Sunteți pe pagina 1din 31

Evidence-based HRM: a Global Forum for Empirical Scholarship

Employment status and job satisfaction


John Sutherland
Article information:
To cite this document:
John Sutherland , (2013),"Employment status and job satisfaction", Evidence-based HRM: a Global Forum
for Empirical Scholarship, Vol. 1 Iss 2 pp. 187 - 216
Permanent link to this document:
http://dx.doi.org/10.1108/EBHRM-08-2012-0008
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

Downloaded on: 14 January 2016, At: 16:12 (PT)


References: this document contains references to 49 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 2856 times since 2013*
Users who downloaded this article also downloaded:
Mark A. Tietjen, Robert M. Myers, (1998),"Motivation and job satisfaction", Management Decision, Vol. 36
Iss 4 pp. 226-231 http://dx.doi.org/10.1108/00251749810211027
Bart Cambr, Evelien Kippers, Marc van Veldhoven, Hans De Witte, (2012),"Jobs and organisations:
Explaining group level differences in job satisfaction in the banking sector", Personnel Review, Vol. 41 Iss 2
pp. 200-215 http://dx.doi.org/10.1108/00483481211200033
Michael W. Graham, Philip E. Messner, (1998),"Principals and job satisfaction", International Journal of
Educational Management, Vol. 12 Iss 5 pp. 196-202 http://dx.doi.org/10.1108/09513549810225925

Access to this document was granted through an Emerald subscription provided by emerald-srm:516270 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.

*Related content and download information correct at time of download.


The current issue and full text archive of this journal is available at
www.emeraldinsight.com/2049-3983.htm

Employment
Employment status and job status and job
satisfaction satisfaction
John Sutherland
Scottish Centre for Employment Research (SCER),
Department of Human Resource Management,
187
University of Strathclyde, Glasgow, UK Received 6 August 2012
Accepted 17 August 2012
Abstract
Purpose Motivated by the concept of procedural utility, which emphasises the salience of process-
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

related job aspects, the purpose of this paper is to addresses three questions: first, is job satisfaction
different for the self-employed with no employees and the self-employed with employees?; second, is
job satisfaction different for managers employed in smaller establishments and managers employed in
larger establishments?; and third, for both the self-employed and those in waged work, is job
satisfaction overall correlated with satisfaction with ten identified job aspects?
Design/methodology/approach A data set which has its origins in the (UK) 2006 Skills survey is
examined, making use of ordinal logit estimations.
Findings There are differences in job satisfaction between the self-employed with no employees and
those with employees, with the latter tending to be more likely to be satisfied. There are differences in
job satisfaction between managers in smaller establishment and managers in larger establishments,
but not for the three process-related job aspects associated with procedural utility. For the self-employed,
there is a predominantly positive and sometimes statistically significant correlation between an
individuals job satisfaction overall and satisfaction with the ten job aspects. For the waged worker, there
is a uniformly positive and predominantly statistically significant correlation between an individuals
job satisfaction overall and satisfaction with the ten job aspects.
Research limitations/implications The self-employed and those in waged work cannot be
assumed to constitute homogeneous groups. Consequently, when future research seeks to examine the
manner in which job satisfaction may differ across employment status groups, these groups cannot be
treated as mere dichotomous dummy variables.
Originality/value This is an empirically based reappraisal of hypotheses associated with
procedural utility which focuses upon within group differences for two sub-populations in the data set,
the self-employed and waged workers.
Keywords Job satisfaction, Self-employment, Personnel economics, HRM
Paper type Research paper

1. Introduction
There is a degree of consensus in the literature about the determinants of job
satisfaction. Most empirical studies identify the central importance of personal
characteristics such as gender, age and education and job characteristics such as
promotion prospects, hours worked, pay and establishment size. Ceteris paribus, males,

The 2006 Skills Survey data set was downloaded from the ESRC Data Archive. The survey was
co-funded by the ESRC, the Department of Education and Skills, the Department for Trade and
Industry, the Learning and Skills Council, the Sector Skills Development Agency, Scottish
Enterprise, Future Skills Wales, the East Midlands Development Agency, Highlands and Islands Evidence-based HRM: a Global
Forum for Empirical Scholarship
Enterprise and the Department for Employment and Learning (Northern Ireland). None of these Vol. 1 No. 2, 2013
parties, however, has any responsibility for the use made of the data set, nor any conclusions pp. 187-216
r Emerald Group Publishing Limited
drawn from the analysis within this paper. The author acknowledges the constructive comments 2049-3983
received from two anonymous referees on a much earlier version of this paper. DOI 10.1108/EBHRM-08-2012-0008
EBHRM individuals in their 30s, with higher levels of education, with no promotion prospects,
1,2 working long hours in larger establishments tend to have relatively lower levels of
job satisfaction. However, there is less consensus about the nature of the relationship
between job satisfaction and employment status, for example whether an individual
is self-employed or in waged work. This lack of consensus becomes especially
apparent when, rather than examining job satisfaction in general, research focuses
188 upon the satisfaction derived from different aspects of the job, for example its
pecuniary aspects, such as pay, and its non-pecuniary aspects, such as the nature
of the work undertaken.
Benz and Frey add a further dimension to this literature. Contextualising their
research in the contrasting work systems of market (epitomised by self-employment),
and hierarchy (in waged work), they argue that what is of especial consequence to
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

the former is neither the job satisfaction to be gained from pecuniary outputs nor the
job satisfaction to be gained from non-pecuniary outputs but the job satisfaction to be
gained from the processes by which these different outputs are generated, attributable
to procedural utility. Moreover, according to Benz and Frey, the job satisfaction an
individual derives from process is not unique to market systems. It is also to be found
in waged work systems. They claim that it is relatively less likely to be found in larger
establishments because of the manner in which decision-making processes there tend
to be more centralised and formalised.
Contrary to Benz and Freys (implicit) assumption, however, not all the self-
employed act as individuals within the market place. Some employ others, thereby
creating small businesses. Hence the self-employed with no employees need to be
distinguished from the self-employed with employees. Accordingly, the first research
question is:

RQ1. Is job satisfaction different for the self-employed with no employees and the
self-employed with employees?

Further, and again contrary to Benz and Freys assumptions, decision making on the
part of those in waged work is more likely to be undertaken by those in managerial
positions. Accordingly, and thereby retaining the salience of size, the second research
question is:

RQ2. Is job satisfaction different for managers employed in smaller establishments


and managers employed in larger establishments?

The analysis undertaken, therefore, is for two mutually exclusive sub-populations in


the data set: those who are self-employed and those who are in waged work. In this
way, the focus is upon differences in job satisfaction within employment status groups
not differences across employment status groups.
First, differences in job satisfaction are examined for ten job aspects, aspects of
equal relevance to individuals in self-employment and waged work and reflecting both
process and product. Latterly, satisfaction with these aspects is examined as they
feature as explanatory variables in the estimated model, generating the third research
question, which is:

RQ3. For both the self-employed and those in waged work, is job satisfaction overall
correlated with satisfaction with ten job aspects?
The paper is structured as follows: first, some literature of relevance is reviewed, to Employment
provide a context; the data set is then described; then the estimated model and the status and job
estimation strategy are outlined; the results are then reported, discussed and evaluated.
A final section concludes. satisfaction
2. The context
In traditional neo-classical economics, work is presented as a source of disutility, with 189
income earned through work treated as compensation for leisure time forgone. Work,
however, need not necessarily be a disutility. Some individuals in some circumstances
may derive degrees of satisfaction from the working situation, for example when it is
viewed as a potential source of personal achievement and self-fulfilment (Spencer,
2009, p. 3).
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

In the context of this latter perspective, job satisfaction may be thought of as a


utility function (Clark and Oswald, 1994), presented in the following form:

U uy; h; i; j 1

where U is the utility from work; y the income; h the hours; i the set of individual
characteristics; and j the set of job-specific characteristics or job aspects[1]. Job
satisfaction is what Freeman (1978) describes as a subjective variable, measuring
what people say rather than what people do (p. 135). Responses to questions about
job satisfaction, therefore, are often recorded making use of Likert-scale configurations
and analysed best by using ordinal logit/probit models.
Traditional studies of the determinants of job satisfaction (e.g. Clark, 1996; Gazioglu
and Tansel, 2006; Hammermesh, 1977; Idson, 1990; Jones et al., 2009; Schmidt, 2007;
Sousa-Poza and Sousa-Poza, 2003; Vila and Garcia-Mora, 2005) conclude that job
satisfaction relates to personal characteristics such as gender, age and education and to
job characteristics such as promotion prospects, hours worked, training provided and
establishment size. Males, individuals in their 30s, with higher levels of education,
working in larger establishments, with no promotion prospects, who have received no
training and who work long hours tend to have relatively lower levels of job satisfaction.
More recent studies of job satisfaction are of three sorts: first, those which seek to
explore further certain features of the determinants of job satisfaction, for example the
respective roles of co-workers wages (Clark et al., 2009); performance pay (Green and
Heywood, 2008); and labour market and welfare regimes (Kaiser, 2007); second, those
which seek to examine some paradoxes which have appeared in the empirical
literature, notably those which relate to part time workers (Booth and van Ours, 2008);
unions (Bryson et al., 2004; Bryson et al., 2010); and why women appear to be so happy
(Clark, 1997); and, third, those which examine the relationship between job satisfaction
and employment status, where the principal focus has been upon differences in job
satisfaction between the self-employed and those in waged work.
The research question which motivates investigations of this third sort is: why are
many individuals willing to enter and remain in self-employment despite receiving
financial remuneration often substantially less than their alternative pay in waged
work? Some, for example Hamilton (2000), examining wage equations for those in self-
employment and waged work, offer responses based upon conjecture. Others, for
example Ajayi-Obe and Parker (2005), making use of more appropriate data sets, are
able to offer more informed responses about the trade-offs which may exist between the
pecuniary and non-pecuniary aspects of jobs.
EBHRM In addressing this question, however, researchers confront the inherent problem of
1,2 sample selectivity (Heckman, 1990). To illustrate, let us assume that an individuals
employment status is a choice variable, determined by the expected utility to be
derived from each potential state. Further, and following although adapting, Taylor
(1999) assumes:

