ArticlePDF Available

Life satisfaction and self-employment : A matching approach


Abstract and Figures

Despite lower incomes, the self-employed consistently report higher satisfaction with their jobs. But are self-employed individuals also happier, more satis ed with their lives as a whole? High job satisfaction might cause them to neglect other important domains of life, such that the ful lling job crowds out other pleasures, leaving the individual on the whole not happier than others. Moreover, self-employment is often chosen to escape unemployment, not for the associated autonomy that seems to account for the high job satisfaction. We apply matching estimators that allow us to better take into account the above-mentioned considerations and construct an appropriate control group. Using the BHPS data set that comprises a large nationally representative sample of the British populace, we nd that in- dividuals who move from regular employment into self-employment experience an increase in life satisfaction (up to two years later), while individuals moving from unemployment to self-employment are not more satis ed than their counterparts moving from unemployment to regular employment. We argue that these groups correspond to \opportunity" and \ne- cessity" entrepreneurship, respectively. These ndings are robust with regard to di erent measures of subjective well-being as well as choice of matching variables, and also robustness exercises involving \simulated confounders".
Content may be subject to copyright.
Max Planck Institute of Economics
Evolutionary Economics Group
Kahlaische Str. 10
07745 Jena, Germany
Fax: ++49-3641-686868
The Papers on Economics and Evolution are edited by the
Evolutionary Economics Group, MPI Jena. For editorial correspondence,
please contact:
ISSN 1430-4716
by the author
# 1020
Life satisfaction and self-employment:
A matching approach
Martin Binder
Alex Coad
Life satisfaction and self-employment: A matching approachI
Martin Binder,a, Alex Coada,b
aMax Planck Institute of Economics, Evolutionary Economics Group, Kahlaische Str.10, 07745 Jena,
bSPRU, University of Sussex, Falmer, Brighton, BN1 9QE, UK
Despite lower incomes, the self-employed consistently report higher satisfaction with their
jobs. But are self-employed individuals also happier, more satisfied with their lives as a
whole? High job satisfaction might cause them to neglect other important domains of life,
such that the fulfilling job crowds out other pleasures, leaving the individual on the whole not
happier than others. Moreover, self-employment is often chosen to escape unemployment,
not for the associated autonomy that seems to account for the high job satisfaction. We
apply matching estimators that allow us to better take into account the above-mentioned
considerations and construct an appropriate control group. Using the BHPS data set that
comprises a large nationally representative sample of the British populace, we find that in-
dividuals who move from regular employment into self-employment experience an increase
in life satisfaction (up to two years later), while individuals moving from unemployment to
self-employment are not more satisfied than their counterparts moving from unemployment
to regular employment. We argue that these groups correspond to “opportunity” and “ne-
cessity” entrepreneurship, respectively. These findings are robust with regard to different
measures of subjective well-being as well as choice of matching variables, and also robustness
exercises involving “simulated confounders”.
Key words: self-employment, happiness, matching estimators, unemployment, BHPS,
necessity entrepreneurship
JEL-classification: J24, J28, C21
INo individuals were mistreated during our matching procedures. We are grateful to Rob Byrne, Jan
Fagerberg, Steffen K¨unn, Ben Martin, Maria Savona, Josh Siepel, Jagannadha Pawan Tamvada, Dagmara
Wechowska, Ulrich Witt and seminar participants at SPRU (University of Sussex), and also to Bram Tim-
mermans for some interesting suggestions, comments etm. The authors are grateful for having been granted
access to the BHPS data set, which was made available through the ESRC Data Archive. The data were
originally collected by the ESRC Research Centre on Micro-Social Change at the University of Essex (now
incorporated within the Institute for Social and Economic Research). Neither the original collectors of the
data nor the Archive bear any responsibility for the analyses or interpretations presented here. Errors are
Corresponding author
Email address: (Martin Binder)
December 21, 2010
1. Introduction
Self-employment is something highly valued by individuals for the self-determination and
autonomy it entails (Benz and Frey, 2008a). Being one’s own boss has been shown to increase
individuals’ satisfaction with their job, despite drawbacks such as initially often decreased
incomes through self-employment (Hamilton, 2000). But are self-employed individuals hap-
pier in a broad sense not only related to their job? Does their attraction to self-employment
possibly crowd out other pleasures of life, leading to overall unhappy “workaholics”? And
how happy are self-employed that are forced to go into self-employment to escape unemploy-
ment? These are questions that have only incompletely been addressed in the literature so
far (e.g., Andersson, 2008).
