifference curves be upward sloping? If
so, what would this tell you about the two goods?
3. Explain why two indifference curves cannot intersect.
4. Jon is always willing to trade one can of Coke for one
can of Sprite, or one can of Sprite for one can of Coke.
a. What can you say about Jon’s marginal rate of
substitution?
. Draw a set of indifference curves for Jon.
c. Draw two budget lines with different slopes and
illustrate the satisfaction-maximizing choice. What
conclusion can you draw?
5. What happens to the marginal rate of substitution as
you move along a convex indifference curve? A linear
indifference curve?
6. Explain why an MRS between two goods must equal
the ratio of the price of the goods for the consumer to
achieve maximum satisfaction.
7. Describe the indifference curves associated with two
goods that are perfect substitutes. What if they are
perfect complements?
8. What is the difference between ordinal utility and
cardinal utility? Explain why the assumption of car-
dinal utility is not needed in order to rank consumer
choices.
9. Upon merging with the West German economy,
East German consumers indicated a preference for
Mercedes-Benz automobiles over Volkswagens.
However, when they converted their savings into
deutsche marks, they flocked to Volkswagen dealer-
ships. How can you explain this apparent paradox?
10. Draw a budget line and then draw an indifference
curve to illustrate the satisfaction-maximizing choice
associated with two products. Use your graph to an-
swer the following questions.
a. Suppose that one of the products is rationed.
Explain why the consumer is likely to be worse off.
. Suppose that the price of one of the products is
fixed at a level below the cu
ent price. As a result,
the consumer is not able to purchase as much as
she would like. Can you tell if the consumer is bet-
ter off or worse off?
11. Describe the equal marginal principle. Explain why this
principle may not hold if increasing marginal utility is
associated with the consumption of one or both goods.
12. The price of computers has fallen substantially over
the past two decades. Use this drop in price to explain
why the Consumer Price Index is likely to overstate
substantially the cost-of-living index for individuals
who use computers intensively.
13. Explain why the Paasche index will generally under-
state the ideal cost-of-living index.
exeRciSeS
1. In this chapter, consumer preferences for various com-
modities did not change during the analysis. In some
situations, however, preferences do change as con-
sumption occurs. Discuss why and how preferences
might change over time with consumption of these
two commodities:
a. cigarettes.
. dinner for the first time at a restaurant with a
special cuisine.
2. Draw indifference curves that represent the follow-
ing individuals’ preferences for hamburgers and soft
drinks. Indicate the direction in which the individuals’
satisfaction (or utility) is increasing.
a. Joe has convex indifference curves and dislikes
oth hamburgers and soft drinks.
. Jane loves hamburgers and dislikes soft drinks. If
she is served a soft drink, she will pour it down the
drain rather than drink it.
c. Bob loves hamburgers and dislikes soft drinks. If he
is served a soft drink, he will drink it to be polite.
d. Molly loves hamburgers and soft drinks, but insists
on consuming exactly one soft drink for every two
hamburgers that she eats.
e. Bill likes hamburgers, but neither likes nor dislikes
soft drinks.
f. Mary always gets twice as much satisfaction from
an extra hamburger as she does from an extra soft
drink.
3. If Jane is cu
ently willing to trade 4 movie tickets for
1 basketball ticket, then she must like basketball better
than movies. True or false? Explain.
4. Janelle and Brian each plan to spend $20,000 on the
styling and gas mileage features of a new car. They
can each choose all styling, all gas mileage, or some
combination of the two. Janelle does not care at all
about styling and wants the best gas mileage possible.
Brian likes both equally and wants to spend an equal
amount on each. Using indifference curves and budget
lines, illustrate the choice that each person will make.
5. Suppose that Bridget and Erin spend their incomes
on two goods, food (F) and clothing (C). Bridget’s
preferences are represented by the utility function
U(F,C) = 10FC, while Erin’s preferences are repre-
sented by the utility function U(F,C) = .20F2C2.
a. With food on the horizontal axis and clothing on
the vertical axis, identify on a graph the set of
ChAPTer 3 CoNSUMer BehAvIor 129
points that give Bridget the same level of utility as
the bundle (10, 5). Do the same for Erin on a sepa-
ate graph.
. On the same two graphs, identify the set of bundles
that give Bridget and Erin the same level of utility
as the bundle (15, 8).
c. Do you think Bridget and Erin have the same pref-
erences or different preferences? Explain.
6. Suppose that Jones and Smith have each decided to
allocate $1000 per year to an entertainment budget
in the form of hockey games or rock concerts. They
oth like hockey games and rock concerts and will
choose to consume positive quantities of both goods.
However, they differ substantially in their preferences
for these two forms of entertainment. Jones prefers
hockey games to rock concerts, while Smith prefers
ock concerts to hockey games.
a. Draw a set of indifference curves for Jones and a
second set for Smith.
. Using the concept of marginal rate of substitution,
explain why the two sets of curves are different
from each other.
7. The price of DVDs (D) is $20 and the price of CDs (C)
is $10. Philip has a budget of $100 to spend on the two
goods. Suppose that he has already bought one DVD
and one CD. In addition, there are 3 more DVDs and 5
more CDs that he would really like to buy.
a. Given the above prices and income, draw his bud-
get line on a graph with CDs on the horizontal axis.
. Considering what he has already purchased and
what he still wants to purchase, identify the three
different bundles of CDs and DVDs that he could
choose. For this part of the question, assume that
he cannot purchase fractional units.
8. Anne has a job that requires her to travel three out of
every four weeks. She has an annual travel budget
and can travel either by train or by plane. The airline
on which she typically flies has a frequent-traveler
program that reduces the cost of her tickets according
to the number of miles she has flown in a given year.
When she reaches 25,000 miles, the airline will reduce
the price of her tickets by 25 percent for the remainder
of the year. When she reaches 50,000 miles, the airline
will reduce the price by 50 percent for the remainder
of the year. Graph Anne’s budget line, with train
miles on the vertical axis and plane miles on the hori-
zontal axis.
9. De
a usually buys a soft drink when she goes to a
movie theater, where she has a choice of three sizes:
the 8-ounce drink costs $1.50, the 12-ounce drink
$2.00, and the 16-ounce drink $2.25. Describe the bud-
get constraint that De
a faces when deciding how
many ounces of the drink to purchase. (Assume that
De
a can costlessly dispose of any of the soft drink
that she does not want.)
10. Antonio buys five new college textbooks during his
first year at school at a cost of $80 each. Used books
cost only $50 each. When the bookstore announces
that there will be a 10 percent increase in the price
of new books and a 5 percent increase in the price of
used books, Antonio’s father offers him $40 extra.
a. What happens to Antonio’s budget line? Illustrate
the change with new books on the vertical axis.
. Is Antonio worse or better off after the price
change? Explain.
11. Consumers in Georgia pay twice as much for avoca-
dos as they do for peaches. However, avocados and
peaches are the same price in California. If consumers
in both states maximize utility, will the marginal rate
of substitution of peaches for avocados be the same for
consumers in both states? If not, which will be higher?
12. Ben allocates his lunch budget between two goods,
pizza and bu
itos.
a. Illustrate Ben’s optimal bundle on a graph with
pizza on the horizontal axis.
. Suppose now that pizza is taxed, causing the price
to increase by 20 percent. Illustrate Ben’s new opti-
mal bundle.
c. Suppose instead that pizza is rationed at a quantity
less than Ben’s desired quantity. Illustrate Ben’s
new optimal bundle.
13. Brenda wants to buy a new car and has a budget of
$25,000. She has just found a magazine that assigns
each car an index for styling and an index for gas mile-
age. Each index runs from 1 to 10, with 10 representing
either the most styling or the best gas mileage. While
looking at the list of cars, Brenda observes that on
average, as the style index increases by one unit, the
price of the car increases by $5000. She also observes
that as the gas-mileage index rises by one unit, the
price of the car increases by $2500.
a. Illustrate the various combinations of style (S) and
gas mileage (G) that Brenda could select with her
$25,000 budget. Place gas mileage on the horizontal
axis.
. Suppose Brenda’s preferences are such that she al-
ways receives three times as much satisfaction from
an extra unit of styling as she does from gas mile-
age. What type of car will Brenda choose?
c. Suppose that Brenda’s marginal rate of substitu-
tion (of gas mileage for styling) is equal to S/(4G).
What value of each index would she like to have
in her car?
d. Suppose that Brenda’s marginal rate of substitution
(of gas mileage for styling) is equal to (3S)/G. What
value of each index would she like to have in her
car?
14. Connie has a monthly income of $200 that she allo-
cates between two goods: meat and potatoes.
a. Suppose meat costs $4 per pound and potatoes $2
per pound. Draw her budget constraint.
. Suppose also that her utility function is given
y the equation U(M, P) = 2M + P. What com-
ination of meat and potatoes should she buy to
130 PArT 2 Producers, Consumers, and Competitive Markets
maximize her utility? (Hint: Meat and potatoes are
perfect substitutes.)
c. Connie’s supermarket has a special promotion. If she
uys 20 pounds of potatoes (at $2 per pound), she gets
the next 10 pounds for free. This offer applies only
to the first 20 pounds she buys. All potatoes in excess
of the first 20 pounds (excluding bonus potatoes) are
still $2 per pound. Draw her budget constraint.
d. An out
eak of potato rot raises the price of po-
tatoes to $4 per pound. The supermarket ends its
promotion. What does her budget constraint look
like now? What combination of meat and potatoes
maximizes her utility?
15. Jane receives utility from days spent traveling on va-
cation domestically (D) and days spent traveling on
vacation in a foreign country (F), as given by the util-
ity function U(D,F) = 10DF. In addition, the price of a
day spent traveling domestically is $100, the price of
a day spent traveling in a foreign country is $400, and
Jane’s annual travel budget is $4000.
a. Illustrate the indifference curve associated with a
utility of 800 and the indifference curve associated
with a utility of 1200.
. Graph Jane’s budget line on the same graph.
c. Can Jane afford any of the bundles that give her a
utility of 800? What about a utility of 1200?
*d. Find Jane’s utility-maximizing choice of days spent
traveling domestically and days spent in a foreign
country.
16. Julio receives utility from consuming food (F)
and clothing (C) as given by the utility function
U(F,C) = FC.
In addition, the price of food is $2 per unit, the price
of clothing is $10 per unit, and Julio’s weekly income
is $50.
a. What is Julio’s marginal rate of substitution of food
for clothing when utility is maximized? Explain.
. Suppose instead that Julio is consuming a bundle
with more food and less clothing than his utility
maximizing bundle. Would his marginal rate of
substitution of food for clothing be greater than or
less than your answer in part a? Explain.
17. The utility that Meredith receives by consuming food
F and clothing C is given by U(F,C) = FC. Suppose
that Meredith’s income in 1990 is $1200 and that the
prices of food and clothing are $1 per unit for each.
By 2000, however, the price of food has increased to
$2 and the price of clothing to $3. Let 100 represent
the cost of living index for 1990. Calculate the ideal
and the Laspeyres cost-of-living index for Meredith
for 2000. (Hint: Meredith will spend equal amounts on
food and clothing with these preferences.)
131
Chapter 3 laid the foundation for the theory of consumer de-mand. We discussed the nature of consumer preferences and saw how, given budget constraints, consumers choose market
askets that maximize utility. From here it’s a short step to analyz-
ing demand and showing how the demand for a good depends on its
price, the prices of other goods, and income.
Our analysis of demand proceeds in six steps:
1. We begin by deriving the demand curve for an individual con-
sumer. Because we know how changes in price and income af-
fect a person’s budget line, we can determine how they affect
consumption choice. We will use this information to see how the
quantity of a good demanded varies in response to price changes
as we move along an individual’s demand curve. We will also
see how this demand curve shifts in response to changes in the
individual’s income.
2. With this foundation, we will examine the effect of a price change
in more detail. When the price of a good goes up, individual de-
mand for it can change in two ways. First, because it has now
ecome more expensive relative to other goods, consumers will
uy less of it and more of other goods. Second, the higher price
educes the consumer’s purchasing power. This reduction is just
like a reduction in income and will lead to a reduction in con-
sumer demand. By analyzing these two distinct effects, we will
etter understand the characteristics of demand.
3. Next, we will see how individual demand curves can be aggre-
gated to determine the market demand curve. We will also study
the characteristics of market demand and see why the demands
for some kinds of goods differ considerably from the demands
for others.
4. We will go on to show how market demand curves can be used
to measure the benefits that people receive when they consume
products, above and beyond the expenditures they make. This
information will be especially important later, when we study
the effects of government intervention in a market.
5. We then describe the effects of network externalities—i.e., what
happens when a person’s demand for a good also depends on
4.1 Consumer Expenditures
in the United States 137
4.2 The Effects of a
Gasoline Tax 142
4.3 The Aggregate Demand
for Wheat 148
4.4 The Demand for Housing 149
4.5 The Long-Run Demand for
Gasoline 151
4.6 The Value of Clean Air 154
4.7 Facebook versus
Google Plus 158
4.8 The Demand for
Ready-to-Eat Cereal 162
LiST oF ExAmPLES
4.1 individual Demand 132
4.2 income and Substitution
Effects 139
4.3 market Demand 144
4.4 Consumer Surplus 152
4.5 Network Externalities 155
*4.6 Empirical Estimation of
Demand 159
Appendix: Demand
Theory—A mathematical
Treatment 169
CHAPTER oUTLiNE
individual and market
Demand
CHAPTER 4
132 PART 2 Producers, Consumers, and Competitive markets
the demands of other people. These effects play a crucial role in the de-
mands for many high-tech products, such as computer hardware and soft-
ware, and telecommunications systems.
6. Finally, we will
iefly describe some of the methods that economists use
to obtain empirical information about demand.
4.1 Individual Demand
This section shows how the demand curve of an individual consumer follows
from the consumption choices that a person makes when faced with a budget
constraint. To illustrate these concepts graphically, we will limit the available
goods to food and clothing, and we will rely on the utility-maximization ap-
proach described in Section 3.3 (page 108).
Price Changes
We begin by examining ways in which the consumption of food and clothing
changes when the price of food changes. Figure 4.1 shows the consumption
choices that a person will make when allocating a fixed amount of income be-
tween the two goods.
Initially, the price of food is $1, the price of clothing $2, and the consum-
er’s income $20. The utility-maximizing consumption choice is at point B in
Figure 4.1 (a). Here, the consumer buys 12 units of food and 4 units of clothing,
thus achieving the level of utility associated with indifference curve U2.
Now look at Figure 4.1 (b), which shows the relationship between the price
of food and the quantity demanded. The horizontal axis measures the quantity
of food consumed, as in Figure 4.1 (a), but the vertical axis now measures the
price of food. Point G in Figure 4.1 (b) co
esponds to point B in Figure 4.1 (a).
At G, the price of food is $1, and the consumer purchases 12 units of food.
Suppose the price of food increases to $2. As we saw in Chapter 3, the budget
line in Figure 4.1 (a) rotates inward about the vertical intercept, becoming twice
as steep as before. The higher relative price of food has increased the magnitude
of the slope of the budget line. The consumer now achieves maximum utility at
A, which is found on a lower indifference curve, U1. Because the price of food
has risen, the consumer’s purchasing power—and thus attainable utility—has
fallen. At A, the consumer chooses 4 units of food and 6 units of clothing. In
Figure 4.1 (b), this modified consumption choice is at E, which shows that at a
price of $2, 4 units of food are demanded.
Finally, what will happen if the price of food decreases to 50 cents? Because
the budget line now rotates outward, the consumer can achieve the higher level
of utility associated with indifference curve U3 in Figure 4.1 (a) by selecting D,
with 20 units of food and 5 units of clothing. Point H in Figure 4.1 (b) shows the
price of 50 cents and the quantity demanded of 20 units of food.
The Individual Demand Curve
We can go on to include all possible changes in the price of food. In Figure 4.1 (a),
the price-consumption curve traces the utility-maximizing combinations of
food and clothing associated with every possible price of food. Note that as the
price of food falls, attainable utility increases and the consumer buys more food.
In §3.3, we explain how
a consumer chooses
the market basket on the
highest indifference curve
that touches the consumer’s
udget line.
In §3.2, we explain how
the budget line shifts in
esponse to a price change.
price-consumption
curve Curve tracing the utility-
maximizing combinations of two
goods as the price of one changes.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 133
This pattern of increasing consumption of a good in response to a decrease in
price almost always holds. But what happens to the consumption of clothing as
the price of food falls? As Figure 4.1 (a) shows, the consumption of clothing
may either increase or decrease. The consumption of both food and clothing can
increase because the decrease in the price of food has increased the consumer’s
ability to purchase both goods.
An individual demand curve relates the quantity of a good that a single
consumer will buy to the price of that good. In Figure 4.1 (b), the individual de-
mand curve relates the quantity of food that the consumer will buy to the price
of food. This demand curve has two important properties:
1. The level of utility that can be attained changes as we move along the
curve. The lower the price of the product, the higher the level of utility.
Note from Figure 4.1 (a) that a higher indifference curve is reached as the
price falls. Again, this result simply reflects the fact that as the price of a
product falls, the consumer’s purchasing power increases.
individual demand
curve Curve relating the
quantity of a good that a single
consumer will buy to its price.
4
6 Price-Consumption Curve
5
204 12
(a)
(b)
1.00
$2.00
Price
of food
Food (units
per month)
Food (units
per month)
Clothing
(units pe
month)
20
Demand Curve
1.50
124
0.50
A
B
D
E
G
H
U1
U2
U3
FIgure 4.1
EffEct of PricE changEs
A reduction in the price of food, with income and the
price of clothing fixed, causes this consumer to choose
a different market basket. In (a), the baskets that maxi-
mize utility for various prices of food (point A, $2; B, $1;
D, $0.50) trace out the price-consumption curve. Part
(b) gives the demand curve, which relates the price of
food to the quantity demanded. (Points E, G, and H
co
espond to points A, B, and D, respectively).
134 PART 2 Producers, Consumers, and Competitive markets
2. At every point on the demand curve, the consumer is maximizing utility
y satisfying the condition that the marginal rate of substitution (MRS)
of food for clothing equals the ratio of the prices of food and clothing. As
the price of food falls, the price ratio and the MRS also fall. In Figure 4.1
(b), the price ratio falls from 1 ($2/$2) at E (because the curve U1 is tan-
gent to a budget line with a slope of -1 at A) to 1/2 ($1/$2) at G, to 1/4
($0.50/$2) at H. Because the consumer is maximizing utility, the MRS of
food for clothing decreases as we move down the demand curve. This
phenomenon makes intuitive sense because it tells us that the relative
value of food falls as the consumer buys more of it.
The fact that the MRS varies along the individual’s demand curve tells us
something about how consumers value the consumption of a good or service.
Suppose we were to ask a consumer how much she would be willing to pay
for an additional unit of food when she is cu
ently consuming 4 units. Point
E on the demand curve in Figure 4.1 (b) provides the answer: $2. Why? As we
pointed out above, because the MRS of food for clothing is 1 at E, one additional
unit of food is worth one additional unit of clothing. But a unit of clothing costs
$2, which is, therefore, the value (or marginal benefit) obtained by consum-
ing an additional unit of food. Thus, as we move down the demand curve in
Figure 4.1 (b), the MRS falls. Likewise, the value that the consumer places on an
additional unit of food falls from $2 to $1 to $0.50.
Income Changes
We have seen what happens to the consumption of food and clothing when the
price of food changes. Now let’s see what happens when income changes.
The effects of a change in income can be analyzed in much the same way as a
price change. Figure 4.2 (a) shows the consumption choices that a consumer will
make when allocating a fixed income to food and clothing when the price of food
is $1 and the price of clothing $2. As in Figure 4.1 (a), the quantity of clothing is
measured on the vertical axis and the quantity of food on the horizontal axis.
Income changes appear as changes in the budget line in Figure 4.2 (a). Initially,
the consumer’s income is $10. The utility-maximizing consumption choice is
then at A, at which point she buys 4 units of food and 3 units of clothing.
This choice of 4 units of food is also shown in Figure 4.2 (b) as E on demand
curve D1. Demand curve D1 is the curve that would be traced out if we held
income fixed at $10 but varied the price of food. Because we are holding the price
of food constant, we will observe only a single point E on this demand curve.
What happens if the consumer’s income is increased to $20? Her budget line
then shifts outward parallel to the original budget line, allowing her to attain
the utility level associated with indifference curve U2. Her optimal consump-
tion choice is now at B, where she buys 10 units of food and 5 units of clothing.
In Figure 4.2 (b) her consumption of food is shown as G on demand curve D2.
D2 is the demand curve that would be traced out if we held income fixed at $20
ut varied the price of food. Finally, note that if her income increases to $30,
she chooses D, with a market basket containing 16 units of food (and 7 units of
clothing), represented by H in Figure 4.2 (b).
We could go on to include all possible changes in income. In Figure 4.2 (a),
the income-consumption curve traces out the utility-maximizing combinations
of food and clothing associated with every income level. The income-consump-
tion curve in Figure 4.2 slopes upward because the consumption of both food
and clothing increases as income increases. Previously, we saw that a change in
the price of a good co
esponds to a movement along a demand curve. Here, the
In §3.1, we introduce the
marginal rate of substitution
(MRS) as a measure of the
maximum amount of one
good that the consumer is
willing to give up in order to
obtain one unit of another
good.
income-consumption
curve Curve tracing the utility-
maximizing combinations of two
goods as a consumer’s income
changes.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 135
situation is different. Because each demand curve is measured for a particular
level of income, any change in income must lead to a shift in the demand curve
itself. Thus A on the income-consumption curve in Figure 4.2 (a) co
esponds
to E on demand curve D1 in Figure 4.2 (b); B co
esponds to G on a different
demand curve D2. The upward-sloping income-consumption curve implies that
an increase in income causes a shift to the right in the demand curve—in this
case from D1 to D2 to D3.
Normal versus Inferior goods
When the income-consumption curve has a positive slope, the quantity de-
manded increases with income. As a result, the income elasticity of demand is
positive. The greater the shifts to the right of the demand curve, the larger the
income elasticity. In this case, the goods are described as normal: Consumers
want to buy more of them as their incomes increase.
In some cases, the quantity demanded falls as income increases; the income
elasticity of demand is negative. We then describe the good as inferior. The term
inferior simply means that consumption falls when income rises. Hamburger,
for example, is inferior for some people: As their income increases, they buy less
hamburger and more steak.
In §2.4, we explain that the
income elasticity of demand
is the percentage change
in the quantity demanded
esulting from a 1-percent
increase in income.
U1
U2
U3
D
Price
of
food
Clothing
(units pe
month)
B
A
E
D1
D2
D3
Food (units
per month)
7
$1.00
4
(a)
10 16
4 10 16
(b)
Food (units
per month)
G H
3
Income-Consumption
Curve
5
FIgure 4.2
EffEct of incomE changEs
An increase in income, with the prices of all goods fixed, causes
consumers to alter their choice of market baskets. In part (a), the
askets that maximize consumer satisfaction for various incomes
(point A, $10; B, $20; D, $30) trace out the income-consumption
curve. The shift to the right of the demand curve in response to
the increases in income is shown in part (b). (Points E, G, and H
co
espond to points A, B, and D, respectively.)
136 PART 2 Producers, Consumers, and Competitive markets
Figure 4.3 shows the income-consumption curve for an inferior good. For
elatively low levels of income, both hamburger and steak are normal goods.
As income rises, however, the income-consumption curve bends backward
(from point B to C). This shift occurs because hamburger has become an inferior
good—its consumption has fallen as income has increased.
engel Curves
Income-consumption curves can be used to construct Engel curves, which
elate the quantity of a good consumed to an individual’s income. Figure 4.4
shows how such curves are constructed for two different goods. Figure 4.4 (a),
Engel curve Curve relating the
quantity of a good consumed to
income.
Income
(dollars pe
month)
Food (units
per month)
0 4 8 12 16
30
20
10
(a)
Engel Curve
Income
(dollars pe
month)
Hamburger (units
per month)
0 10
30
20
10
5
(b)
Inferio
Normal
FIgure 4.4
EngEl curvEs
Engel curves relate the quantity of a good consumed to income. In (a), food is a normal good and
the Engel curve is upward sloping. In (b), however, hamburger is a normal good for income less
than $20 per month and an inferior good for income greater than $20 per month.
U1
U2
U3
C
Steak
(units pe
month)
B
A
Hamburge
(units per month)
10 20 30
15
10
5
5
Income-Consumption
Curve
FIgure 4.3
an infErior good
An increase in a person’s income can
lead to less consumption of one of the
two goods being purchased. Here, ham-
urger, though a normal good between
A and B, becomes an inferior good when
the income-consumption curve bends
ackward between B and C.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 137
which  shows an upward-sloping Engel curve, is derived directly from
Figure 4.2 (a). In both figures, as the individual’s income increases from $10 to
$20 to $30, her consumption of food increases from 4 to 10 to 16 units. Recall
that in Figure 4.2 (a) the vertical axis measured units of clothing consumed per
month and the horizontal axis units of food per month; changes in income were
eflected as shifts in the budget line. In Figures 4.4 (a) and (b), we have replotted
the data to put income on the vertical axis, while keeping food and hamburger
on the horizontal.
The upward-sloping Engel curve in Figure 4.4 (a)—like the upward-sloping
income-consumption curve in Figure 4.2 (a)—applies to all normal goods. Note
that an Engel curve for clothing would have a similar shape (clothing consump-
tion increases from 3 to 5 to 7 units as income increases).
Figure 4.4 (b), derived from Figure 4.3, shows the Engel curve for hamburger.
We see that hamburger consumption increases from 5 to 10 units as income in-
creases from $10 to $20. As income increases further, from $20 to $30, consump-
tion falls to 8 units. The portion of the Engel curve that slopes downward is the
income range within which hamburger is an inferior good.
examPle 4.1 consumEr ExPEnditurEs in thE unitEd statEs
The Engel curves we just exam-
ined apply to individual consum-
ers. However, we can also derive
Engel curves for groups of con-
sumers. This information is partic-
ularly useful if we want to see how
consumer spending varies among
different income groups. Table 4.1
illustrates spending patterns for
several items taken from a survey by the U.S. Bureau
of Labor Statistics. Although the data are averaged
over many households, they can be interpreted as
describing the expenditures of a
typical family.
Note that the data relate expendi-
tures on a particular item rather than
the quantity of the item to income.
The first two items, entertainment
and owned dwellings, are consump-
tion goods for which the income
elasticity of demand is high. Average
family expenditures on entertainment increase almost
fivefold when we move from the lowest to highest in-
come group. The same pattern applies to the purchase
Table 4.1 annual u.s. housEhold consumEr ExPEnditurEs
IncoMe GRoup
expendItuReS
($) on:
LeSS than
$10,000
10,000–
19,999
20,000–
29,999
30,000–
39,999
40,000–
49,999
50,000–
69,999
70,000 and
above
entertainment 1,038 1,165 1,407 1,969 2,131 2,548 4,655
owned dwelling 1,770 2,134 2,795 3,581 4,198 5,556 11,606
Rented dwelling 3,919 3,657 4,054 3,878 4,273 3,812 3,072
health care 1,434 2,319 3,124 3,539 3,709 4,702 6,417
Food 3,466 3,706 4,432 5,194 5,936 6,486 10,116
clothing 798 766 960 1,321 1,518 1,602 2,928
Source: u.S. department of Labor, bureau of Labor Statistics, “consumer expenditure Survey, annual Report 2015.”
138 PART 2 Producers, Consumers, and Competitive markets
Substitutes and Complements
The demand curves that we graphed in Chapter 2 showed the relationship
etween the price of a good and the quantity demanded, with preferences,
income, and the prices of all other goods held constant. For many goods, de-
mand is related to the consumption and prices of other goods. Baseball bats
and baseballs, hot dogs and mustard, and computer hardware and software are
all examples of goods that tend to be used together. Other goods, such as cola
and diet cola, owner-occupied houses and rental apartments, movie tickets and
movie rentals, tend to substitute for one another.
Recall from Section 2.1 (page 44) that two goods are substitutes if an increase
in the price of one leads to an increase in the quantity demanded of the other.
FIgure 4.5
EngEl curvEs
for u.s.
consumErs
Average per-household
expenditures on rented
dwellings, health care,
and entertainment are
plotted as functions of
annual income. Health
care and entertainment
are normal goods, as ex-
penditures increase with
income. Rental housing,
however, is an inferior
good for incomes above
$40,000.
$0
$10,000
$20,000
$30,000
$40,000
$50,000
$60,000
$70,000
$80,000
$0 $1,000 $2,000 $3,000 $4,000 $5,000 $6,000 $7,000
A
nn
ua
l I
nc
om
e
Annual Expenditure
Entertainment Rented Dwelling Health Care
of homes: There is a more than a sixfold increase in
expenditures from the lowest to the highest category.
in contrast, expenditures on rental housing ac-
tually fall as income rises. This pattern reflects the
fact that most higher-income individuals own rather
than rent homes. Thus rental housing is an inferior
good, at least for incomes above $40,000 per year.
Finally, note that health care, food, and cloth-
ing are consumption items for which the income
elasticities are positive, but not as high as for enter-
tainment or owner-occupied housing.
The data in Table 4.1 for rented dwellings, health
care, and entertainment have been plotted in
Figure  4.5. observe in the three Engel curves that
as income rises, expenditures on entertainment and
health care increase rapidly, while expenditures on
ental housing increase when income is low, but de-
crease once income exceeds $40,000.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 139
If  the price of a movie ticket rises, we would expect individuals to rent more
movies, because movie tickets and movie rentals are substitutes. Similarly, two
goods are complements if an increase in the price of one good leads to a decrease
in the quantity demanded of the other. If the price of gasoline goes up, causing
gasoline consumption to fall, we would expect the consumption of motor oil to
fall as well, because gasoline and motor oil are used together. Two goods are
independent if a change in the price of one good has no effect on the quantity
demanded of the other.
One way to see whether two goods are complements or substitutes is to
examine the price-consumption curve. Look again at Figure 4.1 (page 133). Note
that in the downward-sloping portion of the price-consumption curve, food
and clothing are substitutes: The lower price of food leads to a lower consump-
tion of clothing (perhaps because as food expenditures increase, less income is
available to spend on clothing). Similarly, food and clothing are complements
in the upward-sloping portion of the curve: The lower price of food leads to
higher clothing consumption (perhaps because the consumer eats more meals
at restaurants and must be suitably dressed).
The fact that goods can be complements or substitutes suggests that when
studying the effects of price changes in one market, it may be important to look
at the consequences in related markets. (Inte
elationships among markets are
discussed in more detail in Chapter 16.) Determining whether two goods are
complements, substitutes, or independent goods is ultimately an empirical
question. To answer the question, we need to look at the ways in which the de-
mand for the first good shifts (if at all) in response to a change in the price of the
second. This question is more difficult than it sounds because lots of things are
likely to be changing at the same time that the price of the first good changes.
In fact, Section 4.6 of this chapter is devoted to examining ways to distinguish
empirically among the many possible explanations for a change in the demand
for the second good. First, however, it will be useful to undertake a basic theo-
etical exercise. In the next section, we delve into the ways in which a change in
the price of a good can affect consumer demand.
4.2 Income and Substitution effects
A fall in the price of a good has two effects:
1. Consumers will tend to buy more of the good that has become cheaper
and less of those goods that are now relatively more expensive. This re-
sponse to a change in the relative prices of goods is called the substitution
effect.
2. Because one of the goods is now cheaper, consumers enjoy an increase in
eal purchasing power. They are better off because they can buy the same
amount of the good for less money, and thus have money left over for ad-
ditional purchases. The change in demand resulting from this change in
eal purchasing power is called the income effect.
Normally, these two effects occur simultaneously, but it will be useful to
distinguish between them for purposes of analysis. The specifics are illustrated
in Figure 4.6, where the initial budget line is RS and there are two goods, food
and clothing. Here, the consumer maximizes utility by choosing the market
asket at A, thereby obtaining the level of utility associated with the indiffer-
ence curve U1.
140 PART 2 Producers, Consumers, and Competitive markets
Now let’s see what happens if the price of food falls, causing the budget line
to rotate outward to line RT. The consumer now chooses the market basket at
B on indifference curve U2. Because market basket B was chosen even though
market basket A was feasible, we know (from our discussion of revealed prefer-
ence in Section 3.4) that B is prefe
ed to A. Thus, the reduction in the price of
food allows the consumer to increase her level of satisfaction—her purchasing
power has increased. The total change in the consumption of food caused by
the lower price is given by F1F2. Initially, the consumer purchased OF1 units of
food, but after the price change, food consumption has increased to OF2. Line
segment F1F2, therefore, represents the increase in desired food purchases.
Substitution effect
The drop in price has both a substitution effect and an income effect. The
substitution effect is the change in food consumption associated with a change in the
price of food, with the level of utility held constant. The substitution effect captures
the change in food consumption that occurs as a result of the price change that
makes food relatively cheaper than clothing. This substitution is marked by a
movement along an indifference curve. In Figure 4.6, the substitution effect can
e obtained by drawing a budget line which is parallel to the new budget line
RT (reflecting the lower relative price of food), but which is just tangent to the
original indifference curve U1 (holding the level of satisfaction constant). The
new, lower imaginary budget line reflects the fact that nominal income was
educed in order to accomplish our conceptual goal of isolating the substitu-
tion effect. Given that budget line, the consumer chooses market basket D and
consumes OE units of food. The line segment F1E thus represents the substitu-
tion effect.
Figure 4.6 makes it clear that when the price of food declines, the substitu-
tion effect always leads to an increase in the quantity of food demanded. The
In §3.4, we show how infor-
mation about consumer
preferences is revealed by
consumption choices made.
substitution effect Change
in consumption of a good
associated with a change in its
price, with the level of utility held
constant.
Total E�ect
U1
U2
Clothing
(units pe
month)
Food
(units pe
month)
C1
C2
A
D
B
F2 TEF1O
Substitution
E�ect
Income
E�ect
S
R
FIgure 4.6
incomE and substitution
EffEcts: normal good
A decrease in the price of food has both an
income effect and a substitution effect. The
consumer is initially at A, on budget line RS.
When the price of food falls, consumption in-
creases by F1F2 as the consumer moves to B.
The substitution effect F1E (associated with a
move from A to D) changes the relative prices
of food and clothing but keeps real income
(satisfaction) constant. The income effect EF2
(associated with a move from D to B) keeps
elative prices constant but increases purchas-
ing power. Food is a normal good because the
income effect EF2 is positive.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 141
explanation lies in the fourth assumption about consumer preferences dis-
cussed in Section 3.1—namely, that indifference curves are convex. Thus, with
the convex indifference curves shown in the figure, the point that maximizes
satisfaction on the new imaginary budget line parallel to RT must lie below and
to the right of the original point of tangency.
Income effect
Now let’s consider the income effect: the change in food consumption
ought
about by the increase in purchasing power, with relative prices held constant. In
Figure 4.6, we can see the income effect by moving from the imaginary budget
line that passes through point D to the parallel budget line, RT, which passes
through B. The consumer chooses market basket B on indifference curve
U2 (because the lower price of food has increased her level of utility). The
increase in food consumption from OE to OF2 is the measure of the income
effect, which is positive, because food is a normal good (consumers will buy
more of it as their incomes increase). Because it reflects a movement from one
indifference curve to another, the income effect measures the change in the
consumer’s purchasing power.
We have seen in Figure 4.6 that the total effect of a change in price is given
theoretically by the sum of the substitution effect and the income effect:
Total Effect (F1F2) = Substitution Effect (F1E) + Income Effect (EF2)
Recall that the direction of the substitution effect is always the same: A de-
cline in price leads to an increase in consumption of the good. However, the
income effect can move demand in either direction, depending on whether the
good is normal or inferior.
A good is inferior when the income effect is negative: As income rises, con-
sumption falls. Figure 4.7 shows income and substitution effects for an inferior
good. The negative income effect is measured by line segment EF2. Even with
income effect Change in
consumption of a good resulting
from an increase in purchasing
power, with relative prices held
constant.
inferior good A good that has
a negative income effect.
Total E�ect
U1
U2
B
Clothing
(units pe
month)
Food
(units pe
month)
D
A
F2 EF1O
Substitution
E�ect IncomeE�ect
T
R
S
FIgure 4.7
incomE and substitution
EffEcts: infErior good
The consumer is initially at A on budget line
RS. With a decrease in the price of food, the
consumer moves to B. The resulting change in
food purchased can be
oken down into a sub-
stitution effect, F1E (associated with a move from
A to D), and an income effect, EF2 (associated
with a move from D to B). In this case, food is an
inferior good because the income effect is nega-
tive. However, because the substitution effect
exceeds the income effect, the decrease in the
price of food leads to an increase in the quantity
of food demanded.
142 PART 2 Producers, Consumers, and Competitive markets
inferior goods, the income effect is rarely large enough to outweigh the substi-
tution effect. As a result, when the price of an inferior good falls, its consump-
tion almost always increases.
a Special Case: The giffen good
Theoretically, the income effect may be large enough to cause the demand curve
for a good to slope upward. We call such a good a Giffen good, and Figure 4.8
shows its income and substitution effects. Initially, the consumer is at A, con-
suming relatively little clothing and much food. Now the price of food declines.
The decline in the price of food frees enough income so that the consumer de-
sires to buy more clothing and fewer units of food, as illustrated by B. Revealed
preference tells us that the consumer is better off at B rather than A even though
less food is consumed.
Though intriguing, the Giffen good is rarely of practical interest because it
equires a large negative income effect. But the income effect is usually small:
Individually, most goods account for only a small part of a consumer’s budget.
Large income effects are often associated with normal rather than inferior goods
(e.g., total spending on food or housing).
Giffen good Good whose
demand curve slopes upward
ecause the (negative) income
effect is larger than the
substitution effect.
Clothing
(units pe
month)
EF1O F2
U2
U1
Food (units
per month)Substitution E�ect
Income E�ect
Total E�ect
B
A
D
FIgure 4.8
uPward-sloPing dEmand
curvE: thE giffEn good
When food is an inferior good, and when
the income effect is large enough to domi-
nate the substitution effect, the demand
curve will be upward-sloping. The consumer
is initially at point A, but, after the price
of food falls, moves to B and consumes
less food. Because the income effect EF2
is larger than the substitution effect F1E,
the decrease in the price of food leads to a
lower quantity of food demanded.
examPle 4.2 thE EffEcts of a gasolinE tax
in part to conserve energy and in part to raise rev-
enues, the U.S. government has often considered
increasing the federal gasoline tax. in 1993, for
example, a modest 4.3 cent increase was enacted
as part of a larger budget-reform package. This
increase was much less than the increase that
would have been necessary to put U.S. gasoline
prices on a par with those in Europe. Because an
important goal of higher gasoline taxes is to dis-
courage gasoline consumption, the government
has also considered ways of passing the resulting
income back to consumers. one popular sugges-
tion is a rebate program in which tax revenues
would be returned to households on an equal per-
capita basis. What would be the effect of such a
program?