190 EUse f A; X 2

where E(Use) is the expected utility to be derived from self-employment; A the vector of
job aspects; and X the vector of variables reflecting an individuals tastes and
preferences, for example for pecuniary and non-pecuniary job aspects; and:
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

EUww gw; Anw ; X 3

where E(Uww) is the expected utility which accrues from waged work; w the wage
obtainable; Anw the vector of pecuniary and non-pecuniary job aspects; and X the
vector of variables reflecting an individuals tastes and preferences.
On the yet further assumption that individuals receive no utility from the state
of unemployment, therefore when E(Use)4E(Uww), the individual quits waged work
for self-employment. Conversely, when E(Use)oE(Uww) the individual quits the state
of self-employment, the focus of Taylors study[2].
However, if the desire for independence (or any other job aspect) motivates
the choice of individuals to become self-employed (as suggested, inter alia, by
Blanchflower, 2000; Blanchflower and Oswald, 1998; Parker, 2004; Prottas, 2008),
this process of self-selection into particular employment status groups results in
situations where the individuals concerned no longer constitute random samples
of the populations or sub-populations in question (Bazen, 2011). The phenomenon of
selectivity poses particular problems in empirical analysis because it creates a form
of endogeneity in which the dependent variable in the model to be estimated and one
or more of the independent variables are not exogenous. How researchers address this
problem is an especial feature of studies of the differences in job satisfaction between
the self-employed and those in waged work. The use of panel data sets, which
eliminates the problem of unobserved heterogeneity, is one means whereby this
methodological issue may be resolved. Two studies of note make use of the British
Household Panel Survey (BHPS).
In their study of the labour supply of the self-employed and those in waged work,
Ajayi-Obe and Parker (2005) encounter an outstanding puzzle (p. 502), namely the
propensity of the self-employed to work longer hours than those in waged work for
lower rates of pay despite being relatively less satisfied with the hours they work.
Accordingly (using both ordinary least squares and two stage least squares) they
estimate wage elasticities for both groups and identify differences in these estimations.
These differences explain why the self-employed value their leisure time less because
they enjoy their jobs more, at least in terms of the two non-pecuniary job aspects
identified namely the work itself and job satisfaction overall. In contrast, Taylor (2004)
examines differences in job satisfaction between the self-employed and those in
waged work using job satisfaction scores for all five job aspects available within BHPS,
estimating an ordinal probit model to do so, namely, job satisfaction overall; pay;
the work itself; hours worked; and job security. The self-employed are found to be
relatively more satisfied with all job aspects, with the exception of job security.
In the work cited above, authors seek to differentiate between the pecuniary and Employment
non-pecuniary aspect of jobs. Benz and Frey (2008a) make use of the concept of status and job
procedural utility to introduce an alternative categorisation, namely, product and
process. They argue that what is of especial consequence to the self-employed is neither satisfaction
the job satisfaction to be gained from pecuniary outputs nor the job satisfaction to
be gained from non-pecuniary outputs but the job satisfaction to be gained from
the processes by which these different outputs are generated. To quote: People care 191
about the what, but also about the how (Benz and Frey, 2008a, p. 363).
The concept of procedural utility has its origins in the assumption that individuals
have a psychological need for self-determination, in particular that aspect of self-
determination associated with autonomy, what Benz and Frey (2008a) illustrate as the
experience to self-organise ones own actions (p. 364). Procedural goods may be
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

provided by different institutions and they will differ in their quantity and quality
depending upon the different procedures which operate in these different institutions.
Work is one such institution, and following Williamson (1975, 1985), Benz and Frey
(2008a) identify two decision-making systems with respect to the organisation of work
in contemporary society, namely, hierarchy and market. The former implies that work
is integrated into organisations and is characterised by some means of authority.
In contrast, the defining characteristic of the market is autonomous decision making on
the part of self-employed individuals. Self-employment, therefore, generates procedural
utility in a way that waged work cannot. As a consequence, although the self-employed
may not necessarily derive job satisfaction from outcomes such as wages, hours
of work, etc., they do derive job satisfaction from processes such as their decision-
making independence (Frey and Stutzer, 2005).
Procedural utility and process-related job aspects, however, are not the sole
prerogative of the self-employed. An additional feature of Benz and Frey (2008a) is the
assumption that the intensity of hierarchy increases with firm size. With increases
in size, employees are subjected to additional layers of decision making in processes
which are themselves increasingly formal. Therefore, they contend, individuals
working in smaller businesses enjoy higher procedural utility[3].
Benz and Frey (2008a) examine three panel data sets, for Germany, Great Britain
and Switzerland. In each, they demonstrate, first, that the self-employed enjoy higher
satisfaction than those in waged work and, second, for the latter, that satisfaction is
lower the larger the hierarchy to which the individual employee is subject.
Their analysis of the BPHS to undertake a direct test for procedural utility is
especially instructive for the purposes of this paper. The BHPS asks questions about
job satisfaction to all respondents. Benz and Frey argue that two of these questions
are product related i.e. job security and the hours worked and two are related to the
procedural aspects of work i.e. being able to use your own initiative and the actual
work itself. Benz and Frey (2008a) proceed to undertake three weighted ordinal logit
regressions of the fifth job satisfaction question namely, job satisfaction overall.
The first is a base line regression; the second includes the product aspects as
additional independent variables; and the third includes the procedural aspects
as additional independent variables. In the second and third regressions, the product-
related aspects and the procedural-related aspects are statistically significant.
However, in the third regression the value of the coefficient on the self-employed
dummy variable decreases and is no longer statistically significant. From which the
authors conclude: This is strong evidence that procedural utility is the reason
why the self-employed are more satisfied with their jobs than employees (p. 379).
EBHRM A similar exercise is undertaken for those in waged work to examine the firm size-job
1,2 satisfaction relationship and equivalent results are produced thereby demonstrating
again the salience of procedural job aspects.
Focusing upon differences within employment status groups, this paper addresses
three questions. First: for a set of ten job aspects, is job satisfaction different for the
self-employed with no employees and the self-employed with employees, sub groups
192 which Taylor (2004) describes, respectively, as the sole proprietor and the job creator?
Second: for these same job aspects, is job satisfaction different for managers in smaller
establishments and managers in larger establishments? Third: for both the self-
employed and those in waged work, is job satisfaction overall correlated with
satisfaction with these job aspects? The research reported, therefore, is motivated by
the concept of procedural utility in general and Benz and Frey (2008a) in particular.
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

However, it pursues different lines of inquiry and seeks to capitalise upon the richer
set of job aspects available in the data set.

3. The data
The data source used is the 2006 Skills survey, one of several successive surveys in the
UK which have their origins in the surveys associated with the Social Change and
Economic Life Initiative of 1986 (Gallie et al., 1998)[4]. The core sample for the survey is
based on a multi-stage design, with addresses drawn from a random starting point
within the 297 post codes selected across the UK. In all, 4,800 interviews of individuals
aged 20-65 in employment were undertaken over a seven-month period during 2006.
Following boosts taken within five areas, the total number of observations available in
the data set is 7,787. When weighted, the data set is representative of the UK working
population of the age identified. The working data set has 6,187 observations, when
those observations with incomplete information across all the variables used in
the estimations are dropped. Of the 6,187 observations 5,469 are in waged work; the
remainder are self-employed.
The focus of the surveys is the nature and content of the skills individuals possess
and require in their current jobs. Information is collected also about the context e.g. the
working environment in which these skills are acquired, developed and applied
(Felstead et al., 2007). More recently, other related questions have been incorporated, for
example seeking to elicit perspectives on factors such as job quality, job satisfaction
and task discretion. The focus of this paper is on those questions which relate to
job satisfaction.
With reference to the questions on job satisfaction, respondents are asked:
How satisfied are you with this particular aspect of your own present job?, where the
aspects in question are identified in column 1 of Table I. Four of these aspects are
deemed irrelevant to the analysis, being appropriate only to those in waged work
(cf. Table I, column 2). Of the ten aspects which are relevant, seven are deemed
compatible with product where two are similar to the product aspects used by Benz
and Frey (2008a) and three are deemed compatible with process two being similar
to the process aspects used by Benz and Frey (cf. Table I, column 3).
In the questionnaire, there are seven response options, being: 1, completely
satisfied; 2, very satisfied; 3, fairly satisfied; 4, neither satisfied nor dissatisfied;
5, fairly dissatisfied; 6, very dissatisfied; and 7, completely dissatisfied.
To facilitate analysis in the subsequently sub-divided data set, where the number of
observations for the self-employed only total 718, these seven responses are reduced to
three, being: 1, satisfied; 2, neither satisfied nor dissatisfied; and 3, dissatisfied.
Relevant to both those in self- If relevant, process
Employment
Aspect employment and waged work? or product? status and job
satisfaction
Promotion prospects No
Pay Yes Product
Relations with your supervisor/manager No
Job security Yes Product 193
The opportunity to use your abilities Yes Process
Being able to use your own initiative Yes Process
The ability and efficiency of management No
The hours you work Yes Product
Fringe benefits Yes Product
The work itself Yes Process
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

The amount of work Yes Product Table I.