The aim of our paper is thus to assess the satisfaction of the self-employed with their life
in general. Our contribution to the literature is fivefold. First, we focus our analysis on the
satisfaction with life of the self-employed. Most of the previous literature on the other hand
has focused on the relationship between self-employment and the narrower concept of job
satisfaction. By making life satisfaction the dependent variable, which implicitly considers
the trade-off between total income and job satisfaction, we are more interested in the more
global well-being of the self-employed (it is a broader indicator of “total utility”).
Second, we use a large, nationally-representative dataset with a relatively large panel
dimension, where annual responses are recorded for the period 1996-2006. This in itself is
a useful contribution because early work on the topic has often focused on small samples
(Brockhaus, 1980; Cromie and Hayes, 1991), and even in more recent work the data analyzed
is often merely cross-sectional data (Hyytinen and Ruuskanen, 2007; Block and Koellinger,
2009) or data with a limited panel dimension (Bradley and Roberts, 2004; Andersson, 2008).
Third, we apply an appropriate empirical methodology for obtaining estimates of the
causal impact of self-employment on satisfaction, in a context where a comparison of the
treatment group to the control group is not trivial. In recent work, Schjoedt and Shaver
(2007) cast doubt on previous results on the basis of difference-of-means tests relating group
averages for self-employed versus employed individuals (without controlling for other influ-
ences). We argue that this methodology is flawed, because self-employed individuals differ
from other individuals in many ways (see our Table 1), and these differences between the
different employment categories must be controlled for. Multivariate regressions can be a
useful tool here, and have been widely used in the related literature, but also present draw-
backs compared to the matching estimators applied in this paper. Although the researcher is
presumably interested in comparing individuals that have the same values for all covariates,
multivariate regression modelling obscures information on the distribution of covariates in
the treatment versus control groups. Unless there is substantial overlap in the two covariate
distributions, multivariate regression estimates rely heavily on extrapolation, and can there-
fore be misleading (Imbens, 2004; Ichino et al., 2008, p. 312-13). Matching estimators are
preferable because more care is taken to establish an appropriate control group. Another
advantage of matching methods is that they require no assumptions on functional forms
(Hussinger, 2008, p. 730). To our knowledge, however, matching estimators have so far not
been used in the present context.
Fourth, we distinguish between opportunity and necessity entrepreneurship in our anal-
ysis of self-employment and life satisfaction. This distinction stands to be one of the most
important causes for heterogeneity in the group of self-employed, since the former are going
into self-employment voluntarily to pursue entrepreneurial opportunities, while the latter are
forced into self-employment to escape unemployment. As important as this distinction a
priori seems, few studies account for it when analysing the impact of self-employment on
individuals’ satisfaction (but see, e.g., Block and Koellinger, 2009).
A fifth feature of our paper is that we focus on the years of transition into self-employment.
Most previous studies into job satisfaction and the self-employment decision have tended
to pool together new entrants into self-employment and senior self-employed individuals,
implicitly grouping together individuals who have spent greatly different periods of time in
self-employment, an approach which has recently been criticized (Bradley and Roberts, 2004).
In this paper, we focus on the periods of transition into self-employment, thereby focusing on
nascent entrepreneurship (as opposed to individuals who have been self-employed for many
The paper is structured as follows. Section 2 gives the literature background on different
employment types (inter alia self-employment) and their effects on happiness. Section 3
introduces our matching estimators in more technical detail. We then present our data set in
Section 4 and discuss the results of our matching methodology in Section 5. The robustness
of our findings is explored in a variety of ways. Section 6 concludes.