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 143
Let’s begin by focusing on the effect of the pro-
gram over a period of five years. The relevant price
elasticity of demand is about -0.5.1 Suppose that a
low-income consumer uses about 1200 gallons of
gasoline per year, that gasoline costs $1 per gallon,
and that our consumer’s annual income is $9000.
Figure 4.9 shows the effect of the gasoline tax.
(The graph has intentionally been drawn not to
scale so that the effects we are discussing can be
seen more clearly.) The original budget line is AB,
and the consumer maximizes utility (on indiffer-
ence curve U2) by consuming the market basket at
C, buying 1200 gallons of gasoline and spending
$7800 on other goods. if the tax is 50 cents per
gallon, price will increase by 50 percent, shifting
the new budget line to AD.2 (Recall that when price
changes and income stays fixed, the budget line
otates around a pivot point on the unchanged axis.)
With a price elasticity of -0.5, consumption will
decline 25 percent, from 1200 to 900 gallons, as
shown by the utility-maximizing point E on indiffer-
ence curve U1 (for every 1-percent increase in the
price of gasoline, quantity demanded drops by 1/2
percent).
The rebate program, however, partially counters
this effect. Suppose that because the tax revenue
per person is about $450(900 gallons times 50 cents
per gallon), each consumer receives a $450 rebate.
1We saw in Chapter 2 that the price elasticity of demand for gasoline varies substantially from the
short run to the long run.
2To simplify the example, we have assumed that the entire tax is paid by consumers in the form of a
higher price. A
oader analysis of tax shifting is presented in Chapter 9.
U1
U2
Original Budget
Line
C
H
E
After Gasoline Tax
Plus Rebate
Afte
Gasoline
Tax
Expenditures
on othe
goods ($)
F
A
D
Gasoline consumption (gallons per year)
J B1200913.5900
FIgure 4.9
EffEct of a gasolinE tax with a rEbatE
A gasoline tax is imposed when the consumer is initially buying 1200 gallons of gasoline at
point C. After the tax takes effect, the budget line shifts from AB to AD and the consumer
maximizes his preferences by choosing E, with a gasoline consumption of 900 gallons.
However, when the proceeds of the tax are rebated to the consumer, his consumption
increases somewhat, to 913.5 gallons at H. Despite the rebate program, the consumer’s
gasoline consumption has fallen, as has his level of satisfaction.
144 PART 2 Producers, Consumers, and Competitive markets
How does this increased income affect gasoline
consumption? The effect can be shown graphically
y shifting the budget line upward by $450, to line
FJ, which is parallel to AD. How much gasoline
does our consumer buy now? in Chapter 2, we saw
that the income elasticity of demand for gasoline
is approximately 0.3. Because $450 represents a
5- percent increase in income ($450/$9000 = 0.05),
we would expect the rebate to increase consumption
y 1.5 percent (0.3 times 5 percent) of 900 gallons,
or 13.5 gallons. The new utility-maximizing con-
sumption choice at H reflects this expectation. (We
omitted the indifference curve that is tangent at H to
simplify the diagram.) With the rebate program, the
tax would reduce gasoline consumption by 286.5
gallons, from 1200 to 913.5. Because the income
elasticity of demand for gasoline is relatively low, the
income effect of the rebate program is dominated by
the substitution effect, and the program with a rebate
does indeed reduce consumption.
in order to put a real tax-rebate program into
effect, Congress would have to solve a variety of
practical problems. First, incoming tax receipts and
ebate expenditures would vary from year to year,
making it difficult to plan the budgeting process. For
example, the tax rebate of $450 in the first year of
the program is an increase in income. During the
second year, it would lead to some increase in gaso-
line consumption among the low-income consumers
that we are studying. With increased consumption,
however, the tax paid and the rebate received by
an individual will increase in the second year. As
a result, it may be difficult to predict the size of the
program budget.
Figure 4.9 reveals that the gasoline tax program
makes this particular low-income consumer slightly
worse off because H lies just below indifference
curve U2. of course, some low-income consumers
might actually benefit from the program (if, for exam-
ple, they consume less gasoline on average than the
group of consumers whose consumption determines
the selected rebate). Nevertheless, the substitution
effect caused by the tax will make consumers, on
average, worse off.
Why, then, introduce such a program? Those
who support gasoline taxes argue that they pro-
mote national security (by reducing dependence
on foreign oil) and encourage conservation, thus
helping to slow global warming by reducing the
uildup of ca
on dioxide in the atmosphere. We
will further examine the impact of a gasoline tax in
Chapter 9.
4.3 market Demand
So far, we have discussed the demand curve for an individual consumer. Now
we turn to the market demand curve. Recall from Chapter 2 that a market de-
mand curve shows how much of a good consumers overall are willing to buy
as its price changes. In this section, we show how market demand curves can
e derived as the sum of the individual demand curves of all consumers in a
particular market.
From Individual to market Demand
To keep things simple, let’s assume that only three consumers (A, B, and C)
are in the market for coffee. Table 4.2 tabulates several points on each con-
sumer’s demand curve. The market demand, column (5), is found by adding
columns (2), (3), and (4), representing our three consumers, to determine
the  total  quantity  demanded at every price. When the price is $3, for exam-
ple, the total quantity demanded is 2 + 6 + 10, or 18.
Figure 4.10 shows these same three consumers’ demand curves for cof-
fee (labeled DA, DB, and DC). In the graph, the market demand curve is the
horizontal summation of the demands of each consumer. We sum horizontally
to find the total amount that the three consumers will demand at any given
market demand
curve Curve relating the
quantity of a good that all
consumers in a market will buy to
its price.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 145
price. For example, when the price is $4, the quantity demanded by the market
(11 units) is the sum of the quantity demanded by A (no units), by B (4 units),
and by C (7 units). Because all of the individual demand curves slope down-
ward, the market demand curve will also slope downward. However, even
though each of the individual demand curves is a straight line, the market de-
mand curve need not be. In Figure 4.10, for example, the market demand curve
is kinked because one consumer makes no purchases at prices that the other con-
sumers find acceptable (those above $4).
Table 4.2 dEtErmining thE markEt dEmand curvE
(1)
pRIce ($)
(2)
IndIvIduaL a
(unItS)
(3)
IndIvIduaL b
(unItS)
(4)
IndIvIduaL c
(unItS)
(5)
MaRket
(unItS)
1 6 10 16 32
2 4 8 13 25
3 2 6 10 18
4 0 4 7 11
5 0 2 4 6
Price
(dollars pe
unit)
Quantity
5
3
2
1
0 5 10 15 20 25 30
DA DB DC
Market Demand
4
FIgure 4.10
summing to obtain a markEt dEmand curvE
The market demand curve is obtained by summing our three consumers’ demand curves DA,
DB, and DC. At each price, the quantity of coffee demanded by the market is the sum of the
quantities demanded by each consumer. At a price of $4, for example, the quantity demanded
y the market (11 units) is the sum of the quantity demanded by A (no units), B (4 units), and
C (7 units).
146 PART 2 Producers, Consumers, and Competitive markets
Two points should be noted as a result of this analysis:
1. The market demand curve will shift to the right as more consumers enter
the market.
2. Factors that influence the demands of many consumers will also affect
market demand. Suppose, for example, that most consumers in a particu-
lar market earn more income and, as a result, increase their demands for
coffee. Because each consumer’s demand curve shifts to the right, so will
the market demand curve.
The aggregation of individual demands into market demands is not just a
theoretical exercise. It becomes important in practice when market demands are
uilt up from the demands of different demographic groups or from consumers
located in different areas. For example, we might obtain information about the
demand for home computers by adding independently obtained information
about the demands of the following groups:
•     Households with children
•     Households without children
•     Single individuals
Or, we might determine U.S. wheat demand by aggregating domestic demand
(i.e., by U.S. consumers) and export demand (i.e., by foreign consumers), as we
will see in Example 4.3.
elasticity of Demand
Recall from Section 2.4 (page 55) that the price elasticity of demand measures
the percentage change in the quantity demanded resulting from a 1-percent in-
crease in price. Denoting the quantity of a good by Q and its price by P, the price
elasticity of demand is
EP =
∆Q/Q
∆P/P
= a P
Q
b a ∆Q
∆P
b (4.1)
(Here, because ∆ means “a change in,” ∆Q/Q is the percentage change in Q.)
inElastic dEmand When demand is inelastic (i.e., EP is less than 1 in ab-
solute value), the quantity demanded is relatively unresponsive to changes
in price. As a result, total expenditure on the product increases when the
price increases. Suppose, for example, that a family cu
ently uses 1000 gal-
lons of gasoline a year when the price is $1 per gallon; suppose also that our
family’s price elasticity of demand for gasoline is -0.5. If the price of gaso-
line increases to $1.10 (a 10-percent increase), the consumption of gasoline
falls to 950 gallons (a 5-percent decrease). Total expenditure on gasoline,
however, will increase from $1000 (1000 gallons * $1 per gallon) to $1045
(950 gallons * $1.10 per gallon).
Elastic dEmand In contrast, when demand is elastic (EP is greater than 1
in absolute value), total expenditure on the product decreases as the price goes
up. Suppose that a family buys 100 pounds of chicken per year at a price
of $2 per pound; the price elasticity of demand for chicken is -1.5. If the price of
chicken increases to $2.20 (a 10-percent increase), our family’s consumption
of chicken falls to 85 pounds a year (a 15-percent decrease). Total expenditure
In §2.4, we show how the
price elasticity of demand
describes the responsive-
ness of consumer demands
to changes in price.
Recall from §2.4 that
ecause the magnitude
of an elasticity refers to its
absolute value, an elasticity
of -0.5 is less in magnitude
than a -1.0 elasticity.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 147
on chicken will also fall, from $200 (100 pounds * $2 per pound) to $187
(85 pounds * $2.20 per pound).
isoElastic dEmand When the price elasticity of demand is constant all
along the demand curve, we say that the curve is isoelastic. Figure 4.11 shows
an isoelastic demand curve. Note how this demand curve is bowed inward. In
contrast, recall from Section 2.4 what happens to the price elasticity of demand
as we move along a linear demand curve. Although the slope of the linear curve is
constant, the price elasticity of demand is not. It is zero when the price is zero,
and it increases in magnitude until it becomes infinite when the price is suffi-
ciently high for the quantity demanded to become zero.
A special case of the isoelastic curve is the unit-elastic demand curve: a demand
curve with price elasticity always equal to -1, as is the case for the curve in
Figure 4.11. In this case, total expenditure remains the same after a price change.
A price increase, for instance, leads to a decrease in the quantity demanded that
leaves the total expenditure on the good unchanged. Suppose, for example, that
the total expenditure on first-run movies in Berkeley, California, is $5.4 million
per year, regardless of the price of a movie ticket. For all points along the de-
mand curve, the price times the quantity will be $5.4 million. If the price is $6,
the quantity will be 900,000 tickets; if the price increases to $9, the quantity will
drop to 600,000 tickets, as shown in Figure 4.11.
Table 4.3 summarizes the relationship between elasticity and expenditure. It
is useful to review this table from the perspective of the seller of the good rather
isoelastic demand
curve Demand curve with a
constant price elasticity.
In §2.4, we show that when
the demand curve is linear,
demand becomes more
elastic as the price of the
product increases.
9
6
3
600 900 1800
Thousands of movie tickets
Price of
movie
tickets
($)
D
FIgure 4.11
unit-Elastic dEmand
curvE
When the price elasticity of demand is
-1.0 at every price, the total expen-
diture is constant along the demand
curve D.
Table 4.3 PricE Elasticity and consumEr ExPEnditurEs
deMand
IF pRIce IncReaSeS,
expendItuReS
IF pRIce decReaSeS,
expendItuReS
Inelastic Increase decrease
unit elastic are unchanged are unchanged
elastic decrease Increase
148 PART 2 Producers, Consumers, and Competitive markets
than the buyer. (What the seller perceives as total revenue, the consumer views
as total expenditures.) When demand is inelastic, a price increase leads only to a
small decrease in quantity demanded; thus, the seller’s total revenue increases.
But when demand is elastic, a price increase leads to a large decline in quantity
demanded and total revenue falls.
examPle 4.3 thE aggrEgatE dEmand for whEat
in Chapter 2 (Example 2.5—page 59), we explained that the demand for U.S.
wheat has two components: domestic demand (by U.S. consumers) and export
demand (by foreign consumers). Let’s see how the total demand for wheat can
e obtained by aggregating the domestic and foreign demands.
Domestic demand for wheat is given by the equation
QDD = 1430 - 55P
where QDD is the number of bushels (in millions) demanded domestically, and
P is the price in dollars per bushel. Export demand is given by
QDE = 1470 - 70P
where QDE is the number of bushels (in millions) demanded from a
oad. As
shown in Figure 4.12, domestic demand, given by AB, is relatively price in-
elastic. (Statistical studies have shown that price elasticity of domestic demand
is about -0.2 to -0.3.) However, export demand, given by CD, is more price
elastic, with an elasticity of about -0.4. Why? Export demand is more elastic
FIgure 4.12
thE aggrEgatE
dEmand for
whEat
The total world demand
for wheat is the horizontal
sum of the domestic de-
mand AB and the export
demand CD. Even though
each individual demand
curve is linear, the market
demand curve is kinked,
eflecting the fact that
there is no export demand
when the price of wheat
is greater than about $21
per bushel.
Quantity (million bushels per year)
Total Demand
Domestic Demand
Export Demand
C
E
A
DB F
P
ic
e
(d
ol
la
s
p
e

us
he
l)
0
0 500 1000 1500 2000 2500 3000
5
10
15
20
25
30
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 149
Speculative Demand
So far in our treatment of demand, we have assumed that consumers are “ratio-
nal,” in that they allocate their income among various goods and services to max-
imize their overall satisfaction. At times, however, the demands for some goods
are based not on the satisfaction one obtains from actually consuming the good,
ut instead on the belief that the price of the good will rise. In that case, it might
e possible to profit by buying the good and then reselling it later at a higher
price. This speculative demand is partly to blame for the sharp increases in hous-
ing prices that occu
ed in the U.S., Europe, and China during the past decade.
Speculative demand is often (but, as we will explain in Chapter 5, not always)
i
ational. People see that the price of a good has been rising, and somehow
conclude that the price will therefore keep rising. But there is usually no rational
asis for the “therefore,” so that a consumer who buys something because he
believes the price will keep rising is often doing little more than gambling.
speculative demand
Demand driven not by the direct
enefits one obtains from owning
or consuming a good but instead
y an expectation that the price of
the good will increase.
3For a survey of statistical studies of demand and supply elasticities and an analysis of the U.S.
wheat market, see La
y Salathe and Sudchada Langley, “An Empirical Analysis of Alternative
Export Subsidy Programs for U.S. Wheat,” Agricultural Economics Research 38, No. 1 (Winter 1986).
See also Michael J. Roberts and Wolfram Schlenker, “Identifying Supply and Demand Elasticities
of Agricultural Commodities: Implications for the U.S. Ethanol Mandate,” American Economic
Review Vol. 103, No. 6 (October 2013): 2265–95.
than domestic demand because poorer countries that import U.S. wheat turn to
other grains and foodstuffs if wheat prices rise.3
To obtain the world demand for wheat, we set the left side of each demand
equation equal to the quantity of wheat (the variable on the horizontal axis).
We then add the right side of the equations, obtaining
QDD + QDE = (1430 - 55P) + (1470 - 70P)
= 2900 - 125P
This generates the line segment EF in Figure 4.12.
At all prices above point C, however, there is no export demand, so that
world demand and domestic demand are identical. As a result, for all prices
above C, world demand is given by line segment AE. (if we were to add QDE for
prices above C, we would be inco
ectly adding a negative export demand to a
positive domestic demand.) As the figure shows, the resulting total demand for
wheat, given by AEF, is kinked. The kink occurs at point E, the price level above
which there is no export demand.
examPle 4.4 thE dEmand for housing
Housing is typically the most im-
portant single expenditure in a
household’s budget—on average,
households spend 25 percent of
their income on housing. A fam-
ily’s demand for housing de-
pends on the age and status of the
household making the purchasing
decision. one approach to the
housing demand is to relate the
number of rooms per house for each
household (the quantity demanded)
oth to an estimate of the price of
an additional room in a house and
to the household’s family income.
(Prices of rooms vary because of
150 PART 2 Producers, Consumers, and Competitive markets
differences in construction costs, including the price
of land.) Table 4.4 lists price and income elasticities
for different demographic groups.
There are significant differences among subgroups
of the population. For example, families with young
household heads have a price elasticity of -0.25,
which is more price elastic than the demands of fami-
lies with older household heads. Presumably, families
uying houses are more price sensitive when parents
and their children are younger and there may be plans
for more children. Among ma
ied households, the
income elasticity of demand for rooms also increases
with age, which tells us that older households buy
larger houses than younger households.
For poor families, the fraction of income spent on
housing is large. For instance, renters with an income
in the bottom 20 percent of the income distribution
spend roughly 55 percent of their income on hous-
ing, as compared to 2.8 percent of income for house-
holds overall.4 many government programs, such as
subsidies, rent controls, and land-use regulations,
have been proposed to shape the housing market in
ways that might ease the housing burden on the poor.
How effective are income subsidies? if the sub-
sidy increases the demand for housing substantially,
then we can presume that the subsidy will lead to
improved housing for the poor.5 on the other hand,
if the extra money were spent on items other than
housing, the subsidy will have failed to address
policy concerns related to housing.
The evidence indicates that for poor households
(with incomes in the bottom tenth percentile of all
households), the income elasticity of housing is only
about 0.09, which implies that income subsidies
would be spent primarily on items other than hous-
ing. By comparison, the income elasticity for housing
among the wealthiest households (the top 10 percent)
is about 0.54.
This discussion assumes that consumers choose
their expenditures on housing and other goods to max-
imize their overall satisfaction, where the benefits of
housing (and thus the demand for housing) arise from
the amount of living space, the safety of the neighbor-
hood, the quality of schools, etc. in recent years, how-
ever, the demand for housing has been partly driven by
speculative demand: People bought homes under the
assumption that they can re-sell the homes in the future
at a much higher price. Speculative demand—demand
driven not by the direct benefits one obtains from own-
ing a home but instead by an expectation that the price
will increase—has caused housing prices in many
parts of the United States to increase sharply, far more
than could be justified by demographics.
Speculative demand can lead to a bu
le—an
increase in price based not on the fundamentals of
demand, but instead on a belief that the price will
keep going up. Eventually, bu
les burst—the price
stops rising as new buyers stop coming into the mar-
ket, owners of the good become alarmed and start to
sell, the price drops, more people sell, and the price
drops further. As we will see in Chapter 5, bu
les
are problematic because they can distort the func-
tioning of a market and lead to financial dislocations
when they burst. That is what happened to the U.S.
housing market, which experienced a housing price
u
le that finally burst in 2008, leading to mortgage
defaults and contributing to the financial crisis that
hit the U.S. and the global economy in late 2008.
4This is the starting point of the “affordable” housing debate. For an overview, see John Quigley
and Steven Raphael, “Is Housing Unaffordable? Why Isn’t It More Affordable,” Journal of Economic
Perspectives 18 (2004): 191–214.
5Julia L. Hansen, John P. Formby, and W. James Smith, “Estimating the Income Elasticity of Demand
for Housing: A Comparison of Traditional and Lorenz-Concentration Curve Methodologies,”
Journal of Housing Economics 7 (1998): 328–42.
Table 4.4 PricE and incomE ElasticitiEs of thE dEmand for rooms
GRoup pRIce eLaStIcIty IncoMe eLaStIcIty
Single individuals -0.10 0.21
Ma
ied, head of household age less than 30, 1 child -0.25 0.06
Ma
ied, head age 30–39, 2 or more children -0.15 0.12
Ma
ied, head age 50 or older, 1 child -0.08 0.19
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 151
examPle 4.5 thE long-run dEmand for gasolinE
Among industrialized countries, the
United States is unique in that the
price of gasoline is relatively low.
The reason is simple: Europe, Japan,
and other countries have stiff taxes
on gasoline, so that gas prices are
typically double or triple that in the
United States, which imposes very
low taxes on gasoline. many econo-
mists have argued that the United States should sub-
stantially increase its tax on gasoline, because doing
so would lower gasoline consumption and thereby
educe dependence on imported oil and reduce the
greenhouse gas emissions that contribute to global
warming (in addition to providing much-needed rev-
enue to the government). Politicians have resisted,
however, because they fear that a tax increase would
anger voters.
Putting the politics of a gas tax aside, would higher
gasoline prices indeed reduce gasoline consumption,
or are drivers so wedded to big gas-guzzling cars that
higher prices would make little difference? What mat-
ters here is the long-run demand for gasoline, because
we can’t expect drivers to immediately scrap their old
cars and buy new ones following
a price increase. one way to get
at the long-run demand curve is
y looking at per-capital consump-
tion of gasoline in different coun-
tries which historically have had
very different prices (because they
imposed different gasoline taxes).
Figure 4.13 does just that. it plots
the per-capita consumption of gasoline on the vertical
axis and the price in dollars per gallon for 10 countries
on the horizontal axis.6 (Each circle represents the
population of the co
esponding country.)
Note that the United States has had by far the lowest
gasoline prices and also the highest per-capita gasoline
consumption. Australia is roughly in the middle in
terms of prices, and likewise in terms of consumption.
most of the European countries, on the other hand,
have much higher prices and co
espondingly lower
per capita consumption levels. The long-run elasticity
of demand for gasoline turns out to be about -1.4.
Now we come back to our question: Would
higher gasoline prices reduce gasoline consumption?
Figure 4.13 provides a clear answer: most definitely.
6Our thanks to Chris Knittel for providing us with the data for this figure. The figure controls for in-
come differences and is based on Figure 1 in Christopher Knittel, “Reducing Petroleum Consumption
from Transportation,” Journal of Economic Perspectives, 2012. All underlying data are available from
www.worldbank.org.
United States
Australia
Sweden
United Kingdom
Germany
France
Austria
2
100
200
G
as
D
ie
se
l f
o
T
a
ns
po
t
at
io
n
(g
al
lo
ns
y
ea
ca
pi
ta
)
300
400
500
4 6
Gasoline Price
8
Norway
New Zealand
FIgure 4.13
gasolinE PricEs
and PEr caPita
consumPtion in
10 countriEs
The graph plots per capita
consumption of gasoline ver-
sus the price per gallon (con-
verted to U.S. dollars) for 10
countries over the period 2008
to 2010. Each circle represents
the population of the co
e-
sponding country.
http:
www.worldbank.org
152 PART 2 Producers, Consumers, and Competitive markets
4.4 Consumer Surplus
Consumers buy goods because the purchase makes them better off. Consumer
surplus measures how much better off individuals are, in the aggregate, because
they can buy goods in the market. Because different consumers place different
values on the consumption of particular goods, the maximum amount they are
willing to pay for those goods also differs. Individual consumer surplus is the dif-
ference between the maximum amount that a consumer is willing to pay for a good and
the amount that the consumer actually pays. Suppose, for example, that a student
would have been willing to pay $13 for a rock concert ticket even though she
only had to pay $12. The $1 difference is her consumer surplus.7 When we add
the consumer surpluses of all consumers who buy a good, we obtain a measure
of the aggregate consumer surplus.
Consumer Surplus and Demand
Consumer surplus can be calculated easily if we know the demand curve. To
see the relationship between demand and consumer surplus, let’s examine the
individual demand curve for concert tickets shown in Figure 4.14. (Although
the following discussion applies to this particular individual demand curve, a
similar argument also applies to a market demand curve.) Drawing the demand
curve as a staircase rather than a straight line shows us how to measure the
value that our consumer obtains from buying different numbers of tickets.
When deciding how many tickets to buy, our student might reason as fol-
lows: The first ticket costs $14 but is worth $20. This $20 valuation is obtained
consumer surplus Difference
etween what a consumer is
willing to pay for a good and the
amount actually paid.
7Measuring consumer surplus in dollars involves an implicit assumption about the shape of con-
sumers’ indifference curves: namely, that the marginal utility associated with increases in a con-
sumer’s income remains constant within the range of income in question. In many cases, this is a
easonable assumption. It may be suspect, however, when large changes in income are involved.
Price
(dollars pe
ticket)
Rock concert tickets
20
19
18
17
16
15
14
13
0
1 2 3 4 5 6
Consumer Surplus
FIgure 4.14
consumEr surPlus
Consumer surplus is the total bene-
fit from the consumption of a prod-
uct, less the total cost of purchasing
it. Here, the consumer surplus asso-
ciated with six concert tickets (pur-
chased at $14 per ticket) is given by
the yellow-shaded area.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 153
y using the demand curve to find the maximum amount that she will pay
for each additional ticket ($20 being the maximum that she will pay for the first
ticket). The first ticket is worth purchasing because it generates $6 of surplus
value above and beyond its cost. The second ticket is also worth buying because
it generates a surplus of $5 ($19 - $14). The third ticket generates a surplus
of $4. The fourth, however, generates a surplus of only $3, the fifth a surplus of
$2, and the sixth a surplus of just $1. Our student is indifferent about purchas-
ing the seventh ticket (which generates zero surplus) and prefers not to buy any
more than that because the value of each additional ticket is less than its cost. In
Figure 4.14, consumer surplus is found by adding the excess values or surpluses for
all units purchased. In this case, then, consumer surplus equals
$6 + $5 + $4 + $3 + $2 + $1 = $21
To calculate the aggregate consumer surplus in a market, we simply find the
area below the market demand curve and above the price line. For our rock con-
cert example, this principle is illustrated in Figure 4.15. Now, because the num-
er of tickets sold is measured in thousands and individuals’ demand curves
differ, the market demand curve appears as a straight line. Note that the actual
expenditure on tickets is 6500 * $14 = $91,000. Consumer surplus, shown as
the yellow-shaded triangle, is
1/2 * ($20 - $14) * 6500 = $19,500
This amount is the total benefit to consumers, less what they paid for the tickets.
Of course, market demand curves are not always straight lines. Nonetheless,
we can always measure consumer surplus by finding the area below the de-
mand curve and above the price line.
Price
(dollars pe
ticket)
Rock concert tickets (thousands)
Demand Curve
Actual Expenditure
Consume
Surplus
1
19
18
17
16
15
14
13
20
0 2 3 4 5 6 7
Market Price
FIgure 4.15
consumEr surPlus
gEnEralizEd
For the market as a whole, consumer sur-
plus is measured by the area under the
demand curve and above the line rep-
esenting the purchase price of the good.
Here, the consumer surplus is given by
the yellow-shaded triangle and is equal to
1/2 * ($20 - $14) * 6500 = $19,500.
154 PART 2 Producers, Consumers, and Competitive markets
aPPlying consumEr surPlus Consumer surplus has important applica-
tions in economics. When added over many individuals, it measures the aggregate
enefit that consumers obtain from buying goods in a market. When we combine
consumer surplus with the aggregate profits that producers obtain, we can evalu-
ate both the costs and benefits not only of alternative market structures, but of
public policies that alter the behavior of consumers and firms in those markets.
examPle 4.6 thE valuE of clEan ai
Air is free in the sense that we
don’t pay to
eathe it. But the ab-
sence of a market for air may help
explain why the air quality in some
cities has been deteriorating for
decades. To encourage cleaner air,
Congress passed the Clean Air Act
in 1977 and has since amended it
a number of times. in 1990, for ex-
ample, automobile emissions con-
trols were tightened. Were these controls worth it?
Were the benefits of cleaning up the air sufficient to
outweigh the costs imposed directly on car produc-
ers and indirectly on car buyers?
To answer these questions,Congress asked the
National Academy of Sciences to evaluate emissions
controls in a cost-benefit study. Using empirically
determined estimates of the de-
mand for clean air, the benefits
portion of the study determined
how much people value clean air.
Although there is no actual market
for clean air, people do pay more
for houses where the air is clean
than for comparable houses in
areas with dirtier air. This infor-
mation was used to estimate the
demand for clean air.8 Detailed data on house prices
in neighborhoods of Boston and Los Angeles were
compared with the levels of various air pollutants.
The effects of other variables that might affect house
values were taken into account statistically. The study
determined a demand curve for clean air that looked
approximately like the one shown in Figure 4.16.
8The results are summarized in Daniel L. Rubinfeld, “Market Approaches to the Measurement of
the Benefits of Air Pollution Abatement,” in Ann Friedlaender, ed., The Benefits and Costs of Cleaning
the Air (Cam
idge: MIT Press, 1976), 240–73.
A
Value
(dollars per pphm
of reduction)
2000
5 NOX (pphm)
1000
0 10
pollution reduction
FIgure 4.16
valuing clEanEr
ai
The yellow-shaded triangle
gives the consumer surplus
generated when air pollution
is reduced by 5 parts per
100 million of nitrogen oxide
at a cost of $1000 per part
educed. The surplus is cre-
ated because most consum-
ers are willing to pay more
than $1000 for each unit re-
duction of nitrogen oxide.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 155
4.5 Network externalities
So far, we have assumed that people’s demands for a good are independent of
one another. In other words, Tom’s demand for coffee depends on Tom’s tastes
and income, the price of coffee, and perhaps the price of tea. But it does not de-
pend on Dick’s or Ha
y’s demand for coffee. This assumption has enabled us
to obtain the market demand curve simply by summing individuals’ demands.
For some goods, however, one person’s demand also depends on the de-
mands of other people. In particular, a person’s demand may be affected by the
number of other people who have purchased the good. If this is the case, there
exists a network externality. Network externalities can be positive or negative.
A positive network externality exists if the quantity of a good demanded by a typical
consumer increases in response to the growth in purchases of other consumers. If the
quantity demanded decreases, there is a negative network externality.
Positive Network externalities
One example of a positive network externality is word processing. Many stu-
dents use Microsoft Word in part because their friends and many of their pro-
fessors do as well. That allows us to send and receive drafts without the need to
convert from one program to another. The more people use a particular product
or participate in a particular activity, the greater the intrinsic value of that activ-
ity or product to each individual.
Social network websites provide another good example. If I am the only
member of that site, it will have no value to me. But the greater number of
network externality
Situation in which each
individual’s demand depends on
the purchases of other individuals.
The horizontal axis measures the amount of air pol-
lution reduction, as exemplified by a level of nitrogen
oxides (Nox) of 10 parts per 100 million (pphm); the
vertical axis measures the increased value of a home
associated with those reductions. Consider, for ex-
ample, the demand for cleaner air of a homeowner in
a city in which the air is rather dirty. if the family were
equired to pay $1000 for each 1 pphm reduction in
air pollution, it would choose A on the demand curve
in order to obtain a pollution reduction of 5 pphm.
How much is a 50-percent, or 5-pphm, reduction
in pollution worth to this same family? We can mea-
sure this value by calculating the consumer surplus
associated with reducing air pollution. Because the
price for this reduction is $1000 per unit, the family
would pay $5000. However, the family values all but
the last unit of reduction by more than $1000. As a
esult, the yellow-shaded triangle in Figure 4.16 gives
the value of the cleanup (above and beyond the pay-
ment). Because the demand curve is a straight line, the
surplus can be calculated from the area of the triangle
whose height is $1000 ($2000 - $1000) and whose
ase is 5 pphm. Therefore, the value to the household
of the nitrogen oxide pollution reduction is $2500.
A more recent study that focused on sus-
pended particulates also found that households
place substantial value on air pollution reduction.9
A one-milligram per cubic meter reduction in total
suspended particulates (from a mean of about 60
milligrams per cubic meter) was valued at $2,400
per household.
A complete cost-benefit analysis would use a
measure of the total benefit of the cleanup—the ben-
efit per household times the number of households.
This figure could be compared with the total cost
of the cleanup to determine whether such a project
was worthwhile. We will discuss clean air further in
Chapter 18, when we describe the tradeable emis-
sions permits that were introduced by the Clean Air
Act Amendments of 1990.
9Kenneth Y. Chay and Michael Greenstone, “Does Air Quality Matter? Evidence from the Housing
Market,” Journal of Political Economy 113 (2005): 376–424.
156 PART 2 Producers, Consumers, and Competitive markets
people who join the site, the more valuable it will become. If one social network-
ing site has a small advantage in terms of market share early on, the advantage
will grow, because new members will prefer to join the larger site. Hence the
huge success of personal website Facebook and professional website LinkedIn.
A similar story holds for virtual worlds and for multiplayer online games.
Another example of a positive network externality is the bandwagon effect—
the desire to be in style, to possess a good because almost everyone else has it, or
to indulge a fad. The bandwagon effect often arises with children’s toys (video
games, for example). In fact, exploiting this effect is a major objective in market-
ing and advertising toys. Often it is the key to success in selling clothing.
Positive network externalities are illustrated in Figure 4.17, in which the hori-
zontal axis measures the sales of a product in thousands per month. Suppose con-
sumers think that only 20,000 people have purchased a certain product. Because
this is a small number relative to the total population, consumers will have little
incentive to buy the product. Some consumers may still buy it (depending on
price), but only for its intrinsic value. In this case demand is given by the curve
D20. (This hypothetical demand curve assumes that there are no externalities.)
Suppose instead that consumers think 40,000 people have bought the prod-
uct. Now they find it more attractive and want to buy more. The demand curve
is D40, which is to the right of D20. Similarly, if consumers think that 60,000
people have bought the product, the demand curve will be D60, and so on. The
more people consumers believe to have purchased the product, the farther to
the right the demand curve shifts.
Ultimately, consumers will get a good sense of how many people have in fact
purchased a product. This number will depend, of course, on its price. In Figure
4.17, for example, we see that if the price were $30, then 40,000 people would
uy the product. Thus the relevant demand curve would be D40. If the price
were $20, 80,000 people would buy the product and the relevant demand curve
would be D80. The market demand curve is therefore found by joining the points on the
andwagon effect Positive
network externality in which a
consumer wishes to possess a
good in part because others do.
D20
20 40 48 60 80 100
Demand
Pure price
e�ect
Externality
e�ect
30
20
Price
(dollars pe
unit)
D40 D60 D80 D100
Quantity
(thousands
per month)
FIgure 4.17
PositivE nEtwork
ExtErnality
With a positive network externality,
the quantity of a good that an in-
dividual demands grows in response
to the growth of purchases by other
individuals. Here, as the price of the
product falls from $30 to $20, the posi-
tive externality causes the demand for
the good to shift to the right, from D40
to D80.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 157
curves D20, D40, D60, D80, and D100 that co
espond to the quantities 20,000, 40,000,
60,000, 80,000 and 100,000.
Compared with the curves D20, etc., the market demand curve is relatively
elastic. To see why the positive externality leads to a more elastic demand curve,
consider the effect of a drop in price from $30 to $20, with a demand curve of
D40. If there were no externality, the quantity demanded would increase from
40,000 to only 48,000. But as more people buy the product, the positive network
externality increases the quantity demanded further, to 80,000. Thus, the posi-
tive network externality increases the response of demand to price changes—
i.e., it makes demand more elastic. As we’ll see later, this result has important
implications for producers’ pricing strategies.
Negative Network externalities
Network externalities are sometimes negative. Congestion offers one example.
When skiing, I prefer short lines at ski lifts and fewer skiers on the slopes. As a
esult, the value of a lift ticket at a ski resort is lower the more people who bought
the tickets. Likewise for entry to an amusement park, skating rink, or beach.
Another example of a negative network externality is the snob effect—the
desire to own an exclusive or unique good. The quantity demanded of a “snob
good” is higher the fewer people who own it. Rare works of art, specially de-
signed sports cars, and made-to-order clothing are snob goods. The value one
gets from a painting or a sports car is partly the prestige, status, and exclusivity
esulting from the fact that few other people own one like it.
Figure 4.18 illustrates how a negative network externality works. We will as-
sume that the product in question is a snob good, so people value exclusivity.
snob effect Negative network
externality in which a consumer
wishes to own an exclusive or
unique good.
Snob E�ect
30,000
D8
8642
Pure Price E�ect
Net E�ect
Demand
15,000
Price
(dollars pe
unit)
14 Quantity
(thousands
per month)
D6
D4
D2
FIgure 4.18
nEgativE nEtwork
ExtErnality: snob
EffEct
The snob effect is a negative net-
work externality in which the quan-
tity of a good that an individual
demands falls in response to the
growth of purchases by other indi-
viduals. Here, as the price falls from
$30,000 to $15,000 and more peo-
ple buy the good, the snob effect
causes the demand for the good to
shift to the left, from D2 to D6.
158 PART 2 Producers, Consumers, and Competitive markets
In the figure, D2 is the demand curve that would apply if consumers believed
that only 2000 people used the good. If they believe that 4000 people use the
good, it would be less exclusive, and so its value decreases. The quantity de-
manded will therefore be lower; curve D4 applies. Similarly, if consumers be-
lieve that 6000 people use the good, demand is even smaller and D6 applies.
Eventually, consumers learn how widely owned the good actually is. Thus, the
market demand curve is found by joining the points on curves D2, D4, D6, etc.,
that actually co
espond to the quantities 2000, 4000, 6000, etc.
Note that the negative network externality makes market demand less elas-
tic. To see why, suppose the price was initially $30,000 with 2000 people using
the good. What happens when the price is lowered to $15,000? If there were no
externality, the quantity purchased would increase to 14,000 (along curve D2).
But the value of the good is greatly reduced if more people own it. The negative
network externality dampens the increase in the quantity demanded, cutting it
y 8000 units; the net increase in sales is only to 6000 units.
For a variety of goods, marketing and advertising are geared to creating a
snob effect. (Think of Rolex watches.) The goal is a very inelastic demand—
which makes it possible for firms to charge very high prices.
Negative network externalities can arise for other reasons. Consider the ef-
fect of congestion in queues. Because I prefer short lines and fewer skiers on
the slopes, the value I obtain from a lift ticket at a ski resort is lower the more
people there are who have bought tickets. Likewise for entry to an amusement
park, skating rink, or beach.10
10Tastes, of course, differ. Some people associate a positive network externality with skiing or a day
on the beach; they enjoy crowds and may even find the slope or beach lonely without them.
examPle 4.7 facEbook vErsus googlE Plus
The social networking website, Facebook,
egan operation in 2004 and had a mil-
lion users by the end of the year. By
early 2011, with over 600 million users,
Facebook became the world’s second
most visited website (after Google). A
strong positive network externality was
central to Facebook’s success.