The variety in the work Yes Product Job satisfaction aspects;
The training provided Yes Product their relevance to the sub-
The friendliness of the people you populations; and whether
work with No they may be classified as
Overall satisfaction with job Yes process or product

The percentage frequency distribution of these re-coded responses for both sub-
populations is presented in Table II.
The immediate conclusion which may be made from Table II is that of a generally
well-satisfied workforce, substantiating Rose (2005), therefore, and refuting the
despondency thesis of low workforce morale. For individuals in self-employment, more
than nine in every ten are satisfied with four of the ten job aspects. The job aspect
which generates most dissatisfaction is pay (at 14.62 per cent). The job aspect which
generates least satisfaction is the training provided (at 44.83 per cent). For those in
waged work, more than four in every five individuals are satisfied with six of the
ten job aspects. Again, the job aspect which generates most dissatisfaction is pay
(at 22.56 per cent). The job aspect which generates least satisfaction is fringe benefits
(at 44.83 per cent). The job aspect which generates most satisfaction is being able
to use your own initiative (at 88.68 per cent). Significantly, in the context of the
three process aspects central to Benz and Freys (2008a) argument, satisfaction
is proportionately greater for the self-employed than those in waged work in
each instance.
The statistic for job satisfaction overall in the final row of Table II confirms this
picture of satisfaction, with the self-employed recording a higher rate of satisfaction
than those in waged work, by (approximately) eight percentage points.

4. The estimation model and the estimation strategy


Given that the Likert-scale format is retained in the re-coded responses to the job
satisfaction questions, an ordinal logit model is used to analyse the data set.
The ordinal logit model may be presented as a latent variable model. Defining y* as
the latent variable whose values range from Nto N, the structural model is:

yi Xi b ei 4

where X is the vector of variables for observation i, b the set of corresponding


coefficients to be estimated and e is a random error term.
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

1,2

194

Table II.
EBHRM

and waged work


aspects, percentage
Satisfaction with job

frequency distribution,
those in self-employment
In self-employment In waged-work
Neither dissatisfied Neither dissatisfied
Job aspect Dissatisfied nor satisfied Satisfied Dissatisfied nor satisfied Satisfied

Product
Pay 14.62 14.35 71.03 22.56 11.70 65.73
Job security 7.38 20.06 72.56 8.70 8.89 83.41
The hours you work 11.00 9.19 79.81 10.66 8.30 81.04
Fringe benefits 8.91 45.54 45.54 19.25 35.91 44.83
The amount of work 8.77 9.47 81.75 12.51 11.19 76.30
The variety in the
work 2.51 4.18 93.31 6.09 9.07 84.84
The training provided 3.76 44.57 51.67 11.54 21.14 67.32
Process
The opportunity to use
your abilities 2.23 4.18 93.59 6.60 7.22 86.18
Being able to use your
own initiative 1.53 1.95 96.52 5.08 6.24 88.68
The work itself 1.67 3.34 94.99 4.88 7.22 87.90
Job satisfaction overall 2.09 3.06 94.85 7.66 5.69 86.65
The measurement model divides y* into J ordinal categories: Employment
status and job
yi m if tm1 ptm from m 1 to J 5
satisfaction
where threshold points t1-tj1 are estimated.
Accordingly, the measurement model is as follows, where observation i takes the
following values: 195
1 ! dissatisfied if t0 1pyi ot1 5a

2 ! neither dissatisfied nor satisfied if t1 pyi ot2 5b


Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

3 ! satisfied if t2 pyi ot3 1 5c

The sign of the regression parameters (b) is interpreted as determining whether


the latent variable (y*) increases with the regressor. If bi is positive then an increase
in the appropriate variable within the X vector decreases the likelihood of being in
the lowest response category and increases the likelihood of being in the highest
response category (Cameron and Trivedi, 2009; Long and Freese, 2006).
There is a diverse range of potential individual related explanatory variables
available in the Skills survey data set. Consequently, the X vector includes variables
reflecting both personal characteristics, such as gender, age, highest qualification
held, in levels[5]; and characteristics pertaining to work, such as tenure, the number of
hours normally worked each week, whether working fulltime, and the numbers
of years in work since leaving full time education. The full list of the explanatory
variables used is provided in column 1 of Table III.
To address the research questions posed, the variables of importance are those which
relate to an individuals employment status. Two mutually exclusive sub-populations
within the data set are examined: the self-employed and those in waged work.
First, for each sub-population, the ordinal logit model is estimated for each of the
ten job aspects identified in Table II. In the estimations which pertain to the self-
employed group, the self-employed with no employees forms the reference category.
Results for the self-employed with employees are reported in the respective tables.
Consequently, differences between these two categories of self-employed workers in
essence the answer to the first research question may be observed from these tables.
In the corresponding estimations of those in waged work, the reference employment
status category is the employee with no managerial responsibilities. Results for the two
managerial categories i.e. those managing in establishments with o25 employees
and those managing in establishments with more than 24 employees are reported
in the respective tables. Subsequently, Wald tests are undertaken to ascertain whether
satisfaction with the job aspect in question differs between these two managerial
categories in essence the answer to the second research question.
The ordinal logit model is then estimated again for each sub-population for the
dependent variable satisfied overall with the job. This estimation is undertaken twice.
First, the estimation uses the same set of dependent variables which appear in the
earlier estimations of the ten job aspects. Then the model is re-estimated incorporating
satisfaction with these ten job aspects as additional explanatory variables. For
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

1,2

196

Table III.
EBHRM

Ordinal logits: pay


In self-employment In waged work
Variable Coefficient Linearized SE Coefficient Linearized SE

Male 0.1529 0.2798 0.1344 0.0864


Age, in years, in logs 0.4103 1.2963 0.2062 0.2759
Married (or living as) 0.0363 0.2445 0.1896* 0.0985
With financially dependent children aged o16 0.0426 0.2527 0.0587 0.0873
White 0.0640 0.4995 0.3874** 0.1501
Has no qualifications (the reference category)
Has level 1 as highest qualification 0.7943* 0.4750 0.1592 0.1705
Has level 2 as highest qualification 0.4276 0.4265 0.0154 0.1496
Has level 3 as highest qualification 0.8077** 0.3789 0.1024 0.1550
Has level 4 or 5 as highest qualification 0.4579 0.3800 0.1969 0.1491
Holds only 1 job 0.6483* 0.3525 0.3836* 0.2042
Tenure, in years, in logs 0.0710 0.0915 0.0327 0.0339
Number of hours usually worked each week, in logs 0.4388 0.2957 0.0666 0.1509
Works full time 0.0880 0.4527 0.1933 0.1637
Has received training for current job 0.0853 0.2291 0.1699** 0.0813
Number of years in the labour market, in logs 0.3759 0.5346 0.0638 0.0951
Has experienced a spell of unemployment in last 2/3/4/5 years 0.6027* 0.3652 0.3851*** 0.1316
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are
self-employed)
Self-employed with employees 0.2338 0.2331
In waged work as a manager in an establishment with o25 employees 0.4489*** 0.1443
In waged work as a manager in an establishment with 424 employees 0.5327*** 0.1305
Cut 1 0.2643 4.0459 0.0759 1.0592
Cut 2 1.0866 4.0491 0.5419 1.0580
Number of observations 6,187 6,187
Sub-population number of observations 718 5,469
F-statistic (17, 6170) 1.31 (17, 6169) 4.25
Prob4F 0.1742 0.0000
Note: *,**,***Significant at po0.1, po0.05 and po0.01
this exercise, responses to the ten job aspect questions are further re-coded into Employment
dichotomous dummy variables with satisfied 1 and otherwise 0. In this way, status and job
the third research question is addressed. Subsequently, Wald tests are undertaken
to ascertain the joint significance of the product aspects, the process aspects and satisfaction
the product and process aspects together.

5. The results 197


The results are reported, discussed and evaluated in three sub-sections. The first sub-
section relates to the self-employed and examines the results of the estimations of the
ten job aspects; the second sub-section does the same, but for those in waged work; and
the third sub-section examines the results of the estimations of the dependent variable
job satisfaction overall, for the self-employed and those in waged work, first without
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

the ten job aspects as supplementary explanatory variables, then with their inclusion.

5.1 The self-employed


The results of the ordinal logit estimations for the self-employed are presented in
columns 2 and 3 of Tables III-XII.
In the estimation of the process-related job aspect being able to use your own
initiative, 60 observations are completely determined (cf. Table XI). The results
reported, therefore, are problematic and this estimation is ignored in what follows.
For the remaining nine estimations, throughout, the cutpoints are in accordance
with expectations and the majority are statistically significant at po0.1[6]. In four
estimations, namely, pay; fringe benefits; the amount of work; and the variety of
work, the F-statistic for the estimated model as a whole is not statistically significant.
This outcome is most probably attributable to the skewness of the responses
away from dissatisfied, as reported in Table II, and an insufficiency in the number of
observations to compensate for this. Perhaps for similar reasons, very few of the
coefficients of the individual variables in the nine estimations are statistically significant.
Principally making use of the signs of the coefficients of the dependent variables
to do so, it is possible to construct some patterns of the nature of the relationship
between some of these variables and degrees of satisfaction with the job aspect in
question. The coefficients of the variable male are negatively signed across the nine
estimations, with the exception of estimations of the job aspects the hours you work
and the work itself, outcomes in accordance with expectations given the earlier
literature about the determinants of job satisfaction (e.g. Clark, 1996, 1997). Similarly
with the coefficients of the variable the number of hours usually worked each week,
coefficients are also negatively signed across the estimations, with the exception of the
estimation of the amount of work. One outcome which does not accord with the
expectations generated by the literature is the generally positive signs of the
coefficients of the dummy variables associated with highest qualification held, relative
to the reference category of having no qualifications (e.g. Vila and Garcia-Mora, 2005).
When questioned about specific job aspects, for the sub-population of the self-
employed, those with higher levels of qualification are more likely to be satisfied.
Training (cf. Jones et al., 2009) has an impact which varies by estimation. Whereas
the coefficients of the training variable are positively signed in the estimations of
pay; job security; the variety of work; the training provided; and the
opportunities to use your abilities, they are negatively signed in the estimations of
the hours you work; fringe benefits; the amount of work; and the work itself.
The coefficients of the variable depicting labour market experience are positively
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

