Job Satisfaction and Quitting Intentions: A Structural Model of British General Practitioners
| Author | Chris Bojke,Bonnie Sibbald,Steven Simoens,Hugh Gravelle,Anthony Scott |
| DOI | http://doi.org/10.1111/j.1467-8543.2006.00511.x |
| Date | 01 September 2006 |
| Published date | 01 September 2006 |
British Journal of Industrial Relations
44:3 September 2006 0007– 1080 pp. 519– 540
© Blackwell Publishing Ltd/London School of Economics 2006. Published by Blackwell Publishing Ltd,
9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
Blackwell Publishing Ltd.Oxford, UKBJIRBritish Journal of Industrial Relations0007-1080Blackwell Publishing Ltd/London School of Economics 2006September 2006443519540Articles
Job
Satisfaction and Quitting IntentionsBritish Journal of Industrial Relations
Anthony Scott is at the Melbourne Institute of Applied Economic and Social Research, Uni-
versity of Melbourne. Hugh Gravelle is at the National Primary Care Research and Development
Centre, Centre for Health Economics, University of York. Steven Simoens is at the Faculty of
Pharmaceutical Sciences, Catholic University of Leuven. Bonnie Sibbald is at the National
Primary Care Research and Development Centre, University of Manchester.
Job Satisfaction and Quitting Intentions:
A Structural Model of British General
Practitioners
Anthony Scott, Hugh Gravelle, Steven Simoens,
Chris Bojke and Bonnie Sibbald
Abstract
A structural model of job satisfaction and quitting intentions is estimated using
data from a survey of general practitioners in the UK. Previous research has
used reduced form models, making the interpretation of coefficients problematic.
The use of a structural recursive model helps to clarify the relationships between
intentions to quit, overall job satisfaction, domains of job satisfaction and
personal and job characteristics. Job and personal characteristics have a direct
effect on job satisfaction in addition to their indirect impact through job satis-
faction domains. Job satisfaction domains have a direct effect on intentions to
quit, in addition to their effect via overall job satisfaction. The structural
approach provides a richer interpretation of the role and effect of job character-
istics on job satisfaction and intentions to quit than is found in previous research.
This is particularly relevant in some public sector labour markets, where the
opportunity to alter wages to compensate for the relative advantages and dis-
advantages of jobs is limited because of national wage bargaining.
1. Introduction
Lower job satisfaction has been shown to increase the proportion of the
workforce intending to quit (Shields and Ward 2001), to increase actual quits,
to lower participation rates (Akerlof
et al
. 1988; Clark 1996; 2001; Laband
and Lentz 1998) and to increase absenteeism (Clegg 1983). Job satisfaction
data provide information on the non-pecuniary aspects of jobs, which is
520
British Journal of Industrial Relations
© Blackwell Publishing Ltd/London School of Economics 2006.
particularly relevant in many public sector labour markets where wages are
nationally negotiated and cannot be altered by local public employers. Local
public sector employers will, however, be able to alter non-pecuniary aspects
of jobs to attract and retain staff. In public sector jobs intrinsic and public
service motivation may be more important than in the private sector, which
also shifts the focus away from pay to non-pecuniary factors. Lower job
satisfaction reduces the supply of labour at any given wage or, equivalently,
raises the cost of maintaining a workforce of given size. It is necessary to
understand the relationships between job characteristics, including those
amenable to policy, personal characteristics, job satisfaction and intentions
to quit in order to choose the most cost-effective policies to reduce turnover.
In this paper we make a number of contributions to the job satisfaction
literature. We present a general structural model of job satisfaction and
intentions to quit. Personal and job characteristics determine satisfaction
with domains or aspects of the job, such as remuneration or whether there is
recognition for good work. Overall job satisfaction depends on the lower level
domains and on personal and job characteristics. Intentions to quit depend
on overall job satisfaction, the lower level domains and personal and job
characteristics. The model encompasses previous approaches to modelling
overall job satisfaction and intentions to quit.
Most previous studies of the determinants of job satisfaction examine the
influence of personal and job characteristics but do not consider the impact
of satisfaction with lower level job domains (see, e.g., Clark 1997; 2005; Clark
and Oswald 1996; Drakopoulos and Theodossiou 1997; Green and Tsitsianis
2005; Sloane and Williams 1994). These studies use reduced forms to model
these data, with the consequence that the precise relationships between job
characteristics and job satisfaction are unclear.
The Leyden School has developed a model of life satisfaction in which life
satisfaction is a function only of lower level domain satisfactions, such as
satisfaction with jobs, health and housing, with the lower level domains
depending on personal characteristics (Van Praag and Ferrer-i-Carbonell
2004). Recently the model has been extended to three tiers, with job satisfac-
tion depending only on lower level domains of job satisfaction in an overall
job satisfaction model (Van Praag and Ferrer-i-Carbonell 2004). Ward and
Sloane (2000) have estimated a model of overall job satisfaction depending
on lower level domains and on gender and find that gender has a significant
effect on the relationship between overall job satisfaction and lower level
domains for Scottish academics. They do not examine the effect of other
personal and job characteristics on the relationship. Kalleberg and Vaisey
(2005) examine the effect of lower level domains on overall job satisfaction
for US and Canadian machinists and allow for personal characteristics,
although they do not report whether the personal characteristics were signif-
icant or their inclusion had any effect on the coefficients on the domains.
We illustrate our approach by estimating structural models of job satisfac-
tion domains, overall job satisfaction and intentions to quit on a sample of
English and Scottish general practitioners (GPs) in the state-run National
To continue reading
Request your trial