XXXXXXXXXX
Human resource management
practices and workers’ jo
satisfaction
Alina Ileana Petrescu
Lancashire Business School, University of Central Lancashire, Preston, UK, and
Rob Simmons
Department of Economics, The Management School, Lancaster University,
Lancaster, UK
Abstract
Purpose – The purpose of this paper is to investigate the relationship between human resource
management (HRM) practices and workers’ overall job satisfaction and their satisfaction with pay.
Design/methodology/approach – The paper uses British data from two different cross-sectional
datasets. It estimates probit models with overall job satisfaction and satisfaction with pay as
subjective dependent variables.
Findings – After controlling for personal, job and firm characteristics, it is found that several HRM
practices raise workers’ overall job satisfaction and their satisfaction with pay. However, these effects
are only significant for non-union members. Satisfaction with pay is higher where performance-related
pay and seniority-based reward systems are in place. A pay structure that is perceived to be unequal is
associated with a substantial reduction in both non-union members’ overall job satisfaction and thei
satisfaction with pay. Although HRM practices can raise workers’ job satisfaction, if workplace pay
inequality widens as a consequence then non-union members may experience reduced job satisfaction.
Research limitations/implications – The data sets used in the analysis are cross-sectional,
presenting a snapshot of impacts of HRM practices on job satisfaction at a particular point in time.
Dynamic effects are therefore not captured.
Originality/value – The paper adds to the empirical literature on effects of HRM practices,
focussing on impacts on both overall job satisfaction and satisfaction with pay. A novel feature of the
paper is the use of two separate data sets to develop complementary empirical results.
Keywords Job satisfaction, Human resource management, Human resource strategies, Trade unions,
Pay policies
Paper type Research pape
1. Introduction
The past two decades have witnessed a burgeoning literature on the economics of jo
satisfaction[1]. There is also a large human resource management (HRM) literature that
emphasises the influence of so-called “high-performance workplace practices” on jo
satisfaction and hence employee performance. The HRM literature can be sub-divided
into empirical studies, which in the case of the UK are primarily based on the
Workplace Industrial Relations Survey (WIRS) series, and a considerably large
number of studies that typically rely on case study analyses[2]. Relatively few attempts
have been made to combine the job satisfaction and HRM literatures. A primary
objective of this paper is to present new empirical evidence on the impact of HRM
practices on workers’ overall job satisfaction and their satisfaction with their pay.
The cu
ent issue and full text archive of this journal is available at
www.emeraldinsight.com/ XXXXXXXXXXhtm
Workers’ jo
satisfaction
651
International Journal of Manpowe
Vol. 29 No. 7, 2008
pp XXXXXXXXXX
q Emerald Group Publishing Limited
XXXXXXXXXX
DOI XXXXXXXXXX/ XXXXXXXXXX
It is possible that HRM practices are a substitute for unionisation. If HRM policies
aise overall job satisfaction, and (especially) satisfaction with pay, then the motivation
to join unions could be reduced. A second objective of this paper is thus to investigate
whether HRM practices have a different impact on the job satisfaction of union
members as opposed to non-union members.
There is a longstanding interest by economists in the role played by pay and reward
structures in determining worker effort, performance and job satisfaction. However,
the focus of this literature has tended to be on the impact of workers’ own pay or thei
comparison wage. Little is known beyond particular cases about the impact of the
distribution of pay within a firm on worker performance. This is intriguing because
there is a growing literature, which advocates the implementation of contingent, and
implicitly variable, pay structures that encourage wage dispersion[3]. As such, a third
objective of this paper is to analyse the impact of perceived pay inequality on workers’
job satisfaction.
To investigate these issues we analyse two British datasets, the “Changing
Employment Relationships, Employment Contracts and the Future of Work Survey”
(CERS), conducted in 2000, and the 1998 Workplace Employment Relations Survey
(WERS). The CERS and WERS datasets differ and complement each other in that the
former has a larger proportion of workers from small firms (i.e. those with less than ten
employees), whereas WERS excludes establishments with fewer than ten employees.
This means that WERS excludes 73 per cent of the 1.3 million establishments in the UK
(Cully et al., XXXXXXXXXXHowever, WERS covers 82 per cent of employees, making it
epresentative at the employee level, whereas CERS is representative of
establishments. These differences in coverage mean that our two datasets can thus
e seen as complements in the analysis.
To allow a comparison of the findings from the two datasets we construct a set of
covariates that are as consistent as possible. In terms of HRM practices, we identify the
following sets of variables: work organisation, supervision, employee involvement/voice,
ecruitment and selection, training and learning, and pay practices. In addition, we
include variables for the workers’ perception of pay inequality in the workplace and
whether it is unionised. Note that our comparison of the two datasets can only be
performed with respect to the workers’ satisfaction with their pay, since the WERS
dataset does not include a variable for overall job satisfaction. We construct dependent
variables in a simple binary form to ease interpretation of results.