1,2

198

Table IV.
EBHRM

Ordinal logits: job security


In self-employment In waged work
Variable Coefficient Linearized SE Coefficient Linearized SE

Male 0.2802 0.2937 0.3264*** 0.1145


Age, in years, in logs 1.3125 0.9757 0.3918 0.3514
Married (or living as) 0.1571 0.2702 0.0871 0.1057
With financially dependent children aged o16 0.1122 0.2402 0.0530 0.1064
White 0.3969 0.5486 0.3424* 0.1934
Has no qualifications (the reference category)
Has level 1 as highest qualification 0.1581 0.5113 0.3572* 0.2148
Has level 2 as highest qualification 0.1891 0.3949 0.1973 0.1795
Has level 3 as highest qualification 0.0565 0.3572 0.3364* 0.1796
Has level 4 or 5 as highest qualification 0.3051 0.3584 0.5639*** 0.1722
Holds only 1 job 0.9184*** 0.3246 0.3418 0.2528
Tenure, in years, in logs 0.0058 0.0864 0.0623 0.0423
Number of hours usually worked each week, in logs 0.4820* 0.2887 0.1734 0.1811
Works full time 0.3506 0.5011 0.0102 0.2179
Has received training for current job 0.0662 0.2210 0.1436 0.1006
Number of years in the labour market, in logs 0.2303 0.4325 0.0394 0.1202
Has experienced a spell of unemployment in last 2/3/4/5 years 0.8138** 0.4117 0.7493*** 0.1590
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are
self-employed)
Self-employed with employees 0.3885 0.2364
In waged work as a manager in an establishment with o25 employees 0.3080** 0.1453
In waged work as a manager in an establishment with 424 employees 0.1850 0.1280
Cut 1 8.3356*** 2.8440 2.7681** 1.2522
Cut 2 6.6799** 2.8417 1.9571 1.2501
Number of observations 6,187 6,187
Sub-population number of observations 718 5,469
F-statistic (17, 6170) 1.48 (17, 6169) 3.71
Prob4F 0.0903 0.0000
Note: *,**,***Significant at po0.1, po0.05 and po0.01
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

In self-employment In waged work


Variable Coefficient Linearized SE Coefficient Linearized SE

Male 0.3332 0.3620 0.1255 0.1293


Age, in years, in logs 1.5080 1.5500 0.2963 0.3979
Married (or living as) 0.1564 0.2941 0.1204 0.1107
With financially dependent children aged o16 0.2321 0.2627 0.0373 0.1097
White 0.2016 0.6001 0.0346 0.2005
Has no qualifications (the reference category)
Has level 1 as highest qualification 1.2395* 0.6976 0.0727 0.2938
Has level 2 as highest qualification 0.0177 0.4521 0.1808 0.2891
Has level 3 as highest qualification 0.2186 0.4514 0.0368 0.2809
Has level 4 or 5 as highest qualification 0.0096 0.5124 0.3219 0.2869
Holds only 1 job 0.0771 0.4756 0.0019 0.2416
Tenure, in years, in logs 0.1187 0.1224 0.0001 0.0443
Number of hours usually worked each week, in logs 1.9076*** 0.5043 2.2243*** 0.3371
Works full time 0.6074 0.6230 0.5084** 0.2201
Has received training for current job 0.3191 0.2495 0.1755* 0.1021
Number of years in the labour market, in logs 0.7038 0.7494 0.0303 0.1446
Has experienced a spell of unemployment in last 2/3/4/5 years 0.2515 0.5079 0.2706* 0.1633
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are
self-employed)
Self-employed with employees 0.5795** 0.2663
In waged work as a manager in an establishment with o25 employees 0.0214 0.1947
In waged work as a manager in an establishment with 424 employees 0.1494 0.1682
Cut 1 5.8954 4.0209 8.7898*** 1.7369
Cut 2 5.1594 4.0279 8.0334*** 1.7303
Number of observations 6,187 6,187
Sub-population number of observations 718 5,469
F-statistic (17, 6170) 2.71 (18, 6169) 6.02
Prob4F 0.0002 0.0000
Note: *,**,***Significant at po0.1, po0.05 and po0.01
satisfaction
status and job
Employment

199

Ordinal logits: the


Table V.

hours you work


Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

1,2

200

Table VI.
EBHRM

Ordinal logits:
fringe benefits
In self-employment In waged work
Variable Coefficient Linearized SE Coefficient Linearized SE

Male 0.0265 0.2327 0.1788** 0.0789


Age, in years, in logs 1.7623* 0.9609 0.2915 0.2552
Married (or living as) 0.1583 0.2157 0.0225 0.0843
With financially dependent children aged o16 0.2751 0.2138 0.0088 0.0813
White 0.2937 0.4290 0.0337 0.1642
Has no qualifications (the reference category)
Has level 1 as highest qualification 0.3050 0.4015 0.1787 0.1920
Has level 2 as highest qualification 0.1956 0.3637 0.0826 0.1762
Has level 3 as highest qualification 0.0501 0.3247 0.1850 0.1760
Has level 4 or 5 as highest qualification 0.1193 0.3296 0.2592 0.1725
Holds only 1 job 0.5936* 0.3407 0.1928 0.1540
Tenure, in years, in logs 0.0028 0.0791 0.0027 0.0295
Number of hours usually worked each week, in logs 0.5754** 0.2520 0.0605 0.1178
Works full time 0.2828 0.3840 0.0794 0.1257
Has received training for current job 0.1234 0.1928 0.1670** 0.0726
Number of years in the labour market, in logs 0.7634** 0.3808 0.0092 0.0851
Has experienced a spell of unemployment in last 2/3/4/5 years 0.0818 0.3653 0.4650*** 0.1102
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are
self-employed)
Self-employed with employees 0.7243*** 0.2035
In waged work as a manager in an establishment with o25 employees 0.6281*** 0.1435
In waged work as a manager in an establishment with 424 employees 0.5184*** 0.1285
Cut 1 8.2450*** 2.8944 3.1881*** 0.9965
Cut 2 5.6982** 2.8823 1.5417 0.9875
Number of observations 6,187 6,187
Sub-population number of observations 718 6,186
F-statistic (17, 6170) 1.43 (18, 6169) 4.16
Prob4F 0.1139 0.0000
Note: *,**,***Significant at po0.1, po0.05 and po0.01
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

In self-employment In waged work


Variable Coefficient Linearized SE Coefficient Linearized SE

Male 0.2063 0.3162 0.3519*** 0.1000


Age, in years, in logs 0.1632 0.9626 0.1074 0.3349
Married (or living as) 0.0652 0.2964 0.0281 0.1024
With financially dependent children aged o16 0.0322 0.2624 0.0210 0.0929
White 0.1522 0.5915 0.0048 0.1890
Has no qualifications (the reference category)
Has level 1 as highest qualification 0.2255 0.5912 0.4309** 0.2015
Has level 2 as highest qualification 0.0914 0.5053 0.2677 0.1767
Has level 3 as highest qualification 0.4064 0.4864 0.2721 0.1791
Has level 4 or 5 as highest qualification 0.2553 0.4777 0.6177*** 0.1731
Holds only 1 job 0.1062 0.4649 0.0993 0.2010
Tenure, in years, in logs 0.1599* 0.0954 0.0676 0.0413
Number of hours usually worked each week, in logs 0.3371 0.3445 0.7830*** 0.2160
Works full time 0.5518 0.5778 0.1196 0.1776
Has received training for current job 0.5254** 0.2557 0.0462 0.0905
Number of years in the labour market, in logs 0.1180 0.3896 0.0614 0.1266
Has experienced a spell of unemployment in last 2/3/4/5 years 0.3773 0.3790 0.2020 0.1519
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are
self-employed)
Self-employed with employees 0.4251 0.2612
In waged work as a manager in an establishment with o25 employees 0.1313 0.1409
In waged work as a manager in an establishment with 424 employees 0.2579** 0.1212
Cut 1 2.3308 3.0827 4.8707*** 1.2388
Cut 2 1.4202 3.0974 4.0224*** 1.2405
Number of observations 6,187 6,187
Sub-population number of observations 718 5,469
F-statistic (17, 6170) 1.10 (18, 6169) 3.07
Prob4F 0.3425 0.0000
Note: *,**,***Significant at po0.1, po0.05 and po0.01

Ordinal logits:
Table VII.

the amount of work


satisfaction
status and job
Employment

201
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

1,2

202

Table VIII.
EBHRM

Ordinal logits: the


variety of the work
In self-employment In waged work
Variable Coefficient Linearized SE Coefficient Linearized SE