To understand this, just ask yourself
why you would join Facebook rather
than some other social networking site.
You would join because so many other
people have joined. The more friends
that also joined, the more useful the
site becomes for you as a way to share
news and other information with friends.
Conversely, if you are the only one of
your social circle who does not use
Facebook, you may find yourself out of the loop with
espect to news and upcoming events. With more
members, there are more people to meet
or reconnect with, a bigger audience for
your photos and opinions, and generally,
a larger variety of content for you to en-
joy. in Table 4.5, you can see that as the
number of Facebook users has grown,
the time the average user spent on the
site grew as well.
Having observed the success of
Facebook, other companies tried to de-
velop their own competing social web-
sites. most notable was Google, a soft-
ware powerhouse that certainly had the
esources to compete with Facebook. in
2011, Google launched Google Plus,
a social networking site that looked
much like Facebook. At first it seemed
it would be a huge success—more than
25 million people joined in the first month, and
many more users joined later. But then the number
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 159
*4.6 empirical estimation of Demand
Later in this book, we will explain how demand information is used as input
into a firm’s economic decision-making process. General Motors, for example,
must understand automobile demand to decide whether to offer rebates or
elow-market-rate loans for new cars. Knowledge about demand is also impor-
tant for public policy decisions. Understanding the demand for oil, for instance,
can help Congress decide whether to pass an oil import tax. You may wonder
how it is that economists determine the shape of demand curves and how
price and income elasticities of demand are actually calculated. In this sta
ed
11See http:
fortune.com/2015/07/02/google-plus-facebook-privacy/.
of users dropped off dramatically, and by July 2015,
Google Plus had fewer than 20 million users, com-
pared with Facebook’s roughly 1500 million. (See
Table 4.5.) A year later Google gave up on social
networking and no longer used Google Plus as a
social networking site.11
What happened? Why did Google Plus fail so
adly? on its face (so
y for the pun!), it seemed su-
perior to Facebook in that it was easier to customize
and offered better privacy options. And of course it
had the financial resources of Google behind it. But it
lacked something important: a large number of users
elative to Facebook. When Google Plus entered the
market in 2011, Facebook already had close to 800
million users, which gave it an enormous advantage.
Any potential new user will naturally gravitate to the
social networking site with the most users, namely
Facebook. The extremely strong network externality
created a winner-take-all market, and Facebook was
the winner.
Table 4.5 facEbook usErs
yeaR Facebook uSeRS (MILLIonS) houRS peR uSeR peR Month
2004 1
2005 5.5
2006 12 61
2007 50 2
2008 100 3
2009 350 5.5
2010 500 7
2011 766 7.5
2012 980 8.5
2013 1171 9
2014 1334 10
2015 1517 10.5
2016 1654 11
Sources: Facebook, eMarkete
http:
fortune.com/2015/07/02/google-plus-facebook-privacy
160 PART 2 Producers, Consumers, and Competitive markets
section, we will
iefly examine some methods for evaluating and forecasting
demand. The section is sta
ed not only because the material is more advanced,
ut also because it is not essential for much of the later analysis in the book.
Nonetheless, this material is instructive and will help you appreciate the em-
pirical foundation of the theory of consumer behavior. The basic statistical tools
for estimating demand curves and demand elasticities are described in the ap-
pendix to this book, entitled “The Basics of Regression.”
The Statistical approach to Demand estimation
Firms often rely on market information based on actual studies of demand.
Properly applied, the statistical approach to demand estimation can help re-
searchers sort out the effects of variables, such as income and the prices of other
products, on the quantity of a product demanded. Here we outline some of the
conceptual issues involved in the statistical approach.
Table 4.6 shows the quantity of raspbe
ies sold in a market each year.
Information about the market demand for raspbe
ies would be valuable to an
organization representing growers because it would allow them to predict sales
on the basis of their own estimates of price and other demand-determining
variables. Let’s suppose that, focusing on demand, researchers find that the
quantity of raspbe
ies produced is sensitive to weather conditions but not to
the cu
ent market price (because farmers make their planting decisions based
on last year’s price).
The price and quantity data from Table 4.6 are graphed in Figure 4.19.
If we believe that price alone determines demand, it would be plausible to
describe the demand for the product by drawing a straight line (or other ap-
propriate curve), Q = a - bP, which “fit” the points as shown by demand
curve  D. (The “least-squares” method of curve-fitting is described in the
appendix to the book.)
Does curve D (given by the equation Q = 28.2 - 1.00P) really represent the
demand for the product? The answer is yes—but only if no important factors
other than price affect demand. In Table 4.6, however, we have included data
for one other variable: the average income of purchasers of the product. Note
that income (I) has increased twice during the study, suggesting that the de-
mand curve has shifted twice. Thus demand curves d1, d2, and d3 in Figure 4.19
Table 4.6 dEmand data
yeaR QuantIty (Q) pRIce (P) IncoMe (I)
2004 4 24 10
2005 7 20 10
2006 8 17 10
2007 13 17 17
2008 16 10 27
2009 15 15 27
2010 19 12 20
2011 20 9 20
2012 22 5 20
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 161
give a more likely description of demand. This linear demand curve would be
described alge
aically as
Q = a - bP + cI (4.2)
The income term in the demand equation allows the demand curve to shift in a
parallel fashion as income changes. The demand relationship, calculated using
the least-squares method, is given by Q = 8.08 - .49P + .81I.
The Form of the Demand relationship
Because the demand relationships discussed above are straight lines, the ef-
fect of a change in price on quantity demanded is constant. However, the price
elasticity of demand varies with the price level. For the demand equation
Q = a - bP, for example, the price elasticity EP is
EP = (∆Q/∆P)(P/Q) = -b(P/Q) (4.3)
Thus elasticity increases in magnitude as the price increases (and the quan-
tity demanded falls).
Consider, for example, the linear demand for raspbe
ies, which was esti-
mated to be Q = 8.08 - .49P + .81I. The elasticity of demand in 1999 (when
Q = 16 and P = 10) is equal to - .49(10/16) = - .31, whereas the elasticity in
2003 (when Q = 22 and P = 5) is substantially lower: - .11.
There is no reason to expect elasticities of demand to be constant.
Nevertheless, we often find it useful to work with the isoelastic demand curve, in
which the price elasticity and the income elasticity are constant. When written
in its log-linear form, the isoelastic demand curve appears as follows:
log (Q) = a - b log (P) + c log (I) (4.4)
where log ( ) is the logarithmic function and a, b, and c are the constants in the
demand equation. The appeal of the log-linear demand relationship is that
the slope of the line -b is the price elasticity of demand and the constant c is
Price
25
20
15
10
5
0 5 10 15 20 25 Quantity
D
d1
d2
d3
FIgure 4.19
Estimating dEmand
Price and quantity data can be used to deter-
mine the form of a demand relationship. But the
same data could describe a single demand curve
D or three demand curves d1, d2, and d3 that shift
over time.
162 PART 2 Producers, Consumers, and Competitive markets
the income elasticity.12 Using the data in Table 4.5, for example, we obtained the
egression line
log(Q) = -0.23 - 0.34 log(P) + 1.33 log(I)
This relationship tells us that the price elasticity of demand for raspbe
ies is
-0.34 (that is, demand is inelastic), and that the income elasticity is 1.33.
We have seen that it can be useful to distinguish between goods that are com-
plements and goods that are substitutes. Suppose that P2 represents the price of
a second good—one which is believed to be related to the product we are study-
ing. We can then write the demand function in the following form:
log(Q) = a - b log(P) + b2 log(P2) + c log(I)
When b2, the cross-price elasticity, is positive, the two goods are substitutes;
when b2 is negative, the two goods are complements.
The specification and estimation of demand curves has been a rapidly grow-
ing endeavor, not only in marketing, but also in antitrust analyses. It is now
commonplace to use estimated demand relationships to evaluate the likely
effects of mergers.13 What were once prohibitively costly analyses involving
mainframe computers can now be ca
ied out in a few seconds on a personal
computer. Accordingly, governmental competition authorities and economic
and marketing experts in the private sector make frequent use of supermarket
scanner data as inputs for estimating demand relationships. Once the price elas-
ticity of demand for a particular product is known, a firm can decide whether it
is profitable to raise or lower price. Other things being equal, the lower in mag-
nitude the elasticity, the more likely the profitability of a price increase.
12The natural logarithmic function with base e has the property that ∆(log(Q)) = ∆Q/Q for
any change in log(Q). Similarly, ∆(log(P)) = ∆P/P for any change in log(P). It follows that
∆(log(Q)) = ∆Q/Q = -b[∆(log(P))] = -b(∆P/P). Therefore, (∆Q/Q)/(∆P/P) = -b, which is the
price elasticity of demand. By a similar argument, the income elasticity of demand c is given by
(∆Q/Q)/(∆I/I).
13See Jonathan B. Baker and Daniel L. Rubinfeld, “Empirical Methods in Antitrust Litigation:
Review and Critique,” American Law and Economics Review 1 (1999): 386–435.
14State of New York v. Kraft General Foods, Inc., 926 F. Supp. 321, 356 (S.D.N.Y. 1995).
examPle 4.8 thE dEmand for rEady-to-Eat cErEal
The Post Cereals division of
kraft General Foods acquired
the Shredded Wheat cereals of
Nabisco in 1995. The acquisition
aised the legal and economic
question of whether Post would
aise the price of its best-selling
and, Grape Nuts, or the price of
Nabisco’s most successful
and,
Shredded Wheat Spoon Size.14
one important issue in a lawsuit
ought by the
state of New York was whether the two
ands were
close substitutes for one another.
if so, it would be more profitable
for Post to increase the price of
Grape Nuts (or Shredded Wheat)
after rather than before the ac-
quisition. Why? Because after the
acquisition the lost sales from
consumers who switched away
from Grape Nuts (or Shredded
Wheat) would be recovered to
the extent that they switched to the substitute
product.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 163
Interview and experimental approaches
to Demand Determination
Another way to obtain information about demand is through interviews in which
consumers are asked how much of a product they might be willing to buy at a
given price. This approach, however, may not succeed when people lack infor-
mation or interest or even want to mislead the interviewer. Therefore, market
esearchers have designed various indirect survey techniques. Consumers
might be asked, for example, what their cu
ent consumption behavior is and
how they would respond if a certain product were available at, say, a 10- percent
discount. They might be asked how they would expect others to behave.
Although indirect approaches to demand estimation can be fruitful, the difficul-
ties of the interview approach have forced economists and marketing specialists
to look to alternative methods.
In direct marketing experiments, actual sales offers are posed to potential cus-
tomers. An airline, for example, might offer a reduced price on certain flights
for six months, partly to learn how the price change affects demand for flights
and partly to learn how competitors will respond. Alternatively, a cereal
company might test market a new
and in Buffalo, New York, and Omaha,
Ne
aska, with some potential customers being given coupons ranging in
value from 25 cents to $1 per box. The response to the coupon offer tells the
company the shape of the underlying demand curve, helping the marketers
decide whether to market the product nationally and internationally, and at
what price.
Direct experiments are real, not hypothetical, but even so, problems remain.
The wrong experiment can be costly, and even if profits and sales rise, the firm
cannot be entirely sure that these increases resulted from the experimental
change; other factors probably changed at the same time. Moreover, the re-
sponse to experiments—which consumers often recognize as short-lived—may
differ from the response to permanent changes. Finally, a firm can afford to try
only a limited number of experiments.
The extent to which a price increase will cause
consumers to switch is given (in part) by the price
elasticity of demand for Grape Nuts. other things
eing equal, the higher the demand elasticity, the
greater the loss of sales associated with a price in-
crease. The more likely, too, that the price increase
will be unprofitable.
The substitutability of Grape Nuts and Shredded
Wheat can be measured by the cross-price elastic-
ity of demand for Grape Nuts with respect to the
price of Shredded Wheat. The relevant elasticities
were calculated using weekly data obtained from
supermarket scanning of household purchases for
10 cities over a three-year period. one of the esti-
mated isoelastic demand equations appeared in the
following log-linear form:
log(QGN) = 1.998 - 2.085 log(PGN) + 0.62 log(I)
+ 0.14 log(PSW)
where QGN is the amount (in pounds) of Grape
Nuts sold weekly, PGN the price per pound of Grape
Nuts, i real personal income, and PSW the price per
pound of Shredded Wheat Spoon Size.
The demand for Grape Nuts is elastic (at cu
ent
prices), with a price elasticity of about -2. The income
elasticity is 0.62: in other words, increases in income
lead to increases in cereal purchases, but at less than a
1-for-1 rate. Finally, the cross-price elasticity is 0.14. This
figure is consistent with the fact that although the two
cereals are substitutes (the quantity demanded of Grape
Nuts increases in response to an increase in the price of
Shredded Wheat), they are not very close substitutes.
164 PART 2 Producers, Consumers, and Competitive markets
Summary
1. Individual consumers’ demand curves for a com-
modity can be derived from information about their
tastes for all goods and services and from their budget
constraints.
2. Engel curves, which describe the relationship between
the quantity of a good consumed and income, can be
useful in showing how consumer expenditures vary
with income.
3. Two goods are substitutes if an increase in the price of
one leads to an increase in the quantity demanded of
the other. In contrast, two goods are complements if an
increase in the price of one leads to a decrease in the
quantity demanded of the other.
4. The effect of a price change on the quantity demanded
of a good can be
oken into two parts: a substitution
effect, in which the level of utility remains constant
while price changes, and an income effect, in which
the price remains constant while the level of utility
changes. Because the income effect can be positive or
negative, a price change can have a small or a large
effect on quantity demanded. In the unusual case
of a so-called Giffen good, the quantity demanded
may move in the same direction as the price change,
thereby generating an upward-sloping individual de-
mand curve.
5. The market demand curve is the horizontal summa-
tion of the individual demand curves of all consumers
in the market for a good. It can be used to calculate
how much people value the consumption of particular
goods and services.
6. Demand is price inelastic when a 1-percent increase in
price leads to a less than 1-percent decrease in quan-
tity demanded, thereby increasing the consumer’s
expenditure. Demand is price elastic when a 1- percent
increase in price leads to a more than 1-percent de-
crease in quantity demanded, thereby decreasing the
consumer’s expenditure. Demand is unit elastic when
a 1-percent increase in price leads to a 1-percent de-
crease in quantity demanded.
7. The concept of consumer surplus can be useful in
determining the benefits that people receive from the
consumption of a product. Consumer surplus is the
difference between the maximum amount a consumer
is willing to pay for a good and what he actually pays
for it.
8. In some instances demand will be speculative, driven
not by the direct benefits one obtains from owning or
consuming a good but instead by an expectation that
the price of the good will increase.
9. A network externality occurs when one person’s de-
mand is affected directly by the purchasing or usage
decisions of other consumers. There is a positive net-
work externality when a typical consumer’s quantity
demanded increases because others have purchased
or are using the product or service. Conversely, there
is a negative network externality when quantity de-
manded increases because fewer people own or use
the product or service.
10. A number of methods can be used to obtain in-
formation about consumer demand. These include
interview and experimental approaches, direct mar-
keting experiments, and the more indirect statistical
approach. The statistical approach can be very power-
ful in its application, but it is necessary to determine
the appropriate variables that affect demand before
the statistical work is done.
QuEStionS for rEviEw
1. Explain the difference between each of the following
terms:
a. a price consumption curve and a demand curve
. an individual demand curve and a market demand
curve
c. an Engel curve and a demand curve
d. an income effect and a substitution effect
2. Suppose that an individual allocates his or her entire
udget between two goods, food and clothing. Can
oth goods be inferior? Explain.
3. Explain whether the following statements are true or
false:
a. The marginal rate of substitution diminishes as an
individual moves downward along the demand
curve.
. The level of utility increases as an individual moves
downward along the demand curve.
c. Engel curves always slope upward.
4. Tickets to a rock concert sell for $10. But at that price,
the demand is substantially greater than the available
number of tickets. Is the value or marginal benefit of
an additional ticket greater than, less than, or equal to
$10? How might you determine that value?
5. Which of the following combinations of goods are
complements and which are substitutes? Can they be
either in different circumstances? Discuss.
a. a mathematics class and an economics class
. tennis balls and a tennis racket
c. steak and lobste
d. a plane trip and a train trip to the same destination
e. bacon and eggs
6. Suppose that a consumer spends a fixed amount of
income per month on the following pairs of goods:
a. tortilla chips and salsa
. tortilla chips and potato chips
c. movie tickets and gourmet coffee
d. travel by bus and travel by subway
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 165
If the price of one of the goods increases, explain the
effect on the quantity demanded of each of the goods.
In each pair, which are likely to be complements and
which are likely to be substitutes?
7. Which of the following events would cause a move-
ment along the demand curve for U.S. produced cloth-
ing, and which would cause a shift in the demand
curve?
a. the removal of quotas on the importation of foreign
clothes
. an increase in the income of U.S. citizens
c. a cut in the industry’s costs of producing domestic
clothes that is passed on to the market in the form
of lower prices
8. For which of the following goods is a price increase
likely to lead to a substantial income (as well as sub-
stitution) effect?
a. salt
. housing
c. theater tickets
d. food
9. Suppose that the average household in a state con-
sumes 800 gallons of gasoline per year. A 20-cent gaso-
line tax is introduced, coupled with a $160 annual tax
ebate per household. Will the household be better or
worse off under the new program?
10. Which of the following three groups is likely to
have the most, and which the least, price-elastic de-
mand for membership in the Association of Business
Economists?
a. students
. junior executives
c. senior executives
11. Explain which of the following items in each pair is
more price elastic.
a. The demand for a specific
and of toothpaste and
the demand for toothpaste in general
. The demand for gasoline in the short run and the
demand for gasoline in the long run
12. Explain the difference between a positive and a
negative network externality and give an example
of each.
ExErciSES
1. An individual sets aside a certain amount of his in-
come per month to spend on his two ho
ies, collect-
ing wine and collecting books. Given the information
elow, illustrate both the price-consumption curve
associated with changes in the price of wine and the
demand curve for wine.
pRIce
WIne
pRIce
ook
QuantIty
WIne
QuantIty
ook budGet
$10 $10 7 8 $150
$12 $10 5 9 $150
$15 $10 4 9 $150
$20 $10 2 11 $150
2. An individual consumes two goods, clothing and
food. Given the information below, illustrate both the
income-consumption curve and the Engel curve for
clothing and food.
pRIce
cLothInG
pRIce
Food
QuantIty
cLothInG
QuantIty
Food IncoMe
$10 $2 6 20 $100
$10 $2 8 35 $150
$10 $2 11 45 $200
$10 $2 15 50 $250
3. Jane always gets twice as much utility from an extra
allet ticket as she does from an extra basketball ticket,
egardless of how many tickets of either type she has.
Draw Jane’s income-consumption curve and her Engel
curve for ballet tickets.
4. a. Orange juice and apple juice are known to be
perfect substitutes. Draw the appropriate price-
consumption curve (for a variable price of orange
juice) and income-consumption curve.
. Left shoes and right shoes are perfect comple-
ments. Draw the appropriate price-consumption
and income-consumption curves.
5. Each week, Bill, Mary, and Jane select the quantity of
two goods, x1 and x2, that they will consume in order
to maximize their respective utilities. They each spend
their entire weekly income on these two goods.
a. Suppose you are given the following information
about the choices that Bill makes over a three-week
period:
x1 x2 P1 P2 I
Week 1 10 20 2 1 40
Week 2 7 19 3 1 40
Week 3 8 31 3 1 55
Did Bill’s utility increase or decrease between week
1 and week 2? Between week 1 and week 3?
Explain using a graph to support your answer.
166 PART 2 Producers, Consumers, and Competitive markets
. Now consider the following information about the
choices that Mary makes:
x1 x2 P1 P2 I
Week 1 10 20 2 1 40
Week 2 6 14 2 2 40
Week 3 20 10 2 2 60
Did Mary’s utility increase or decrease between
week 1 and week 3? Does Mary consider both
goods to be normal goods? Explain.
*c. Finally, examine the following information about
Jane’s choices:
x1 x2 P1 P2 I
Week 1 12 24 2 1 48
Week 2 16 32 1 1 48
Week 3 12 24 1 1 36
Draw a budget line-indifference curve graph that
illustrates Jane’s three chosen bundles. What can
you say about Jane’s preferences in this case?
Identify the income and substitution effects that
esult from a change in the price of good x1.
6. Two individuals, Sam and Ba
, derive utility from the
hours of leisure (L) they consume and from the amount
of goods (G) they consume. In order to maximize
utility, they need to allocate the 24 hours in the day
etween leisure hours and work hours. Assume that
all hours not spent working are leisure hours. The price
of a good is equal to $1 and the price of leisure is equal
to the hourly wage. We observe the following informa-
tion about the choices that the two individuals make:
SaM baRb SaM baR
pRIce
oF G
pRIce
oF L
L
(houRS)
L
(houRS) G ($) G ($)
1 8 16 14 64 80
1 9 15 14 81 90
1 10 14 15 100 90
1 11 14 16 110 88
Graphically illustrate Sam’s leisure demand curve and
Ba
’s leisure demand curve. Place price on the verti-
cal axis and leisure on the horizontal axis. Given that
they both maximize utility, how can you explain the
difference in their leisure demand curves?
7. The director of a theater company in a small col-
lege town is considering changing the way he prices
tickets. He has hired an economic consulting firm to
estimate the demand for tickets. The firm has classi-
fied people who go to the theater into two groups and
has come up with two demand functions. The demand
curves for the general public (Qgp) and students (Qs)
are given below:
Qgp = 500 - 5P
Qs = 200 - 4P
a. Graph the two demand curves on one graph, with
P on the vertical axis and Q on the horizontal axis.
If the cu
ent price of tickets is $35, identify the
quantity demanded by each group.
. Find the price elasticity of demand for each group
at the cu
ent price and quantity.
c. Is the director maximizing the revenue he collects
from ticket sales by charging $35 for each ticket?
Explain.
d. What price should he charge each group if he
wants to maximize revenue collected from ticket
sales?
8. Judy has decided to allocate exactly $500 to college
textbooks every year, even though she knows that the
prices are likely to increase by 5 to 10 percent per year
and that she will be getting a substantial monetary gift
from her grandparents next year. What is Judy’s price
elasticity of demand for textbooks? Income elasticity?
9. The ACME Corporation determines that at cu
ent
prices, the demand for its computer chips has a price
elasticity of -2 in the short run, while the price elastic-
ity for its disk drives is -1.
a. If the corporation decides to raise the price of both
products by 10 percent, what will happen to its
sales? To its sales revenue?
. Can you tell from the available information which
product will generate the most revenue? If yes,
why? If not, what additional information do you
need?
10. By observing an individual’s behavior in the situations
outlined below, determine the relevant income elastici-
ties of demand for each good (i.e., whether it is normal
or inferior). If you cannot determine the income elastic-
ity, what additional information do you need?
a. Bill spends all his income on books and coffee. He
finds $20 while rummaging through a used paper-
ack bin at the bookstore. He immediately buys a
new hardcover book of poetry.
. Bill loses $10 he was going to use to buy a double
espresso. He decides to sell his new book at a dis-
count to a friend and use the money to buy coffee.
c. Being bohemian becomes the latest teen fad. As a
esult, coffee and book prices rise by 25 percent. Bill
lowers his consumption of both goods by the same
percentage.
d. Bill drops out of art school and gets an M.B.A. in-
stead. He stops reading books and drinking coffee.
Now he reads the Wall Street Journal and drinks
ottled mineral water.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 167
11. Suppose the income elasticity of demand for food is
0.5 and the price elasticity of demand is -1.0. Suppose
also that Felicia spends $10,000 a year on food, the
price of food is $2, and that her income is $25,000.
a. If a sales tax on food caused the price of food to in-
crease to $2.50, what would happen to her consump-
tion of food? (Hint: Because a large price change is
involved, you should assume that the price elastic-
ity measures an arc elasticity, rather than a point
elasticity.)
. Suppose that Felicia gets a tax rebate of $2500 to
ease the effect of the sales tax. What would her con-
sumption of food be now?
c. Is she better or worse off when given a rebate equal
to the sales tax payments? Draw a graph and explain.
12. You run a small business and would like to pre-
dict what will happen to the quantity demanded for
your product if you raise your price. While you do
not know the exact demand curve for your product,
you do know that in the first year you charged $45
and sold 1200 units and that in the second year you
charged $30 and sold 1800 units.
a. If you plan to raise your price by 10 percent, what
would be a reasonable estimate of what will hap-
pen to quantity demanded in percentage terms?
. If you raise your price by 10 percent, will revenue
increase or decrease?
13. Suppose you are in charge of a toll
idge that costs
essentially nothing to operate. The demand for
idge
crossings Q is given by P = 15 - (1/2)Q.
a. Draw the demand curve for
idge crossings.
. How many people would cross the
idge if there
were no toll?
c. What is the loss of consumer surplus associated
with a
idge toll of $5?
d. The toll-
idge operator is considering an increase
in the toll to $7. At this higher price, how many peo-
ple would cross the
idge? Would the toll-
idge
evenue increase or decrease? What does your an-
swer tell you about the elasticity of demand?
e. Find the lost consumer surplus associated with the
increase in the price of the toll from $5 to $7.
14. Vera has decided to upgrade the operating system
on her new PC. She hears that the new Linux operat-
ing system is technologically superior to Windows
and substantially lower in price. However, when she
asks her friends, it turns out they all use PCs with
Windows. They agree that Linux is more appealing
ut add that they see relatively few copies of Linux on
sale at local stores. Vera chooses Windows. Can you
explain her decision?
15. Suppose that you are the consultant to an agricultural
cooperative that is deciding whether members should
cut their production of cotton in half next year. The
cooperative wants your advice as to whether this ac-
tion will increase members’ revenues. Knowing that
cotton (C) and soybeans (S) both compete for agricul-
tural land in the South, you estimate the demand for
cotton to be C = 3.5 - 1.0PC + 0.25PS + 0.50I, where
PC is the price of cotton, PS the price of soybeans, and
I income. Should you support or oppose the plan? Is
there any additional information that would help you
to provide a definitive answer?
16. A consumer lives on a diet of solely steak and po-
tatoes. Her budget is $30 for every 10 days, and she
must buy enough potatoes to eat at least two potatoes
per day.
a. A potato costs $0.50 and the price of a steak is $10.
How much will the consumer purchase of each
good?
. Now suppose that the price of a potato increases to
$1. How much will the consumer purchase of each
good?
c. Now suppose that the price of a potato increases
to $1.25. How much will the consumer purchase of
each good?
d. What kind of good is the potato?
e. Would you expect the demand curve for potatoes
to continue to follow this trend indefinitely? Why
or why not?
This page intentionally left blank
APPENDix To CHAPTER 4
Demand Theory
—a Mathematical Treatment
This appendix presents a mathematical treatment of the basics of demand
theory. Our goal is to provide a short overview of the theory of demand for
students who have some familiarity with the use of calculus. To do this, we will
explain and then apply the concept of constrained optimization.
utility maximization
The theory of consumer behavior is based on the assumption that consum-
ers maximize utility subject to the constraint of a limited budget. We saw in
Chapter 3 that for each consumer, we can define a utility function that attaches
a level of utility to each market basket. We also saw that the marginal utility of
a good is defined as the change in utility associated with a one-unit increase
in the consumption of the good. Using calculus, as we do in this appendix, we
measure marginal utility as the utility change that results from a very small in-
crease in consumption.
Suppose, for example, that Bob’s utility function is given by U(X, Y) = log X +
log Y, where, for the sake of generality, X is now used to represent food and
Y represents clothing. In that case, the marginal utility associated with the ad-
ditional consumption of X is given by the partial derivative of the utility function
with respect to good X. Here, MUX, representing the marginal utility of good X, is
given by

0U(X, Y)
0X
=
0( log X + log Y)
0X
=
1
X

In the following analysis, we will assume, as in Chapter 3, that while the
level of utility is an increasing function of the quantities of goods consumed,
marginal utility decreases with consumption. When there are two goods, X and
Y, the consumer’s optimization problem may thus be written as
Maximize U(X, Y) (A4.1)
subject to the constraint that all income is spent on the two goods:
PXX + PYY = 1 (A4.2)
Here, U( ) is the utility function, X and Y the quantities of the two goods pur-
chased, PX and PY the prices of the goods, and I income.1
To determine the individual consumer’s demand for the two goods, we
choose those values of X and Y that maximize (A4.1) subject to (A4.2). When
we know the particular form of the utility function, we can solve to find the
In §3.1, we explain that a
utility function is a formula
that assigns a level of utility
to each market basket.
In §3.5, marginal utility is
described as the additional
satisfaction obtained by
consuming an additional
amount of a good.
1To simplify the mathematics, we assume that the utility function is continuous (with continuous
derivatives) and that goods are infinitely divisible. The logarithmic function log (.) measures the
natural logarithm of a number.
169
170 PART 2 Producers, Consumers, and Competitive markets
consumer’s demand for X and Y directly. However, even if we write the utility
function in its general form U(X, Y), the technique of constrained optimization can
e used to describe the conditions that must hold if the consumer is maximizing
utility.
The method of lagrange multipliers
The method of Lagrange multipliers is a technique that can be used to maxi-
mize or minimize a function subject to one or more constraints. Because we will
use this technique to analyze production and cost issues later in the book, we
will provide a step-by-step application of the method to the problem of finding
the consumer’s optimization given by equations (A4.1) and (A4.2).
1. Stating the Problem First, we write the Lagrangian for the problem. The
Lagrangian is the function to be maximized or minimized (here, utility
is being maximized), plus a variable which we call l times the constraint
(here, the consumer’s budget constraint). We will interpret the meaning of
l in a moment. The Lagrangian is then
Φ = U(X, Y) - l (PXX + PYY - I) (A4.3)
Note that we have written the budget constraint as
PXX + PYY - I = 0
i.e., as a sum of terms that is equal to zero. We then insert this sum into the
Lagrangian.
2. Differentiating the Lagrangian If we choose values of X and Y that sat-
isfy the budget constraint, then the second term in equation (A4.3) will be
zero. Maximizing will therefore be equivalent to maximizing U(X, Y). By
differentiating Φ with respect to X, Y, and l and then equating the deriva-
tives to zero, we can obtain the necessary conditions for a maximum.2 The
esulting equations are

0Φ
0X
= MUX(X, Y) - lPX = 0

0Φ
0Y
= MUY(X, Y) - lPY = 0 (A4.4)

0Φ
0l
= I - PXX - PYY = 0
Here as before, MU is short for marginal utility: In other words,
MUX(X, Y) = 0U(X, Y)/0X, the change in utility from a very small increase in
the consumption of good X.
method of Lagrange
multipliers Technique
to maximize or minimize a
function subject to one or more
constraints.
Lagrangian Function to be
maximized or minimized, plus a
variable (the Lagrange multiplier)
multiplied by the constraint.
2These conditions are necessary for an “interior” solution in which the consumer consumes posi-
tive amounts of both goods. The solution, however, could be a “corner” solution in which all of one
good and none of the other is consumed.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 171
3. Solving the Resulting Equations The three equations in (A4.4) can be
ewritten as
MUX = l PX
MUY = l PY
PXX + PYY = I
Now we can solve these three equations for the three unknowns. The resulting
values of X and Y are the solution to the consumer’s optimization problem: They
are the utility-maximizing quantities.
The equal marginal Principle
The third equation above is the consumer’s budget constraint with which we
started. The first two equations tell us that each good will be consumed up to
the point at which the marginal utility from consumption is a multiple (l) of the
price of the good. To see the implication of this, we combine the first two condi-
tions to obtain the equal marginal principle:
l =
MUX(X, Y)
PX
=
MUY(X, Y)
PY
(A4.5)
In other words, the marginal utility of each good divided by its price is the
same. To optimize, the consumer must get the same utility from the last dollar spent
y consuming either X or Y. If this were not the case, consuming more of one good
and less of the other would increase utility.
To characterize the individual’s optimum in more detail, we can rewrite the
information in (A4.5) to obtain

MUX(X, Y)
MUY(X, Y)
=
PX
PY
(A4.6)
In other words, the ratio of the marginal utilities is equal to the ratio of the prices.
marginal rate of Substitution
We can use equation (A4.6) to see the link between utility functions and indif-
ference curves that was spelled out in Chapter 3. An indifference curve repre-
sents all market baskets that give the consumer the same level of utility. If U* is
a fixed utility level, the indifference curve that co
esponds to that utility level
is given by
U(X, Y) = U*
As the market baskets are changed by adding small amounts of X and
subtracting small amounts of Y, the total change in utility must equal zero.
Therefore,
MUX(X, Y)dX + MUY(X, Y)dY = dU* = 0 (A4.7)
In §3.5, we show that the
marginal rate of substitution
is equal to the ratio of the
marginal utilities of the two
goods being consumed.
172 PART 2 Producers, Consumers, and Competitive markets
Rea
anging,
-dY/dX = MUX(X, Y)/MUY(X, Y) = MRSXY (A4.8)
where MRSXY represents the individual’s marginal rate of substitution of X for
Y. Because the left-hand side of (A4.8) represents the negative of the slope of the
indifference curve, it follows that at the point of tangency, the individual’s mar-
ginal rate of substitution (which trades off goods while keeping utility constant)
is equal to the individual’s ratio of marginal utilities, which in turn is equal to
the ratio of the prices of the two goods, from (A4.6).3
When the individual indifference curves are convex, the tangency of the in-
difference curve to the budget line solves the consumer’s optimization problem.
This principle was illustrated by Figure 3.13 (page 108) in Chapter 3.
marginal utility of Income
Whatever the form of the utility function, the Lagrange multiplier l represents
the extra utility generated when the budget constraint is relaxed—in this case
y adding one dollar to the budget. To show how the principle works, we dif-
ferentiate the utility function U(X, Y) totally with respect to I:
dU/dI = MUX(X, Y)(dX/dI) + MUY(X, Y)(dY/dI) (A4.9)
Because any increment in income must be divided between the two goods, it
follows that
dI = PXdX + PYdY (A4.10)
Substituting from (A4.5) into (A4.9), we get
dU/dI = lPX(dX/dI) + lPY(dY/dI) = l(PXdX + PYdY)/dI (A4.11)
and substituting (A4.10) into (A4.11), we get
dU/dI = l(PXdX + PYdY)/(PXdX + PYdY) = l (A4.12)
Thus the Lagrange multiplier is the extra utility that results from an extra dollar
of income.
Going back to our original analysis of the conditions for utility maximiza-
tion, we see from equation (A4.5) that maximization requires the utility ob-
tained from the consumption of every good, per dollar spent on that good, to
e equal to the marginal utility of an additional dollar of income. If this were
not the case, utility could be increased by spending more on the good with the
higher ratio of marginal utility to price and less on the other good.
3We implicitly assume that the “second-order conditions” for a utility maximum hold. The con-
sumer, therefore, is maximizing rather than minimizing utility. The convexity condition is suf-
ficient for the second-order conditions to be satisfied. In mathematical terms, the condition is
that d(MRS)/dX 6 0 or that dY2/dX2 7 0 where -dY/dX is the slope of the indifference curve.
Remember: diminishing marginal utility is not sufficient to ensure that indifference curves are
convex.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 173
an example
In general, the three equations in (A4.4) can be solved to determine the three
unknowns X, Y, and l as a function of the two prices and income. Substitution
for l then allows us to solve for the demand for each of the two goods in terms
of income and the prices of the two commodities. This principle can be most
easily seen in terms of an example.
A frequently used utility function is the Co
-Douglas utility function,
which can be represented in two forms:
U(X, Y) = a log(X) + (1 - a) log(Y)
and
U(X, Y) = XaY1 - a
For the purposes of demand theory, these two forms are equivalent because they
oth yield the identical demand functions for goods X and Y. We will derive the
demand functions for the first form and leave the second as an exercise for the
student.
To find the demand functions for X and Y, given the usual budget constraint,
we first write the Lagrangian:
Φ = a log (X) + (1 - a) log (Y) - l(PXX + PYY - I)
Now differentiating with respect to X, Y, and l and setting the derivatives equal
to zero, we obtain
0Φ/0X = a/X - lPX = 0
0Φ/0Y = (1 - a)/Y - lPY = 0
0Φ/0l = PXX + PYY - I = 0
The first two conditions imply that
PXX = a/l (A4.13)
PYY = (1 - a)/l (A4.14)
Combining these expressions with the last condition (the budget constraint)
gives us
a/l + (1 - a)/l - I = 0
or l = 1/I. Now we can substitute this expression for l back into (A4.13) and
(A4.14) to obtain the demand functions:
X = (a/PX)I
Y = [(1 - a)/PY]I
In this example, the demand for each good depends only on the price of that
good and on income, not on the price of the other good. Thus, the cross-price
elasticities of demand are 0.
co
-Douglas utility
function Utility function
U(X, Y) = XaY1 - a, where X and Y
are two goods and a is a constant.
In §2.4, we explain that
the cross-price elasticity of
demand refers to the per-
centage change in the quan-
tity demanded of one good
that results from a 1-percent
increase in the price of
another good.
174 PART 2 Producers, Consumers, and Competitive markets
We can also use this example to review the meaning of Lagrange multipliers.
To do so, let’s substitute specific values for each of the parameters in the problem.
Let a = 1/2, PX = $1, PY = $2, and I = $100. In this case, the choices that maxi-
mize utility are X = 50 and Y = 25. Also note that l = 1/100. The Lagrange
multiplier tells us that if an additional dollar of income were available to the
consumer, the level of utility achieved would increase by 1/100. This conclusion
is relatively easy to check. With an income of $101, the maximizing choices of the
two goods are X = 50.5 and Y = 25.25. A bit of arithmetic tells us that the origi-
nal level of utility is 3.565 and the new level of utility 3.575. As we can see, the
additional dollar of income has indeed increased utility by .01, or 1/100.
Duality in Consumer Theory
There are two different ways of looking at the consumer’s optimization deci-
sion. The optimum choice of X and Y can be analyzed not only as the problem
of choosing the highest indifference curve—the maximum value of U( )—that
touches the budget line, but also as the problem of choosing the lowest budget
line—the minimum budget expenditure—that touches a given indifference
curve. We use the term duality to refer to these two perspectives. To see how
this principle works, consider the following dual consumer optimization prob-
lem: the problem of minimizing the cost of achieving a particular level of utility:
Minimize PXX + PYY
subject to the constraint that
U(X, Y) = U*
The co
esponding Lagrangian is given by
Φ = PXX + PYY - µ(U(X, Y) - U*) (A4.15)
where µ is the Lagrange multiplier. Differentiating Φ with respect to X, Y, and µ
and setting the derivatives equal to zero, we find the following necessary condi-
tions for expenditure minimization:
PX - µMUX(X, Y) = 0
PY - µMUY(X, Y) = 0
and
U(X, Y) = U*
By solving the first two equations, and recalling (A4.5), we see that
µ = [PX/MUX(X, Y)] = [PY/MUY(X, Y)] = 1/l
Because it is also true that
MUX(X, Y)/MUY(X, Y) = MRSXY = PX/PY
duality Alternative way
of looking at the consumer’s
utility maximization decision:
Rather than choosing the highest
indifference curve, given a
udget constraint, the consumer
chooses the lowest budget line
that touches a given indifference
curve.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 175
the cost-minimizing choice of X and Y must occur at the point of tangency of
the budget line and the indifference curve that generates utility U*. Because
this is the same point that maximized utility in our original problem, the dual
expenditure-minimization problem yields the same demand functions that are
obtained from the direct utility-maximization problem.