The remainder of the paper is
oken down as follows. In section 2 we
iefly review
the theoretical and empirical literature on job satisfaction and then discuss the
literature on the impact of HRM practices on worker performance. The issue of pay
inequality is also introduced. Section 3 describes the two datasets used in the analysis
together with our econometric methodology. In section 4 we discuss the findings of ou
analysis of the impact of HRM practices on workers’ overall satisfaction with the jo
and satisfaction with pay. This section also examines the differences in outcome
etween union and non-union members, followed by a discussion of the impact of pay
inequality. Our conclusions follow in section 5.
2. Job satisfaction and HRM practices: theory and literature
Various theories of job satisfaction have been developed by psychologists and
management scholars. They tend to assign different degrees of importance to sources
IJM
29,7
652
of satisfaction[4], which can be classified as either intrinsic or extrinsic. Intrinsic
sources depend on the individual characteristics of the person, such as attitudes.
Extrinsic sources are situational, and depend on the environment, such as workplace
climate. Theories which rely on extrinsic sources are more typically adopted by
economists, albeit by reference to a different terminology, whereas intrinsic sources are
more commonly associated with other social sciences (Luchak, 2003).
Traditionally, economists have em
aced job satisfaction with “professional
suspicion” (Freeman, 1978 p. 135) because it is a subjective variable. In 2000, it was still
possible to state that the study of job satisfaction by economists is “still in its infancy”
(Blanchflower and Oswald, 2000, p. 8)[5]. However, the empirical analysis of jo
satisfaction either implicitly or explicitly draws on the theoretical models discussed
above, and in so doing job satisfaction is specified as a function of several individual
and job characteristics, and ultimately interpreted as a utility function (Clark and
Oswald, 1996; Easterlin, 2001).
Some studies show that certain HRM practices, such as working in teams, greate
discretion and autonomy in the workplace and various employee involvement and pay
schemes, do motivate workers and hence generate higher labour productivity (Cully
et al., 1999; Boselie and Van der Wiele, XXXXXXXXXXHowever, overall job satisfaction need not
increase if effort is a “bad” and the aim of workers is to maximise the returns from the
exerted effort.
In terms of the relationship between pay and job satisfaction, Clark and Oswald
(1996) show that a workers’ reported level of well-being is weakly co
elated with thei
income, whereas Belfield and Ha
is XXXXXXXXXXfind no evidence of such a relationship fo
those working in higher education. There is mounting research into the weakness of
the link between income and life satisfaction, as in the happiness studies of Layard
(2003, 2006) and Clark XXXXXXXXXXshowing that despite rising wages there are stagnant
levels in job satisfaction. Other studies show that it is relative income, rather than own
income, that matters for job satisfaction (Clark and Oswald, XXXXXXXXXXStill, some studies
contest the importance of relative income at lower pay levels (e.g. McBride, 2001), o
highlight the importance of real income (Greene and Nelson, 2007).
A wider literature exists on the effects of introducing new pay practices in
companies (see, e.g. Black and Lynch, 2004; Booth and Frank, 1999; Cappelli and
Neumark, 1999; McCausland et al., 2005; Lazear, XXXXXXXXXXYet, empirical evidence is
lacking on the relationship between such practices and job satisfaction. There are also
very few studies that seek to examine the relationship between the pay distribution
within a firm, including the perception of that distribution by a worker, and individual
worker performance or their job satisfaction[6]. An exception is Bloom and Michel
(2002), who discuss the advantages and disadvantages of dispersed and compressed
“actual” pay structures. Dispersed pay structures may induce higher levels of
performance as employees have to work harder to move up the pay ladder. This is
consistent with the notion of promoting the “star” workers in a competitive
environment and the provision of compensating differentials for high-risk jobs.
However, consistent with the prediction of tournament theory (Bloom and Michel,
2002), dispersed pay systems may also be linked to workforce instability and highe
turnover. On the other hand, compressed pay promotes team effort and cooperation by
creating a more egalitarian workplace, which tends to reduce turnover (Beaumont and
Ha
is, XXXXXXXXXXHowever, it may discourage effort above a certain minimal necessary
Workers’ jo
satisfaction
653
level, and may be perceived as unfair, not least because of free-rider problems. Hence, it
is usually difficult to accurately identify the effect of the pay distribution within a firm
on workers’ job satisfaction.
3. Data and methodology
We use two British datasets for our empirical analysis. The CERS was commissioned
y the Policy Studies Institute as part of the Future of Work research programme. The
data were collected between July 2000 and January 2001, and the main aim of the
Survey was to identify and describe the key changes in British employee relations.
Two data collection methods were used: interviews and self-completion questionnaires.
The one-hour interviews were personal, paper-based, conducted in the home, and
totalled 2,466 responses. Self-completion questionnaires were returned by 2,349
espondents, representing a 95-per cent response rate. Once we omit respondents with
missing values on key variables and the self-employed (334), the sample size reduces to
1,518. The WERS contains a much larger sample of workers (19,890 after allowing fo
missing data), and has the advantage that responses are obtained from both employees
and their managers.
The CERS