Male 1.2405** 0.5462 0.2340* 0.1361


Age, in years, in logs 2.8302** 1.3264 0.9743** 0.4063
Married (or living as) 0.1447 0.4298 0.1252 0.1257
With financially dependent children aged o16 0.0543 0.4175 0.0273 0.1193
White 0.4316 0.7436 0.1011 0.2027
Has no qualifications (the reference category)
Has level 1 as highest qualification 0.8413 0.8918 0.0668 0.2283
Has level 2 as highest qualification 0.0032 0.6118 0.0711 0.1900
Has level 3 as highest qualification 0.1411 0.5947 0.1143 0.1980
Has level 4 or 5 as highest qualification 0.0109 0.5419 0.2194 0.2051
Holds only 1 job 0.9019 1.2133 0.0333 0.3603
Tenure, in years, in logs 0.0819 0.1473 0.0333 0.0511
Number of hours usually worked each week, in logs 0.7055 0.5905 0.2905 0.1944
Works full time 0.6807 0.8781 0.1379 0.2184
Has received training for current job 0.5945 0.4126 0.5590 *** 0.1084
Number of years in the labour market, in logs 1.0489* 0.6025 0.1312 0.1391
Has experienced a spell of unemployment in last 2/3/4/5 years 0.2645 0.5870 0.4284** 0.1932
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are
self-employed)
Self-employed with employees 0.2610 0.4367
In waged work as a manager in an establishment with o25 employees 0.2511 0.1911
In waged work as a manager in an establishment with 424 employees 0.7120*** 0.1679
Cut 1 15.0094*** 4.0816 1.3543 1.3917
Cut 2 13.8705*** 4.0602 2.4364* 1.3869
Number of observations 6,187 6,187
Sub-population number of observations 718 5,469
F-statistic (17, 6170) 1.23 (18, 6169) 6.55
Prob4F 0.2295 0.0000
Note: *,**,***Significant at po0.1, po0.05 and po0.01
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

In self-employment In waged work


Variable Coefficient Linearized SE Coefficient Linearized SE

Male 0.1857 0.2415 0.1745* 0.0890


Age, in years, in logs 3.2499*** 0.8620 0.3576 0.2978
Married (or living as) 0.0054 0.2106 0.1168 0.0976
With financially dependent children aged o16 0.4789 0.2179 0.0066 0.0892
White 0.0746 0.4674 0.0275 0.1765
Has no qualifications (the reference category)
Has level 1 as highest qualification 0.4401 0.3847 0.0812 0.1715
Has level 2 as highest qualification 0.7066** 0.3393 0.1861 0.1442
Has level 3 as highest qualification 0.1621 0.3203 0.0532 0.1498
Has level 4 or 5 as highest qualification 0.2413 0.3127 0.0806 0.1417
Holds only 1 job 0.3325 0.3164 0.0205 0.2011
Tenure, in years, in logs 0.0423 0.0785 0.0212 0.0353
Number of hours usually worked each week, in logs 0.2196 0.3025 0.0453 0.1465
Works full time 0.1514 0.3747 0.0208 0.1585
Has received training for current job 0.6172*** 0.1973 0.7718*** 0.0791
Number of years in the labour market, in logs 0.9848*** 0.3536 0.0031 0.1055
Has experienced a spell of unemployment in last 2/3/4/5 years 0.2121 0.3328 0.4785*** 0.1334
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are
self-employed)
Self-employed with employees 0.5187** 0.2170
In waged work as a manager in an establishment with o25 employees 0.1289 0.1308
In waged work as a manager in an establishment with 424 employees 0.0391 0.1177
Cut 1 13.3910*** 2.6780 2.9355*** 1.1300
Cut 2 9.8137*** 2.6222 1.5875 1.1265
Number of observations 6,187 6,187
Sub-population number of observations 718 5,469
F-statistic (17, 6170) 2.35 (18, 6169) 8.71
Prob4F 0.0014 0.0000
Note: *,**,***Significant at po0.1, po0.05 and po0.01

Ordinal logits: the


satisfaction

Table IX.

training provided
status and job
Employment

203
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

1,2

204

Table X.
EBHRM

Ordinal logits:

use your abilities


the opportunity to
In self-employment In waged work
Variable Coefficient Linearized SE Coefficient Linearized SE

Male 1.5899** 0.6730 0.2885** 0.1397


Age, in years, in logs 1.4880 1.5308 0.2145 0.3734
Married (or living as) 0.0415 0.4895 0.1600 0.1322
With financially dependent children aged o16 0.3278 0.4283 0.0541 0.1193
White 0.6040 0.8349 0.0110 0.2306
Has no qualifications (the reference category)
Has level 1 as highest qualification 1.6187* 0.9589 0.3507 0.2422
Has level 2 as highest qualification 0.2380 0.6972 0.2735 0.1966
Has level 3 as highest qualification 0.7902 0.6191 0.4421** 0.2001
Has level 4 or 5 as highest qualification 0.0918 0.6598 0.3822* 0.2098
Holds only 1 job 0.6699 0.7423 0.1197 0.2822
Tenure, in years, in logs 0.0249 0.1618 0.1787*** 0.0496
Number of hours usually worked each week, in logs 0.0583 0.5668 0.3232 0.2064
Works full time 0.0579 0.8479 0.0332 0.2295
Has received training for current job 1.1233*** 0.4172 0.6727*** 0.1151
Number of years in the labour market, in logs 0.9220 0.5805 0.1179 0.1325
Has experienced a spell of unemployment in last 2/3/4/5 years 0.7190 0.5190 0.3509* 0.1881
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are self-employed)
Self-employed with employees 0.2839 0.4163
In waged work as a manager in an establishment with o25 employees 0.4480** 0.2050
In waged work as a manager in an establishment with 424 employees 0.6884*** 0.1782
Cut 1 7.5229* 4.5186 2.2356* 1.3504
Cut 2 6.4425 4.5314 1.3268 1.3525
Number of observations 6,187 6,187
Sub-population number of observations 718 5,469
F-statistic (17, 6170) 3.41 (18, 6169) 6.84
Prob4F 0.0000 0.0000
Note: *,**,***Significant at po0.1, po0.05 and po0.01
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

In self-employment In waged work


Variable Coefficient Linearized SE Coefficient Linearized SE

Male 3.5621*** 0.9698 0.0776 0.1531


Age, in years, in logs 2.5899 1.7951 0.2637 0.4037
Married (or living as) 0.4613 0.7635 0.2174 0.1428
With financially dependent children aged o16 0.2455 0.5802 0.1084 0.1278
White 0.1244 0.7900 0.3666* 0.2158
Has no qualifications (the reference category)
Has level 1 as highest qualification 15.4284*** 0.9010 0.0509 0.2578
Has level 2 as highest qualification 0.1635 1.1413 0.0214 0.2060
Has level 3 as highest qualification 0.0606 0.8011 0.1035 0.2131
Has level 4 or 5 as highest qualification 0.5356 0.9436 0.0825 0.2295
Holds only 1 job 0.2365 1.2000 0.3574 0.3177
Tenure, in years, in logs 0.1516 0.2055 0.1700** 0.0561
Number of hours usually worked each week, in logs 1.1028 0.9394 0.1381 0.2054
Works full time 1.2242 1.1188 0.0693 0.2524
Has received training for current job 0.5491 0.5177 0.6725*** 0.1320
Number of years in the labour market, in logs 0.9616 0.6847 0.1711 0.1396
Has experienced a spell of unemployment in last 2/3/4/5 years 0.9147 0.6584 0.2559 0.2197
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are
self-employed)
Self-employed with employees 0.0338 0.5096
In waged work as a manager in an establishment with o25 employees 0.4829** 0.2307
In waged work as a manager in an establishment with 424 employees 0.7431*** 0.2038
Cut 1 17.4963*** 6.4385 2.1326 1.4592
Cut 2 16.4923** 6.3675 1.2050 1.4577
Number of observations 6,187 6,187
Sub-population number of observations 718 5,469
F-statistic (17, 6170) 110.22 (18, 6169) 6.32
Prob4F 0.0000 0.0000
Notes: In the estimation of the self-employed sub-population, 60 observations were completely determined. Stata reports: Standard errors are questionable.
*,**,***Significant at po0.1, po0.05 and po0.01
satisfaction

Ordinal logits: being able


Table XI.

to use your own initiative


status and job
Employment

205
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

1,2

206

Table XII.
EBHRM

Ordinal logits:
the work itself
In self-employment In waged work
Variable Coefficient Linearized SE Coefficient Linearized SE

Male 0.0903 0.5838 0.2946** 0.1339


Age, in years, in logs 2.5848 1.7120 0.3004 0.4100
Married (or living as) 0.3754 0.5317 0.2821** 0.1378
With financially dependent children aged o16 0.4421 0.4744 0.0925 0.1209
White 0.2660 0.7665 0.1752 0.2304
Has no qualifications (the reference category)
Has level 1 as highest qualification 1.30006 1.0059 0.3379 0.2495
Has level 2 as highest qualification 0.0719 0.6772 0.2783 0.2094
Has level 3 as highest qualification 1.2354* 0.6669 0.3499* 0.2108
Has level 4 or 5 as highest qualification 0.4475 0.6869 0.3690* 0.2235
Holds only 1 job 1.0396 0.7169 0.0019 0.2460
Tenure, in years, in logs 0.4153** 0.1934 0.0465 0.0588
Number of hours usually worked each week, in logs 0.8841 0.6472 0.1543 0.2304
Works full time 1.2853 1.1484 0.2400 0.2344
Has received training for current job 0.0001 0.4068 0.4678*** 0.1217
Number of years in the labour market, in logs 1.4246** 0.6689 0.1338 0.1499
Has experienced a spell of unemployment in last 2/3/4/5 years 0.9954* 0.5907 0.3644* 0.1951
In waged work as employee (the reference category for those in waged work)
Self-employed with no employees (the reference category for those who are
self-employed)
Self-employed with employees 0.1221 0.4506
In waged work as a manager in an establishment with o25 employees 0.2747 0.2138
In waged work as a manager in an establishment with 424 employees 0.6086*** 0.1798
Cut 1 13.0877** 5.0967 1.7045 1.4850
Cut 2 12.087** 5.0693 0.6795 1.4832
Number of observations 6,187 6,187
Sub-population number of observations 718 5,469
F-statistic (17, 6170) 2.51 (18, 6169) 4.64
Prob4F 0.0006 0.0000
Note: *,**,***Significant at po0.1, po0.05 and po0.01
signed, with the exception of the estimation of the hours you work. Surprisingly, this Employment
latter results contrast with those of tenure, which may be equated with the spell status and job
duration of self-employment in this instance. Here, most of the coefficients are
negatively signed. Finally, the possibility of the some scarring effects from satisfaction
unemployment (cf. Arulampalam et al., 2001) on satisfaction with some job aspects
are apparent for this sub-population, notably in the context of pay and job security.
Perhaps this is indicative of the transition into self-employment not being preceded by 207
well-formed plans (Henley, 2007) or that the motivating aspirations are indeed not
pecuniary (e.g. Benz and Frey, 2008a).
However, although these patterns of the nature of the relationship between some of
the independent variables and degrees of satisfaction with the job aspect in question
are discernible from the estimation results, it should be re-emphasised that there is
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

little that is statistically significant in them.