To see how the dual approach works, let’s reconsider our Co
-Douglas
example. The alge
a is somewhat easier to follow if we use the exponential
form of the Co
-Douglas utility function, U(X, Y) = XaY1 - a. In this case, the
Lagrangian is given by
Φ = PXX + PYY - µ [XaY1 - a - U*] (A4.16)
Differentiating with respect to X, Y, and µ and equating to zero, we obtain
PX = µ aU*/X
PY = µ (1 - a)U*/Y
Multiplying the first equation by X and the second by Y and adding, we get
PXX + PYY = µ U*
First, we let I be the cost-minimizing expenditure (if the individual did not
spend all of his income to get utility level U*, U* would not have maximized
utility in the original problem). Then it follows that µ = I/U*. Substituting in
the equations above, we obtain
X = aI/PX and Y = (1 - a)I/PY
These are the same demand functions that we obtained before.
Income and Substitution effects
The demand function tells us how any individual’s utility-maximizing choices
espond to changes in both income and the prices of goods. It is important,
however, to distinguish that portion of any price change that involves movement
along an indifference curve from that portion which involves movement to a dif-
ferent indifference curve (and therefore a change in purchasing power). To make
this distinction, we consider what happens to the demand for good X when
the price of X changes. As we explained in Section 4.2, the change in demand
can be divided into a substitution effect (the change in quantity demanded when
the level of utility is fixed) and an income effect (the change in the quantity de-
manded with the level of utility changing but the relative price of good X un-
changed). We denote the change in X that results from a unit change in the price
of X, holding utility constant, by
0X/0PX �U = U*
Thus the total change in the quantity demanded of X resulting from a unit
change in PX is
dX/dPX = 0X/0PX �U = U* + (0X/0I)(0I/0PX) (A4.17)
In §4.2, the effect of a price
change is divided into an
income effect and a substitu-
tion effect.
176 PART 2 Producers, Consumers, and Competitive markets
The first term on the right side of equation (A4.17) is the substitution effect
(because utility is fixed); the second term is the income effect (because income
increases).
From the consumer’s budget constraint, I = PXX + PYY, we know by dif-
ferentiation that
0I/0PX = X (A4.18)
Suppose for the moment that the consumer owned goods X and Y. In that case,
equation (A4.18) would tell us that when the price of good X increases by $1,
the amount of income that the consumer can obtain by selling the good in-
creases by $X. In our theory of consumer behavior, however, the consumer does
not own the good. As a result, equation (A4.18) tells us how much additional in-
come the consumer would need in order to be as well off after the price change
as he or she was before. For this reason, it is customary to write the income ef-
fect as negative (reflecting a loss of purchasing power) rather than as a positive.
Equation (A4.17) then appears as follows:
dX/dPX = 0X/0PX�U = U* - X(0X/0I) (A4.19)
In this new form, called the Slutsky equation, the first term represents the
substitution effect: the change in demand for good X obtained by keeping util-
ity fixed. The second term is the income effect: the change in purchasing power
esulting from the price change times the change in demand resulting from a
change in purchasing power.
An alternative way to decompose a price change into substitution and in-
come effects, which is usually attributed to John Hicks, does not involve indif-
ference curves. In Figure A4.1, the consumer initially chooses market basket A
on budget line RS. Suppose that after the price of food falls (and the budget line
Slutsky equation Formula
for decomposing the effects of a
price change into substitution and
income effects.
Clothing
(units pe
month)
A
B
S T Food
(units per month)
R′
R
T′
FIgure a4.1
hicksian substitution
EffEct
The individual initially consumes market
asket A. A decrease in the price of food
shifts the budget line from RS to RT. If a suf-
ficient amount of income is taken away to
make the individual no better off than he or
she was at A, two conditions must be met:
The new market basket chosen must lie on
line segment BT’ of budget line R’ T’ (which
intersects RS to the right of A), and the
quantity of food consumed must be greater
than at A.
CHAPTER 4 iNDiViDUAL AND mARkET DEmAND 177
ExErciSES
1. Which of the following utility functions are consistent
with convex indifference curves and which are not?
a. U(X, Y) = 2X + 5Y
. U(X, Y) = (XY)5
c. U(X, Y) = Min (X, Y), where Min is the minimum
of the two values of X and Y.
2. Show that the two utility functions given below gener-
ate identical demand functions for goods X and Y:
a. U(X, Y) = log (X) + log (Y)
. U(X, Y) = (XY)5
3. Assume that a utility function is given by Min(X, Y),
as in Exercise 1(c). What is the Slutsky equation that
decomposes the change in the demand for X in re-
sponse to a change in its price? What is the income
effect? What is the substitution effect?
4. Sharon has the following utility function:
U(X, Y) = 1X + 1Y
where X is her consumption of candy bars, with price
PX = $1, and Y is her consumption of espressos, with
PY = $3.
a. Derive Sharon’s demand for candy bars and
espresso.
. Assume that her income I = $100. How many
candy bars and how many espressos will Sharon
consume?
c. What is the marginal utility of income?
5. Maurice has the following utility function:
U(X, Y) = 20X + 80Y - X2 - 2Y2
where X is his consumption of CDs with a price of $1
and Y is his consumption of movie videos, with a rental
price of $2. He plans to spend $41 on both forms of en-
tertainment. Determine the number of CDs and video
entals that will maximize Maurice’s utility.
moves to RT), we take away enough income so that the individual is no better
off (and no worse off) than he was before. To do so, we draw a budget line par-
allel to RT. If the budget line passed through A, the consumer would be at least
as satisfied as he was before the price change: He still has the option to purchase
market basket A if he wishes. According to the Hicksian substitution effect,
therefore, the budget line that leaves him equally well off must be a line such as
R’T’, which is parallel to RT and which intersects RS at a point B below and to
the right of point A.
Revealed preference tells us that the newly chosen market basket must lie on
line segment BT’. Why? Because all market baskets on line segment R’ B could
have been chosen but were not when the original budget line was RS. (Recall
that the consumer prefe
ed basket A to any other feasible market basket.)
Now note that all points on line segment BT’ involve more food consumption
than does basket A. It follows that the quantity of food demanded increases
whenever there is a decrease in the price of food with utility held constant.
This negative substitution effect holds for all price changes and does not
ely on the assumption of convexity of indifference curves that we made in
Section 3.1 (page 91).
Hicksian substitution
effect Alternative to the Slutsky
equation for decomposing price
changes without recourse to
indifference curves.
In §3.1, we explain that an
indifference curve is convex
if the marginal rate of sub-
stitution diminishes as we
move down along the curve.
In §3.4, we explain how
information about consumer
preferences is revealed
through the consumption
choices that consumers
make.
This page intentionally left blank
179
So far, we have assumed that prices, incomes, and other variables are known with certainty. However, many of the choices that people make involve considerable uncertainty. Most people, for
example, bo
ow to finance large purchases, such as a house or a col-
lege education, and plan to pay for them out of future income. But for
most of us, future incomes are uncertain. Our earnings can go up or
down; we can be promoted or demoted, or even lose our jobs. And if
we delay buying a house or investing in a college education, we risk
price increases that could make such purchases less affordable. How
should we take these uncertainties into account when making major
consumption or investment decisions?
Sometimes we must choose how much risk to bear. What, for exam-
ple, should you do with your savings? Should you invest your money
in something safe, such as a savings account, or something riskier but
potentially more lucrative, such as the stock market? Another example
is the choice of a job or career. Is it better to work for a large, stable
company with job security but slim chance for advancement, or is it
etter to join (or form) a new venture that offers less job security but
more opportunity for advancement?
To answer such questions, we must examine the ways that people
can compare and choose among risky alternatives. We will do this by
taking the following steps:
1. In order to compare the riskiness of alternative choices, we need
to quantify risk. We therefore begin this chapter by discussing
measures of risk.
2. We will examine people’s preferences toward risk. Most people
find risk undesirable, but some people find it more undesirable
than others.
3. We will see how people can sometimes reduce or eliminate risk.
Sometimes risk can be reduced by diversification, by buying in-
surance, or by investing in additional information.
4. In some situations, people must choose the amount of risk
they wish to bear. A good example is investing in stocks or
onds. We will see that such investments involve trade-offs
etween the monetary gain that one can expect and the riski-
ness of that gain.
5.1 Dete
ing Crime 184
5.2 Business Executives
and the Choice of Risk 189
5.3 The Value of Title
Insurance When Buying
a House 193
5.4 The Value of Information
in an Online Consumer
Electronics Market 195
5.5 Doctors, Patients,
and the Value of
Information 195
5.6 Investing in the Stock
Market 203
LIST Of ExaMPLES
5.1 Describing Risk 180
5.2 Preferences Toward Risk 185
5.3 Reducing Risk 190
*5.4 The Demand for Risky
assets 196
CHaPTER OuTLInE
uncertainty and
Consumer Behavio
CHaPTER 5
180 PaRT 2 Producers, Consumers, and Competitive Markets
In a world of uncertainty, individual behavior may sometimes seem
unpredictable, even i
ational, and perhaps contrary to the basic assumptions
of consumer theory. In the final chapter of this book, we offer an overview of the
flourishing field of behavioral economics, which, by introducing important ideas
from psychology, has
oadened and enriched the study of microeconomics.
5.1 Describing Risk
To describe risk quantitatively, we begin by listing all the possible outcomes
of a particular action or event, as well as the likelihood that each outcome will
occur.1 Suppose, for example, that you are considering investing in a company
that explores for offshore oil. If the exploration effort is successful, the com-
pany’s stock will increase from $30 to $40 per share; if not, the price will fall
to $20 per share. Thus there are two possible future outcomes: a $40-per-share
price and a $20-per-share price.
Probability
Probability is the likelihood that a given outcome will occur. In our example,
the probability that the oil exploration project will be successful might be 1/4
and the probability that it is unsuccessful 3/4. (Note that the probabilities for all
possible events must add up to 1.)
Our interpretation of probability can depend on the nature of the uncertain
event, on the beliefs of the people involved, or both. One objective interpretation
of probability relies on the frequency with which certain events tend to occur.
Suppose we know that of the last 100 offshore oil explorations, 25 have suc-
ceeded and 75 failed. In that case, the probability of success of 1/4 is objective
ecause it is based directly on the frequency of similar experiences.
But what if there are no similar past experiences to help measure probability?
In such instances, objective measures of probability cannot be deduced and more
subjective measures are needed. Subjective probability is the perception that an out-
come will occur. This perception may be based on a person’s judgment or experi-
ence, but not necessarily on the frequency with which a particular outcome has
actually occu
ed in the past. When probabilities are subjectively determined, dif-
ferent people may attach different probabilities to different outcomes and thereby
make different choices. For example, if the search for oil were to take place in an
area where no previous searches had ever occu
ed, I might attach a higher sub-
jective probability than you to the chance that the project will succeed: Perhaps
I know more about the project or I have a better understanding of the oil business
and can therefore make better use of our common information. Either different
information or different abilities to process the same information can cause sub-
jective probabilities to vary among individuals.
Regardless of the interpretation of probability, it is used in calculating two
important measures that help us describe and compare risky choices. One
measure tells us the expected value and the other the variability of the possible
outcomes.
probability Likelihood that a
given outcome will occur.
1Some people distinguish between uncertainty and risk along the lines suggested some 60 years ago
y economist Frank Knight. Uncertainty can refer to situations in which many outcomes are possible
ut the likelihood of each is unknown. Risk then refers to situations in which we can list all possible
outcomes and know the likelihood of each occu
ing. In this chapter, we will always refer to risky
situations, but will simplify the discussion by using uncertainty and risk interchangeably.
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 181
Expected Value
The expected value associated with an uncertain situation is a weighted
average of the payoffs or values associated with all possible outcomes.
The probabilities of each outcome are used as weights. Thus the expected
value measures the central tendency—the payoff or value that we would expect
on average.
Our offshore oil exploration example had two possible outcomes: Success
yields a payoff of $40 per share, failure a payoff of $20 per share. Denoting
“probability of” by Pr, we express the expected value in this case as
Expected value = Pr(success)($40/share) + Pr(failure)($20/share)
= (1/4)($40/share) + (3/4)($20/share) = $25/share
More generally, if there are two possible outcomes having payoffs X1 and X2
and if the probabilities of each outcome are given by Pr1 and Pr2, then the
expected value is
E(X) = Pr1X1 + Pr2X2
When there are n possible outcomes, the expected value becomes
E(X) = Pr1X1 + Pr2X2 + g + PrnXn
Variability
Variability is the extent to which the possible outcomes of an uncertain
situation differ. To see why variability is important, suppose you are choos-
ing between two part-time summer sales jobs that have the same expected
income ($1500). The first job is based entirely on commission—the income
earned depends on how much you sell. There are two equally likely payoffs
for this job: $2000 for a successful sales effort and $1000 for one that is less
successful. The second job is salaried. It is very likely (.99 probability) that
you will earn $1510, but there is a .01 probability that the company will
go out of business, in which case you would earn only $510 in severance
pay. Table 5.1 summarizes these possible outcomes, their payoffs, and their
probabilities.
Note that these two jobs have the same expected income. For Job 1, expected
income is .5($2000) + .5($1000) = $1500; for Job 2, it is .99($1510) + .01($510) =
$1500. However, the variability of the possible payoffs is different. We measure
variability by recognizing that large differences between actual and expected
payoffs (whether positive or negative) imply greater risk. We call these differences
deviations. Table 5.2 shows the deviations of the possible income from the ex-
pected income from each job.
expected value Probability-
weighted average of the payoffs
associated with all possible
outcomes.
payoff Value associated with
a possible outcome.
variability Extent to which
possible outcomes of an uncertain
event differ.
deviation Difference between
expected payoff and actual payoff.
Table 5.1 Income from SaleS JobS
OutcOme 1 OutcOme 2
expected
IncOme ($)prObabIlIty IncOme ($) prObabIlIty IncOme ($)
Job 1: commission .5 2000 .5 1000 1500
Job 2: Fixed Salary .99 1510 .01 510 1500
182 PaRT 2 Producers, Consumers, and Competitive Markets
By themselves, deviations do not provide a measure of variability. Why?
Because they are sometimes positive and sometimes negative, and as you
can see from Table 5.2, the average of the probability-weighted deviations
is always 0.2 To get around this problem, we square each deviation, yielding
numbers that are always positive. We then measure variability by calculating
the standard deviation: the square root of the average of the squares of the
deviations of the payoffs associated with each outcome from their expected
values.3
Table 5.3 shows the calculation of the standard deviation for our example.
Note that the average of the squared deviations under Job 1 is given by
.5($250,000) + .5($250,000) = $250,000
The standard deviation is therefore equal to the square root of $250,000, or
$500. Likewise, the probability-weighted average of the squared deviations
under Job 2 is
.99($100) + .01($980,100) = $9900
The standard deviation is the square root of $9900, or $99.50. Thus the second
job is much less risky than the first; the standard deviation of the incomes is
much lower.4
The concept of standard deviation applies equally well when there are many
outcomes rather than just two. Suppose, for example, that the first summer job
yields incomes ranging from $1000 to $2000 in increments of $100 that are all
equally likely. The second job yields incomes from $1300 to $1700 (again in in-
crements of $100) that are also equally likely. Figure 5.1 shows the alternatives
graphically. (If there had been only two equally probable outcomes, then the
figure would be drawn as two vertical lines, each with a height of 0.5.)
standard deviation Square
oot of the weighted average of
the squares of the deviations of
the payoffs associated with each
outcome from their expected
values.
Table 5.2 DevIatIonS from expecteD Income ($)
OutcOme 1 devIatIOn OutcOme 2 devIatIOn
Job 1 2000 500 1000 -500
Job 2 1510 10 510 -990
Table 5.3 calculatIng varIance ($)
OutcOme 1
devIatIOn
Squared OutcOme 2
devIatIOn
Squared
WeIghted average
devIatIOn Squared
Standard
devIatIOn
Job 1 2000 250,000 1000 250,000 250,000 500
Job 2 1510 100 510 980,100 9900 99.50
2For Job 1, the average deviation is .5($500) + .5(- $500) = 0; for Job 2 it is .99($10) + .01(- $990) = 0.
3Another measure of variability, variance, is the square of the standard deviation.
4In general, when there are two outcomes with payoffs X1 and X2, occu
ing with probability Pr1
and Pr2, and E(X) is the expected value of the outcomes, the standard deviation is given by s, where
s2 = Pr1[(X1 - E(X))2] + Pr2[(X2 - E(X))2]
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 183
You can see from Figure 5.1 that the first job is riskier than the second. The
“spread” of possible payoffs for the first job is much greater than the spread for
the second. As a result, the standard deviation of the payoffs associated with the
first job is greater than that associated with the second.
In this particular example, all payoffs are equally likely. Thus the curves
describing the probabilities for each job are flat. In many cases, however, some
payoffs are more likely than others. Figure 5.2 shows a situation in which the
most extreme payoffs are the least likely. Again, the salary from Job 1 has a
greater standard deviation. From this point on, we will use the standard devia-
tion of payoffs to measure the degree of risk.
Decision Making
Suppose you are choosing between the two sales jobs described in our original ex-
ample. Which job would you take? If you dislike risk, you will take the second job:
It offers the same expected income as the first but with less risk. But suppose we
add $100 to each of the payoffs in the first job, so that the expected payoff increases
from $1500 to $1600. Table 5.4 gives the new earnings and the squared deviations.
FiguRE 5.1
outcome probabIlItIeS
for two JobS
The distribution of payoffs associated with Job
1 has a greater spread and a greater standard
deviation than the distribution of payoffs as-
sociated with Job 2. Both distributions are flat
ecause all outcomes are equally likely.
Probability
$1000 $1500 $2000 Income
Job 1
Job 2
0.2
0.1
FiguRE 5.2
unequal probabIlIty
outcomeS
The distribution of payoffs associated with Job 1
has a greater spread and a greater standard
deviation than the distribution of payoffs associ-
ated with Job 2. Both distributions are peaked
ecause the extreme payoffs are less likely than
those near the middle of the distribution.
Probability
0.3
0.2
0.1
Income$2000$1500$1000
Job 1
Job 2
184 PaRT 2 Producers, Consumers, and Competitive Markets
The two jobs can now be described as follows:
Job 1: Expected Income = $1600 Standard Deviation = $500
Job 2: Expected Income = $1500 Standard Deviation = $99.50
Job 1 offers a higher expected income but is much riskier than Job 2. Which job
is prefe
ed depends on the individual. While an aggressive entrepreneur who
doesn’t mind taking risks might choose Job 1, with the higher expected income
and higher standard deviation, a more conservative person might choose the
second job.
People’s attitudes toward risk affect many of the decisions they make. In
Example 5.1 we will see how attitudes toward risk affect people’s willingness
to
eak the law, and how this has implications for the fines that should be set
for various violations. Then in Section 5.2, we will further develop our theory
of consumer choice by examining people’s risk preferences in greater detail.
Table 5.4 IncomeS from SaleS JobS—moDIfIeD ($)
OutcOme 1
devIatIOn
Squared OutcOme 2
devIatIOn
Squared
expected
IncOme
Standard
devIatIOn
Job 1 2100 250,000 1100 250,000 1600 500
Job 2 1510 100 510 980,100 1500 99.50
5This discussion builds indirectly on Gary S. Becker, “Crime and Punishment: An Economic
Approach,” Journal of Political Economy (March/April 1968): 169–217. See also A. Mitchell Polinsky
and Steven Shavell, “The Optimal Tradeoff Between the Probability and the Magnitude of Fines,”
American Economic Review 69 (December 1979): 880–91.
ExaMPlE 5.1 Dete
Ing crIme
fines may be better than incarceration in dete
ing
certain types of crimes, such as speeding, double-
parking, tax evasion, and air polluting.5 a person
choosing to violate the law in these ways has good
information and can reasonably be assumed to be
ehaving rationally.
Other things being equal, the greater the fine, the
more a potential criminal will be discouraged from
committing the crime. for example, if it cost nothing
to catch criminals, and if the crime imposed a calcu-
lable cost of $1000 on society, we might choose to
catch all violations and impose a fine of $1000 on
each. This practice would discourage people whose
enefit from engaging in the activity was less than the
$1000 fine.
In practice, however, it is very costly to catch
law
eakers. Therefore, we save on administrative
costs by imposing relatively high fines (which are
no more costly to collect than low fines), while
allocating resources so that only a fraction of the
violators are apprehended. Thus the size of the fine
that must be imposed to discourage criminal behav-
ior depends on the attitudes toward risk of potential
violators.
Suppose that a city wants to deter people from
double-parking. By double-parking, a typical resi-
dent saves $5 in terms of his own time for engaging
in activities that are more pleasant than searching for
a parking space. If it costs nothing to catch a double-
parker, a fine of just over $5—say, $6—should be
assessed every time he double-parks. This policy will
ensure that the net benefit of double-parking (the $5
enefit less the $6 fine) would be less than zero. Our
citizen will therefore choose to obey the law. In fact,
all potential violators whose benefit was less than
or equal to $5 would be discouraged, while a few
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 185
whose benefit was greater than $5 (say, someone
who double-parks because of an emergency) would
violate the law.
In practice, it is too costly to catch all violators.
fortunately, it’s also unnecessary. The same deter-
ence effect can be obtained by assessing a fine of
$50 and catching only one in ten violators (or per-
haps a fine of $500 with a one-in-100 chance of
eing caught). In each case, the expected penalty is $5,
i.e., [$50][.1] or [$500][.01]. a policy that combines
a high fine and a low probability of apprehension is
likely to reduce enforcement costs. This approach is
especially effective if drivers don’t like to take risks. In
our example, a $50 fine with a .1 probability of being
caught might discourage most people from violating
the law. We will examine attitudes toward risk in the
next section.
a new type of crime that has become a serious
problem for music and movie producers is digital
piracy; it is particularly difficult to catch and fines
are rarely imposed. nevertheless, fines that are lev-
ied are often very high. In 2009, a woman was fined
$1.9 million for illegally downloading 24 songs. That
amounts to a fine of $80,000 per song.
5.2 Preferences Toward Risk
We used a job example to show how people might evaluate risky outcomes, but
the principles apply equally well to other choices. In this section, we concen-
trate on consumer choices generally and on the utility that consumers obtain
from choosing among risky alternatives. To simplify things, we’ll consider the
utility that a consumer gets from his or her income—or, more appropriately, the
market basket that the consumer’s income can buy. We now measure payoffs,
therefore, in terms of utility rather than dollars.
Figure 5.3 (a) shows how we can describe one woman’s preferences toward
isk. The curve 0E, which gives her utility function, tells us the level of utility
(on the vertical axis) that she can attain for each level of income (measured in
thousands of dollars on the horizontal axis). The level of utility increases from
10 to 16 to 18 as income increases from $10,000 to $20,000 to $30,000. But note
that marginal utility is diminishing, falling from 10 when income increases from
0 to $10,000, to 6 when income increases from $10,000 to $20,000, and to 2 when
income increases from $20,000 to $30,000.
Now suppose that our consumer has an income of $15,000 and is considering
a new but risky sales job that will either double her income to $30,000 or cause it
to fall to $10,000. Each possibility has a probability of .5. As Figure 5.3 (a) shows,
the utility level associated with an income of $10,000 is 10 (at point A) and the
utility level associated with an income of $30,000 is 18 (at E). The risky job must
e compared with the cu
ent $15,000 job, for which the utility is 13.5 (at B).
To evaluate the new job, she can calculate the expected value of the resulting
income. Because we are measuring value in terms of her utility, we must calcu-
late the expected utility E(u) that she can obtain. The expected utility is the sum
of the utilities associated with all possible outcomes, weighted by the probability that
each outcome will occur. In this case expected utility is
E(u) = (1/2)u($10,000) + (1/2)u($30,000) = 0.5(10) + 0.5(18) = 14
The risky new job is thus prefe
ed to the original job because the expected
utility of 14 is greater than the original utility of 13.5.
The old job involved no risk—it guaranteed an income of $15,000 and a util-
ity level of 13.5. The new job is risky but offers both a higher expected income
($20,000) and, more importantly, a higher expected utility. If the woman wishes
to increase her expected utility, she will take the risky job.
In §3.1, we explained that
a utility function assigns a
level of utility to each pos-
sible market basket.
In §3.5, marginal utility is
described as the additional
satisfaction obtained by
consuming an additional
amount of a good.
expected utility Sum of
the utilities associated with all
possible outcomes, weighted by
the probability that each outcome
will occur.
186 PaRT 2 Producers, Consumers, and Competitive Markets
Different Preferences Toward Risk
People differ in their willingness to bear risk. Some are risk averse, some risk
loving, and some risk neutral. An individual who is risk averse prefers a certain
given income to a risky income with the same expected value. (Such a person
has a diminishing marginal utility of income.) Risk aversion is the most com-
mon attitude toward risk. To see that most people are risk averse most of the
time, note that most people not only buy life insurance, health insurance, and
car insurance, but also seek occupations with relatively stable wages.
Figure 5.3 (a) applies to a woman who is risk averse. Suppose hypothetically
that she can have either a certain income of $20,000, or a job yielding an income
of $30,000 with probability .5 and an income of $10,000 with probability .5 (so
that the expected income is also $20,000). As we saw, the expected utility of the
uncertain income is 14—an average of the utility at point A(10) and the utility at
E(18)—and is shown by F. Now we can compare the expected utility associated
with the risky job to the utility generated if $20,000 were earned without risk.
This latter utility level, 16, is given by D in Figure 5.3 (a). It is clearly greater
than the expected utility of 14 associated with the risky job.
For a risk-averse person, losses are more important (in terms of the change in
utility) than gains. Again, this can be seen from Figure 5.3 (a). A $10,000 increase
in income, from $20,000 to $30,000, generates an increase in utility of two units; a
$10,000 decrease in income, from $20,000 to $10,000, creates a loss of utility of six
units.
A person who is risk neutral is indifferent between a certain income and an un-
certain income with the same expected value. In Figure 5.3 (c) the utility associated
with a job generating an income of either $10,000 or $30,000 with equal probability
is 12, as is the utility of receiving a certain income of $20,000. As you can see from
the figure, the marginal utility of income is constant for a risk-neutral person.6
Finally, an individual who is risk loving prefers an uncertain income to a
certain one, even if the expected value of the uncertain income is less than that
of the certain income. Figure 5.3 (b) shows this third possibility. In this case, the
expected utility of an uncertain income, which will be either $10,000 with prob-
ability .5 or $30,000 with probability .5, is higher than the utility associated with
a certain income of $20,000. Numerically,
E(u) = .5u($10,000) + .5u($30,000) = .5(3) + .5(18) = 10.5 7 u($20,000) = 8
Of course, some people may be averse to some risks and act like risk lovers with
espect to others. For example, many people purchase life insurance and are
conservative with respect to their choice of jobs, but still enjoy gambling. Some
criminologists might describe criminals as risk lovers, especially if they commit
crimes despite a high prospect of apprehension and punishment. Except for
such special cases, however, few people are risk loving, at least with respect to
major purchases or large amounts of income or wealth.
ISk premIum The risk premium is the maximum amount of money that
a risk-averse person will pay to avoid taking a risk. In general, the magnitude
of the risk premium depends on the risky alternatives that the person faces.
isk averse Condition of
prefe
ing a certain income to
a risky income with the same
expected value.
isk neutral Condition of
eing indifferent between a
certain income and an uncertain
income with the same expected
value.
isk loving Condition of
prefe
ing a risky income to a
certain income with the same
expected value.
isk premium Maximum
amount of money that a risk-
averse person will pay to avoid
taking a risk.
6Thus, when people are risk neutral, the income they earn can be used as an indicator of well-being.
A government policy that doubles incomes would then also double their utility. At the same time,
government policies that alter the risks that people face, without changing their expected incomes,
would not affect their well-being. Risk neutrality allows a person to avoid the complications that
might be associated with the effects of governmental actions on the riskiness of outcomes.
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 187
To  determine the risk premium, we have reproduced the utility function of
Figure 5.3 (a) in Figure 5.4 and extended it to an income of $40,000. Recall that
an expected utility of 14 is achieved by a woman who is going to take a risky
job with an expected income of $20,000. This outcome is shown graphically by
drawing a horizontal line to the vertical axis from point F, which bisects straight
FiguRE 5.3
ISk averSe, rISk lovIng, anD rISk neutral
People differ in their preferences toward risk. In (a), a consumer’s marginal utility diminishes as income in-
creases. The consumer is risk averse because she would prefer a certain income of $20,000 (with a utility of 16)
to a gamble with a .5 probability of $10,000 and a .5 probability of $30,000 (and expected utility of 14). In (b),
the consumer is risk loving: She would prefer the same gamble (with expected utility of 10.5) to the certain
income (with a utility of 8). Finally, the consumer in (c) is risk neutral and indifferent between certain and uncer-
tain events with the same expected income.
Utility
Income ($1000)
302010
18
0
E
C
A
12
6
Utility
Income ($1000)
302010
18
8
3
0
E
C
A
Utility
Income ($1000)
3020161510
D
E
C
FB
A
10
13.5
14
16
18
0
(a)
(b) (c)
188 PaRT 2 Producers, Consumers, and Competitive Markets
line AE (thus representing an average of $10,000 and $30,000). But the utility
level of 14 can also be achieved if the woman has a certain income of $16,000,
as shown by dropping a vertical line from point C. Thus, the risk premium of
$4000, given by line segment CF, is the amount of expected income ($20,000 mi-
nus $16,000) that she would give up in order to remain indifferent between the
isky job and a hypothetical job that would pay her a certain income of $16,000.
ISk averSIon anD Income The extent of an individual’s risk aversion de-
pends on the nature of the risk and on the person’s income. Other things being
equal, risk-averse people prefer a smaller variability of outcomes. We saw that
when there are two outcomes—an income of $10,000 and an income of $30,000—
the risk premium is $4000. Now consider a second risky job, also illustrated in
Figure 5.4. With this job, there is a .5 probability of receiving an income of $40,000,
with a utility level of 20, and a .5 probability of getting an income of $0, with a utility
level of 0. The expected income is again $20,000, but the expected utility is only 10:
Expected utility = .5u($0) + .5u($40,000) = 0 + .5(20) = 10
Compared to a hypothetical job that pays $20,000 with certainty, the per-
son holding this risky job gets 6 fewer units of expected utility: 10 rather than
16 units. At the same time, however, this person could also get 10 units of utility
from a job that pays $10,000 with certainty. Thus the risk premium in this case is
$10,000, because this person would be willing to give up $10,000 of her $20,000
expected income to avoid bearing the risk of an uncertain income. The greater
the variability of income, the more the person would be willing to pay to avoid
the risky situation.
FiguRE 5.4
ISk premIum
The risk premium, CF, measures the amount of income that an individual would give up
to leave her indifferent between a risky choice and a certain one. Here, the risk premium is
$4000 because a certain income of $16,000 (at point C ) gives her the same expected utility
(14) as the uncertain income (a .5 probability of being at point A and a .5 probability of being
at point E) that has an expected value of $20,000.
10 16 20 30 40
Risk Premium
18
20
14
Utility
Income ($1000)
10
A
C
F
E
G
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 189
ISk averSIon anD InDIfference curveS We can also describe the
extent of a person’s risk aversion in terms of indifference curves that relate ex-
pected income to the variability of income, where the latter is measured by the
standard deviation. Figure 5.5 shows such indifference curves for two individu-
als, one who is highly risk averse and another who is only slightly risk averse.
Each indifference curve shows the combinations of expected income and stan-
dard deviation of income that give the individual the same amount of utility.
Observe that all of the indifference curves are upward sloping: Because risk
is undesirable, the greater the amount of risk, the greater the expected income
needed to make the individual equally well off.
Figure 5.5 (a) describes an individual who is highly risk averse. Observe
that in order to leave this person equally well off, an increase in the stan-
dard deviation of income requires a large increase in expected income.
FiguRE 5.5
ISk averSIon anD InDIfference curveS
Part (a) applies to a person who is highly risk averse: An increase in this individual’s stan-
dard deviation of income requires a large increase in expected income if he or she is to
emain equally well off. Part (b) applies to a person who is only slightly risk averse: An
increase in the standard deviation of income requires only a small increase in expected
income if he or she is to remain equally well off.
U3
U2
U1
Expected
income
Standard deviation of income
(a)
U3
U2
U1
Expected
income
Standard deviation of income
(b)
In §3.1, we define an
indifference curve as all
market baskets that gener-
ate the same level of satis-
faction for a consumer.
ExaMPlE 5.2 buSIneSS executIveS anD the choIce of rISk
are business executives more risk loving than most
people? When they are presented with alternative
strategies, some risky, some safe, which do they
choose? In one study, 464 executives were asked to
espond to a questionnaire describing risky situations
that an individual might face as vice president of a
hypothetical company.7 Respondents were presented
with four risky events, each of which had a given
7This example is based on Kenneth R. MacCrimmon and Donald A. Wehrung, “The Risk In-Basket,”
Journal of Business 57 (1984): 367–87.
190 PaRT 2 Producers, Consumers, and Competitive Markets
Figure  5.5  (b) applies to a slightly risk-averse person. In this case, a large
increase in the standard deviation of income requires only a small increase in
expected income.
We will return to the use of indifference curves as a means of describing risk
aversion in Section 5.4, where we discuss the demand for risky assets. First,
however, we will turn to the ways in which an individual can reduce risk.
5.3 Reducing Risk
As the recent growth in state lotteries shows, people sometimes choose risky
alternatives that suggest risk-loving rather than risk-averse behavior. Most
people, however, spend relatively small amounts on lottery tickets and casinos.
When more important decisions are involved, they are generally risk averse. In
this section, we describe three ways by which both consumers and businesses
commonly reduce risks: diversification, insurance, and obtaining more information
about choices and payoffs.
Diversification
Recall the old saying, “Don’t put all your eggs in one basket.” Ignoring this
advice is unnecessarily risky: If your basket turns out to be a bad bet, all will
e lost. Instead, you can reduce risk through diversification: allocating your
esources to a variety of activities whose outcomes are not closely related.
Suppose, for example, that you plan to take a part-time job selling appliances
on a commission basis. You can decide to sell only air conditioners or only heat-
ers, or you can spend half your time selling each. Of course, you can’t be sure
how hot or cold the weather will be next year. How should you apportion your
time in order to minimize the risk involved?
diversification Practice
of reducing risk by allocating
esources to a variety of activities
whose outcomes are not closely
elated.
probability of a favorable and unfavorable outcome.
The payoffs and probabilities were chosen so that
each event had the same expected value. In increas-
ing order of the risk involved (as measured by the
difference between the favorable and unfavorable
outcomes), the four items were:
1. a lawsuit involving a patent violation
2. a customer threatening to buy from a competito
3. a union dispute
4. a joint venture with a competito
To gauge their willingness to take or avoid risks,
esearchers asked respondents a series of questions
egarding business strategy. In one situation, they
could pursue a risky strategy with the possibility of
a high return right away or delay making a choice
until the outcomes became more certain and the risk
was reduced. In another situation, respondents could
opt for an immediately risky but potentially profit-
able strategy that could lead to a promotion, or they
could delegate the decision to someone else, which
would protect their job but eliminate the promotion
possibility.
The study found that executives vary substantially
in their preferences toward risk. Roughly 20  percent
indicated that they were relatively risk neutral;
40 percent opted for the more risky alternatives; and
20 percent were clearly risk averse (20 percent did
not respond). More importantly, executives (includ-
ing those who chose risky alternatives) typically
made efforts to reduce or eliminate risk, usually by
delaying decisions and collecting more information.
Some have argued that a cause of the financial
crisis of 2008 was excessive risk-taking by bankers
and Wall Street executives who could earn huge bo-
nuses if their ventures succeeded but faced very little
downside if the ventures failed. The u.S. Treasury
Department’s Troubled asset Relief Program (TaRP)
ailed out some of the banks, but so far has been
unable to impose constraints on “unnecessary and
excessive” risk-taking by banks’ executives.
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 191
Table 5.5 Income from SaleS of applIanceS ($)
hOt Weather cOld Weathe
air conditioner sales 30,000 12,000
heater sales 12,000 30,000
Risk can be minimized by diversification—by allocating your time so that you
sell two or more products (whose sales are not closely related) rather than a
single product. Suppose there is a 0.5 probability that it will be a relatively hot
year, and a 0.5 probability that it will be cold. Table 5.5 gives the earnings that
you can make selling air conditioners and heaters.
If you sell only air conditioners or only heaters, your actual in-
come will be either $12,000 or $30,000, but your expected income will be
$21,000 (.5 [$30,000] + .5 [$12,000]). But suppose you diversify by dividing
your time evenly between the two products. In that case, your income will cer-
tainly be $21,000, regardless of the weather. If the weather is hot, you will earn
$15,000 from air conditioner sales and $6000 from heater sales; if it is cold, you
will earn $6000 from air conditioners and $15,000 from heaters. In this instance,
diversification eliminates all risk.
Of course, diversification is not always this easy. In our example, heater and
air conditioner sales are negatively co
elated variables—they tend to move
in opposite directions; whenever sales of one are strong, sales of the other are
weak. But the principle of diversification is a general one: As long as you can
allocate your resources toward a variety of activities whose outcomes are not
closely related, you can eliminate some risk.
the Stock market Diversification is especially important for people who
invest in the stock market. On any given day, the price of an individual stock
can go up or down by a large amount, but some stocks rise in price while oth-
ers fall. An individual who invests all her money in a single stock (i.e., puts
all her eggs in one basket) is therefore taking much more risk than necessary.
Risk can be reduced—although not eliminated—by investing in a portfolio of
ten or twenty different stocks. Likewise, you can diversify by buying shares in
mutual funds: organizations that pool funds of individual investors to buy a
large number of different stocks. There are thousands of mutual funds available
today for both stocks and bonds. These funds are popular because they reduce
isk through diversification and because their fees are typically much lower
than the cost of assembling one’s own portfolio of stocks.