2. Literature review
Work is an important facet of human life and it has strong effects on individuals’ sat-
isfaction with life or happiness (which we will use synonymously here). This relationship
is especially strong and clear for unemployment, which makes individuals unhappier than
can explained by only the effect of loss of income. Effects are consistently negative across a
wide range of studies (e.g., Clark and Oswald, 1994; Di Tella et al., 2001; Helliwell, 2003).
Moreover, males are more strongly affected by unemployment and there seems to be only
incomplete adaptation to continued unemployment for them (Clark, 2003; Lucas et al., 2004).
These effects are robust in panel studies that control for selection effects, i.e. the relation-
ship is not due to unhappy individuals that self-select into unemployment (Winkelmann and
Winkelmann, 1998; Lucas et al., 2004; Oswald and Powdthavee, 2008).
On the other hand, the relationship between self-employment and happiness is less clear.2
We have “rather robust finding[s] across the nations on which data are available” that self-
employment is related to higher job satisfaction (Blanchflower, 2004), this being the case e.g.
in the US (Blanchflower and Oswald, 1998; Kawaguchi, 2008) and for other OECD countries
(Blanchflower, 2000; Blanchflower et al., 2001). In contrast to this finding, however, one
must also take into account the robust finding that the returns to self-employment are lower,
1Our focus on nascent entrepreneurship bears similarities to some previous work (Bradley and Roberts,
2004; Schjoedt and Shaver, 2007; Fuchs-Schundeln, 2009); see also Andersson (2008) who focuses on changes
between two cross-sections (1991 and 2000).
2Van Praag and Versloot (2007, pp. 375-6) provide a brief overview over some contributions entrepreneur-
ship has on the utility levels of entrepreneurs and their employees.
on average, than those obtained from employment (Hamilton, 2000).3,4The self-employed
generally have lower pay than the employed, but this does not mean that the self-employed
are not interested in financial rewards — in fact, it has been observed that financial success
is the single most important variable associated with start-up satisfaction among a group of
self-employed individuals (Block and Koellinger, 2009). In addition to lower pay, there is also
evidence that the self-employed have longer working weeks than paid employees (Hyytinen
and Ruuskanen, 2007). Interestingly enough, it has even been observed that, among the
self-employed, the number of hours worked for the start-up business is positively correlated
with start-up satisfaction (Block and Koellinger, 2009). Taken together, these results suggest
that the self-employed derive utility from their job (known as “procedural utility”, Benz and
Frey, 2008a,b) that cannot simply be expressed in terms of the “output” (remuneration,
hours worked) associated with their jobs.5
Self-employed individuals obtain satisfaction from leading an independent lifestyle and
“being their own bosses”. Hundley (2001) finds that the self-employed are more satisfied
with their jobs mainly because of greater autonomy, but also because of more flexibility, skill
utilization and, to some extent, higher (perceived) job security. Relatedly, empirical work
has shown that employees have a lower job satisfaction in large firms compared to small
firms (Idson, 1990; Benz and Frey, 2008a), and this can be explained to a large extent by
“procedural” aspects of work such as the nature of the work tasks and the ability to use of
one’s own initiative (Benz and Frey, 2008a).6
Other researchers have found that self-employment can be associated with a dissatisfaction
with previous circumstances. For example, Kawaguchi (2008) observes that job quitting
tends to follow low job satisfaction. Noorderhaven et al. (2004) observe that the levels of
“dissatisfaction with life” observed in a society are positively associated with self-employment
Having a higher job satisfaction, however, does not necessarily translate into self-employed
individuals being overall more satisfied with their lives as a whole. Life satisfaction in itself
is a much more global evaluation of individual’s actual state of being, being influenced not
only by job satisfaction but a complex and interacting web of influences (Binder and Coad,
2010a,b). Since individuals might be able to compensate high achievement in some domains
of life with low achievements otherwise, a high job satisfaction might be counterbalanced
by lower satisfaction in the family domain, or social life more generally, or, as mentioned
3In addition, it has been shown that individuals in small businesses have fewer fringe benefits compared
to their counterparts in large firms (Storey, 1994, Ch. 6).