The objective of this sub-section is to address the first research question: is job
satisfaction different for the self-employed with no employees and the self-employed
with employees?, categories which Taylor (2004) describes as the sole proprietors and
job creators, respectively.
In the estimation results reported in columns 2 and 3 of Tables III-XII, the self-
employed with no employees constitutes the reference category. Consequently, the
results reported for the variable self-employed with employees establishes whether
or not there is a difference, and whether this difference is statistically significant.
Table XIII collates the results. There are statistically significant differences in three
instances: the hours you work (where those with employees are less likely to be
satisfied); fringe benefits (where those with employees are more likely to be satisfied);
and the training provided (where those with employees are more likely to be
satisfied). In all, the sign of the coefficient on the variable self-employed with employees
is positive for seven of the nine job aspects. The second negatively signed job aspect
is the amount of work.

Sign of the Level of statistical


Job aspect coefficient significance

Pay
Job security
The hours you work  **
Fringe benefits ***
The amount of work 
The variety of the work
The training provided **
The opportunity to use your abilities
Being able to use your own initiativea
The work itself
Job satisfaction overall (without the supplementary Table XIII.
variables)  The sign and level of
Job satisfaction overall (with the supplementary significance of the variable
variables) self-employed with
employees relative to the
Notes: aRefer back to the footnote in Table XI, with respect to the quality of the original estimation. reference category
**,***Significant at po0.05 and po0.01 variable self-employed
Source: Extracted from previous Tables III-XII without employees
EBHRM For the nine discrete job aspects examined, therefore, there are differences in job
1,2 satisfaction between the self-employed with no employees and those with employees,
with the latter tending to be the more likely to be satisfied. Some of these differences
are statistically significant, but not, notably, for the three process-related job aspects.

5.2 The waged workers


208 The results of the ordinal logit estimations for the self-employed are presented in
columns 4 and 5 of Tables III-XII.
Throughout, the cutpoints are in accordance with expectations, although only a
minority are statistically significant. The F-statistic for the estimated model as a whole
is statistically significant in all ten estimations. The skewness of the responses away
from dissatisfied, as reported in Table II, remains. For this sub-population, however,
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

there is a sufficiency in the number of observations to compensate for this.


Proportionately more of the coefficients of the individual variables in the estimations
are statistically significant. As a consequence, patterns of the nature of the relationship
between some of these dependent variables and degrees of satisfaction with the job
aspect in question have more substance.
Constituting as they do the vast majority of the employed workforce, it is perhaps
not unexpected that the results of these ten estimations for waged workers tend to
substantiate most of the earlier research about the determinants of job satisfaction (e.g.
Clark, 1996; Gazioglu and Tansel, 2006; Jones et al., 2009; Sousa-Poza and Sousa-Poza,
2003; Vila and Garcia-Mora, 2005). Males are more likely to be dissatisfied than females.
This is apparent in the results for the following job aspects: job security; the variety
of the work; the training provided; the opportunity to use your abilities; and the
work itself. By contrast, males are more likely to be satisfied with fringe benefits and
the amount of work. Relative to those with no qualifications, those with different levels
of qualifications are more likely to be dissatisfied. This is apparent in the results for the
following job aspects: job security; the amount of work; the opportunity to use
your abilities; and the work itself. Only in the estimation of the job aspect the variety
of the work is there positively signed coefficients associated with the higher levels
of highest qualification held. These pertain only to levels 3 and 4/5, and neither result
is statistically significant. Relative to those who have received no training, those who
have received training are more likely to be satisfied. The signs of the coefficient of the
training variable are positive throughout and statistically significant on eight occasions.
One novel result relates to those who have had a recent experience of unemployment.
The signs of the coefficients of this variable are uniformly negative for all ten job aspect
estimations and statistically significant in eight of these. In addition to having a negative
impact on future earnings and employment probabilities as studies suggest, therefore,
unemployment appears to have manifold scarring effects in the context of job satisfaction.
The objective of this sub-section is to address the second research question: is job
satisfaction different for managers employed in smaller establishments and managers
employed in larger establishments?
The results for the dummy variables denoting the two categories of managers,
relative to the reference waged work category of those without managerial status, are
presented in columns 4 and 5 of Tables III-XII. To establish difference between the two
managerial categories, Wald tests of the hypothesis that the coefficient of the variable
a manager in an establishment with less than 25 employees equals the coefficient
of the variable a manager in an establishment with more than 24 employees are
undertaken. The results of these tests are reported in Table XIV.
Equality (at po0.1) i.e. no difference is established for four job aspects, namely, the Employment
variety of the work; the opportunity to use your abilities; being able to use your status and job
own initiative; and the work itself. Notably, all three process-related job aspects are
included in this list. This outcome, therefore, contrasts with that reported in Benz and Frey satisfaction
(2008a). For managers, satisfaction with the process-related job aspects identified is not
related to establishment size. Differences, however, do exist for the other six job aspects.
There is an interesting addendum to these results for waged workers who are 209
managers. Again relative to the same reference category, the only job aspect for which
managers are more likely to be satisfied is job security. Although satisfaction with
the ten job aspects may differ or be the same between the two categories of managers,
all managers are more likely to be dissatisfied with the diverse aspects of their jobs,
including those job aspects which reflect process.
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

5.3 Job satisfaction overall


The objective of this sub-section is to address the third research question: for both
the self-employed and those in waged work, is job satisfaction overall correlated with
satisfaction with the aforementioned ten job aspects?
For both sub-populations, two ordinal logit estimations of the dependent variable
satisfied overall are undertaken. The first is a baseline estimation, making use of
the dependent variables previously used to examine satisfaction with the ten job
aspects. The second estimation includes dummy variables denoting satisfaction
with these ten job aspects (re-coded, as described above) as additional independent
variables. The research question, therefore, is addressed by means of inspecting
each of the coefficients of the variables denoting satisfaction with the ten separate
job aspects in the second estimation. A series of Wald tests is also undertaken to
examine the joint significance of these job aspects, from the perspective of product
and process. The results for the estimations of the sub-population of the self-employed
are reported in columns 2, 3, 4 and 5 of Table XV. The corresponding results of
the sub-population of those in waged work are reported in columns 6, 7, 8 and 9 of the
same table.
In the first, baseline estimation of satisfaction overall for the self-employed
sub-population the cutpoints accord with expectations and their coefficients are
statistically significant. The F-statistic for the estimation as a whole is also statistically

Job aspect F-statistic Prob4F

Pay 0.89 0.3468


Job security 1.12 0.2903
The hours you work 1.14 0.2864
Fringe benefits 2.00 0.1573
The amount of work 1.50 0.2208 Table XIV.
The variety of the work 12.29 0.0005 Wald test of the
The training provided 1.00 0.3167 hypothesis: the coefficient
The opportunity to use your abilities 3.06 0.0805 of the variable a manager
Being able to use your own initiative 2.90 0.0889 in an establishment with
The work itself 5.33 0.0210 o25 employees equals
Job satisfaction overall (without the supplementary variables) 2.84 0.0921 the coefficient of the
Job satisfaction overall (with the supplementary variables) 0.54 0.4627 variable a manager in an
establishment with 424
Note: Ordinal logit estimation results of Tables III-XIII employees
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

1,2

210

Table XV.
EBHRM

Ordinal logits:
satisfied overall
Self-employed Waged work
Coefficient Linearized Coefficient Linearized Coefficient Linearized Coefficient Linearized
Variable (1) (2) SE (3) (4) SE (5) (6) SE (7) (8) SE (9)

Male 0.3155 0.6851 1.3462* 0.7171 0.1006 0.1219 0.1569 0.1515


Age, in years, in logs 2.2129 1.9435 0.5286 3.1951 1.0098** 0.4137 1.7235*** 0.5073
Married (or living as) 0.2952 0.4728 0.5745 0.6098 0.1447 0.1285 0.1982 0.1488
With financially dependent children aged o16 0.6585 0.4320 0.6276 0.8205 0.1039 0.1141 0.1549 0.1404
White 0.4384 0.9027 0.6250 0.9473 0.0453 0.2076 0.1064 0.2823
Has no qualifications (the reference category)
Has level 1 as highest qualification 1.1396 1.0501 0.2284 1.4913 0.4723* 0.2502 0.5290 0.3475
Has level 2 as highest qualification 0.0797 0.7658 0.2202 1.1276 0.4679** 0.2038 0.7121*** 0.2678
Has level 3 as highest qualification 1.2136 0.8260 0.4423 1.2643 0.4343** 0.2083 0.5604** 0.2805
Has level 4 or 5 as highest qualification 0.0025 0.8926 1.6475 1.2749 0.6443*** 0.2101 0.8161*** 0.2645
Holds only 1 job 0.5311 0.6339 1.2401 1.1881 0.0268 0.2580 0.0584 0.3412
Tenure, in years, in logs 0.0493 0.1729 0.0119 0.2787 0.0351 0.0493 0.0332 0.0607
Number of hours usually worked each week, in logs 0.9912 0.6136 0.5722 1.0762 0.1504 0.2043 0.3598 0.2611
Works full time 0.4520 0.9766 1.9798 1.2927 0.0997 0.2254 0.0365 0.2915
Has received training for current job 0.3347 0.4555 1.4620** 0.6283 0.5330*** 0.1129 0.1515 0.1431
Number of years in the labour market, in logs 0.8430 0.7640 0.7034 1.0212 0.2623* 0.1585 0.5387*** 0.1857
Has experienced a spell of unemployment in last 2/3/4/5 years 0.3394 0.6601 1.6386 1.0226 0.6435*** 0.1793 0.3593* 0.2179
In waged work as employee (the reference category for those in
waged work)
Self-employed with no employees (the reference category for
those who are self-employed)
Self-employed with employees 0.4521 0.4094 0.9409 0.5918
In waged work as a manager in an establishment with o25
employees 0.3174* 0.1915 0.0243 0.2229
In waged work as a manager in an establishment with 424
employees 0.5319*** 0.1685 0.0867 0.1994
Satisfied with Pay 1.2122** 0.5875 0.8071*** 0.1431
Satisfied with Job Security 1.7198** 0.7076 0.7178*** 0.1612
Satisfied with Hours of Work 0.6313 0.7034 0.9146*** 0.1508
(continued)
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