In the case of the stock market, not all risk is diversifiable. Although some
stocks go up in price when others go down, stock prices are to some extent
positively co
elated variables: They tend to move in the same direction in
esponse to changes in economic conditions. For example, the onset of a severe
ecession, which is likely to reduce the profits of many companies, may be ac-
companied by a decline in the overall market. Even with a diversified portfolio
of stocks, therefore, you still face some risk.
insurance
We have seen that risk-averse people are willing to pay to avoid risk. In fact, if
the cost of insurance is equal to the expected loss (e.g., a policy with an expected
loss of $1000 will cost $1000), risk-averse people will buy enough insurance to
ecover fully from any financial losses they might suffer.
negatively co
elated
variables Variables having
a tendency to move in opposite
directions.
mutual fund Organization
that pools funds of individual
investors to buy a large number of
different stocks or other financial
assets.
positively co
elated
variables Variables having a
tendency to move in the same
direction.
192 PaRT 2 Producers, Consumers, and Competitive Markets
Why? The answer is implicit in our discussion of risk aversion. Buying insur-
ance assures a person of having the same income whether or not there is a loss.
Because the insurance cost is equal to the expected loss, this certain income is
equal to the expected income from the risky situation. For a risk-averse con-
sumer, the guarantee of the same income regardless of the outcome generates
more utility than would be the case if that person had a high income when there
was no loss and a low income when a loss occu
ed.
To clarify this point, let’s suppose a homeowner faces a 10-percent prob-
ability that his house will be burglarized and he will suffer a $10,000 loss. Let’s
assume he has $50,000 worth of property. Table 5.6 shows his wealth in two
situations—with insurance costing $1000 and without insurance.
Note that expected wealth is the same ($49,000) in both situations. The
variability, however, is quite different. As the table shows, with no insurance
the standard deviation of wealth is $3000; with insurance, it is 0. If there is no
urglary, the uninsured homeowner gains $1000 relative to the insured home-
owner. But with a burglary, the uninsured homeowner loses $9000 relative to
the insured homeowner. Remember: for a risk-averse individual, losses count
more (in terms of changes in utility) than gains. A risk-averse homeowner,
therefore, will enjoy higher utility by purchasing insurance.
the law of large numberS Consumers usually buy insurance from com-
panies that specialize in selling it. Insurance companies are firms that offer in-
surance because they know that when they sell a large number of policies, they
face relatively little risk. The ability to avoid risk by operating on a large scale
is based on the law of large numbers, which tells us that although single events
may be random and largely unpredictable, the average outcome of many simi-
lar events can be predicted. For example, I may not be able to predict whether
a coin toss will come out heads or tails, but I know that when many coins are
flipped, approximately half will turn up heads and half tails. Likewise, if I am
selling automobile insurance, I cannot predict whether a particular driver will
have an accident, but I can be reasonably sure, judging from past experience,
what fraction of a large group of drivers will have accidents.
actuarIal faIrneSS By operating on a large scale, insurance companies
can be sure that over a sufficiently large number of events, total premiums
paid in will be equal to the total amount of money paid out. Let’s return to
our burglary example. A man knows that there is a 10-percent probability that
his house will be burgled; if it is, he will suffer a $10,000 loss. Prior to facing
this risk, he calculates the expected loss to be $1000 (.10 * $10,000). The risk
involved is considerable, however, because there is a 10-percent probability of
a large loss. Now suppose that 100 people are similarly situated and that all of
them buy burglary insurance from the same company. Because they all face a
10-percent probability of a $10,000 loss, the insurance company might charge
Table 5.6 the DecISIon to InSure ($)
InSurance
urglary
(pr = .1)
nO burglary
(pr = .9)
expected
Wealth
Standard
devIatIOn
no 40,000 50,000 49,000 3000
yes 49,000 49,000 49,000 0
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 193
each of them a premium of $1000. This $1000 premium generates an insurance
fund of $100,000 from which losses can be paid. The insurance company can
ely on the law of large numbers, which holds that the expected loss to the 100
individuals as a whole is likely to be very close to $1000 each. The total payout,
therefore, will be close to $100,000, and the company need not wo
y about los-
ing more than that.
When the insurance premium is equal to the expected payout, as in the
example above, we say that the insurance is actuarially fair. But because they
must cover administrative costs and make some profit, insurance companies
typically charge premiums above expected losses. If there are a sufficient num-
er of insurance companies to make the market competitive, these premiums
will be close to actuarially fair levels. In some states, however, insurance premi-
ums are regulated in order to protect consumers from “excessive” premiums.
We will examine government regulation of markets in detail in Chapters 9 and
10 of this book.
In recent years, some insurance companies have come to the view that
catastrophic disasters such as earthquakes are so unique and unpredictable
that they cannot be viewed as diversifiable risks. Indeed, as a result of losses
from past disasters, these companies do not feel that they can determine ac-
tuarially fair insurance rates. In California, for example, the state itself has
had to enter the insurance business to fill the gap created when private com-
panies refused to sell earthquake insurance. The state-run pool offers less
insurance coverage at higher rates than was previously offered by private
insurers.
actuarially fair
Characterizing a situation in
which an insurance premium is
equal to the expected payout.
ExaMPlE 5.3 the value of tItle InSurance when buyIng a houSe
Suppose you are buying your first
house. To close the sale, you will
need a deed that gives you clear
“ title.” Without such a clear title,
there is always a chance that the
seller of the house is not its true
owner. Of course, the seller could
e engaging in fraud, but it is more
likely that the seller is unaware of
the exact nature of his or her ownership rights. for
example, the owner may have bo
owed heavily,
using the house as “collateral” for a loan. Or the
property might ca
y with it a legal requirement
that limits the use to which it may be put.
Suppose you are willing to pay $300,000 for
the house, but you believe there is a one-in-twenty
chance that careful research will reveal that the seller
does not actually own the property. The property
would then be worth nothing. If there were no insur-
ance available, a risk-neutral person would bid at most
$285,000 for the property (.95[$300,000] + .05[0]).
However, if you expect to tie up
most of your assets in the house,
you would probably be risk averse
and, therefore, bid much less—
say, $230,000.
In situations such as this, it
is clearly in the interest of the
uyer to be sure that there is no
isk of a lack of full ownership.
The buyer does this by purchasing “title insurance.”
The title insurance company researches the history
of the property, checks to see whether any legal
liabilities are attached to it, and generally assures
itself that there is no ownership problem. The insur-
ance company then agrees to bear any remaining
isk that might exist.
Because the title insurance company is a special-
ist in such insurance and can collect the relevant
information relatively easily, the cost of title insur-
ance is often less than the expected value of the loss
involved. a fee of $1500 for title insurance is not
194 PaRT 2 Producers, Consumers, and Competitive Markets
The Value of information
People often make decisions based on limited information. If more information
were available, one could make better predictions and reduce risk. Because
information is a valuable commodity, people will pay for it. The value of
complete information is the difference between the expected value of a choice
when there is complete information and the expected value when information
is incomplete.
To see how information can be valuable, suppose you manage a clothing
store and must decide how many suits to order for the fall season. If you or-
der 100 suits, your cost is $180 per suit. If you order only 50 suits, your cost
increases to $200. You know that you will be selling suits for $300 each, but you
are not sure how many you can sell. All suits not sold can be returned, but for
only half of what you paid for them. Without additional information, you will
act on your belief that there is a .5 probability that you will sell 100 suits and a .5
probability that you will sell 50. Table 5.7 gives the profit that you would earn in
each of these two cases.
Without additional information, you would choose to buy 100 suits if you
were risk neutral, taking the chance that your profit might be either $12,000
or $1500. But if you were risk averse, you might buy 50 suits: In that case, you
would know for sure that your profit would be $5000.
With complete information, you can place the co
ect order regardless of
future sales. If sales were going to be 50 and you ordered 50 suits, your prof-
its would be $5000. If, on the other hand, sales were going to be 100 and you
ordered 100 suits, your profits would be $12,000. Because both outcomes are
equally likely, your expected profit with complete information would be $8500.
The value of information is computed as
Expected value with complete information: $8500
Less: Expected value with uncertainty (buy 100 suits): -6750
Equals: Value of complete information $1750
Thus it is worth paying up to $1750 to obtain an accurate prediction of
sales. Even though forecasting is inevitably imperfect, it may be worth
investing in a marketing study that provides a reasonable forecast of next
year’s sales.
value of complete
information Difference
etween the expected value of a
choice when there is complete
information and the expected
value when information is
incomplete.
Table 5.7 profItS from SaleS of SuItS ($)
SaleS OF 50 SaleS OF 100 expected prOFIt
uy 50 suits 5000 5000 5000
uy 100 suits 1500 12,000 6750
unusual, even though the expected loss can be much
higher. It is also in the interest of sellers to provide
title insurance, because all but the most risk-loving
uyers will pay much more for a house when it is in-
sured than when it is not. In fact, most states require
sellers to provide title insurance before a sale can
e completed. In addition, because mortgage lend-
ers are all concerned about such risks, they usually
equire new buyers to have title insurance before is-
suing a mortgage.
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 195
You might think that more information is always a good thing. As the following
example shows, however, that is not always the case.
8Michael Baye, John Morgan, and Patrick Scholten, “The Value of Information in an Online Electronics
Market.”Journal of Public Policy and Marketing 22 (2003): 17–25.
ExaMPlE 5.4 the value of InformatIon In an onlIne conSumer
electronIcS market
Internet-based price comparison sites offer a valu-
able informational resource to consumers, as shown
y a study of a leading price-comparison website,
Shopper.com. Researchers studied price information
provided to consumers on over 1,000 top-selling
electronics products for an 8-month period. They
found that consumers saved about 16% when using
this website versus shopping in the store, because the
website significantly reduced the cost of finding the
lowest priced product.8
The value of price comparison information is
not the same for everyone and for every product.
Competition matters. The study found that when only
two firms list prices on Shopper.com, consumers
save 11%. But the savings increase with the number
of competitors, jumping to 20% when more than 30
companies list prices.
One might think that the Internet will generate so
much information about prices that only the lowest-
price products will be sold in the long run, causing
the value of such information to eventually decline
to zero. So far, this has not been the case. There are
fixed costs for parties to both transmit and to acquire
information over the Internet. These include the costs
of maintaining servers and the fees that sites such
as Shopper.com charge to list prices at their sites.
The result is that prices are likely to continue to vary
widely as the Internet continues to grow and mature.
ExaMPlE 5.5 DoctorS, patIentS, anD the value of InformatIon
Suppose you were seriously ill and re-
quired major surgery. assuming you
wanted to get the best care possible, how
would you go about choosing a surgeon
and a hospital to provide that care? Many
people would ask their friends or their
primary-care physician for a recommen-
dation. although this might be helpful, a
truly informed decision would probably
equire more detailed information. for
example, how successful has a recom-
mended surgeon and her affiliated hos-
pital been in performing the particular
operation that you need? How many of her patients
have died or had serious complications from the
operation, and how do these numbers compare with
those for other surgeons and hospitals? This kind of
information is likely to be difficult or impossible for
most patients to obtain. Would patients be better off
if detailed information about the performance re-
cords of doctors and hospitals were readily available?
not necessarily. More informa-
tion is often, but not always, better.
Interestingly in this case, access to
performance information could actu-
ally lead to worse health outcomes.
Why? Because access to such informa-
tion would create two different incen-
tives that would affect the behavior
of both doctors and patients. first, it
would allow patients to choose doc-
tors with better performance records,
which creates an incentive for doctors
to perform better. That is a good thing.
But second, it would encourage doctors to limit
their practices to patients who are in relatively good
health. The reason is that very old or very sick pa-
tients are more likely to have complications or die as
a result of treatment; doctors who treat such patients
are likely to have worse performance records (other
factors being equal). To the extent that doctors would
e judged according to performance, they would
196 PaRT 2 Producers, Consumers, and Competitive Markets
*5.4 The Demand for Risky assets
Most people are risk averse. Given a choice, they prefer fixed monthly incomes
to those which, though equally large on average, fluctuate randomly from
month to month. Yet many of these same people will invest all or part of their
savings in stocks, bonds, and other assets that ca
y some risk. Why do risk-
averse people invest in the stock market and thereby risk losing part or all of
their investments?10 How do people decide how much risk to bear when mak-
ing investments and planning for the future? To answer these questions, we
must examine the demand for risky assets.
assets
An asset is something that provides a f low of money or services to its owner. A
home, an apartment building, a savings account, or shares of General Motors
stock are all assets. A home, for example, provides a flow of housing services
to its owner, and, if the owner did not wish to live there, could be rented out,
thereby providing a monetary flow. Likewise, apartments can be rented out,
providing a flow of rental income to the owner of the building. A savings
account pays interest (usually every day or every month), which is usually
einvested in the account.
asset Something that provides
a flow of money or services to its
owner.
have an incentive to avoid treating very old or sick
patients. as a result, such patients would find it dif-
ficult or impossible to obtain treatment.
Whether more information is better depends on
which effect dominates—the ability of patients to
make more informed choices versus the incentive
for doctors to avoid very sick patients. In a recent
study, economists examined the impact of the man-
datory “report cards” introduced in new york and
Pennsylvania in the early 1990s to evaluate outcomes
of coronary bypass surgeries.9 They analyzed hospital
choices and outcomes for all elderly heart attack pa-
tients and patients receiving coronary bypass surgery
in the united States from 1987 through 1994. By
comparing trends in new york and Pennsylvania to
the trends in other states, they could determine the
effect of the increased information made possible
y the availability of report cards. They found that
although report cards improved matching of patients
with hospitals and doctors, they also caused a shift
in treatment from sicker patients towards healthier
ones. Overall, this led to worse outcomes, especially
among sicker patients. Thus the study concluded that
eport cards reduced welfare.
The medical profession has responded to this
problem to some extent. for example, in 2010, car-
diac surgery programs across the country voluntarily
eported the results of coronary-artery bypass graft-
ing procedures. Each program was rated with one
to three stars, but this time the ratings were “risk ad-
justed” to reduce the incentive for doctors to choose
less risky patients.
More information often improves welfare because
it allows people to reduce risk and to take actions that
might reduce the effect of bad outcomes. However,
as this example makes clear, information can cause
people to change their behavior in undesirable ways.
We will discuss this issue further in Chapter 17.
9David Dranove, Daniel Kessler, Mark McClennan, and Mark Satterthwaite, “Is More Information
Better? The Effects of ’Report Cards’ on Health Care Providers,” Journal of Political Economy 3 (June
2003): 555–558.
10Most Americans have at least some money invested in stocks or other risky assets, though often
indirectly. For example, many people who hold full-time jobs have shares in pension funds under-
written in part by their own salary contributions and in part by employer contributions. Usually
such funds are partly invested in the stock market.
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 197
The monetary flow that one receives from asset ownership can take the form
of an explicit payment, such as the rental income from an apartment building:
Every month, the landlord receives rent checks from the tenants. Another form
of explicit payment is the dividend on shares of common stock: Every three
months, the owner of a share of General Motors stock receives a quarterly divi-
dend payment.
But sometimes the monetary flow from ownership of an asset is implicit: It
takes the form of an increase or decrease in the price or value of the asset. An
increase in the value of an asset is a capital gain; a decrease is a capital loss. For
example, as the population of a city grows, the value of an apartment build-
ing may increase. The owner of the building will then earn a capital gain
eyond the rental income. The capital gain is unrealized until the building is
sold because no money is actually received until then. There is, however, an
implicit monetary flow because the building could be sold at any time. The
monetary flow from owning General Motors stock is also partly implicit. The
price of the stock changes from day to day, and each time it does, owners
gain or lose.
Risky and Riskless assets
A risky asset provides a monetary flow that is at least in part random. In other
words, the monetary flow is not known with certainty in advance. A share of
General Motors stock is an obvious example of a risky asset: You cannot know
whether the price of the stock will rise or fall over time, nor can you even be
sure that the company will continue to pay the same (or any) dividend per
share. Although people often associate risk with the stock market, most other
assets are also risky.
An apartment building is one example. You cannot know how much land
values will rise or fall, whether the building will be fully rented all the time,
or even whether the tenants will pay their rents promptly. Corporate bonds
are another example—the issuing corporation could go bankrupt and fail to
pay bond owners their interest and principal. Even long-term U.S. govern-
ment bonds that mature in 10 or 20 years are risky. Although it is highly
unlikely that the federal government will go bankrupt, the rate of inflation
could unexpectedly increase and make future interest payments and the
eventual repayment of principal worth less in real terms, thereby reducing
the value of the bonds.
In contrast, a riskless (or risk-free) asset pays a monetary flow that is known
with certainty. Short-term U.S. government bonds—called Treasury bills—are
iskless, or almost riskless. Because they mature in a few months, there is very
little risk from an unexpected increase in the rate of inflation. You can also be
easonably confident that the U.S. government will not default on the bond (i.e.,
efuse to pay back the holder when the bond comes due). Other examples of
iskless or almost riskless assets include passbook savings accounts and short-
term certificates of deposit.
asset Returns
People buy and hold assets because of the monetary flows they provide. To
compare assets with each other, it helps to think of this monetary flow relative
to an asset’s price or value. The return on an asset is the total monetary flow it
yields—including capital gains or losses—as a fraction of its price. For example, a
ond worth $1000 today that pays out $100 this year (and every year) has a
isky asset asset that provides
an uncertain flow of money or
services to its owner.
iskless (or risk-free)
asset asset that provides a
flow of money or services that is
known with certainty.
eturn Total monetary flow of
an asset as a fraction of its price.
198 PaRT 2 Producers, Consumers, and Competitive Markets
eturn of 10 percent.11 If an apartment building was worth $10 million last year,
increased in value to $11 million this year, and also provided rental income (af-
ter expenses) of $0.5 million, it would have yielded a return of 15 percent over
the past year. If a share of General Motors stock was worth $80 at the beginning
of the year, fell to $72 by the end of the year, and paid a dividend of $4, it will
have yielded a return of -5 percent (the dividend yield of 5 percent less the
capital loss of 10 percent).
When people invest their savings in stocks, bonds, land, or other assets, they
usually hope to earn a return that exceeds the rate of inflation. Thus, by delay-
ing consumption, they can buy more in the future than they can by spending
all their income now. Consequently, we often express the return on an asset in
eal—i.e., inflation-adjusted—terms. The real return on an asset is its simple (or
nominal) return less the rate of inflation. For example, with an annual inflation
ate of 5 percent, our bond, apartment building, and share of GM stock have
yielded real returns of 5 percent, 10 percent, and -10 percent, respectively.
expecteD verSuS actual returnS Because most assets are risky, an inves-
tor cannot know in advance what returns they will yield over the coming year.
For example, our apartment building might have depreciated in value instead
of appreciating, and the price of GM stock might have risen instead of fallen.
However, we can still compare assets by looking at their expected returns. The
expected return on an asset is the expected value of its return, i.e., the return that
it should earn on average. In some years, an asset’s actual return may be much
higher than its expected return and in some years much lower. Over a long pe-
iod, however, the average return should be close to the expected return.
Different assets have different expected returns. Table 5.8, for example,
shows that while the expected real return of a U.S. Treasury bill has been less
than 1 percent, the expected real return on a group of representative stocks
on the New York Stock Exchange has been close to 9 percent.12 Why would
anyone buy a Treasury bill when the expected return on stocks is so much
higher? Because the demand for an asset depends not just on its expected
eturn, but also on its risk: Although stocks have a higher expected return
than Treasury bills, they also ca
y much more risk. One measure of risk, the
standard deviation of the real annual return, is equal to about 20 percent for
common stocks, 8.4 percent for corporate bonds, and only 3.1 percent for U.S.
Treasury bills.
The numbers in Table 5.8 suggest that the higher the expected return on an
investment, the greater the risk involved. Assuming that one’s investments are
well diversified, this is indeed the case.13 As a result, the risk-averse investor
eal return Simple (or
nominal) return on an asset, less
the rate of inflation.
expected return Return that
an asset should earn on average.
actual return Return that an
asset earns.
11The price of a bond often changes during the course of a year. If the bond appreciates (or depreci-
ates) in value during the year, its return will be greater (or less) than 10 percent. In addition, the
definition of return given above should not be confused with the “internal rate of return,” which
is sometimes used to compare monetary flows occu
ing over a period of time. We discuss other
eturn measures in Chapter 15, when we deal with present discounted values.
12For some stocks, the expected return is higher, and for some it is lower. Stocks of smaller compa-
nies (e.g., some of those traded on the NASDAQ) have higher expected rates of return—and higher
eturn standard deviations.
13It is nondiversifiable risk that matters. An individual stock may be very risky but still have a low
expected return because most of the risk could be diversified away by holding a large number of
such stocks. Nondiversifiable risk, which arises from the fact that individual stock prices are co
e-
lated with the overall stock market, is the risk that remains even if one holds a diversified portfolio
of stocks. We discuss this point in detail in the context of the capital asset pricing model in Chapter 15.
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 199
must balance expected return against risk. We examine this trade-off in more
detail in the next section.
The Trade-Off Between Risk and Return
Suppose a woman wants to invest her savings in two assets—Treasury bills,
which are almost risk free, and a representative group of stocks. She must de-
cide how much to invest in each asset. She might, for instance, invest only in
Treasury bills, only in stocks, or in some combination of the two. As we will see,
this problem is analogous to the consumer’s problem of allocating a budget be-
tween purchases of food and clothing.
Let’s denote the risk-free return on the Treasury bill by Rf. Because the
eturn is risk free, the expected and actual returns are the same. In addi-
tion, let the expected return from investing in the stock market be Rm and
the actual return be rm. The actual return is risky. At the time of the invest-
ment decision, we know the set of possible outcomes and the likelihood of
each, but we do not know what particular outcome will occur. The risky
asset will have a higher expected return than the risk-free asset (Rm 7 Rf).
Otherwise, risk-averse investors would buy only Treasury bills and no
stocks would be sold.
the InveStment portfolIo To determine how much money the investor
should put in each asset, let’s set b equal to the fraction of her savings placed in
the stock market and (1 - b) the fraction used to purchase Treasury bills. The
expected return on her total portfolio, Rp, is a weighted average of the expected
eturn on the two assets:14
Rp = bRm + (1 - b)Rf (5.1)
Suppose, for example, that Treasury bills pay 4 percent (Rf = .04), the stock
market’s expected return is 12 percent (Rm = .12), and b = 1/2. Then Rp = 8
percent. How risky is this portfolio? One measure of riskiness is the standard
Table 5.8 InveStmentS – rISk anD return (1926–2014)
average rate OF
eturn (%)
average real
ate OF return (%)
ISk (Standard
devIatIOn)
Common stocks
(S&P 500)
12.1 8.8 20.1
Long-term
corporate bonds
6.4 3.3 8.4
U.S. Treasury bills 3.5 0.5 3.1
14The expected value of the sum of two variables is the sum of the expected values. Therefore
Rp = E [
m] + E[(1 - b)Rf] = bE[rm] + (1 - b)Rf = bRm + (1 - b)Rf
Source: © 2015 morningstar, Inc. all rights reserved. reproduced with permission.
200 PaRT 2 Producers, Consumers, and Competitive Markets
deviation of its return. We will denote the standard deviation of the risky stock
market investment by sm. With some alge
a, we can show that the standard
deviation of the portfolio, sp (with one risky and one risk-free asset) is the fraction
of the portfolio invested in the risky asset times the standard deviation of that
asset:15
sp = bsm (5.2)
The investor’s Choice Problem
We have still not determined how the investor should choose this fraction b. To
do so, we must first show that she faces a risk-return trade-off analogous to a
consumer’s budget line. To identify this trade-off, note that equation (5.1) for
the expected return on the portfolio can be rewritten as
Rp = Rf + b(Rm - Rf)
Now, from equation (5.2) we see that b = sp/sm, so that
Rp = Rf +
(Rm - Rf)
sm
sp (5.3)
ISk anD the buDget lIne This equation is a budget line because it de-
scribes the trade-off between risk (sp) and expected return (Rp). Note that it is
the equation for a straight line: Because Rm, Rf, and sm are constants, the slope
(Rm - Rf)/sm is a constant, as is the intercept, Rf. The equation says that the ex-
pected return on the portfolio Rp increases as the standard deviation of that return sp
increases. We call the slope of this budget line, (Rm - Rf)/sm, the price of risk,
ecause it tells us how much extra risk an investor must incur to enjoy a higher
expected return.
The budget line is drawn in Figure 5.6. If our investor wants no risk, she
can invest all her funds in Treasury bills (b = 0) and earn an expected return
Rf. To receive a higher expected return, she must incur some risk. For ex-
ample, she could invest all her funds in stocks (b = 1), earning an expected
eturn Rm but incu
ing a standard deviation sm. Or she might invest some
fraction of her funds in each type of asset, earning an expected return some-
where between Rf and Rm and facing a standard deviation less than sm but
greater than zero.
ISk anD InDIfference curveS Figure 5.6 also shows the solution to
the investor’s problem. Three indifference curves are drawn in the figure.
Each curve describes combinations of risk and return that leave the investor
equally satisfied. The curves are upward-sloping because risk is undesirable.
Thus, with a greater amount of risk, it takes a greater expected return to make
the investor equally well-off. Curve U3 yields the greatest amount of satisfac-
tion and U1 the least amount: For a given amount of risk, the investor earns a
In §3.2 we explain how a
udget line is determined
from an individual’s income
and the prices of the avail-
able goods.
Price of risk Extra risk that
an investor must incur to enjoy a
higher expected return.
15To see why, we observe from footnote 4 that we can write the variance of the portfolio return as
sp
2 = E[
m + (1 - b)Rf - Rp]2
Substituting equation (5.1) for the expected return on the portfolio, Rp, we have
sp
2 = E[
m + (1 - b)Rf - bRm - (1 - b)Rf]2 = E[b(rm - Rm)]2 = b2sm2
Because the standard deviation of a random variable is the square root of its variance, sp = bsm.
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 201
higher expected return on U3 than on U2 and a higher expected return on U2
than on U1.
Of the three indifference curves, the investor would prefer to be on U3. This
position, however, is not feasible, because U3 does not touch the budget line.
Curve U1 is feasible, but the investor can do better. Like the consumer choosing
quantities of food and clothing, our investor does best by choosing a combina-
tion of risk and return at the point where an indifference curve (in this case U2)
is tangent to the budget line. At that point, the investor’s return has an expected
value R* and a standard deviation s*.
Naturally, people differ in their attitudes toward risk. This fact is illustrated
in Figure 5.7, which shows how two different investors choose their portfolios.
Investor A is quite risk averse. Because his indifference curve UA is tangent to
the budget line at a point of low risk, he will invest almost all of his funds in
Treasury bills and earn an expected return RA just slightly larger than the risk-
free return Rf. Investor B is less risk averse. She will invest most of her funds in
stocks, and while the return on her portfolio will have a higher expected value
RB, it will also have a higher standard deviation sB.
U3
U2
U1
Rm
s* sm
Budget Line
0 Standard
deviation of
eturn, sp
Expected
eturn, Rp
Rf
R*
FiguRE 5.6
chooSIng between rISk anD return
An investor is dividing her funds between two assets—Treasury bills, which are risk
free, and stocks. The budget line describes the trade-off between the expected
eturn and its riskiness, as measured by the standard deviation of the return. The
slope of the budget line is (Rm - Rf)/sm, which is the price of risk. Three indiffer-
ence curves are drawn, each showing combinations of risk and return that leave
an investor equally satisfied. The curves are upward-sloping because a risk-averse
investor will require a higher expected return if she is to bear a greater amount of
isk. The utility-maximizing investment portfolio is at the point where indifference
curve U2 is tangent to the budget line.
202 PaRT 2 Producers, Consumers, and Competitive Markets
If Investor B has a sufficiently low level of risk aversion, she might buy
stocks on margin: that is, she would bo
ow money from a
okerage firm in or-
der to invest more than she actually owns in the stock market. In effect, a person
who buys stocks on margin holds a portfolio with more than 100 percent of the
portfolio’s value invested in stocks. This situation is illustrated in Figure 5.8,
which shows indifference curves for two investors. Investor A, who is relatively
Expected
eturn, Rp
Rm
Budget Line
Standard
deviation of
eturn, �p
�m0 �B�A
UB
RB
UA
Rf
RA
RB
UA
UB
Rm
RA
Rf
0
Budget
Line
sA sm sB
FiguRE 5.7
the choIceS of two
DIfferent InveStorS
Investor A is highly risk averse.
Because his portfolio will consist
mostly of the risk-free asset, his ex-
pected return RA will be only slightly
greater than the risk-free return.
His risk sA, however, will be small.
Investor B is less risk averse. She will
invest a large fraction of her funds in
stocks. Although the expected re-
turn on her portfolio RB will be larger,
it will also be riskier.
FiguRE 5.8
uyIng StockS
on margIn
Because Investor A is risk averse, his
portfolio contains a mixture of stocks
and risk-free Treasury bills. Investor B,
however, has a very low degree of risk
aversion. Her indifference curve, UB, is
tangent to the budget line at a point
where the expected return and stan-
dard deviation for her portfolio exceed
those for the stock market overall. This
implies that she would like to invest
more than 100 percent of her wealth
in the stock market. She does so by
uying stocks on margin—i.e., by bor-
owing from a
okerage firm to help
finance her investment.
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 203
isk-averse, invests about half of his funds in stocks. Investor B, however, has
an indifference curve that is relatively flat and tangent with the budget line at
a point where the expected return on the portfolio exceeds the expected return
on the stock market. In order to hold this portfolio, the investor must bo
ow
money because she wants to invest more than 100 percent of her wealth in the
stock market. Buying stocks on margin in this way is a form of leverage: the in-
vestor increases her expected return above that for the overall stock market, but
at the cost of increased risk.
In Chapters 3 and 4, we simplified the problem of consumer choice by as-
suming that the consumer had only two goods from which to choose—food and
clothing. In the same spirit, we have simplified the investor’s choice by limiting
it to Treasury bills and stocks. The basic principles, however, would be the same
if we had more assets (e.g., corporate bonds, land, and different types of stocks).
Every investor faces a trade-off between risk and return.16 The degree of extra
isk that each is willing to bear in order to earn a higher expected return de-
pends on how risk averse he or she is. Less risk-averse investors tend to include
a larger fraction of risky assets in their portfolios.
ExaMPlE 5.6 InveStIng In the Stock market
The 1990s witnessed a shift in the
investing behavior of americans. first,
many people started investing in the
stock market for the first time. In
1989, about 32 percent of families
in the united States had part of their
wealth invested in the stock market,
either directly (by owning individual
stocks) or indirectly (through mutual
funds or pension plans invested in stocks). By 1998,
that fraction had risen to 49 percent. In addition, the
share of wealth invested in stocks increased from
about 26 percent to about 54 percent during the
same period.17 Much of this shift is attributable to
younger investors. for those under the age of 35, par-
ticipation in the stock market increased from about
22 percent in 1989 to about 41 percent in 1998.
In most respects, household investing behavior has
stabilized after the 1990s shift. The percent of fami-
lies with investments in the stock market was 51.1%
in 2007. However, older americans have become
much more active. By 2007, 40
percent of people over age 75
held stocks, up from 29 percent
in 1998.
Why have more people started
investing in the stock market?
One reason is the advent of online
trading, which has made investing
much easier. another reason may
e the considerable increase in stock prices that oc-
cu
ed during the late 1990s, driven in part by the so-
called “dot com euphoria.” These increases may have
convinced some investors that prices could only con-
tinue to rise in the future. as one analyst put it, “The
market’s relentless seven-year climb, the popularity of
mutual funds, the shift by employers to self-directed
etirement plans, and the avalanche of do-it-yourself
investment publications all have combined to create a
nation of financial know-it-alls.”18
figure 5.9 shows the dividend yield and price
earnings (P/E) ratio for the S&P 500 (an index of
16As mentioned earlier, what matters is nondiversifiable risk, because investors can eliminate di-
versifiable risk by holding many different stocks (e.g., via mutual funds). We discuss diversifiable
versus nondiversifiable risk in Chapter 15.
17Data are from the Federal Reserve Bulletin, January 2000, and the Survey of Consumer Finances, 2011.
18”Investors Ignore Brokers, Dishing Out Their Own Tips” Wall Street Journal, September 12, 1997.
204 PaRT 2 Producers, Consumers, and Competitive Markets
stocks of 500 large corporations) over the period
1970 to 2016. Observe that the dividend yield (the
annual dividend divided by the stock price) fell
from about 5 percent in 1980 to below 2 percent
y 2000. Meanwhile, however, the price/earnings
atio (the share price divided by annual earnings
per share) increased from about 8 in 1980 to over
40 in 2002, before falling to around 20 between
2005 and 2007 and then increasing through 2016.
In retrospect, the increase in the P/E ratio could
only have occu
ed if investors believed that corpo-
ate profits would continue to grow rapidly in the
coming decade. This suggests that in the late 1990s,
many investors had a low degree of risk aversion,
were quite optimistic about the economy, or both.
alternatively, some economists have argued that
P/E Ratio
Dividend Yield
0
5
10
15
20
25
30
35
40
45
50
1970 1974 1978 1982 1986 1990
Yea
1994 1998 2002 2006 2010 2014
P
E
R
at
io
0
1
2
3
4
5
6
7
D
ivid
end
Y
ield
(percent)
the run-up of stock prices during the 1990s was the
esult of “herd behavior,” in which investors rushed
to get into the market after hearing of the successful
experiences of others.19
The psychological motivations that explain herd
ehavior can help to explain stock market bu
les.
However, they go far beyond the stock market.
They also apply to the behavior of consumers and
firm managers in a wide variety of settings. Such
ehavior cannot always be captured by the sim-
plified assumptions that we have made up to this
point about consumer choice. In Chapter 19, we
will discuss these aspects of behavior in detail, and
we will see how the traditional models of Chapters
3 and 4 can be expanded to help us understand this
ehavior.
FiguRE 5.9
DIvIDenD yIelD anD p/e ratIo for S&p 500
The dividend yield for the S&P 500 (the annual dividend divided by the stock price) has fallen dra-
matically, while the price/earnings ratio (the stock price divided by the annual earnings-per-share)
ose from 1980 to 2002 and then dropped.
19See, for example, Robert Shiller, I
ational Exuberance, Princeton University Press, 2000.
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 205
Summary
1. Consumers and managers frequently make decisions
in which there is uncertainty about the future. This
uncertainty is characterized by the term risk, which
applies when each of the possible outcomes and its
probability of occu
ence is known.
2. Consumers and investors are concerned about the
expected value and the variability of uncertain out-
comes. The expected value is a measure of the central
tendency of the values of risky outcomes. Variability
is frequently measured by the standard deviation of
outcomes, which is the square root of the probability-
weighted average of the squares of the deviation from
the expected value of each possible outcome.
3. Facing uncertain choices, consumers maximize their
expected utility—an average of the utility associated
with each outcome—with the associated probabilities
serving as weights.
4. A person who would prefer a certain return of
a given amount to a risky investment with the
same expected return is risk averse. The maximum
amount of money that a risk-averse person would
pay to avoid taking a risk is called the risk premium.
A person who is indifferent between a risky invest-
ment and the certain receipt of the expected return
on that investment is risk neutral. A risk-loving
consumer would prefer a risky investment with a
given expected return to the certain receipt of that
expected return.
5. Risk can be reduced by (a) diversification, (b) insur-
ance, and (c) additional information.
6. The law of large numbers enables insurance companies
to provide insurance for which the premiums paid
equal the expected value of the losses being insured
against. We call such insurance actuarially fair.
7. Consumer theory can be applied to decisions to invest
in risky assets. The budget line reflects the price of
isk, and consumers’ indifference curves reflect their
attitudes toward risk.
QueStionS for review
1. What does it mean to say that a person is risk averse?
Why are some people likely to be risk averse while
others are risk lovers?
2. Why is the variance a better measure of variability
than the range?
3. George has $5000 to invest in a mutual fund. The
expected return on mutual fund A is 15 percent and
the expected return on mutual fund B is 10 percent.
Should George pick mutual fund A or fund B?
4. What does it mean for consumers to maximize ex-
pected utility? Can you think of a case in which a per-
son might not maximize expected utility?
5. Why do people often want to insure fully against
uncertain situations even when the premium paid
exceeds the expected value of the loss being insured
against?
6. Why is an insurance company likely to behave as if it
were risk neutral even if its managers are risk-averse
individuals?
7. When is it worth paying to obtain more information to
educe uncertainty?
8. How does the diversification of an investor’s portfolio
avoid risk?
9. Why do some investors put a large portion of their
portfolios into risky assets while others invest
largely in risk-free alternatives? (Hint: Do the two
investors receive exactly the same return on average?
If so, why?)
exerciSeS
1. Consider a lottery with three possible outcomes:
•     $125 will be received with probability .2
•     $100 will be received with probability .3
•     $50 will be received with probability .5
a. What is the expected value of the lottery?
. What is the variance of the outcomes?
c. What would a risk-neutral person pay to play the
lottery?
2. Suppose you have invested in a new computer com-
pany whose profitability depends on two factors: (1)
whether the U.S. Congress passes a tariff raising the
cost of Japanese computers and (2) whether the U.S.
economy grows slowly or quickly. What are the four
mutually exclusive states of the world that you should
e concerned about?
3. Richard is deciding whether to buy a state lottery
ticket. Each ticket costs $1, and the probability of win-
ning payoffs is given as follows:
prObabIlIty return
.5 $0.00
.25 $1.00
.2 $2.00
.05 $7.50
206 PaRT 2 Producers, Consumers, and Competitive Markets
a. What is the expected value of Richard’s payoff if he
uys a lottery ticket? What is the variance?
. Richard’s nickname is “No-Risk Rick” because he is
an extremely risk-averse individual. Would he buy
the ticket?
c. Richard has been given 1000 lottery tickets. Discuss
how you would determine the smallest amount for
which he would be willing to sell all 1000 tickets.
d. In the long run, given the price of the lottery tick-
ets and the probability
eturn table, what do you
think the state would do about the lottery?
4. Suppose an investor is concerned about a business
choice in which there are three prospects—the prob-
ability and returns are given below:
prObabIlIty return
.4 $100
.3 30
.3 -30
What is the expected value of the uncertain invest-
ment? What is the variance?
5. You are an insurance agent who must write a policy
for a new client named Sam. His company, Society
for Creative Alternatives to Mayonnaise (SCAM), is
working on a low-fat, low-cholesterol mayonnaise
substitute for the sandwich-condiment industry. The
sandwich industry will pay top dollar to the first in-
ventor to patent such a mayonnaise substitute. Sam’s
SCAM seems like a very risky proposition to you. You
have calculated his possible returns table as follows:
prObabIlIty return OutcOme
.999 - $1,000,000 (he fails)
.001 $1,000,000,000 (he succeeds and
sells his formula)
a. What is the expected return of Sam’s project? What
is the variance?