Self-employed Waged work


Coefficient Linearized Coefficient Linearized Coefficient Linearized Coefficient Linearized
Variable (1) (2) SE (3) (4) SE (5) (6) SE (7) (8) SE (9)

Satisfied with Fringe Benefits 1.9740 1.3638 0.1052 0.1506


Satisfied with Amount of Work 2.2974*** 0.7307 0.4015** 0.1627
Satisfied with Variety in the Work 1.6515 1.0560 0.5128*** 0.1762
Satisfied with the training provided 1.5116* 0.8523 0.5389*** 0.1366
Satisfied with the Opportunity to use of your abilities 1.0474 0.7845 0.8707*** 0.1764
Satisfied with being able to use your own initiative 2.2518 1.2104 0.5485*** 0.1916
Satisfied with the work itself 3.9515*** 0.9125 1.9687*** 0.1755
Cut1 11.8976** 5.4456 1.4737 12.0169 0.5302 1.4417 5.2656*** 1.8145
Cut 2 11.187** 5.5071 3.0366 11.998 0.1162 1.4400 6.3709*** 1.8099
Number of observations 6,187 6,187 6,187 6,187
Sub-population number of observations 718 718 5,469 5,469
F-statistic (17, 6,170) 1.46 (27, 6160) 5.43 (18, 6169) 4.49 (28, 6159) 33.98
Prob4F 0.0980 0.0000 0.0000 0.0000
Notes: Results of Wald tests. For the self-employed: for the joint significance of the supplementary product variables: F(7, 6180) 2.70: Prob4F 0.0086;
for the joint significance of the supplementary process variables: F(3, 6184) 9.23: Prob4F 0.0000; for the joint significance of both the product and
process variables: F(10, 6177) 6.95: Prob4F 0.0000. For the waged worked: for the joint significance of the supplementary product variables:
F(7, 6180) 27.58: Prob4F 0.0000; for the joint significance of the supplementary process variables: F(3, 6184) 81.27: Prob4F 0.0000; for the joint
significance of both the product and process variables: F(10, 6177) 81.66: Prob4F 0.0000. *,**,***Significant at po0.1, po0.05 and po0.01
satisfaction
status and job

Table XV.
Employment

211
EBHRM significant. None of the independent variables is statistically significant as tended to be
1,2 the case in the previous estimations of the ten job aspects for this sub-population and,
generally, the signs of the coefficients of the variables remain the same. The coefficient of
the variable identifying self-employed with employees is negative.
In the second estimation, again the cutpoints accord with expectations, although
neither coefficient is statistically significant. The F-statistic for the estimation as a
212 whole is statistically significant. Two of the variables which featured in the baseline
estimation are now statistically significant: male and having received training.
Further, the coefficients of the qualifications variables are now negatively signed.
The coefficient of the variable identifying self-employed with employees is again
negative. With respect to the additional variables denoting satisfaction with the ten job
aspects, eight of the respective coefficients are positively signed, of which five are
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

statistically significant. The two job aspects for which the coefficients are negatively
signed are: satisfied with the hours of work and satisfied with variety in the work.
Neither is statistically significant. The results of the Wald tests show that the seven
product-related variables, the three process-related variables and the ten job aspect
variables in all are each jointly significant.
To address the third question from the perspective of the self-employed sub-
population, therefore, there is a predominantly positive and sometimes statistically
significant correlation between an individuals job satisfaction overall and satisfaction
with the ten different job aspects. Any distinction between product and process job
aspects would appear to be of no consequence in this context.
In the first baseline estimation of satisfaction overall for the waged work sub-
population the cutpoints accord with expectations, although neither coefficient is
statistically significant. The F-statistic for the estimation as a whole is statistically
significant. Several of the coefficients of original independent variables are statistically
significant as tended to be the case in the estimations of the ten job aspects for this
sub-population. The coefficients of the two variables identifying having managerial
status are both negatively signed, and both are statistically significant.
In the second estimation, again the cutpoints accord with expectation and this
time both are statistically significant. The F-statistic for the equation as a whole is
also statistically significant. Once more, several of the coefficients of the original
independent variables are statistically significant. Neither of the coefficients denoting
having managerial status is statistically significant on this occasion, however, and the
sign of the coefficient denoting a manager in an establishment with o25 employees is
now positively signed. With respect to the additional variables denoting satisfaction
with the ten job aspects in question, all of the coefficients are positively signed and nine
are statistically significant. Again, the results of the Wald tests show that the seven
product-related variables, the three process-related variables and the ten job aspect
variables in all are each jointly significant.
To address the third question from the perspective of the waged work sub-
population, therefore, there is a positive and predominantly statistically significant
correlation between an individuals job satisfaction overall and satisfaction with the
ten different job aspects. The distinction between product and process job aspects is
again of no consequence in this context.

6. Conclusion
Using data from the 2006 Skills survey, this paper focused upon differences within
two employment status groups namely the self-employed and those in waged work.
It addressed three questions. First: for a set of ten job aspects, was job satisfaction Employment
different for the self-employed with no employees and the self-employed with status and job
employees? Second: for these same job aspects, was job satisfaction different for
managers in smaller establishments and managers in larger establishments? Third: satisfaction
for both the self-employed and those in waged work, was job satisfaction overall
correlated with satisfaction with these job aspects?
To address these three questions in turn: first, there were differences in job 213
satisfaction between the self-employed with no employees and those with employees.
The latter tended to be more likely to be satisfied in that the sign of the coefficient on
the variable self-employed with employees was positively signed for seven of the
nine job aspects examined. These included the two process-related job aspects of the
opportunity to use your abilities and the work itself. Three results were statistically
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

significant: fringe benefits, where those with employees were more likely to be
satisfied; the training provided, where those with employees were more likely to
be satisfied; and the hours you work, where those with employees were less likely
to be satisfied. These results suggest that although a search for procedural utility
may motivate the transition into self-employment, relative to his employment status
equivalent, the job creator, the sole proprietor may not necessarily derive the job
satisfaction sought from this transition. In contrast, the job creator would appear to
obtain the advantages and disadvantages, which may accrue to the owner-manager
of a small business: the fringes, the training opportunities, but also the long hours.
Second, there were differences between managers in smaller establishment and
managers in larger establishments for six of the ten job aspects examined. However,
there were no differences for the following: the variety of the work; the opportunity
to use your abilities; being able to use your own initiative; and the work itself.
Notably, and an outcome which contrasts with the job satisfaction-establishment size
conclusion of Benz and Frey, all three process-related job aspects were included in this
list of no difference. A secondary but nonetheless important finding to emanate from
this examination of those in waged work was that managers irrespective of the size of
establishment with which they worked were less likely to be satisfied than employees
who had no managerial responsibilities across nine of the ten job aspects examined.
Hierarchy has a literal meaning in addition to the interpretation given to it by
Williamson. These results suggest that the relationship between status differences
within the waged work and differences in job satisfaction warrants further investigation.
Finally, for the self-employed there was a predominantly positive and sometimes
statistically significant correlation between an individuals job satisfaction overall and
satisfaction with the ten job aspects. For the waged worker, there was a uniformly
positive and predominantly statistically significant correlation between an individuals
job satisfaction overall and satisfaction with the ten job aspects. Benz and Freys
distinction between product and process job aspects, therefore, was of no consequence
in either context.
The focus of this paper has been upon differences in job satisfaction within
employment status groups i.e. within the self-employed and within those in waged
work not differences across these employment status groups. It has been the latter
which has tended to dominate research investigations. One by-product of this paper,
however, has been the insight that the self-employed and those in waged work cannot
be assumed to constitute homogeneous groups. Consequently, when future research
seeks to examine the manner in which job satisfaction may differ across employment
status groups, these groups cannot be treated as mere dichotomous dummy variables.
EBHRM Notes
1,2 1. Job satisfaction, therefore, needs to be distinguished from subjective well-being, an inclusive
term referring to life satisfaction, happiness, etc. (cf. Bowling et al., 2010; Frey and Stutzer,
2002; Layard, 2005). Although job satisfaction may well impact upon subjective well-being
(and vice versa) they are different concepts (Andersson, 2008; Binder and Coad, 2013;
Helliwell and Huang, 2010).
214 2. The study of entry to and exit from self-employment by Taylor (2004) and the study of exit from
self-employment by Georgellis et al. (2007) provide some empirical support for this model.
3. Illustrative additional references which apply the concept of procedural utility although
not all in the context of comparing job satisfaction across employment status groups would
include the following: Benz and Frey (2008b), Block and Koellinger (2009) and Fuchs-
Schundeln (2009).
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

4. There is no panel element in these surveys.


5. The qualification mapping follows convention and is, broadly: level 4 or level 5 equates with
degrees and postgraduate degrees and their professional and vocational equivalents; level 3
equates with sub degree academic qualifications and their professional and vocational
equivalents, usually gained in further education, post 18 years of age; level 2 equates with higher
school academic qualifications obtained, usually, at the age of 18/19 and their professional and
vocational equivalents; and level 1 equates with lower school academic qualifications obtained,
again usually, at the age of 16, the end of mandatory formal education in the UK.
6. A level of significance of ( po0.1) is the criterion of statistical significance used throughout
the paper.