. What is the most that Sam is willing to pay for in-
surance? Assume Sam is risk neutral.
c. Suppose you found out that the Japanese are on
the verge of introducing their own mayonnaise
substitute next month. Sam does not know this and
has just turned down your final offer of $1000 for
the insurance. Assume that Sam tells you SCAM is
only six months away from perfecting its mayon-
naise substitute and that you know what you know
about the Japanese. Would you raise or lower your
policy premium on any subsequent proposal to
Sam? Based on his information, would Sam accept?
6. Suppose that Natasha’s utility function is given by
u(I) = 110I, where I represents annual income in
thousands of dollars.
a. Is Natasha risk loving, risk neutral, or risk averse?
Explain.
. Suppose that Natasha is cu
ently earning an in-
come of $40,000 (I = 40) and can earn that income
next year with certainty. She is offered a chance to
take a new job that offers a .6 probability of earn-
ing $44,000 and a .4 probability of earning $33,000.
Should she take the new job?
c. In (b), would Natasha be willing to buy insurance to
protect against the variable income associated with the
new job? If so, how much would she be willing to pay
for that insurance? (Hint: What is the risk premium?)
7. Suppose that two investments have the same three
payoffs, but the probability associated with each pay-
off differs, as illustrated in the table below:
payOFF
prObabIlIty
(InveStment a)
prObabIlIty
(InveStment b)
$300 0.10 0.30
$250 0.80 0.40
$200 0.10 0.30
a. Find the expected return and standard deviation of
each investment.
. Jill has the utility function U = 5I, where I denotes
the payoff. Which investment will she choose?
c. Ken has the utility function U = 51I. Which in-
vestment will he choose?
d. Laura has the utility function U = 5I 2. Which in-
vestment will she choose?
8. As the owner of a family farm whose wealth is
$250,000, you must choose between sitting this sea-
son out and investing last year’s earnings ($200,000)
in a safe money market fund paying 5.0 percent or
planting summer corn. Planting costs $200,000, with
a six-month time to harvest. If there is rain, planting
summer corn will yield $500,000 in revenues at har-
vest. If there is a drought, planting will yield $50,000
in revenues. As a third choice, you can purchase
AgriCorp drought-resistant summer corn at a cost of
$250,000 that will yield $500,000 in revenues at har-
vest if there is rain, and $350,000 in revenues if there
is a drought. You are risk averse, and your preference
for family wealth (W) is specified by the relationship
U(W) = 1W. The probability of a summer drought is
0.30, while the probability of summer rain is 0.70.
Which of the three options should you choose?
Explain.
9. Draw a utility function over income u(I) that describes
a man who is a risk lover when his income is low but
isk averse when his income is high. Can you explain
why such a utility function might reasonably describe
a person’s preferences?
10. A city is considering how much to spend to hire
people to monitor its parking meters. The following
information is available to the city manager:
CHaPTER 5 unCERTaInTy anD COnSuMER BEHaVIOR 207
•     Hiring each meter monitor costs $10,000 per year.
•     With one monitoring person hired, the probability
of a driver getting a ticket each time he or she parks
illegally is equal to .25.
•     With two monitors, the probability of getting a
ticket is .5; with three monitors, the probability is
.75; and with four, it’s equal to 1.
•     With two monitors hired, the cu
ent fine for over-
time parking is $20.
a. Assume first that all drivers are risk neutral. What
parking fine would you levy, and how many meter
monitors would you hire (1, 2, 3, or 4) to achieve
the cu
ent level of dete
ence against illegal park-
ing at the minimum cost?
. Now assume that drivers are highly risk averse.
How would your answer to (a) change?
c. (For discussion) What if drivers could insure
themselves against the risk of parking fines?
Would it make good public policy to permit such
insurance?
11. A moderately risk-averse investor has 50 percent of
her portfolio invested in stocks and 50 percent in risk-
free Treasury bills. Show how each of the following
events will affect the investor’s budget line and the
proportion of stocks in her portfolio:
a. The standard deviation of the return on the stock
market increases, but the expected return on the
stock market remains the same.
. The expected return on the stock market increases,
ut the standard deviation of the stock market re-
mains the same.
c. The return on risk-free Treasury bills increases.
This page intentionally left blank
209
In the last three chapters, we focused on the demand side of the market—the preferences and behavior of consumers. Now we turn to the supply side and examine the behavior of producers. We will see how firms
can produce efficiently and how their costs of production change with
changes in both input prices and the level of output. We will also see
that there are strong similarities between the optimizing decisions made
y firms and those made by consumers. In other words, understanding
consumer behavior will help us understand producer behavior.
In this chapter and the next we discuss the theory of the firm, which
describes how a firm makes cost-minimizing production decisions and
how the firm’s resulting cost varies with its output. Our knowledge of
production and cost will help us understand the characteristics of mar-
ket supply. It will also prove useful for dealing with problems that arise
egularly in business. To see this, just consider some of the problems of-
ten faced by a company like General Motors. How much assembly-line
machinery and how much labor should it use in its new automobile
plants? If it wants to increase production, should it hire more work-
ers, construct new plants, or both? Does it make more sense for one
automobile plant to produce different models, or should each model be
manufactured in a separate plant? What should GM expect its costs to
e during the coming year? How are these costs likely to change over
time and be affected by the level of production? These questions apply
not only to business firms but also to other producers of goods and ser-
vices, such as governments and nonprofit agencies.
The Production Decisions of a Firm
In Chapters 3 and 4, we studied consumer behavior by
eaking it
down into three steps. First, we explained how to describe consumer
preferences. Second, we accounted for the fact that consumers face
udget constraints. Third, we saw how, given their preferences and
udget constraints, consumers can choose combinations of goods to
maximize their satisfaction. The production decisions of firms are
analogous to the purchasing decisions of consumers, and can likewise
e understood in three steps:
1. Production Technology: We need a practical way of describing
how inputs (such as labor, capital, and raw materials) can be
6.1 The Authors Debate a
Production Function
for Health Care 219
6.2 Malthus and the Food
Crisis 221
6.3 Labor Productivity and
the Standard of Living 223
6.4 A Production Function
for Wheat 230
6.5 Returns to Scale in
the Carpet Industry 233
LIST oF ExAMPLES
6.1 Firms and Their Production
Decisions 210
6.2 Production with one
Variable Input (Labor) 214
6.3 Production with Two
Variable Inputs 224
6.4 Returns to Scale 231
CHAPTER ouTLInE
Production
CHAPTER 6
210 PART 2 Producers, Consumers, and Competitive Markets
transformed into outputs (such as cars and televisions). Just as a consumer
can reach a level of satisfaction from buying different combinations of
goods, the firm can produce a particular level of output by using different
combinations of inputs. For example, an electronics firm might produce
10,000 televisions per month by using a substantial amount of labor (e.g.,
workers assembling the televisions by hand) and very little capital, or
y building a highly automated capital-intensive factory and using very
little labor.
2. Cost Constraints: Firms must take into account the prices of labor, capital,
and other inputs. Just as a consumer is constrained by a limited budget,
the firm will be concerned about its cost of production. For example, the
firm that produces 10,000 televisions per month will want to do so in a
way that minimizes its total production cost, which is determined in part
y the prices of the inputs it uses.
3. Input Choices: Given its production technology and the prices of labor,
capital, and other inputs, the firm must choose how much of each input to
use in producing its output. Just as a consumer takes account of the prices
of different goods when deciding how much of each good to buy, the firm
must take into account the prices of different inputs when deciding how
much of each input to use. If our electronics firm operates in a country
with low wage rates, it may decide to produce televisions by using a large
amount of labor, thereby using very little capital.
These three steps are the building blocks of the theory of the firm, and we
will discuss them in detail in this chapter and the next. We will also address
other important aspects of firm behavior. For example, assuming that the firm is
always using a cost-minimizing combination of inputs, we will see how its total
cost of production varies with the quantity it produces and how it can choose
that quantity to maximize its profit.
We begin this chapter by discussing the nature of the firm and asking why
firms exist in the first place. Next, we explain how the firm’s production tech-
nology can be represented in the form of a production function—a compact
description of how inputs are turned into output. We then use the production
function to show how the firm’s output changes when just one of its inputs
(labor) is varied, holding the other inputs fixed. Next, we turn to the more gen-
eral case in which the firm can vary all of its inputs, and we show how the firm
chooses a cost-minimizing combination of inputs to produce its output. We will
e particularly concerned with the scale of the firm’s operation. Are there, for
example, any technological advantages that make the firm more productive as
its scale increases?
6.1 Firms and Their Production
Decisions
Firms as we know them today are a relatively new invention. Prior to the
mid-1800s, almost all production was done by farmers, craftsmen, indi-
viduals who wove cloth and made clothing, and merchants and traders
who bought and sold various goods. This was true in the U.S., Europe,
and everywhere else in the world. The concept of a firm—run by managers
separate from the firm’s owners, and who hire and manage a large number of
theory of the firm
Explanation of how a firm makes
cost-minimizing production
decisions and how its cost varies
with its output.
CHAPTER 6 PRoDuCTIon 211
workers—did not even exist. Modern corporations emerged only in the latter
part of the 19th century.1
Today we take firms for granted. It is hard for us to imagine the production
of automobiles without large companies like Ford and Toyota, the production
of oil and natural gas without companies like Exxon-Mobil and Shell, or even
the production of
eakfast cereal without companies like Kellogg and General
Mills. But stop for a minute and ask yourself whether we really need firms to
produce the goods and services that we consume regularly. This was the ques-
tion raised by Ronald Coase in a famous 1937 article: If markets work so well in
allocating resources, why do we need firms?2
Why Do Firms Exist?
Do we really need firms to produce cars? Why couldn’t cars be produced by
a collection of individuals who worked independently and contracted with
each other when appropriate, rather than being employed by General Motors?
Couldn’t some people design a car (for a fee), other people buy steel, rent the
equipment needed to stamp the steel into the shapes called for in the design,
and then do the stamping (also for negotiated fees), other people make steering
wheels and radiators, still other people assemble the various parts, and so on,
where again, every task would be performed for a negotiated fee?
Or take another example: We—the authors of this book—work for universi-
ties, which are essentially firms that provide educational services along with
esearch. We are paid monthly salaries and in return are expected to teach
egularly (to students recruited by our “firms” and in classrooms the “firms”
provide), do research and write (in the offices our “firms” give us), and ca
y
out administrative tasks. Couldn’t we simply bypass the universities and offer
our teaching services on an hourly basis in rented classrooms to students who
show up and pay us, and likewise do research on a paid piecemeal basis? Do we
eally need colleges and universities with all their overhead costs?
In principle, cars could indeed be produced by a large number of independent
workers, and an education could be produced by a number of independent
teachers. These independent workers would offer their services for negotiated
fees, and those fees would be determined by market supply and demand. It
shouldn’t take you long, however, to realize that such a system of production
would be extremely inefficient. Think about how difficult it would be for inde-
pendent workers to decide who will do what to produce cars, and negotiate the
fees that each worker will charge for each task. And if there were any change in
the design of the car, all of these tasks and fees would have to be renegotiated.
For cars produced this way, the quality would likely be abysmal, and the cost
astronomical.
Firms offer a means of coordination that is extremely important and would be
sorely missing if workers operated independently. Firms eliminate the need for
every worker to negotiate every task that he or she will perform, and bargain
over the fees that will be paid for those tasks. Firms can avoid this kind of bar-
gaining by having managers that direct the production of salaried workers—they
1The classic history of the development of the modern corporation is Alfred Chandler, Jr., The Visible
Hand: The Managerial Revolution in American Business, Cam
idge: Harvard University Press, 1977.
2Ronald Coase, “The Nature of the Firm,” Economica 4 (1937): 386–405. Coase won a Nobel Prize in
Economics in 1991.
212 PART 2 Producers, Consumers, and Competitive Markets
tell workers what to do and when to do it, and the workers (as well as the man-
agers themselves) are simply paid a weekly or monthly salary.
There is no guarantee, of course, that a firm will operate efficiently, and there
are many examples of firms that operate very inefficiently. Managers cannot
always monitor what workers are doing, and managers themselves sometimes
make decisions that are in their interest, but not in the firm’s best interest. As
a result, the theory of the firm (and more
oadly, organizational economics) has
ecome an important area of microeconomic research. The theory has both posi-
tive aspects (explaining why managers and workers behave the way they do)
and normative aspects (explaining how firms can be best organized so that they
operate as efficiently as possible).3 We will discuss some aspects of the theory
later in this book. At this point we simply stress that firms exist because they
allow goods and services to be produced far more efficiently than would be
possible without them.
The Technology of Production
What do firms do? We have seen that firms organize and coordinate the ac-
tivities of large numbers of workers and managers. But to what purpose? At
the most fundamental level, firms take inputs and turn them into outputs (or
products). This production process, turning inputs into outputs, is the essence
of what a firm does. Inputs, which are also called factors of production, include
anything that the firm must use as part of the production process. In a bakery,
for example, inputs include the labor of its workers; raw materials, such as flour
and sugar; and the capital invested in its ovens, mixers, and other equipment
needed to produce such outputs as
ead, cakes, and pastries.
As you can see, we can divide inputs into the
oad categories of labor, mate-
ials, and capital, each of which might include more na
ow subdivisions. Labor
inputs include skilled workers (carpenters, engineers) and unskilled workers
(agricultural workers), as well as the entrepreneurial efforts of the firm’s man-
agers. Materials include steel, plastics, electricity, water, and any other goods
that the firm buys and transforms into final products. Capital includes land,
uildings, machinery and other equipment, as well as inventories.
The Production Function
Firms can turn inputs into outputs in a variety of ways, using various combina-
tions of labor, materials, and capital. We can describe the relationship between
the inputs into the production process and the resulting output by a production
function. A production function indicates the highest output q that a firm can
produce for every specified combination of inputs.4 Although in practice firms
use a wide variety of inputs, we will keep our analysis simple by focusing on
only two, labor L and capital K. We can then write the production function as
q = F(K, L) (6.1)
This equation relates the quantity of output to the quantities of the two in-
puts, capital and labor. For example, the production function might describe
factors of production Inputs
into the production process (e.g.,
labor, capital, and materials).
production function
Function showing the highest
output that a firm can produce for
every specified combination of
inputs.
3The literature on the theory of the firm is vast. One of the classics is Oliver Williamson, Markets and
Hierarchies: Analysis and Antitrust Implications, New York: Free Press, 1975. (Williamson won a Nobel
Prize for his work in 2009.)
4In this chapter and those that follow, we will use the variable q for the output of the firm, and Q for
the output of the industry.
CHAPTER 6 PRoDuCTIon 213
the number of personal computers that can be produced each year with a
10,000-square-foot plant and a specific amount of assembly-line labor. Or it
might describe the crop that a farmer can obtain using specific amounts of ma-
chinery and workers.
It is important to keep in mind that inputs and outputs are flows. For ex-
ample, our PC manufacturer uses a certain amount of labor each year to produce
some number of computers over that year. Although it might own its plant and
machinery, we can think of the firm as paying a cost for the use of that plant
and machinery over the year. To simplify things, we will frequently ignore the
eference to time and refer only to amounts of labor, capital, and output. Unless
otherwise indicated, however, we mean the amount of labor and capital used
each year and the amount of output produced each year.
Because the production function allows inputs to be combined in varying
proportions, output can be produced in many ways. For the production func-
tion in equation (6.1), this could mean using more capital and less labor, or vice
versa. For example, wine can be produced in a labor-intensive way using many
workers, or in a capital-intensive way using machines and only a few workers.
Note that equation (6.1) applies to a given technology—that is, to a given
state of knowledge about the various methods that might be used to trans-
form inputs into outputs. As the technology becomes more advanced and the
production function changes, a firm can obtain more output for a given set of
inputs. For example, a new, faster assembly line may allow a hardware manu-
facturer to produce more high-speed computers in a given period of time.
Production functions describe what is technically feasible when the firm op-
erates efficiently—that is, when the firm uses each combination of inputs as
effectively as possible. The presumption that production is always technically
efficient need not always hold, but it is reasonable to expect that profit-seeking
firms will not waste resources.
The Short Run versus the Long Run
It takes time for a firm to adjust its inputs to produce its product with differing
amounts of labor and capital. A new factory must be planned and built, and
machinery and other capital equipment must be ordered and delivered. Such
activities can easily take a year or more to complete. As a result, if we are look-
ing at production decisions over a short period of time, such as a month or two,
the firm is unlikely to be able to substitute very much capital for labor.
Because firms must consider whether or not inputs can be varied, and if they
can, over what period of time, it is important to distinguish between the short
and long run when analyzing production. The short run refers to a period of
time in which the quantities of one or more factors of production cannot be
changed. In other words, in the short run there is at least one factor that cannot
e varied; such a factor is called a fixed input. The long run is the amount of
time needed to make all inputs variable.
As you might expect, the kinds of decisions that firms can make are very
different in the short run than those made in the long run. In the short run, firms
vary the intensity with which they utilize a given plant and machinery; in the
long run, they vary the size of the plant. All fixed inputs in the short run repre-
sent the outcomes of previous long-run decisions based on estimates of what a
firm could profitably produce and sell.
There is no specific time period, such as one year, that separates the short run
from the long run. Rather, one must distinguish them on a case-by-case basis.
For example, the long run can be as
ief as a day or two for a child’s lemonade
short run Period of time in
which quantities of one or more
production factors cannot be
changed.
fixed input Production factor
that cannot be varied.
long run Amount of time
needed to make all production
inputs variable.
214 PART 2 Producers, Consumers, and Competitive Markets
stand or as long as five or ten years for a petrochemical producer or an automo-
ile manufacturer.
We will see that in the long run firms can vary the amounts of all their inputs to
minimize the cost of production. Before treating this general case, however, we be-
gin with an analysis of the short run, in which only one input to the production pro-
cess can be varied. We assume that capital is the fixed input, and labor is variable.
6.2 Production with One Variable
Input (Labor)
When deciding how much of a particular input to buy, a firm has to compare
the benefit that will result with the cost of that input. Sometimes it is useful to
look at the benefit and the cost on an incremental basis by focusing on the ad-
ditional output that results from an incremental addition to an input. In other
situations, it is useful to make the comparison on an average basis by consider-
ing the result of substantially increasing an input. We will look at benefits and
costs in both ways.
When capital is fixed but labor is variable, the only way the firm can produce
more output is by increasing its labor input. Imagine, for example, that you are
managing a clothing factory. Although you have a fixed amount of equipment,
you can hire more or less labor to sew and to run the machines. You must de-
cide how much labor to hire and how much clothing to produce. To make the
decision, you will need to know how the amount of output q increases (if at all)
as the input of labor L increases.
Table 6.1 gives this information. The first three columns show the amount of
output that can be produced in one month with different amounts of labor and
capital fixed at 10 units. The first column shows the amount of labor, the second
the fixed amount of capital, and the third total output. When labor input is zero,
output is also zero. Output then increases as labor is increased up to an input
of 9 units. Beyond that point, total output declines: Although initially each unit
of labor can take greater and greater advantage of the existing machinery and
plant, after a certain point, additional labor is no longer useful and indeed can
e counterproductive. Five people can run an assembly line better than two, but
twelve people may get in one another’s way.
Average and Marginal Products
The contribution that labor makes to the production process can be described
on both an average and a marginal (i.e., incremental) basis. The fourth column
in Table 6.1 shows the average product of labor (APL ), which is the output per
unit of labor input. The average product is calculated by dividing the total out-
put q by the total input of labor L. The average product of labor measures the
productivity of the firm’s workforce in terms of how much output each worker
produces on average. In our example, the average product increases initially
ut falls when the labor input becomes greater than four.
The fifth column of Table 6.1 shows the marginal product of labor (MPL).
This is the additional output produced as the labor input is increased by 1 unit.
For example, with capital fixed at 10 units, when the labor input increases from
2 to 3, total output increases from 40 to 69, creating an additional output of 29 (i.e.,
69–40) units. The marginal product of labor can be written as ∆q/∆L—in other
words, the change in output ∆q resulting from a 1-unit increase in labor input ∆L.
average product output per
unit of a particular input.
marginal product Additional
output produced as an input is
increased by one unit.
CHAPTER 6 PRoDuCTIon 215
Remember that the marginal product of labor depends on the amount of
capital used. If the capital input increased from 10 to 20, the marginal product of
labor most likely would increase. Why? Because additional workers are likely
to be more productive if they have more capital to use. Like the average prod-
uct, the marginal product first increases then falls—in this case, after the third
unit of labor.
To summarize:
Average product of labor = Output/labor input = q/L
Marginal product of labor = Change in output/change in labor input
= ∆q/∆L
The Slopes of the Product Curve
Figure 6.1 plots the information contained in Table 6.1. (We have connected all
the points in the figure with solid lines.) Figure 6.1(a) shows that as labor is
increased, output increases until it reaches the maximum output of 153; there-
after, it falls. The portion of the total output curve that is declining is drawn
with a dashed line to denote that producing with more than nine workers is not
economically rational; it can never be profitable to use additional amounts of a
costly input to produce less output.
Figure 6.1 (b) shows the average and marginal product curves. (The units on
the vertical axis have changed from output per month to output per worker per
month.) Note that the marginal product is positive as long as output is increas-
ing but becomes negative when output is decreasing.
It is no coincidence that the marginal product curve crosses the horizontal
axis of the graph at the point of maximum total product. This happens because
Table 6.1 Production with one Variable inPut
Amount of
LAbor (L)
Amount of
CApitAL (K)
totAL
output (q)
AverAge
produCt (q/L)
mArginAL
produCt (∆q/∆L)
0 10 0 — —
1 10 15 15 15
2 10 40 20 25
3 10 69 23 29
4 10 96 24 27
5 10 120 24 24
6 10 138 23 18
7 10 147 21 9
8 10 152 19 5
9 10 153 17 1
10 10 150 15 -3
11 10 143 13 -7
12 10 133 11.08 -10
216 PART 2 Producers, Consumers, and Competitive Markets
adding a worker in a manner that slows production and decreases total output
implies a negative marginal product for that worker.
The average product and marginal product curves are closely related. When
the marginal product is greater than the average product, the average product is increas-
ing. This is the case for labor inputs up to 5 in Figure 6.1 (b). If the output of an
additional worker is greater than the average output of each existing worker (i.e.,
the marginal product is greater than the average product), then adding the worker
causes average output to increase. In Table 6.1, two workers produce 40 units of
output, for an average product of 20 units per worker. Adding a third worker in-
creases output by 29 units (to 69), which raises the average product from 20 to 23.
Similarly, when the marginal product is less than the average product, the average
product is decreasing. This is the case when the labor input is greater than 5 in
1 2 10 11 12
1 2 43 5 6 7
D
8 9 10 11 120
120
140
160
60
80
100
40
A
B
C
20
0
30
20
25
10
15
5
0
O
ut
pu
t p
e
m
on
th
O
ut
pu
t p
e
w
o
ke

pe

m
on
th
0
Labor pe
month
(a)
(b)
Total Product
Average
Product
Marginal
Product
3 4 5 6 7 8
Labor pe
month
9
FIguRE 6.1
Production with one
Variable inPut
The total output curve in (a) shows the
output produced for different amounts
of labor input. The average and mar-
ginal products in (b) can be obtained
(using the data in Table 6.1) from the
total product curve. At point A in (a),
with 3 units of labor, the marginal prod-
uct is 29 because the tangent to the
total product curve has a slope of 29.
The average product of labor, however,
is 23, which is the slope of the line
from the origin to point A. Also, the
marginal product of labor reaches its
maximum at this point. At point B, with
5 units of labor, the marginal product
of labor has dropped to 24 and is
equal to the average product of labor.
Thus, in (b), the average and marginal
product curves intersect (at point D).
Note that when the marginal product
curve is above the average product,
the average product is increasing.
When the labor input is greater than
5 units, the marginal product is below
the average product, so the average
product is falling. Once the labor input
exceeds 9 units, the marginal product
ecomes negative, so that total output
falls as more labor is added.
CHAPTER 6 PRoDuCTIon 217
Figure 6.1 (b). In Table 6.1, six workers produce 138 units of output, for an aver-
age product of 23. Adding a seventh worker contributes a marginal product of
only 9 units (less than the average product), reducing the average product to 21.
We have seen that the marginal product is above the average product when
the average product is increasing and below the average product when the
average product is decreasing. It follows, therefore, that the marginal product
must equal the average product when the average product reaches its maxi-
mum. This happens at point D in Figure 6.1 (b).
Why, in practice, should we expect the marginal product curve to rise and then
fall? Think of a television assembly plant. Fewer than ten workers might be insuf-
ficient to operate the assembly line at all. Ten to fifteen workers might be able to
un the assembly line, but not very efficiently. If adding a few more workers al-
lowed the assembly line to operate much more efficiently, the marginal product
of those workers would be very high. This added efficiency, however, might start
to diminish once there were more than 20 workers. The marginal product of the
twenty-second worker, for example, might still be very high (and above the aver-
age product), but not as high as the marginal product of the nineteenth or twenti-
eth worker. The marginal product of the twenty-fifth worker might be lower still,
perhaps equal to the average product. With 30 workers, adding one more worker
would yield more output, but not very much more (so that the marginal product,
while positive, would be below the average product). Once there were more than
40 workers, additional workers would simply get in each other’s way and actu-
ally reduce output (so that the marginal product would be negative).
The Average Product of Labor Curve
The geometric relationship between the total product and the average and mar-
ginal product curves is shown in Figure 6.1 (a). The average product of labor
is the total product divided by the quantity of labor input. At A, for example,
the average product is equal to the output of 69 divided by the input of 3, or
23 units of output per unit of labor input. This ratio, however, is exactly the
slope of the line running from the origin to A in Figure 6.1 (a). In general,
the average product of labor is given by the slope of the line drawn from the origin to the
co
esponding point on the total product curve.
The Marginal Product of Labor Curve
As we have seen, the marginal product of labor is the change in the total prod-
uct resulting from an increase of one unit of labor. At B, for example, the mar-
ginal product is 24 because the tangent to the total product curve has a slope
of 24. In general, the marginal product of labor at a point is given by the slope of the
total product at that point. We can see in Figure 6.1 (b) that the marginal product
of labor increases initially, peaks at an input of 3, and then declines as we move
up the total product curve to B and C. At C, when total output is maximized, the
slope of the tangent to the total product curve is 0, as is the marginal product.
Beyond that point, the marginal product becomes negative.
the relationshiP between the aVerage and Marginal Products
Note the graphical relationship between average and marginal products in
Figure 6.1 (a). At A, the marginal product of labor (the slope of the tangent to
the total product curve at A—not shown explicitly) is greater than the average
product (dashed line 0A). As a result, the average product of labor increases
as we move from A to B. At B, the average and marginal products of labor are
218 PART 2 Producers, Consumers, and Competitive Markets
equal: While the average product is the slope of the line from the origin, 0B, the
marginal product is the tangent to the total product curve at B (note the equal-
ity of the average and marginal products at point D in Figure 6.1 (b)). Finally,
as we move beyond B toward C, the marginal product falls below the average
product; you can check that the slope of the tangent to the total product curve
at any point between B and C is lower than the slope of the line from the origin.
The Law of Diminishing Marginal Returns
A diminishing marginal product of labor (as well as a diminishing marginal prod-
uct of other inputs) holds for most production processes. The law of diminishing
marginal returns states that as the use of an input increases in equal increments
(with other inputs fixed), a point will eventually be reached at which the resulting
additions to output decrease. When the labor input is small (and capital is fixed),
extra labor adds considerably to output, often because workers are allowed to
devote themselves to specialized tasks. Eventually, however, the law of diminish-
ing marginal returns applies: When there are too many workers, some workers
ecome ineffective and the marginal product of labor falls.
The law of diminishing marginal returns usually applies to the short run when
at least one input is fixed. However, it can also apply to the long run. Even though
inputs are variable in the long run, a manager may still want to analyze produc-
tion choices for which one or more inputs are unchanged. Suppose, for example,
that only two plant sizes are feasible and that management must decide which
to build. In that case, management would want to know when diminishing mar-
ginal returns will set in for each of the two options.
Do not confuse the law of diminishing marginal returns with possible
changes in the quality of labor as labor inputs are increased (as would likely
occur, for example, if the most highly qualified laborers are hired first and the
least qualified last). In our analysis of production, we have assumed that all la-
or inputs are of equal quality; diminishing marginal returns results from limi-
tations on the use of other fixed inputs (e.g., machinery), not from declines in
worker quality. In addition, do not confuse diminishing marginal returns with
negative returns. The law of diminishing marginal returns describes a declining
marginal product but not necessarily a negative one.
The law of diminishing marginal returns applies to a given production tech-
nology. Over time, however, inventions and other improvements in technology
may allow the entire total product curve in Figure 6.1 (a) to shift upward, so
that more output can be produced with the same inputs. Figure 6.2 illustrates
this principle. Initially the output curve is given by O1, but improvements in
technology may allow the curve to shift upward, first to O2, and later to O3.
Suppose, for example, that over time, as labor is increased in agricultural
production, technological improvements are being made. These improvements
might include genetically engineered pest-resistant seeds, more powerful and
effective fertilizers, and better farm equipment. As a result, output changes
from A (with an input of 6 on curve O1) to B (with an input of 7 on curve O2) to
C (with an input of 8 on curve O3).
The move from A to B to C relates an increase in labor input to an increase
in output and makes it appear that there are no diminishing marginal returns
when in fact there are. Indeed, the shifting of the total product curve suggests
that there may be no negative long-run implications for economic growth.
In  fact, as we can see in Example 6.1, the failure to account for long-run im-
provements in technology led British economist Thomas Malthus wrongly to
predict dire consequences from continued population growth.
law of diminishing marginal
eturns Principle that as the
use of an input increases with
other inputs fixed, the resulting
additions to output will eventually
decrease.
CHAPTER 6 PRoDuCTIon 219
109876543210
50
100
Output
pe
time
period
O3
O2
O1
Labor per time period
C
B
A
FIguRE 6.2
the effect of technological
iMProVeMent
Labor productivity (output per unit of labor) can in-
crease if there are improvements in technology, even
though any given production process exhibits di-
minishing returns to labor. As we move from point A
on curve O1 to B on curve O2 to C on curve O3 over
time, labor productivity increases.
ExAMPLE 6.1 the authors debate a Production function for health care
Expenditures on health care
have increased rapidly in many
countries. This is especially true
in the united States, which has
een spending 15% of its GDP
on health care in recent years.
But other countries also devote
substantial resources to health
care (e.g., 11% of GDP in France
and Germany and 8% of GDP in
Japan and the united Kingdom).
Do these increased expenditures reflect increases in
output or do they reflect inefficiencies in the produc-
tion process?
Figure 6.3 shows a production function for health
care in the united States.5 The vertical axis utilizes
one possible measure of health output, the average
increase in life expectancy for the population.
(Another measure of output might be reductions
in the average numbers of heart attacks or strokes.)
The hori zontal axis measures
thousands of dollars spent on
health care inputs, which in-
clude expenditures on doctors,
nurses, administrators, hospital
equipment, and drugs. The pro-
duction function represents the
maximum achievable health
outcome for the population
as a whole, as a function of
the dollars spent per capita on
health care inputs. Points on the production function
such as A, B, and C are by construction inputs that
are being used as efficiently as possible to produce
output. Point D, which lies below the production
function, is inefficient in that the health care inputs
associated with D do not generate the maximum pos-
sible health output.
“The primary explanation for increased spending
on health care is the fact that the production function
5This example is based on Alan M. Ga
er and Jonathan Skinner, “Is American Health Care
Uniquely Inefficient?” Journal of Economic Perspectives, Vol. 22, No. 4 (Fall 2008): 27–50.
220 PART 2 Producers, Consumers, and Competitive Markets
8
7
Increased
Life
Expectancy
(years)
6
4 A
D
C
B
0 10 30 50
Input Expenditures per person ($000)
FIguRE 6.3
a Production
function for
health care
Additional expenditures on health
care (inputs) increase life expec-
tancy (output) along the produc-
tion frontier. Points A, B, and C
epresent points at which inputs
are efficiently utilized, although
there are diminishing returns when
moving from B to C. Point D is a
point of input inefficiency.
for health care exhibits diminishing returns,” claims
Bob. “Figure 6.3 offers a simplified view, but it is
highly instructive. notice that output at point B in
Figure 6.3 is quite a bit higher than the output at
point A. Starting at point A, an additional $20,000
of health expenditures (from $10,000 to $30,000) in-
creases life expectancy by 3 years. However, it’s likely
that the u.S. reached a point such as B decades ago.
As you can see from the graph, a move from B to
C, an additional $20,000 of health expenditures in-
creases life expectancy by only 1 year. Why is this?
The answer is that given medical technologies, ad-
ditional expenditures on medical procedures and/or
the use of newer drugs has only a minimal effect on
life expectancy. As a result, the marginal productiv-
ity of dollars expended on health has become less
and less effective over time as expenditures have
increased.”
“I have a different view,” responds Dan. “The
eality is that the production of health care in the
united States is inefficient, i.e., higher medical
outputs could be achieved with the same or similar
input expenditures if those expenditures were more
effectively utilized. This is shown in the figure as a
move from point D to point B. The figure suggests
that life expectancy could be increased with no ad-
ditional expenditures if the health care system were
un more efficiently.”
“of course, there have been improvements in
technology in the past several decades,” counters
Bob. “not only have new drugs been developed, but
newer medical devices have been designed and pro-
cedures for diagnostics have been improved.”
“That may be true,” responds Dan. “But the inef-
ficiencies in u.S. health care have been prominent
and undeniable. To illustrate, the billing, insurance,
and credentialing system is more complex and bur-
densome in the united States than in many other
countries, so the number of health care administra-
tive personnel per capita is greater. To see the im-
plications of the inefficiencies in u.S. health care,
note that in 2015 the u.S. ranked 31st in the world
in average life expectancy (79.3 years), way be-
hind Switzerland (83.2). Yet the u.S. was spending
$8,713 per capita on health expenditures, whereas
Switzerland’s was only $6,466.”
CHAPTER 6 PRoDuCTIon 221
ExAMPLE 6.2 Malthus and the food crisis
The law of diminishing marginal returns was central to
the thinking of political economist Thomas Malthus
(1766–1834).6 Malthus believed that the world’s limited
amount of land would not be able to supply enough
food as the population grew. He predicted that as both
the marginal and average productivity of labor fell and
there were more mouths to feed, mass hunger and star-
vation would result. Fortunately, Malthus was wrong
(although he was right about the diminishing marginal
eturns to labor).
over the past century, technological improve-
ments have dramatically altered food production in
most countries (including developing countries, such
as India). As a result, the average product of labor
and total food output have increased. These improve-
ments include new high-yielding, disease-resistant
strains of seeds, better fertilizers, and better harvesting
equipment. As the food production index in Table 6.2
shows, overall food production throughout the world
has outpaced population growth continually since
1960.7 This increase in world agricultural productivity
is also illustrated in Figure 6.4, which shows average
cereal yields from 1970 through 2005, along with a
world price index for food.8 note that cereal yields
have increased steadily over the period. Through the
latter part of the twentieth century, growth in agri-
cultural productivity led to increases in food supplies
that outstripped the growth of demand, and food
prices (on average) declined. However, in the twenty-
first century, changing climatic conditions and cut-
acks in food exports have caused demand to outstrip
supply and for world food prices to increase.
6Thomas Malthus, Essay on the Principle of Population, 1798.
7World per capita food production data are from the United Nations Food and Agriculture
Organization (FAO). See also http:
faostat.fao.org.
8Data are from the United Nations Food and Agriculture Organization and the World Bank.
Table 6.2 index of world food
Production Per caPita
YeAr index
1961–64 100
1965 101
1970 105
1975 106
1980 109
1985 115
1990 117
1995 119
2000 127
2005 135
2010 146
2013 151
“I think we’ll have to agree that both explanations
have some validity,” comments Bob. “It is likely that
the united States indeed suffers from inefficiency in
health care production, but is is also likely that as
u.S. incomes have grown, people have demanded
more and more health care relative to other goods,
so that with diminishing returns, the incremental
health benefits have been limited.”
Hunger remains a severe problem in some areas,
such as the Sahel region of Africa, in part because of
the low productivity of labor there. Although other
countries produce an agricultural surplus, mass hun-
ger still occurs because of the difficulty of redistribut-
ing food from more to less productive regions of the
world and because of the low incomes of those less
productive regions.
http:
faostat.fao.org
222 PART 2 Producers, Consumers, and Competitive Markets
Labor Productivity
Although this is a textbook in microeconomics, many of the concepts developed
here provide a foundation for macroeconomic analysis. Macroeconomists are
particularly concerned with labor productivity—the average product of labor
for an entire industry or for the economy as a whole. In this subsection we dis-
cuss labor productivity in the United States and a number of foreign countries.
This topic is interesting in its own right, but will also help to illustrate one of the
links between micro- and macroeconomics.
Because the average product measures output per unit of labor input, it is
elatively easy to measure (total labor input and total output are the only pieces
of information you need). Labor productivity can provide useful comparisons
across industries and for one industry over a long period. But labor productiv-
ity is especially important because it determines the real standard of living that a
country can achieve for its citizens.
ProductiVity and the standard of liVing There is a simple link
etween labor productivity and the standard of living. In any particular year,
the aggregate value of goods and services produced by an economy is equal to
the payments made to all factors of production, including wages, rental pay-
ments to capital, and profit to firms. Consumers ultimately receive these factor
labor productivity Average
product of labor for an entire
industry or for the economy as
a whole.
1.8
1.6
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
4.0
80
100
120
140
160
180
1970 1975 1980 1985 1990 1995 2000 2005 2010
C
ereal Y
ields (m
etric tons per hectare)F
oo
d
P
ic
e
In
de
x
(2
00
2–
20
04
=
1
00
)
Cereal Yield
Food Price Index
FIguRE 6.4
cereal yields and the world Price of food
Cereal yields have increased. The average world price of food decreased in the latter part of the
twentieth century but has been generally increasing in the twenty-first century.
CHAPTER 6 PRoDuCTIon 223
payments in the form of wages, salaries, dividends, or interest payments. As a
esult, consumers in the aggregate can increase their rate of consumption in the
long run only by increasing the total amount they produce.