References
Ajayi-Obe, O. and Parker, S. (2005), The changing nature of work among the self-employed in
the 1990s: evidence from Britain, Journal of Labor Research, Vol. XXVI No. 3, pp. 501-517.
Andersson, P. (2008), Happiness and health: well-being among the self-employed, Journal of
Socio-Economics, Vol. 37 No. 1, pp. 213-236.
Arulampalam, W., Gregg, P. and Gregory, M. (2001), Unemployment scarring, Economic
Journal, Vol. 111 No. 475, pp. 577-584.
Bazen, S. (2011), Practical Econometrics: Econometrics Methods for Labour Economists, Oxford
University Press, Oxford.
Benz, M. and Frey, B.S. (2008a), Being independent is a great thing: subjective evaluations of
self-employment and hierarchy, Economica, Vol. 75 No. 298, pp. 362-383.
Benz, M. and Frey, B.S. (2008b), The value of doing what you like: evidence from the
self-employed in 23 countries, Journal of Economic Behaviour & Organisation, Vol. 68
Nos 3-4, pp. 445-455.
Binder, M. and Coad, A. (2013), Life satisfaction and self-employment: a matching approach,
Small Business Economics, Vol. 40 No. 4, pp. 1009-1033.
Blanchflower, D. (2000), Self-employment in OECD countries, Labour Economics, Vol. 7 No. 5,
pp. 471-505.
Blanchflower, D. and Oswald, A. (1998), What makes an entrepreneur?, Journal of Labor
Economics, Vol. 14 No. 1, pp. 571-602.
Block, J. and Koellinger, P. (2009), I cant get no satisfaction necessity entrepreneurship and
procedural utility, Kyklos, Vol. 62 No. 2, pp. 191-209.
Booth, A.L. and van Ours, J.C. (2008), Job satisfaction and family happiness: the part time
puzzle, Economic Journal, Vol. 118 No. 526, pp. F77-F99.
Bowling, N.A., Eschleman, K.J. and Wang, Q. (2010), A meta-analytic examination of the Employment
relationship between job satisfaction and subjective well-being, Journal of Occupational
and Organisational Psychology, Vol. 83 No. 4, pp. 915-934. status and job
Bryson, A., Cappellari, L. and Lucifora, C. (2004), Does union membership really reduce job satisfaction
satisfaction?, British Journal of Industrial Relations, Vol. 42 No. 3, pp. 439-459.
Bryson, A., Cappellari, L. and Lucifora, C. (2010), Why so unhappy? The effects of unionisation
on job satisfaction, Oxford Bulletin of Economics and Statistics, Vol. 72 No. 3, pp. 357-380. 215
Cameron, A.C. and Trivedi, P.K. (2009), Microeconometrics Using Stata, Stata Press, College
Station, TX.
Clark, A.E. (1996), Job satisfaction in Britain, British Journal of Industrial Relations, Vol. 34
No. 2, pp. 189-217.
Clark, A.E. (1997), Job satisfaction and gender: why are women so happy at work?, Labour
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

Economics, Vol. 4 No. 4, pp. 341-372.


Clark, A.E. and Oswald, A.J. (1994), Unhappiness and unemployment, Economic Journal,
Vol. 104 No. 424, pp. 648-659.
Clark, A.E., Kristensen, N. and Westergard-Nielsen, N. (2009), Job satisfaction and co-worker
wages: status or signal?, Economic Journal, Vol. 119 No. 536, pp. 430-447.
Felstead, A., Gallie, D., Green, F. and Zhou, Y. (2007), Skills at Work, 1986-2006, ESRC Centre on
Skills, Knowledge and Organisational Performance (SCOPE), Cardiff and Oxford.
Freeman, R.B. (1978), Job satisfaction as an economic variable, American Economic Review,
Vol. 68 No. 2, pp. 135-141.
Frey, B.S. and Stutzer, A. (2002), What can economists learn from happiness research?, Journal
of Economic Literature, Vol. XL No. 2, pp. 402-435.
Frey, B.S. and Stutzer, A. (2005), Beyond outcomes: measuring procedural utility, Oxford
Economic Papers, Vol. 57 No. 1, pp. 90-111.
Fuchs-Schundeln, N. (2009), On preferences for being self-employed, Journal of Economic
Behaviour & Organisation, Vol. 71 No. 2, pp. 162-171.
Gallie, D., White, M., Cheng, Y. and Tomlinson, M. (1998), Restructuring the Employment
Relationship, Oxford University Press, Oxford.
Gazioglu, S. and Tansel, A. (2006), Job satisfaction in Britain: individual and job related factors,
Applied Economics, Vol. 38 No. 10, pp. 1163-1171.
Georgellis, Y., Sessions, J. and Tsitsianis, N. (2007), Pecuniary and non-pecuniary aspects of
self-employment survival, The Quarterly Review of Economics and Finance, Vol. 47 No. 1,
pp. 94-112.
Green, C. and Heywood, J.S. (2008), Does performance pay increase job satisfaction?,
Economica, Vol. 75, pp. 710-728.
Hamilton, B.H. (2000), Does entrepreneurship pay? An empirical analysis of the returns to
self-employment, Journal of Political Economy, Vol. 108 No. 3, pp. 604-631.
Hammermesh, D.S. (1977), Economic aspects of job satisfaction, in Ashenfelter, O.C. and Oates, W.E.
(Eds), Essays in Labor Market Analysis, Wiley, New York, NY, pp. 53-72.
Heckman, J. (1990), Varieties of selection bias, American Economic Review, Vol. 80 No. 2,
pp. 313-318.
Helliwell, J.F. and Huang, H. (2010), Hows the job? Well-being and social capital in the
workplace, Industrial and Labor Relations Review, Vol. 63 No. 2, pp. 205-227.
Henley, A. (2007), Entrepreneurial aspiration and transition into self-employment: some
evidence from British longitudinal data, Entrepreneurship and Regional Development,
Vol. 19 No. 3, pp. 253-280.
EBHRM Idson, T. (1990), Firm size, job satisfaction and the structure of work, Applied Economics,
Vol. 22 No. 8, pp. 1007-1018.
1,2
Jones, M.K., Jones, J.J., Latreille, P. and Sloane, P.J. (2009), Training, job satisfaction and
workplace performance in Britain, Labour, Vol. 23 No. S1, pp. 139-175.
Kaiser, L.C. (2007), Gender-job satisfaction differences across Europe: an indicator of labour
market modernisation, International Journal of Manpower, Vol. 28 No. 1, pp. 75-94.
216 Layard, R. (2005), Happiness: Lessons from a New Science, Penguin Books, London.
Long, J.S. and Freese, J. (2006), Regression Models for Categorical Dependent Variables Using
Stata, Stata Press, College Station, TX.
Parker, S. (2004), The Economics of Self-Employment and Entrepreneurship, Cambridge
University Press, Cambridge.
Downloaded by Universiti Kebangsaan Malaysia At 16:12 14 January 2016 (PT)

Prottas, D. (2008), Do the self-employed value autonomy more than employees?, Career
Development International, Vol. 13 No. 1, pp. 33-45.
Rose, M. (2005), Job satisfaction in Britain: coping with complexity, British Journal of Industrial
Relations, Vol. 43 No. 3, pp. 455-467.
Schmidt, S.W. (2007), The Relationship between satisfaction with workplace training and
overall job satisfaction, Human Resource Development Quarterly, Vol. 18 No. 4,
pp. 481-498.
Sousa-Poza, A. and Sousa-Poza, A.A. (2003), Gender differences in job satisfaction in Great
Britain, 1991-2000: permanent or transitory?, Applied Economics Letters, Vol. 10 No. 11,
pp. 691-694.
Spencer, D. (2009), The Political Economy of Work, Routledge, London.
Taylor, M. (2004), Self-employment in Britain: when, who and why?, Swedish Economic Policy
Review, Vol. 11 No. 1, pp. 139-173.
Taylor, M.T. (1999), Survival of the fittest: an analysis of self-employment duration in Britain,
Economic Journal, Vol. 109 No. 454, pp. C140-C155.
Vila, L.E. and Garcia-Mora, B. (2005), Education and the determinants of job satisfaction,
Education Economics, Vol. 13 No. 4, pp. 409-425.
Williamson, O.E. (1975), Markets and Hierarchies: Analysis and Anti-Trust Implications, Free
Press, New York, NY.
Williamson, O.E. (1985), The Economic Institutions of Capitalism: Firms, Markets and Relational
Contracting, Free Press, New York, NY.

About the author


Dr John Sutherland is a Visiting Researcher at the Scottish Centre for Employment Research
(SCER) in the Department of Human Resource Management at the University of Strathclyde,
Glasgow. Between 2007 and 2012 he was an Honorary Research Fellow at the Centre for Public
Policy for Regions (CPPR), at the University of Glasgow. He is an Economist with research
interests in contemporary issues, problems and policies which relate to the world of work. John
Sutherland can be contacted at: john.sutherland@strath.ac.uk

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com


Or visit our web site for further details: www.emeraldinsight.com/reprints

S-ar putea să vă placă și