Understanding the causes of productivity growth is an important area of
esearch in economics. We do know that one of the most important sources of
growth in labor productivity is growth in the stock of capital—i.e., the total
amount of capital available for use in production. Because an increase in capital
means more and better machinery, each worker can produce more output for
each hour worked. Another important source of growth in labor productivity is
technological change—i.e., the development of new technologies that allow la-
or (and other factors of production) to be used more effectively and to produce
new and higher-quality goods.
As Example 6.3 shows, levels of labor productivity have differed consider-
ably across countries, as have rates of growth of productivity. Given the central
ole that productivity has in affecting our standards of living, understanding
these differences is important.
stock of capital Total amount
of capital available for use in
production.
technological change
Development of new technologies
allowing factors of production to
e used more effectively.
ExAMPLE 6.3 labor ProductiVity and the standard of liVing
Will the standard of living in the
united States, Europe, and Japan con-
tinue to improve, or will these econo-
mies barely keep future generations
from being worse off than they are
today? Because the real incomes of
consumers in these countries increase
only as fast as productivity does, the
answer depends on the labor produc-
tivity of workers.
As Table 6.3 shows, the level of
output per employed person in the united States
in 2010 was higher than in other industrial coun-
tries. But two patterns over the post–World War II
period have been distu
ing. First, until the 1990s,
productivity in the united States grew on average
less rapidly than productivity in most other devel-
oped nations. Second, productivity growth during
1980–2014 was much lower in all developed coun-
tries than it had been in the past.9
Throughout most of the 1970–1999 period, Japan
had the highest rate of productivity growth, fol-
lowed by Germany and France. u.S. productivity
growth was the lowest, even somewhat lower than
that of the united Kingdom. This is partly due to
differences in rates of investment and growth in the
stock of capital in each country. The
greatest capital growth during the
postwar period was in Japan, France,
and Germany, which were rebuilt
substantially after World War II. To
some extent, therefore, the lower
ate of growth of productivity in the
united States, when compared to
that of Japan, France, and Germany,
is the result of these countries catch-
ing up after the war.
Productivity growth is also tied to the natural
esource sector of the economy. As oil and other
esources began to be depleted, output per worker
fell. Environmental regulations (e.g., the need to re-
store land to its original condition after strip-mining
for coal) magnified this effect as the public became
more concerned with the importance of cleaner air
and water.
observe from Table 6.3 that productivity growth
in the united States accelerated in the 1990s and
2000s. Some economists believe that information
and communication technology (ICT) has been
the key impetus for this growth. However, sluggish
growth in more recent years suggests that ICT’s con-
tribution may have already peaked.
9Recent growth numbers on GDP, employment, and GDP per hour worked are available at
http:
www.oecd.org.
http:
www.oecd.org
224 PART 2 Producers, Consumers, and Competitive Markets
6.3 Production with Two Variable Inputs
We have completed our analysis of the short-run production function in which
one input, labor, is variable, and the other, capital, is fixed. Now we turn to the
long run, for which both labor and capital are variable. The firm can now pro-
duce its output in a variety of ways by combining different amounts of labor
and capital. In this section, we will see how a firm can choose among combina-
tions of labor and capital that generate the same output. In the first subsection,
we will examine the scale of the production process, analyzing how output
changes as input combinations are doubled, tripled, and so on.
Isoquants
Let’s begin by examining the production technology of a firm that uses two
inputs and can vary both of them. Suppose that the inputs are labor and capital
and that they are used to produce food. Table 6.4 tabulates the output achiev-
able for various combinations of inputs.
Table 6.4 Production with two Variable inPuts
Labor Input
CApitAL input 1 2 3 4 5
1 20 40 55 65 75
2 40 60 75 85 90
3 55 75 90 100 105
4 65 85 100 110 115
5 75 90 105 115 120
Table 6.3 labor ProductiVity in deVeloPed countries
unIted
StateS Japan France Germany
unIted
KInGdom
Gdp per Hour WorKed (In 2010 u.S. doLLarS)
$62.41 $39.39 $60.28 $58.92 $47.39
YeArs AnnuAL rAte of growth of LAbor produCtivitY (%)
1970–1979 1.7 4.5 4.3 4.1 3.2
1980–1989 1.4 3.8 2.9 2.1 2.2
1990–1999 1.7 2.4 2.0 2.3 2.3
2000–2009 2.1 1.3 1.2 1.1 1.5
2010–2014 0.7 1.2 0.9 1.2 0.5
CHAPTER 6 PRoDuCTIon 225
Labor inputs are listed across the top row, capital inputs down the column
on the left. Each entry in the table is the maximum (technically efficient) output
that can be produced each year with each combination of labor and capital used
over that year. For example, 4 units of labor per year and 2 units of capital per
year yield 85 units of food per year. Reading along each row, we see that out-
put increases as labor inputs are increased, while capital inputs remain fixed.
Reading down each column, we see that output also increases as capital inputs
are increased, while labor inputs remain fixed.
The information in Table 6.4 can also be represented graphically using iso-
quants. An isoquant is a curve that shows all the possible combinations of inputs
that yield the same output. Figure 6.5 shows three isoquants. (Each axis in the
figure measures the quantity of inputs.) These isoquants are based on the data
in Table 6.4, but are drawn as smooth curves to allow for the use of fractional
amounts of inputs.
For example, isoquant q1 shows all combinations of labor and capital per year
that together yield 55 units of output per year. Two of these points, A and D,
co
espond to Table 6.4. At A, 1 unit of labor and 3 units of capital yield 55 units
of output; at D, the same output is produced from 3 units of labor and 1 unit
of capital. Isoquant q2 shows all combinations of inputs that yield 75 units of
output and co
esponds to the four combinations of labor and capital circled in
the table (e.g., at B, where 2 units of labor and 3 units of capital are combined).
Isoquant q2 lies above and to the right of q1 because obtaining a higher level of
output requires more labor and capital. Finally, isoquant q3 shows labor-capital
combinations that yield 90 units of output. Point C, for example, involves 3
units of labor and 3 units of capital, whereas Point E involves 2 units of labor
and 5 units of capital.
isoquant MaPs When a number of isoquants are combined in a single
graph, we call the graph an isoquant map. Figure 6.5 shows three of the
many isoquants that make up an isoquant map. An isoquant map is another
way of describing a production function, just as an indifference map is a way
isoquant Curve showing all
possible combinations of inputs
that yield the same output.
isoquant map Graph
combining a number of isoquants,
used to describe a production
function.
Capital
pe
yea
Labor per yea
A
D
1 2 3 4
1
2
3
4
5 E
q3 � 90
q2 � 75
q1 � 55
5
B C
FIguRE 6.5
Production with two
Variable inPuts
Production isoquants show the various com-
inations of inputs necessary for the firm to
produce a given output. A set of isoquants, or
isoquant map, describes the firm’s production
function. Output increases as we move from
isoquant q1 (at which 55 units per year are pro-
duced at points such as A and D), to isoquant
q2 (75 units per year at points such as B), and to
isoquant q3 (90 units per year at points such as
C and E).
226 PART 2 Producers, Consumers, and Competitive Markets
of describing a utility function. Each isoquant co
esponds to a different level
of output, and the level of output increases as we move up and to the right in
the figure.
Input Flexibility
Isoquants show the flexibility that firms have when making production deci-
sions: They can usually obtain a particular output by substituting one input
for another. It is important for managers to understand the nature of this
flexibility. For example, fast-food restaurants have recently faced shortages of
young, low-wage employees. Companies have responded by automating—
adding self-service salad bars and introducing more sophisticated cooking
equipment. They have also recruited older people to fill positions. As we will
see in Chapters 7 and 8, by taking into account this flexibility in the produc-
tion process, managers can choose input combinations that minimize cost and
maximize profit.
Diminishing Marginal Returns
Even though both labor and capital are variable in the long run, it is useful
for a firm that is choosing the optimal mix of inputs to ask what happens to
output as each input is increased, with the other input held fixed. The out-
come of this exercise is described in Figure 6.5, which reflects diminishing
marginal returns to both labor and capital. We can see why there are dimin-
ishing marginal returns to labor by drawing a horizontal line at a particular
level of capital—say, 3. Reading the levels of output from each isoquant as
labor is increased, we note that each additional unit of labor generates less
and less additional output. For example, when labor is increased from 1 unit
to 2 (from A to B), output increases by 20 (from 55 to 75). However, when la-
or is increased by an additional unit (from B to C), output increases by only
15 (from 75 to 90). Thus there are diminishing marginal returns to labor both
in the long and short run. Because adding one factor while holding the other
factor constant eventually leads to lower and lower incremental output, the
isoquant must become steeper as more capital is added in place of labor and
flatter when labor is added in place of capital.
There are also diminishing marginal returns to capital. With labor fixed, the
marginal product of capital decreases as capital is increased. For example, when
capital is increased from 1 to 2 and labor is held constant at 3, the marginal
product of capital is initially 20 (75 – 55) but falls to 15 (90 – 75) when capital is
increased from 2 to 3.
Substitution Among Inputs
With two inputs that can be varied, a manager will want to consider substitut-
ing one input for another. The slope of each isoquant indicates how the quantity
of one input can be traded off against the quantity of the other, while output
is held constant. When the negative sign is removed, we call the slope the
marginal rate of technical substitution (MRTS). The marginal rate of technical
substitution of labor for capital is the amount by which the input of capital can be
educed when one extra unit of labor is used, so that output remains constant.
This is analogous to the marginal rate of substitution (MRS) in consumer theory.
Recall from Section 3.1 that the MRS describes how consumers substitute
marginal rate of
technical substitution
(MRTS) Amount by which
the quantity of one input can
e reduced when one extra unit
of another input is used, so that
output remains constant.
CHAPTER 6 PRoDuCTIon 227
among two goods while holding the level of satisfaction constant. Like the
MRS, the MRTS is always measured as a positive quantity:
MRTS = -Change in capital input/change in labor input
= - ∆K/∆L(for a fixed level of q)
where ∆K and ∆L are small changes in capital and labor along an isoquant.
In Figure 6.6 the MRTS is equal to 2 when labor increases from 1 unit to 2
and output is fixed at 75. However, the MRTS falls to 1 when labor is increased
from 2 units to 3, and then declines to 2/3 and to 1/3. Clearly, as more and more
labor replaces capital, labor becomes less productive and capital becomes rela-
tively more productive. Therefore, we need less capital to keep output constant,
and the isoquant becomes flatter.
diMinishing Mrts We assume that there is a diminishing MRTS. In other
words, the MRTS falls as we move down along an isoquant. The mathemati-
cal implication is that isoquants, like indifference curves, are convex, or bowed
inward. This is indeed the case for most production technologies. The di-
minishing MRTS tells us that the productivity of any one input is limited. As
more and more labor is added to the production process in place of capital,
the productivity of labor falls. Similarly, when more capital is added in place
of labor, the productivity of capital falls. Production needs a balanced mix of
oth inputs.
As our discussion has just suggested, the MRTS is closely related to the mar-
ginal products of labor MPL and capital MPK. To see how, imagine adding some
labor and reducing the amount of capital sufficient to keep output constant. The
additional output resulting from the increased labor input is equal to the addi-
tional output per unit of additional labor (the marginal product of labor) times
the number of units of additional labor:
Additional output from increased use of labor = (MPL)(∆L)
in §3.1, we explain that the
marginal rate of substitution
is the maximum amount of
one good that the consumer
is willing to give up to obtain
one unit of another good.
in §3.1, we explain that an
indifference curve is convex
if the marginal rate of sub-
stitution diminishes as we
move down along the curve.
1 2 3 4 50
1
2
3
4
5
Capital
pe
yea
Labor per yea
q3 = 90
q2 = 75
q1 = 55
�K = 2
�L = 1
�K = 1
�L = 1 �L = 1
�L = 1
�K = 2 3
�K = 1 3
FIguRE 6.6
Marginal rate of
technical substitution
Like indifference curves, isoquants are down-
ward sloping and convex. The slope of the iso-
quant at any point measures the marginal rate
of technical substitution—the ability of the firm
to replace capital with labor while maintaining
the same level of output. On isoquant q2, the
MRTS falls from 2 to 1 to 2/3 to 1/3.
228 PART 2 Producers, Consumers, and Competitive Markets
Similarly, the decrease in output resulting from the reduction in capital is the
loss of output per unit reduction in capital (the marginal product of capital)
times the number of units of capital reduction:
Reduction in output from decreased use of capital = (MPK)(∆K)
Because we are keeping output constant by moving along an isoquant, the total
change in output must be zero. Thus,
(MPL)(∆L) + (MPK)(∆K) = 0
Now, by rea
anging terms we see that
(MPL)/(MPK) = -(∆K/∆L) = MRTS (6.2)
Equation (6.2) tells us that the marginal rate of technical substitution between two
inputs is equal to the ratio of the marginal products of the inputs. This formula will be
useful when we look at the firm’s cost-minimizing choice of inputs in Chapter 7.
Production Functions—Two Special Cases
Two extreme cases of production functions show the possible range of input
substitution in the production process. In the first case, shown in Figure 6.7,
inputs to production are perfect substitutes for one another. Here the MRTS is
constant at all points on an isoquant. As a result, the same output (say q3) can be
produced with mostly capital (at A), with mostly labor (at C), or with a balanced
combination of both (at B). For example, musical instruments can be manu-
factured almost entirely with machine tools or with very few tools and highly
skilled labor.
Figure 6.8 illustrates the opposite extreme, the fixed-proportions produc-
tion function, sometimes called a Leontief production function. In this case,
it is impossible to make any substitution among inputs. Each level of out-
put requires a specific combination of labor and capital: Additional output
in §3.1, we explain that two
goods are perfect substi-
tutes if the marginal rate of
substitution of one for the
other is a constant.
fixed-proportions
production function
Production function with
L-shaped isoquants, so that only
one combination of labor and
capital can be used to produce
each level of output.
Capital
pe
yea
A
B
C
Labor per yea
q1 q2 q3
FIguRE 6.7
isoquants when inPuts are
Perfect substitutes
When the isoquants are straight lines, the MRTS is
constant. Thus the rate at which capital and labor
can be substituted for each other is the same no
matter what level of inputs is being used. Points
A, B, and C represent three different capital-labor
combinations that generate the same output q3.
CHAPTER 6 PRoDuCTIon 229
cannot be obtained unless more capital and labor are added in specific
proportions. As a result, the isoquants are L-shaped, just as indifference
curves are L-shaped when two goods are perfect complements. An example
is the reconstruction of concrete sidewalks using jackhammers. It takes one
person to use a jackhammer—neither two people and one jackhammer nor
one person and two jackhammers will increase production. As another ex-
ample, suppose that a cereal company offers a new
eakfast cereal, Nutty
Oat Crunch, whose two inputs, not surprisingly, are oats and nuts. The se-
cret formula for the cereal requires exactly one ounce of nuts for every four
ounces of oats in every serving. If the company were to purchase additional
nuts but not additional oats, the output of cereal would remain unchanged,
since the nuts must be combined with the oats in a fixed proportion.
Similarly, purchasing additional oats without additional nuts would also be
unproductive.
In Figure 6.8 points A, B, and C represent technically efficient combinations
of inputs. For example, to produce output q1, a quantity of labor L1 and capital
K1 can be used, as at A. If capital stays fixed at K1, adding more labor does not
change output. Nor does adding capital with labor fixed at L1. Thus, on the
vertical and the horizontal segments of the L-shaped isoquants, either the mar-
ginal product of capital or the marginal product of labor is zero. Higher output
esults only when both labor and capital are added, as in the move from input
combination A to input combination B.
The fixed-proportions production function describes situations in which
methods of production are limited. For example, the production of a television
show might involve a certain mix of capital (camera and sound equipment, etc.)
and labor (producer, director, actors, etc.). To make more television shows, all
inputs to production must be increased proportionally. In particular, it would
e difficult to increase capital inputs at the expense of labor, because actors are
necessary inputs to production (except perhaps for animated films). Likewise,
it would be difficult to substitute labor for capital, because filmmaking today
equires sophisticated film equipment.
in §3.1, we explain that two
goods are perfect comple-
ments when the indifference
curves for the goods are
shaped as right angles.
Capital
pe
yea
K1
Labor per yea
A
B
C
L1
q1
q2
q3
FIguRE 6.8
fixed-ProPortions Production
function
When the isoquants are L-shaped, only one combination
of labor and capital can be used to produce a given out-
put (as at point A on isoquant q1, point B on isoquant q2,
and point C on isoquant q3). Adding more labor alone
does not increase output, nor does adding more capital
alone.
230 PART 2 Producers, Consumers, and Competitive Markets
ExAMPLE 6.4 a Production function for wheat
Capital
(machine
hours pe
year)
120
100
90
80
40
Output = 13,800 Bushels
per Yea
500 760 1000250
Labor (hours per year)
A
BDK = 210
DL = 260
FIguRE 6.9
isoquant
describing the
Production of
wheat
A wheat output of 13,800 bushels
per year can be produced with dif-
ferent combinations of labor and
capital. The more capital-intensive
production process is shown as
point A, the more labor-intensive
process as point B. The marginal
ate of technical substitution be-
tween A and B is 10/260 = 0.04.
Crops can be produced using differ-
ent methods. Food grown on large
farms in the united States is usually
produced with a capital-intensive
technology, which involves sub-
stantial investments in capital,
such as buildings and equipment,
and relatively little input of labor.
However, food can also be pro-
duced using very little capital (a hoe) and a lot of
labor (several people with the patience and stamina
to work the soil). one way to describe the agricul-
tural production process is to show one isoquant (or
more) that describes the combination of inputs which
generates a given level of output (or several output
levels). The description that follows comes from a
production function for wheat that was estimated
statistically.10
Figure 6.9 shows one isoquant,
associated with the production
function, co
esponding to an out-
put of 13,800 bushels of wheat
per year. The manager of the farm
can use this isoquant to decide
whether it is profitable to hire
more labor or use more machin-
ery. Assume the farm is cu
ently
operating at A, with a labor input L of 500 hours and
a capital input K of 100 machine hours. The manager
decides to experiment by using only 90 hours of ma-
chine time. To produce the same crop per year, he
finds that he needs to replace this machine time by
adding 260 hours of labor.
The results of this experiment tell the manager about
the shape of the wheat production isoquant. When he
compares points A (where L = 500 and K = 100)
10The food production function on which this example is based is given by the equation
q = 100(K8L2), where q is the rate of output in bushels of wheat per year, K is the quantity of ma-
chines in use per year, and L is the number of hours of labor per year.
CHAPTER 6 PRoDuCTIon 231
6.4 Returns to Scale
Our analysis of input substitution in the production process has shown us what
happens when a firm substitutes one input for another while keeping output
constant. However, in the long run, with all inputs variable, the firm must also
consider the best way to increase output. One way to do so is to change the
scale of the operation by increasing all of the inputs to production in proportion. If
it takes one farmer working with one harvesting machine on one acre of land
to produce 100 bushels of wheat, what will happen to output if we put two
farmers to work with two machines on two acres of land? Output will almost
certainly increase, but will it double, more than double, or less than double?
Returns to scale is the rate at which output increases as inputs are increased
proportionately. We will examine three different cases: increasing, constant, and
decreasing returns to scale.
increasing returns to scale If output more than doubles when in-
puts are doubled, there are increasing returns to scale. This might arise be-
cause the larger scale of operation allows managers and workers to specialize
in their tasks and to make use of more sophisticated, large-scale factories and
equipment. The automobile assembly line is a famous example of increasing
eturns.
The prospect of increasing returns to scale is an important issue from a
public–policy perspective. If there are increasing returns, then it is economi-
cally advantageous to have one large firm producing (at relatively low cost)
ather than to have many small firms (at relatively high cost). Because this large
firm can control the price that it sets, it may need to be regulated. For example,
eturns to scale Rate at
which output increases as inputs
are increased proportionately.
increasing returns to
scale Situation in which output
more than doubles when all
inputs are doubled.
and B (where L = 760 and K = 90) in Figure 6.9,
oth of which are on the same isoquant, the manager
finds that the marginal rate of technical substitution is
equal to 0.04 (- ∆K/∆L) = - (-10)/260 = .04).
The MRTS reveals the nature of the trade-off
involved in adding labor and reducing the use of
farm machinery. Because the MRTS is substan-
tially less than 1 in value, the manager knows that
when the wage of a laborer is equal to the cost of
unning a machine, he ought to use more capital.
(At his cu
ent level of production, he needs 260
units of labor to substitute for 10 units of capital.)
In fact, he knows that unless labor is much less ex-
pensive than the use of a machine, his production
process ought to become more capital-intensive.
The decision about how many laborers to hire
and machines to use cannot be fully resolved until
we discuss the costs of production in the next chap-
ter. However, this example illustrates how knowl-
edge about production isoquants and the marginal
ate of technical substitution can help a manager. It
also suggests why most farms in the united States
and Canada, where labor is relatively expensive,
operate in the range of production in which the
MRTS is relatively high (with a high capital-to-labor
atio), whereas farms in developing countries, in
which labor is cheap, operate with a lower MRTS
(and a lower capital-to-labor ratio).11 The exact
labo
capital combination to use depends on input
prices, a subject that we discuss in Chapter 7.
11With the production function given in footnote 6, it is not difficult (using calculus) to show that
the marginal rate of technical substitution is given by MRTS = (MPL/MPK) = (1/4)(K/L). Thus, the
MRTS decreases as the capital-to-labor ratio falls. For an interesting study of agricultural produc-
tion in Israel, see Richard E. Just, David Zilberman, and Eithan Hochman, “Estimation of Multicrop
Production Functions,” American Journal of Agricultural Economics 65 (1983): 770–80.
232 PART 2 Producers, Consumers, and Competitive Markets
increasing returns in the provision of electricity is one reason why we have
large, regulated power companies.
constant returns to scale A second possibility with respect to the scale
of production is that output may double when inputs are doubled. In this case,
we say there are constant returns to scale. With constant returns to scale, the size
of the firm’s operation does not affect the productivity of its factors: Because one
plant using a particular production process can easily be replicated, two plants
produce twice as much output. For example, a large travel agency might provide
the same service per client and use the same ratio of capital (office space) and
labor (travel agents) as a small agency that services fewer clients.
decreasing returns to scale Finally, output may less than double when
all inputs double. This case of decreasing returns to scale applies to some firms
with large-scale operations. Eventually, difficulties in organizing and running
a large-scale operation may lead to decreased productivity of both labor and
capital. Communication between workers and managers can become difficult
to monitor as the workplace becomes more impersonal. Thus, the decreasing-
eturns case is likely to be associated with the problems of coordinating tasks and
maintaining a useful line of communication between management and workers.
Describing Returns to Scale
Returns to scale need not be uniform across all possible levels of output. For ex-
ample, at lower levels of output, the firm could have increasing returns to scale,
ut constant and eventually decreasing returns at higher levels of output.
The presence or absence of returns to scale is seen graphically in the two
parts of Figure 6.10. The line 0A from the origin in each panel describes a
constant returns to scale
Situation in which output doubles
when all inputs are doubled.
decreasing returns to scale
Situation in which output less
than doubles when all inputs are
doubled.
Capital
(machine
hours)
Capital
(machine
hours)
6
4
2
0
2
4
5 10
30
20
10
30
20
10
15 5 10
Labor (hours) Labor (hours)
(a) (b)
A A
0
FIguRE 6.10
eturns to scale
When a firm’s production process exhibits constant returns to scale as shown by a movement along
line 0A in part (a), the isoquants are equally spaced as output increases proportionally. However, when
there are increasing returns to scale as shown in (b), the isoquants move closer together as inputs are
increased along the line.
CHAPTER 6 PRoDuCTIon 233
ExAMPLE 6.5 returns to scale in the carPet industry
The carpet industry in the united
States centers on the town of
Dalton in northern Georgia. From a
elatively small industry with many
small firms in the first half of the
twentieth century, it grew rapidly
and became a major industry with
a large number of firms of all sizes.
Cu
ently, there are three
elatively large manufacturers
(Shaw, Mohawk, and Beaulieu),
along with a number of smaller
producers. There are also many
etailers, wholesale distributors,
buying groups, and national re-
tail chains. The carpet industry has
grown rapidly for several reasons.
Consumer demand for wool, ny-
lon, and polypropylene carpets in commercial and
esidential uses has skyrocketed. In addition, inno-
vations such as the introduction of larger, faster, and
more efficient carpet-tufting machines have reduced
costs and greatly increased car-
pet production. Along with the in-
crease in production, innovation
and competition have worked to-
gether to reduce real carpet prices.
To what extent, if any, can the
growth of the carpet industry be ex-
plained by the presence of returns
to scale? There have certainly been
substantial improvements in the
processing of key production inputs
(such as stain-resistant yarn) and in
the distribution of carpets to retail-
ers and consumers. But what about
the production of carpets? Carpet
production is capital intensive—
manufacturing plants require heavy
investments in high-speed tufting
machines that turn various types of yarn into carpet,
as well as machines that put the backings onto the
carpets, cut the carpets into appropriate sizes, and
package, label, and distribute them.
production process in which labor and capital are used as inputs to produce
various levels of output in the ratio of 5 hours of labor to 2 hours of machine
time. In Figure 6.10 (a), the firm’s production function exhibits constant returns
to scale. When 5 hours of labor and 2 hours of machine time are used, an out-
put of 10 units is produced. When both inputs double, output doubles from 10
to 20 units; when both inputs triple, output triples, from 10 to 30 units. Put dif-
ferently, twice as much of both inputs is needed to produce 20 units, and three
times as much is needed to produce 30 units.
In Figure 6.10 (b), the firm’s production function exhibits increasing returns
to scale. Now the isoquants come closer together as we move away from the
origin along 0A. As a result, less than twice the amount of both inputs is needed
to increase production from 10 units to 20; substantially less than three times
the inputs are needed to produce 30 units. The reverse would be true if the pro-
duction function exhibited decreasing returns to scale (not shown here). With
decreasing returns, the isoquants are increasingly distant from one another as
output levels increase proportionally.
Returns to scale vary considerably across firms and industries. Other
things being equal, the greater the returns to scale, the larger the firms in an
industry are likely to be. Because manufacturing involves large investments in
capital equipment, manufacturing industries are more likely to have increas-
ing returns to scale than service-oriented industries. Services are more labor-
intensive and can usually be provided as efficiently in small quantities as they
can on a large scale.
234 PART 2 Producers, Consumers, and Competitive Markets
overall, physical capital (including plant and
equipment) accounts for about 77 percent of a
typical carpet manufacturer’s costs, while labor
accounts for the remaining 23 percent. over time,
the major carpet manufacturers have increased the
scale of their operations by putting larger and more
efficient tufting machines into larger plants. At the
same time, the use of labor in these plants has also
increased significantly. The result? Proportional in-
creases in inputs have resulted in a more than
proportional increase in output for these larger
plants. For example, a doubling of capital and la-
or inputs might lead to a 110-percent increase in
output. This pattern has not, however, been uniform
across the industry. Most smaller carpet manufac-
turers have found that small changes in scale have
little or no effect on output; i.e., small proportional
increases in inputs have only increased output
proportionally.
We can therefore characterize the carpet industry
as one in which there are constant returns to scale
for relatively small plants but increasing returns
to scale for larger plants. These increasing returns,
however, are limited, and we can expect that if plant
size were increased further, there would eventually
e decreasing returns to scale.
QueSTionS foR Review
1. What is a production function? How does a long-run
production function differ from a short-run produc-
tion function?
2. Why is the marginal product of labor likely to increase
initially in the short run as more of the variable input
is hired?
3. Why does production eventually experience diminish-
ing marginal returns to labor in the short run?
SuMMaRy
1. A production function describes the maximum output
that a firm can produce for each specified combination
of inputs.
2. In the short run, one or more inputs to the production
process are fixed. In the long run, all inputs are poten-
tially variable.
3. Production with one variable input, labor, can be use-
fully described in terms of the average product of labor
(which measures output per unit of labor input) and
the marginal product of labor (which measures the ad-
ditional output as labor is increased by 1 unit).
4. According to the law of diminishing marginal returns,
when one or more inputs are fixed, a variable input
(usually labor) is likely to have a marginal prod-
uct that eventually diminishes as the level of input
increases.
5. An isoquant is a curve that shows all combinations
of inputs that yield a given level of output. A firm’s
production function can be represented by a series of
isoquants associated with different levels of output.
6. Isoquants always slope downward because the mar-
ginal product of all inputs is positive. The shape of
each isoquant can be described by the marginal rate
of technical substitution at each point on the isoquant.
The marginal rate of technical substitution of labor for capi-
tal (MRTS) is the amount by which the input of capital
can be reduced when one extra unit of labor is used so
that output remains constant.
7. The standard of living that a country can attain for its
citizens is closely related to its level of labor produc-
tivity. Decreases in the rate of productivity growth
in developed countries are due in part to the lack of
growth of capital investment.
8. The possibilities for substitution among inputs in the
production process range from a production function
in which inputs are perfect substitutes to one in which
the proportions of inputs to be used are fixed (a fixed-
proportions production function).
9. In long-run analysis, we tend to focus on the firm’s
choice of its scale or size of operation. Constant returns
to scale means that doubling all inputs leads to doubling
output. Increasing returns to scale occurs when output
more than doubles when inputs are doubled; decreasing
eturns to scale applies when output less than doubles.
4. You are an employer seeking to fill a vacant position
on an assembly line. Are you more concerned with
the average product of labor or the marginal product
of labor for the last person hired? If you observe that
your average product is just beginning to decline,
should you hire any more workers? What does this
situation imply about the marginal product of your
last worker hired?
CHAPTER 6 PRoDuCTIon 235
exeRciSeS
1. The menu at Joe’s coffee shop consists of a variety of
coffee drinks, pastries, and sandwiches. The marginal
product of an additional worker can be defined as
the number of customers that can be served by that
worker in a given time period. Joe has been employing
one worker, but is considering hiring a second and a
third. Explain why the marginal product of the second
and third workers might be higher than the first. Why
might you expect the marginal product of additional
workers to diminish eventually?
2. Suppose a chair manufacturer is producing in the short
un (with its existing plant and equipment). The manu-
facturer has observed the following levels of produc-
tion co
esponding to different numbers of workers:
Number of Workers Number of Chairs
1 10
2 18
3 24
4 28
5 30
6 28
7 25
a. Calculate the marginal and average product of
labor for this production function.
. Does this production function exhibit diminishing
eturns to labor? Explain.
c. Explain intuitively what might cause the marginal
product of labor to become negative.
5. What is the difference between a production function
and an isoquant?
6. Faced with constantly changing conditions, why
would a firm ever keep any factors fixed? What
criteria determine whether a factor is fixed or
variable?
7. Isoquants can be convex, linear, or L-shaped. What
does each of these shapes tell you about the nature
of the production function? What does each of these
shapes tell you about the MRTS?
8. Can an isoquant ever slope upward? Explain.
9. Explain the term “marginal rate of technical substitu-
tion.” What does a MRTS = 4 mean?
3. Fill in the gaps in the table below.
QuaNtity
of Variable
iNput
total
output
margiNal
produCt
of Variable
iNput
aVerage
produCt
of Variable
iNput
0 0 — —
1 225
2 300
3 300
4 1140
5 225
6 225
4. A political campaign manager must decide whether
to emphasize television advertisements or letters to
potential voters in a reelection campaign. Describe
the production function for campaign votes. How
might information about this function (such as the
shape of the isoquants) help the campaign manager
to plan strategy?
5. For each of the following examples, draw a represen-
tative isoquant. What can you say about the marginal
ate of technical substitution in each case?
a. A firm can hire only full-time employees to produce
its output, or it can hire some combination of full-
time and part-time employees. For each full-time
worker let go, the firm must hire an increasing num-
er of temporary employees to maintain the same
level of output.
10. Explain why the marginal rate of technical substitu-
tion is likely to diminish as more and more labor is
substituted for capital.
11. Is it possible to have diminishing returns to a single
factor of production and constant returns to scale at
the same time? Discuss.
12. Can a firm have a production function that exhibits in-
creasing returns to scale, constant returns to scale, and
decreasing returns to scale as output increases? Discuss.
13. Suppose that output q is a function of a single in-
put, labor (L). Describe the returns to scale associ-
ated with each of the following production functions:
(a) q = L/2 (b) q = L2 + L (c) q = log (L).
236 PART 2 Producers, Consumers, and Competitive Markets
. A firm finds that it can always trade two units of
labor for one unit of capital and still keep output
constant.
c. A firm requires exactly two full-time workers to op-
erate each piece of machinery in the factory.
6. A firm has a production process in which the inputs to
production are perfectly substitutable in the long run.
Can you tell whether the marginal rate of technical
substitution is high or low, or is further information
necessary? Discuss.
7. The marginal product of labor in the production of
computer chips is 50 chips per hour. The marginal rate
of technical substitution of hours of labor for hours of
machine capital is 1/4. What is the marginal product
of capital?
8. Do the following functions exhibit increasing, con-
stant, or decreasing returns to scale? What happens to
the marginal product of each individual factor as that
factor is increased and the other factor held constant?
a. q = 3L + 2K
. q = (2L + 2K)1/2
c. q = 3LK2
d. q = L1/2K1/2
e. q = 4L1/2 + 4K
9. The production function for the personal computers of
DISK, Inc., is given by
q = 10K0.5L0.5
where q is the number of computers produced per day,
K is hours of machine time, and L is hours of labor in-
put. DISK’s competitor, FLOPPY, Inc., is using the pro-
duction function
q = 10K0.6L0.4
a. If both companies use the same amounts of capital
and labor, which will generate more output?
. Assume that capital is limited to 9 machine
hours, but labor is unlimited in supply. In which
company is the marginal product of labor greater?
Explain.
10. In Example 6.4, wheat is produced according to the
production function
q = 100(K0.8L0.2)
a. Beginning with a capital input of 4 and a labor input
of 49, show that the marginal product of labor and
the marginal product of capital are both decreasing.
. Does this production function exhibit increasing,
decreasing, or constant returns to scale?
11. Suppose life expectancy in years (L) is a function of
two inputs, health expenditures (H) and nutrition
expenditures (N) in hundreds of dollars per year. The
production function is L = c H0.8N0.2.
a. Beginning with a health input of $400 per year
(H = 4) and a nutrition input of $4900 per year
(N = 49), show that the marginal product of health
expenditures and the marginal product of nutrition
expenditures are both decreasing.
. Does this production function exhibit increasing,
decreasing, or constant returns to scale?
c. Suppose that in a country suffering from famine,
N is fixed at 2 and that c = 20. Plot the production
function for life expectancy as a function of health
expenditures, with L on the vertical axis and H on
the horizontal axis.
d. Now suppose another nation provides food aid
to the country suffering from famine so that N in-
creases to 4. Plot the new production function.
e. Now suppose that N = 4 and H = 2. You run a
charity that can provide either food aid or health
aid to this country. Which would provide a greater
enefit: increasing H by 1 or N by 1?
237
In the last chapter, we examined the firm’s production technology— the relationship that shows how factor inputs can be transformed into outputs. Now we will see how the production technology, to-
gether with the prices of factor inputs, determines the firm’s cost of
production.
Given a firm’s production technology, managers must decide how to
produce. As we saw, inputs can be combined in different ways to yield
the same amount of output. For example, one can produce a certain
output with a lot of labor and very little capital, with very little labor
and a lot of capital, or with some other combination of the two. In this
chapter we see how the optimal—i.e., cost-minimizing—combination
of inputs is chosen. We will also see how a firm’s costs depend on its
ate of output and show how these costs are likely to change over time.
We begin by explaining how cost is defined and measured, distin-
guishing between the concept of cost used by economists, who are
concerned about the firm’s future performance, and by accountants,
who focus on the firm’s financial statements. We then examine how
the characteristics of the firm’s production technology affect costs,
oth in the short run, when the firm can do little to change its capi-
tal stock, and in the long run, when the firm can change all its factor
inputs.
We then show how the concept of returns to scale can be general-
ized to allow for both changes in the mix of inputs and the production
of many different outputs. We also show how cost sometimes falls
over time as managers and workers learn from experience and make
production processes more efficient. Finally, we show how empirical
information can be used to estimate cost functions and predict future
costs.
7.1 Measuring Cost: Which
Costs Matter?
Before we can analyze how firms minimize costs, we must clarify
what we mean by cost in the first place and how we should measure it.
What items, for example, should be included as part of a firm’s cost?
Cost obviously includes the wages that a firm pays its workers and the
ent that it pays for office space. But what if the firm already owns an
7.1 Choosing the Location for a
New Law School Building 240
7.2 Sunk, Fixed, and Variable
Costs: Computers,
Software, and Pizzas 243
7.3 The Short-Run Cost of
Aluminum Smelting 248
7.4 The Effect of Effluent
Fees on Input Choices 255
7.5 Reducing the Use
of Energy 259
7.6 Tesla’s Battery Costs 265
7.7 Economies of Scope in
the Trucking Industry 269
7.8 The Learning Curve
in Practice 273
7.9 Cost Functions for
Electric Power 277
LIST oF ExAmPLES
7.1 measuring Cost: Which Costs
matter? 237
7.2 Cost in the Short Run 245
7.3 Cost in the Long Run 251
7.4 Long-Run versus
Short-Run Cost Curves 261
7.5 Production with Two
outputs—Economies of
Scope 267
ChAPTER oUTLINE
The Cost of
Production
ChAPTER 7
*7.6 Dynamic Changes in
Costs—The Learning Curve 270
*7.7 Estimating and
Predicting Cost 275
Appendix: Production and
Cost Theory—A mathematical
Treatment 283
238 PART 2 Producers, Consumers, and Competitive markets
office building and doesn’t have to pay rent? How should we treat money that
the firm spent two or three years ago (and can’t recover) for equipment or for
esearch and development? We’ll answer questions such as these in the context
of the economic decisions that managers make.
Economic Cost versus Accounting Cost
Economists think of cost differently from financial accountants, who are usually
concerned with keeping track of assets and liabilities and reporting past perfor-
mance for external use, as in annual reports. Financial accountants tend to take
a retrospective view of the firm’s finances and operations. As a result account-
ing cost—the cost that financial accountants measure—can include items that
an economist would not include and may not include items that economists
usually do include. For example, accounting cost includes actual expenses plus
depreciation expenses for capital equipment, which are determined on the basis
of the allowable tax treatment by the Internal Revenue Service.
Economists—and we hope managers—take a forward-looking view. They
are concerned with the allocation of scarce resources. Therefore, they care about
what cost is likely to be in the future and about ways in which the firm might be
able to rea
ange its resources to lower its costs and improve its profitability. As
we will see, economists are therefore concerned with economic cost, which is
the cost of utilizing resources in production. What kinds of resources are part of
economic cost? The word economic tells us to distinguish between the costs the
firm can control and those it cannot. It also tells us to consider all costs relevant
to production. Clearly capital, labor, and raw materials are resources whose
costs should be included. But the firm might use other resources with costs that
are less obvious, but equally important. Here the concept of opportunity cost
plays an important role.
Opportunity Cost
Opportunity cost is the cost associated with opportunities that are forgone
y not putting the firm’s resources to their best alternative use. This is easiest
to understand through an example. Consider a firm that owns a building and
therefore pays no rent for office space. Does this mean the cost of office space
is zero? The firm’s managers and accountant might say yes, but an economist
would disagree. The economist would note that the firm could have earned
ent on the office space by leasing it to another company. Leasing the office
space would mean putting this resource to an alternative use, a use that would
have provided the firm with rental income. This forgone rent is the opportunity
cost of utilizing the office space. And because the office space is a resource that
the firm is utilizing, this opportunity cost is also an economic cost of doing
usiness.
What about the wages and salaries paid to the firm’s workers? This is clearly
an economic cost of doing business, but if you think about it, you will see that
it is also an opportunity cost. The reason is that the money paid to the workers
could have been put to some alternative use instead. Perhaps the firm could
have used some or all of that money to buy more labor-saving machines, or
even to produce a different product altogether. Thus we see that economic cost
and opportunity cost actually boil down to the same thing. As long as we ac-
count for and measure all of the firm’s resources properly, we will find that:
Economic cost = Opportunity cost
accounting cost Actual
expenses plus depreciation
charges for capital equipment.
economic cost Cost to a firm
of utilizing economic resources in
production.
opportunity cost Cost
associated with opportunities
forgone when a firm’s resources
are not put to their best
alternative use.
ChAPTER 7 ThE CoST oF PRoDUCTIoN 239
While economic cost and opportunity cost both describe the same thing, the
concept of opportunity cost is particularly useful in situations where alternatives
that are forgone do not reflect monetary outlays. Let’s take a more detailed look
at opportunity cost to see how it can make economic cost differ from accounting
cost in the treatment of wages, and then in the cost of production inputs. Consider
an owner that manages her own retail toy store and does not pay herself a salary.
(We’ll put aside the rent that she pays for the office space just to simplify the discus-
sion.) Had our toy store owner chosen to work elsewhere she would have been able
to find a job that paid $60,000 per year for essentially the same effort. In this case the
opportunity cost of the time she spends working in her toy store business is $60,000.
Now suppose that last year she acquired an inventory of toys for which she
paid $1 million. She hopes to be able to sell those toys during the holiday season
for a substantial markup over her acquisition cost. However, early in the fall she
eceives an offer from another toy retailer to acquire her inventory for $1.5 mil-
lion. Should she sell her inventory or not? The answer depends in part on her
usiness prospects, but it also depends on the opportunity cost of acquiring a
toy inventory. Assuming that it would cost $1.5 million to acquire the new in-
ventory all over again, the opportunity cost of keeping it is $1.5 million, not the
$1.0 million she originally paid.
You might ask why the opportunity cost isn’t just $500,000, since that is the
difference between the market value of the inventory and the cost of its acquisi-
tion. The key is that when the owner is deciding what to do with the inventory,
she is deciding what is best for her business in the future. To do so, she needs to
account for the fact that if she keeps the inventory for her own use, she would
e sacrificing the $1.5 million that she could have received by selling the inven-
tory to another firm.1
Note that an accountant may not see things this way. The accountant might tell
the toy store owner that the cost of utilizing the inventory is just the $1 million that
she paid for it. But we hope that you understand why this would be misleading.
The actual economic cost of keeping and utilizing that inventory is the $1.5 million
that the owner could have obtained by instead selling it to another retailer.
Accountants and economists will also sometimes differ in their treatment of
depreciation. When estimating the future profitability of a business, economists
and managers are concerned with the capital cost of plant and machinery. This
cost involves not only the monetary outlay for buying and then running the ma-
chinery, but also the cost associated with wear and tear. When evaluating past
performance, cost accountants use tax rules that apply to
oadly defined types
of assets to determine allowable depreciation in their cost and profit calcula-
tions. But these depreciation allowances need not reflect the actual wear and
tear on the equipment, which is likely to vary asset by asset.
Sunk Costs
Although an opportunity cost is often hidden, it should be taken into account
when making economic decisions. Just the opposite is true of a sunk cost: an
expenditure that has been made and cannot be recovered. A sunk cost is usually
visible, but after it has been incu
ed it should always be ignored when making
future economic decisions.
sunk cost Expenditure that
has been made and cannot be
ecovered.
1Of course, opportunity cost will change from circumstance to circumstance and from one time
period to the next. If the value of our retailer’s inventory suddenly increased to $1.7 million because
that inventory included some holiday products that were in great demand, the opportunity cost of
keeping and using the inventory would increase to $1.7 million.
240 PART 2 Producers, Consumers, and Competitive markets
Because a sunk cost cannot be recovered, it should not influence the firm’s
decisions. For example, consider the purchase of specialized equipment for a
plant. Suppose the equipment can be used to do only what it was originally
designed for and cannot be converted for alternative use. The expenditure on
this equipment is a sunk cost. Because it has no alternative use, its opportunity cost
is zero. Thus it should not be included as part of the firm’s economic costs. The
decision to buy this equipment may have been good or bad. It doesn’t matter.
It’s water under the
idge and shouldn’t affect cu
ent decisions.
What if, instead, the equipment could be put to other use or could be sold
or rented to another firm? In that case, its use would involve an economic
cost—namely, the opportunity cost of using it rather than selling or renting it to
another firm.
Now consider a prospective sunk cost. Suppose, for example, that the firm has
not yet bought the specialized equipment but is merely considering whether
to do so. A prospective sunk cost is an investment. Here the firm must decide
whether that investment in specialized equipment is economical—i.e., whether it
will lead to a flow of revenues large enough to justify its cost. In Chapter 15, we
explain in detail how to make investment decisions of this kind.
As an example, suppose a firm is considering moving its headquarters to a
new city. Last year it paid $500,000 for an option to buy a building in the city.
The option gives the firm the right to buy the building at a cost of $5,000,000, so
that if it ultimately makes the purchase its total expenditure will be $5,500,000.
Now it finds that a comparable building has become available in the same city
at a price of $5,250,000. Which building should it buy? The answer is the origi-
nal building. The $500,000 option is a cost that has been sunk and thus should
not affect the firm’s cu
ent decision. What’s at issue is spending an additional
$5,000,000 or an additional $5,250,000. Because the economic analysis removes
the sunk cost of the option from the analysis, the economic cost of the original
property is $5,000,000. The newer property, meanwhile, has an economic cost of
$5,250,000. Of course, if the new building costs $4,900,000, the firm should buy
it and forgo its option.
ExAMplE 7.1 Choosing the LoCation for a new Law sChooL BuiLding
The Northwestern University Law School has long
een located in Chicago, along the shores of Lake
michigan. however, the main campus of the university
is located in the subu
of Evanston. In the mid-1970s,
the law school began planning the construction of a
new building and needed to decide on an appropriate
location. Should it be built on the cu
ent site, where
it would remain near downtown Chicago law firms?
or should it be moved to Evanston, where it would
e physically integrated with the rest of the university?
The downtown location had many prominent sup-
porters. They argued in part that it was cost-effective
to locate the new building in the city because the
university already owned the land. A large parcel of
land would have to be purchased in Evanston if the
uilding were to be built there. Does this argument
make economic sense?
No. It makes the common mistake of failing to ap-
preciate opportunity cost. From an economic point
of view, it is very expensive to locate downtown be-
cause the opportunity cost of the valuable lakeshore
location is high: That property could have been sold
for enough money to buy the Evanston land with sub-
stantial funds left over.
In the end, Northwestern decided to keep the law
school in Chicago. This was a costly decision. It may
have been appropriate if the Chicago location was
particularly valuable to the law school, but it was
inappropriate if it was made on the presumption that
the downtown land had no cost.
ChAPTER 7 ThE CoST oF PRoDUCTIoN 241
Fixed Costs and Variable Costs
Some costs vary with output, while others remain unchanged as long as the
firm is producing any output at all. This distinction will be important when we
examine the firm’s profit-maximizing choice of output in the next chapter. We
therefore divide total cost (TC or C)—the total economic cost of production—
into two components.
•     Fixed cost (FC): A cost that does not vary with the level of output and that
can be eliminated only by going out of business.
•     Variable cost (VC): A cost that varies as output varies.
Depending on circumstances, fixed costs may include expenditures for plant
maintenance, insurance, heat and electricity, and perhaps a minimal number of
employees. They remain the same no matter how much output the firm pro-
duces. Variable costs, which include expenditures for wages, salaries, and raw
materials used for production, increase as output increases.
Fixed cost does not vary with the level of output—it must be paid even if
there is no output. The only way that a firm can eliminate its fixed costs is by shutting
down.
shutting down Shutting down doesn’t necessarily mean going out of
usiness. Suppose a clothing company owns several factories, is experiencing
declining demand, and wants to reduce output and costs as much as possible at
one factory. By reducing the output of that factory to zero, the company could
eliminate the costs of raw materials and much of the labor, but it would still
incur the fixed costs of paying the factory’s managers, security guards, and
ongoing maintenance. The only way to eliminate those fixed costs would be
to close the doors, turn off the electricity, and perhaps even sell off or scrap the
machinery. The company would still remain in business and could operate its
emaining factories. It might even be able to re-open the factory it had closed,
although doing so could be costly if it involved buying new machinery or refur-
ishing the old machinery.
fixed or VariaBLe? How do we know which costs are fixed and which are
variable? The answer depends on the time horizon that we are considering.
Over a very short time horizon—say, a few months—most costs are fixed. Over
such a short period, a firm is usually obligated to pay for contracted shipments
of materials and cannot easily lay off workers, no matter how much or how
little the firm produces.
On the other hand, over a longer time period—say, two or three years—
many costs become variable. Over this time horizon, if the firm wants to
educe its output, it can reduce its workforce, purchase fewer raw materials,
and perhaps even sell off some of its machinery. Over a very long time hori-
zon—say, ten years—nearly all costs are variable. Workers and managers can
e laid off (or employment can be reduced by attrition), and much of the ma-
chinery can be sold off or not replaced as it becomes obsolete and is scrapped.
Knowing which costs are fixed and which are variable is important for the
management of a firm. When a firm plans to increase or decrease its production,
it will want to know how that change will affect its costs. Consider, for example,
a problem that Delta Air Lines faced. Delta wanted to know how its costs would
change if it reduced the number of its scheduled flights by 10 percent. The
answer depends on whether we are considering the short run or the long run.
Over the short run—say six months—schedules are fixed and it is difficult to lay
total cost (TC or C) Total
economic cost of production,
consisting of fixed and variable
costs.
fixed cost (FC) Cost that does
not vary with the level of output
and that can be eliminated only
y shutting down.
variable cost (VC) Cost that
varies as output varies.
242 PART 2 Producers, Consumers, and Competitive markets
off or discharge workers. As a result, most of Delta’s short-run costs are fixed
and won’t be reduced significantly with the flight reduction. In the long run—
say two years or more—the situation is quite different. Delta has sufficient time
to sell or lease planes that are not needed and to discharge unneeded workers.
In this case, most of Delta’s costs are variable and thus can be reduced signifi-
cantly if a 10-percent flight reduction is put in place.
Fixed versus Sunk Costs
People often confuse fixed and sunk costs. As we just explained, fixed costs are
costs that are paid by a firm that is operating, regardless of the level of output
it produces. Such costs can include, for example, the salaries of the key execu-
tives and expenses for their office space and support staff, as well as insurance
and the costs of plant maintenance. Fixed costs can be avoided if the firm shuts
down a plant or goes out of business—the key executives and their support
staff, for example, will no longer be needed.
Sunk costs, on the other hand, are costs that have been incu
ed and can-
not be recovered. An example is the cost of R&D to a pharmaceutical company
to develop and test a new drug and then, if the drug has been proven to be
safe and effective, the cost of marketing it. Whether the drug is a success or a
failure, these costs cannot be recovered and thus are sunk. Another example
is the cost of a chip-fa
ication plant to produce microprocessors for use in
computers. Because the plant’s equipment is too specialized to be of use in
any other industry, most if not all of this expenditure is sunk, i.e., cannot be
ecovered. (Some small part of the cost might be recovered if the equipment
is sold for scrap.)
Suppose, on the other hand, that a firm had agreed to make annual pay-
ments into an employee retirement plan as long as the firm was in operation,
egardless of its output or its profitability. These payments could cease only if
the firm went out of business. In this case, the payments should be viewed as
a fixed cost.
Why distinguish between fixed and sunk costs? Because fixed costs affect
the firm’s decisions looking forward, whereas sunk costs do not. Fixed costs
that are high relative to revenue and cannot be reduced might lead a firm to
shut down—eliminating those fixed costs and earning zero profit might be
etter than incu
ing ongoing losses. Incu
ing a high sunk cost might later
turn out to be a bad decision (for example, the unsuccessful development
of a new product), but the expenditure is gone and cannot be recovered
y shutting down. Of course a prospective sunk cost is different and, as we
mentioned earlier, would certainly affect the firm’s decisions looking for-
ward. (Should the firm, for example, undertake the development of that new
product?)
amortizing sunk Costs In practice, many firms don’t always distin-
guish between sunk and fixed costs. For example, the semiconductor com-
pany that spent $600 million for a chip-fa
ication plant (clearly a sunk cost)
might amortize the expenditure over six years and treat it as a fixed cost of
$100 million per year. This is fine as long as the firm’s managers understand
that shutting down will not make the $100 million annual cost go away. In
fact, amortizing capital expenditures this way—spreading them out over
many years and treating them as fixed costs—can be a useful way of evaluat-
ing the firm’s long-term profitability.
amortization Policy of
treating a one-time expenditure
as an annual cost spread out over
some number of years.
ChAPTER 7 ThE CoST oF PRoDUCTIoN 243
Amortizing large capital expenditures and treating them as ongoing fixed
costs can also simplify the economic analysis of a firm’s operation. As we will
see, for example, treating capital expenditures this way can make it easier to
understand the tradeoff that a firm faces in its use of labor versus capital. For
simplicity, we will usually treat sunk costs in this way as we examine the firm’s
production decisions. When distinguishing sunk from fixed costs does become
essential to the economic analysis, we will let you know.
ExAMplE 7.2 sunk, fixed, and VariaBLe Costs: Computers,
software, and pizzas
As you progress through this book, you will see that
a firm’s pricing and production decisions—and its
profitability—depend strongly on the structure of its
costs. It is therefore important for managers to under-
stand the characteristics of production costs and to
e able to identify which costs are fixed, which are
variable, and which are sunk. The relative sizes of
these different cost components can vary consider-
ably across industries. Good examples include the
personal computer industry (where most costs are
variable), the computer software industry (where
most costs are sunk), and the pizzeria business
(where most costs are fixed). Let’s look at each of
these in turn.
Computers: Companies like Dell, Lenovo, and
hewlett-Packard produce millions of personal com-
puters every year. Because computers are very simi-
lar, competition is intense, and profitability depends
critically on the ability to keep costs down. most of
these costs are variable—they increase in proportion
to the number of computers produced each year.
most important is the cost of components: the micro-
processor that does much of the actual computation,
memory chips, hard disk drives and other storage
devices, video and sound cards, etc. Typically, the
majority of these components are purchased from
outside suppliers in quantities that depend on the
number of computers to be produced.
Another important variable cost is labor: Workers
are needed to assemble computers and then package
and ship them. There is little in the way of sunk costs
ecause factories cost little relative to the value of
the company’s annual output. Likewise, there is little
in the way of fixed costs—perhaps the salaries of the
top executives, some security guards, and electricity.
Thus, when Dell and hewlett-Packard think about
ways of reducing cost, they focus largely on get-
ting better prices for components or reducing labor
equirements—both of which are ways of reducing
variable cost.
Software: What about the software programs that
un on these personal computers? microsoft pro-
duces the Windows operating system as well as a
variety of applications such as Word, Excel, and
PowerPoint. But many other firms—some large and
some small—also produce software programs that
un on personal computers. For such firms, produc-
tion costs are quite different from those facing hard-
ware manufacturers. In software production, most
costs are sunk. Typically, a software firm will spend a
large amount of money to develop a new application
program. These expenditures cannot be recovered.
once the program is completed, the company
can try to recoup its investment (and make a profit
as well) by selling as many copies of the program as
possible. The variable cost of producing copies of the
program is very small—largely the cost of copying
the program to CDs and then packaging and shipping
the product. Likewise, the fixed cost of production is
small. Because most costs are sunk, entering the soft-
ware business can involve considerable risk. Until
the development money has been spent and the
product has been released for sale, an entrepreneur is
unlikely to know how many copies can be sold and
whether or not he will be able to make money.
pizzas: Now let’s turn to your neighborhood
pizzeria. For the pizzeria, the largest component of
cost is fixed. Sunk costs are fairly low because pizza
ovens, chairs, tables, and dishes can be resold if the
pizzeria goes out of business. Variable costs are also
fairly low—mainly the ingredients for pizza (flour, to-
mato sauce, cheese, and pepperoni for a typical large
pizza might cost $1 or $2) and perhaps wages for a
couple of workers to help produce, serve, and deliver
pizzas. most of the cost is fixed—the opportunity
cost of the owner’s time (he might typically work a
244 PART 2 Producers, Consumers, and Competitive markets
Marginal and Average Cost
To complete our discussion of costs, we now turn to the distinction between
marginal and average cost. In explaining this distinction, we use a specific nu-
merical example of a cost function (the relationship between cost and output)
that typifies the cost situation of many firms. The example is shown in Table 7.1.
After we explain the concepts of marginal and average cost, we will consider
how the analysis of costs differs between the short run and the long run.
marginaL Cost (mC) Marginal cost—sometimes called incremental cost—is
the increase in cost that results from producing one extra unit of output. Because
fixed cost does not change as the firm’s level of output changes, marginal cost
marginal cost (MC) Increase
in cost resulting from the
production of one extra unit of
output.
Table 7.1 a firm’s Costs
Rate of
output
(units peR
YeaR)
fixed
Cost
(dollaRs
peR YeaR)
VaRiable
Cost
(dollaRs
peR YeaR)
total
Cost
(dollaRs
peR YeaR)
MaRginal
Cost
(dollaRs
peR unit)
aVeRage
fixed Cost
(dollaRs
peR unit)
aVeRage
VaRiable Cost
(dollaRs peR
unit)
aVeRage
total Cost
(dollaRs
peR unit)
(fC) (1) (VC) (2) (tC) (3) (MC) (4) (afC) (5) (aVC) (6) (atC) (7)
0 50 0 50 — — — —
1 50 50 100 50 50 50 100
2 50 78 128 28 25 39 64
3 50 98 148 20 16.7 32.7 49.3
4 50 112 162 14 12.5 28 40.5
5 50 130 180 18 10 26 36
6 50 150 200 20 8.3 25 33.3
7 50 175 225 25 7.1 25 32.1
8 50 204 254 29 6.3 25.5 31.8
9 50 242 292 38 5.6 26.9 32.4
10 50 300 350 58 5 30 35
11 50 385 435 85 4.5 35 39.5
60- or 70-hour week), rent, and utilities. Because of
these high fixed costs, most pizzerias (which might
charge $12 for a large pizza costing about $3 in vari-
able cost to produce) don’t make very high profits.
This textbook: Finally, let’s consider the cost of
this wonderful textbook that you’ve enjoyed read-
ing so much. What do you think is the variable
cost of production, i.e., the cost of producing one
additional book? Ignoring royalties to the authors
(which are a percentage of the wholesale selling
price) and assuming you have the hardbound ver-
sion of the book, the cost is only about $5 to $10.
most of the costs are sunk: the opportunity cost of
the authors’ time spent writing (and revising) the
ook and the costs of the publisher for copyedit-
ing, typesetting, and proofing. As with computer
software, textbook production costs need have
little connection to the price you paid for the
ook. You’ll learn more about that when you get to
Chapter 12 and read Example 12.5.
ChAPTER 7 ThE CoST oF PRoDUCTIoN 245
is equal to the increase in variable cost or the increase in total cost that results
from an extra unit of output. We can therefore write marginal cost as
MC = ∆VC>∆q = ∆TC>∆q
Marginal cost tells us how much it will cost to expand output by one unit. In
Table 7.1, marginal cost is calculated from either the variable cost (column 2) or
the total cost (column 3). For example, the marginal cost of increasing output
from 2 to 3 units is $20 because the variable cost of the firm increases from $78
to $98. (The total cost of production also increases by $20, from $128 to $148.
Total cost differs from variable cost only by the fixed cost, which by definition
does not change as output changes.)
aVerage totaL Cost (atC) Average total cost, used interchangeably
with AC and average economic cost, is the firm’s total cost divided by its level of
output, TC/q. Thus the average total cost of producing at a rate of five units
is $36—that is, $180/5. Basically, average total cost tells us the per-unit cost of
production.
ATC has two components. Average fixed cost (AFC) is the fixed cost (column
1 of Table 7.1) divided by the level of output, FC/q. For example, the average
fixed cost of producing 4 units of output is $12.50 ($50/4). Because fixed cost
is constant, average fixed cost declines as the rate of output increases. Average
variable cost (AVC) is variable cost divided by the level of output, VC/q. The
average variable cost of producing 5 units of output is $26—that is, $130/5.
We have now discussed all of the different types of costs that are relevant to
production decisions in both competitive and non-competitive markets. Now
we turn to how costs differ in the short run versus the long run. This is particu-
larly important for fixed costs. Costs that are fixed in the very short run, e.g.,
the wages of employees under fixed-term contracts—may not be fixed over a
longer time horizon. Similarly, the fixed capital costs of plant and equipment
ecome variable if the time horizon is sufficiently long to allow the firm to pur-
chase new equipment and build a new plant. Fixed costs, however, need not
disappear, even in the long run. Suppose, for example, that a firm has been con-
tributing to an employee retirement program. Its obligations, which are fixed
in part, may remain even in the long run; they might only disappear if the firm
were to declare bankruptcy.
7.2 Cost in the Short Run
In this section we focus our attention on short-run costs. We turn to long-run
costs in Section 7.3.
The Determinants of Short-Run Cost
The data in Table 7.1 show how variable and total costs increase with output
in the short run. The rate at which these costs increase depends on the nature
of the production process and, in particular, on the extent to which production
involves diminishing marginal returns to variable factors. Recall from Chapter
6 that diminishing marginal returns to labor occur when the marginal product
of labor is decreasing. If labor is the only input, what happens as we increase
the firm’s output? To produce more output, the firm must hire more labor.
Then, if the marginal product of labor decreases as the amount of labor hired
average total cost
(ATC) Firm’s total cost divided
y its level of output.
average fixed cost
(AFC) Fixed cost divided
y the level of output.
average variable cost
(AVC) Variable cost divided by
the level of output.
in §6.2, we explain that
diminishing marginal returns
occurs when additional
inputs result in decreasing
additions to output.
246 PART 2 Producers, Consumers, and Competitive markets
is increased (owing to diminishing returns), successively greater expenditures
must be made to produce output at the higher rate. As a result, variable and
total costs increase as the rate of output is increased. On the other hand, if the
marginal product of labor decreases only slightly as the amount of labor is in-
creased, costs will not rise so quickly when the rate of output is increased.2
Let’s look at the relationship between production and cost in more detail by
concentrating on the costs of a firm that can hire as much labor as it wishes at
a fixed wage w. Recall that marginal cost MC is the change in variable cost for
a 1-unit change in output (i.e., ∆VC>∆q). But the change in variable cost is the
per-unit cost of the extra labor w times the amount of extra labor needed to pro-
duce the extra output ∆L. Because ∆VC = w∆L, it follows that
MC = ∆VC>∆q = w∆L>∆q
Recall from Chapter 6 that the marginal product of labor MPL is the change in
output resulting from a 1-unit change in labor input, or ∆q>∆L. Therefore, the ex-
tra labor needed to obtain an extra unit of output is ∆L>∆q = 1>MPL. As a result,
MC = w>MPL (7.1)
Equation (7.1) states that when there is only one variable input, the marginal
cost is equal to the price of the input divided by its marginal product. Suppose,
for example, that the marginal product of labor is 3 and the wage rate is $30 per
hour. In that case, 1 hour of labor will increase output by 3 units, so that 1 unit
of output will require 1/3 additional hour of labor and will cost $10. The mar-
ginal cost of producing that unit of output is $10, which is equal to the wage,
$30, divided by the marginal product of labor, 3. A low marginal product of
labor means that a large amount of additional labor is needed to produce more
output—a fact that leads, in turn, to a high marginal cost. Conversely, a high
marginal product means that the labor requirement is low, as is the marginal
cost. More generally, whenever the marginal product of labor decreases, the
marginal cost of production increases, and vice versa.3
diminishing marginaL returns and marginaL Cost Diminishing
marginal returns means that the marginal product of labor declines as the
quantity of labor employed increases. As a result, when there are diminishing
marginal returns, marginal cost will increase as output increases. This can be
seen by looking at the numbers for marginal cost in Table 7.1. For output levels
from 0 through 4, marginal cost is declining; for output levels from 4 through
11, however, marginal cost is increasing—a reflection of the presence of dimin-
ishing marginal returns.
The Shapes of the Cost Curves
Figure 7.1 illustrates how various cost measures change as output changes. The
top part of the figure shows total cost and its two components, variable cost and
fixed cost; the bottom part shows marginal cost and average costs. These cost
curves, which are based on the information in Table 7.1, provide different kinds
of information.
the marginal product of
labor is discussed in §6.2.
2We are implicitly assuming that because labor is hired in competitive markets, the payment per
unit of labor used is the same regardless of the firm’s output.
3With two or more variable inputs, the relationship is more complex. The basic principle, however,
still holds: The greater the productivity of factors, the less the variable cost that the firm must incur
to produce any given level of output.
ChAPTER 7 ThE CoST oF PRoDUCTIoN 247
Observe in Figure 7.1 (a) that fixed cost FC does not vary with output—it is
shown as a horizontal line at $50. Variable cost VC is zero when output is zero and
then increases continuously as output increases. The total cost curve TC is deter-
mined by vertically adding the fixed cost curve to the variable cost curve. Because
fixed cost is constant, the vertical distance between the two curves is always $50.
Figure 7.1 (b) shows the co
esponding set of marginal and average variable cost
curves.4 Because total fixed cost is $50, the average fixed cost curve AFC falls con-
tinuously from $50 when output is 1, toward zero for large output. The shapes of
the remaining curves are determined by the relationship between the marginal and
average cost curves. Whenever marginal cost lies below average cost, the average
cost curve falls. Whenever marginal cost lies above average cost, the average cost
curve rises. When average cost is at a minimum, marginal cost equals average cost.
the aVerage-marginaL reLationship Marginal and average costs are
another example of the average-marginal relationship described in Chapter 6
1
Cost
(dollars
pe
year)
FC
Output (units per year)
2 3 4 5 6 7 8 9 10 11
VC
TC
A
1
Cost
(dollars
pe
unit)
Output (units per year)
100
75
50
25
0
2 3 4 5 6 7 8 9 10 11
MC
AVC
ATC
AFC
(a)
(b)
400
300
100
0
175
FiguRE 7.1
Cost CurVes
for a firm
In (a) total cost TC is the vertical
sum of fixed cost FC and vari-
able cost VC. In (b) average total
cost ATC is the sum of average
variable cost AVC and average
fixed cost AFC. Marginal cost
MC crosses the average vari-
able cost and average total cost
curves at their minimum points.
4The curves do not exactly match the numbers in Table 7.1. Because marginal cost represents the
change in cost associated with a change in output, we have plotted the MC curve for the first unit
of output by setting output equal to 12, for the second unit by setting output equal to 1
1
2, and so on.
248 PART 2 Producers, Consumers, and Competitive markets
(with respect to marginal and average product). At an output of 5 in Table 7.1, for
example, the marginal cost of $18 is below the average variable cost of $26; thus
the average is lowered in response to increases in output. But when marginal cost
is $29, which is greater than average variable cost ($25.5), the average increases
as output increases. Finally, when marginal cost ($25) and average variable cost
($25) are nearly the same, average variable cost increases only slightly.
The ATC curve shows the average total cost of production. Because average
total cost is the sum of average variable cost and average fixed cost and the AFC
curve declines everywhere, the vertical distance between the ATC and AVC curves
decreases as output increases. The AVC cost curve reaches its minimum point at a
lower output than the ATC curve. This follows because MC = AVC at its mini-
mum point and MC = ATC at its minimum point. Because ATC is always greater
than AVC and the marginal cost curve MC is rising, the minimum point of the
ATC curve must lie above and to the right of the minimum point of the AVC curve.
Another way to see the relationship between the total cost curves and the
average and marginal cost curves is to consider the line drawn from origin to
point A in Figure 7.1 (a). In that figure, the slope of the line measures average
variable cost (a total cost of $175 divided by an output of 7, or a cost per unit
of $25). Because the slope of the VC curve is the marginal cost (it measures the
change in variable cost as output increases by 1 unit), the tangent to the VC
curve at A is the marginal cost of production when output is 7. At A, this mar-
ginal cost of $25 is equal to the average variable cost of $25 because average
variable cost is minimized at this output.
totaL Cost as a fLow Note that the firm’s output is measured as a flow: The
firm produces a certain number of units per year. Thus its total cost is a flow—for
example, some number of dollars per year. (Average and marginal costs, however,
are measured in dollars per unit.) For simplicity, we will often drop the time refer-
ence, and refer to total cost in dollars and output in units. But you should remem-
er that a firm’s production of output and expenditure of cost occur over some
time period. In addition, we will often use cost (C) to refer to total cost. Likewise,
unless noted otherwise, we will use average cost (AC) to refer to average total cost.
Marginal and average cost are very important concepts. As we will see in
Chapter 8, they enter critically into the firm’s choice of output level. Knowledge
of short-run costs is particularly important for firms that operate in an environ-
ment in which demand conditions fluctuate considerably. If the firm is cu
ently
producing at a level of output at which marginal cost is sharply increasing, and
if demand may increase in the future, management might want to expand pro-
duction capacity to avoid higher costs.
ExAMplE 7.3 the short-run Cost
of aLuminum smeLting
Aluminum is a lightweight, versatile metal used in a wide variety of applica-
tions, including airplanes, automobiles, packaging, and building materials. The
production of aluminum begins with the mining of bauxite in such countries
as Australia, Brazil, Guinea, Jamaica, and Suriname. Bauxite is an ore that
contains a relatively high concentration of alumina (aluminum oxide), which
is separated from the bauxite through a chemical refining process. The alumina
ChAPTER 7 ThE CoST oF PRoDUCTIoN 249
5This example is based on Kenneth S. Corts, “The Aluminum Industry in 1994,” Harvard Business
School Case N9-799-129, April 1999.
is then converted to aluminum through a smelting process in which an electric
cu
ent is used to separate the oxygen atoms from the aluminum oxide mol-
ecules. It is this smelting process—which is the most costly step in producing
aluminum—that we focus on here.
All of the major aluminum producers, including UC RUSAL, Alcoa, Alcan,
Chalco, and hydro Aluminum, operate smelting plants. A typical smelting
plant will have two production lines, each of which produces approximately
300 to 400 tons of aluminum per day. We will examine the short-run cost of
production. Thus we consider the cost of operating an existing plant because
there is insufficient time in the short run to build additional plants. (It takes
about four years to plan, build, and fully equip an aluminum smelting plant.)
Although the cost of a smelting plant is substantial (over $1 billion), we will
assume that the plant cannot be sold; the expenditure is therefore sunk and can
e ignored. Furthermore, because fixed costs, which are largely for adminis-
trative expenses, are relatively small, we will ignore them also. Thus we can
focus entirely on short-run variable costs. Table 7.2 shows the average (per-ton)
production costs for a typical aluminum smelter.5 The cost numbers apply to a
plant that runs two shifts per day to produce 600 tons of aluminum per day. If
prices were sufficiently high, the firm could choose to operate the plant on a
three-shifts-per-day basis by asking workers to work overtime. however, wage
and maintenance costs would likely increase about 50 percent for this third
shift because of the need to pay higher overtime wages. We have divided the
cost components in Table 7.2 into two groups. The first group includes those
costs that would remain the same at any output level; the second includes costs
that would increase if output exceeded 600 tons per day.
Table 7.2 produCtion Costs for aLuminum smeLting
($ / ton) (Based on an output of 600 tons / day)
peR-ton Costs that aRe Constant
foR all output leVels
output … 600
tons/daY
output 7 600
tons/daY
electricity $316 $316
alumina 369 369
other raw materials 125 125
plant power and fuel 10 10
subtotal $820 $820
peR-ton Costs that inCRease when
output exCeeds 600 tons/daY
labor $150 $225
Maintenance 120 180
freight 50 75
subtotal $320 $480
Total Per-ton Production Costs $1140 $1300
250 PART 2 Producers, Consumers, and Competitive markets
Note that the largest cost components for an aluminum smelter are elec-
tricity and the cost of alumina; together, they represent about 60 percent of
total production costs. Because electricity, alumina, and other raw materials
are used in direct proportion to the amount of aluminum produced, they
epresent per-ton production costs that are constant with respect to the level
of output. The costs of labor, maintenance, and freight are also proportional
to the level of output, but only when the plant operates two shifts per day. To
increase output above 600 tons per day, a third shift would be necessary and
would result in a 50-percent increase in the per-ton costs of labor, mainte-
nance, and freight.
The short-run marginal cost and average variable cost curves for the smelt-
ing plant are shown in Figure 7.2. For an output q up to 600 tons per day, total
variable cost is $1140q, so marginal cost and average variable cost are constant
at $1140 per ton. If we increase production beyond 600 tons per day by means
of a third shift, the marginal cost of labor, maintenance, and freight increases
from $320 per ton to $480 per ton, which causes marginal cost as a whole to
increase from $1140 per ton to $1300 per ton.
What happens to average variable cost when output is greater than 600 tons
per day? When q 7 600, total variable cost is given by:
TVC = (1140)(600) + 1300(q - 600) = 1300q - 96,000
Therefore average variable cost is
AVC = 1300 -
96,000
q

As Figure 7.2 shows, when output reaches 900 tons per day, an absolute
capacity constraint is reached, at which point the marginal and average costs of
production become infinite.
300 600 900
Output (tons per day)
1300
1200
1100
1140
Cost
(dollars per ton)
MC
AVC
FiguRE 7.2
the short-run VariaBLe
Costs of aLuminum
smeLting
The short-run average variable cost of smelt-
ing is constant for output levels using up to
two labor shifts. When a third shift is added,
marginal cost and average variable cost in-
crease until maximum capacity is reached.
ChAPTER 7 ThE CoST oF PRoDUCTIoN 251
7.3 Cost in the long Run
In the long run, a firm has much more flexibility. It can expand its capacity by
expanding existing factories or building new ones; it can expand or contract its la-
or force, and in some cases, it can change the design of its products or introduce
new products. In this section, we show how a firm can choose its combination of
inputs to minimize its cost of producing a given output. We will also examine the
elationship between long-run cost and the level of output. We begin by taking a
careful look at the cost of using capital equipment. We then show how this cost,
along with the cost of labor, enters into the production decision.
The user Cost of Capital
Firms often rent or lease equipment, buildings, and other capital used in the
production process. On other occasions, the capital is purchased. In our analy-
sis, however, it will be useful to treat capital as though it were rented even if it
was purchased. An illustration will help to explain how and why we do this.
Let’s suppose that Delta Airlines is thinking about purchasing a new Boeing 777
airplane for $150 million. Even though Delta would pay a large sum for the air-
plane now, for economic purposes the purchase price can be allocated or amor-
tized across the life of the airplane. This will allow Delta to compare its revenues
and costs on an annual flow basis. We will assume that the life of the airplane is
30 years; the amortized cost is therefore $5 million per year. The $5 million can
e viewed as the annual economic depreciation for the airplane.
So far, we have ignored the fact that had the firm not purchased the airplane,
it could have earned interest on its $150 million. This forgone interest is an
opportunity cost that must be accounted for. Therefore, the user cost of capital—
the annual cost of owning and using the airplane instead of selling it or never
uying it in the first place—is given by the sum of the economic depreciation and
the interest (i.e., the financial return) that could have been earned had the money been
invested elsewhere.6 Formally,
User Cost of Capital = Economic Depreciation + (Interest Rate)
(Value of Capital)
In our example, economic depreciation on the airplane is $5 million per year.
Suppose Delta could have earned a return of 10 percent had it invested its
money elsewhere. In that case, the user cost of capital is $5 million + (.10)
($150 million - depreciation). As the plane depreciates over time, its value
declines, as does the opportunity cost of the financial capital that is invested
in it. For example, at the time of purchase, looking forward for the first
year, the user cost of capital is $5 million + (.10)($150 million) = $20 million.
In the tenth year of ownership, the airplane, which will have depreciated by
$50 million, will be worth $100 million. At that point, the user cost of capital
will be $5 million + (.10)($100 million) = $15 million per year.
We can also express the user cost of capital as a rate per dollar of capital:
= Depreciation rate + Interest rate
user cost of capital Annual
cost of owning and using a
capital asset, equal to economic
depreciation plus forgone interest.
6More precisely, the financial return should reflect an investment with similar risk. The interest rate,
therefore, should include a risk premium. We discuss this point in Chapter 15. Note also that the
user cost of capital is not adjusted for taxes; when taxes are taken into account, revenues and costs
should be measured on an after-tax basis.
252 PART 2 Producers, Consumers, and Competitive markets
For our airplane example, the depreciation rate is 1>30 = 3.33 percent per year.
If Delta could have earned a rate of return of 10 percent per year, its user cost of
capital would be r = 3.33 + 10 = 13.33 percent per year.
As we’ve already pointed out, in the long run the firm can change all of its
inputs. We will now show how the firm chooses the combination of inputs
that minimizes the cost of producing a certain output, given information about
wages and the user cost of capital. We will then examine the relationship be-
tween long-run cost and the level of output.
The Cost-Minimizing input Choice
We now turn to a fundamental problem that all firms face: how to select inputs
to produce a given output at minimum cost. For simplicity, we will work with two
variable inputs: labor (measured in hours of work per year) and capital (mea-
sured in hours of use of machinery per year).
The amount of labor and capital that the firm uses will depend, of course,
on the prices of these inputs. We will assume that because there are com-
petitive markets for both inputs, their prices are unaffected by what the firm
does. (In Chapter 14 we will examine labor markets that are not competitive.)
In this case, the price of labor is simply the wage rate, w. But what about the
price of capital?
the priCe of CapitaL In the long run, the firm can adjust the amount of
capital it uses....