4Interestingly enough, Cooper and Artz (1995, p. 452) observe that female entrepreneurs had a lower
financial performance than their male counterparts, but that controlling for these differences in income level,
the female entrepreneurs indicated marginally higher levels of satisfaction.
5Cooper and Artz (1995) found that entrepreneurs with initially high expectations for their business
venture performance turned out to be more satisfied than other entrepreneurs, suggesting that these more
satisfied individuals have some more optimistic personality traits that influence their subsequent job satis-
faction. This finding does probably only pertain to those entrepreneurs that create their business out of
opportunity, not to escape unemployment.
6The positive effect of being self-employed on job satisfaction diminishes markedly when taking into
account the heterogeneity of the control group of the employed in terms of the size of the firm they are
working in (Benz and Frey, 2008a, p. 374).
above, in the income domain (etc.). If the satisfying work the self-employed enjoy crowds
out pleasures from other domains of life, the overall life satisfaction of the self-employed
could actually be not as high as one might expect based on their job satisfaction assessment
alone. And indeed, there is scant evidence so far on the relationship between happiness and
self-employment (Andersson, 2008, p. 231): Blanchflower and Oswald (1998) report for cross-
sectional data from the US that young self-employed are happier and in a similar vein Craig
et al. (2007) provide some evidence for this relationship from Australian small businesses.
Looking at European countries, Blanchflower (2004) fails to find overly strong effects of self-
employment on life satisfaction (only for subgroups, self-employment is significantly related
to life satisfaction; and strongly depending on the data set used).
The empirically weak association between happiness and self-employment, however, could
also be explained by a different, methodological phenomenon: it could be due to the fact
that the self-employed are a quite heterogeneous group (Santarelli and Vivarelli, 2007).7It
has been argued that, while some individuals would gladly self-select into self-employment,
others who are forced into self-employment might not appreciate the self-employed lifestyle
(Fuchs-Schundeln, 2009). Vivarelli (1991) writes that entry into self-employment cannot be
seen merely in terms of pull factors such as expected profits (as predicted by the tradi-
tional industrial economics perspective), but there are also important push factors such as
previous unemployment. In other words, in the light of the terminology in Reynolds et al.
(2005), we should distinguish between necessity entrepreneurship (such as the flight from
unemployment) and opportunity entrepreneurship (such as the exploitation of new business
opportunities). Block and Koellinger (2009) distinguish between necessity entrepreneurship
and opportunity entrepreneurship, and observe that necessity entrepreneurs have a lower
average satisfaction with their startup than opportunity entrepreneurs, and also that a long
period of unemployment is negatively related to startup satisfaction.
If one does not make this distinction of types of self-employment, one thus might lump
together widely different individuals in the regression exercise and thus not be able to find any
robust relationship between life satisfaction and self-employment. In order to separate the two
possible explanations for the scant empirical evidence for a positive relationship between the
two variables, we thus differentiate between opportunity and necessity self-employment and
use a regression methodology that is better suited to deal with this individual heterogeneity
than regression techniques usually employed in this context. We now turn to an exposition
of this matching technique.
3. Matching methodology
“How happy would I be if I had not chosen to be self-employed?” To answer this kind
of question, one must consider a counterfactual. The main problem is that if an individual
chooses to be self-employed, then there is no data on exactly what would have happened had
they not chosen to be self-employed.
In the case of a randomized laboratory experiment, such as a clinical trial, an accurate
counterfactual can be established by referring to a control group that was not exposed to
7Another source of heterogeneity might stem from distinguishing entrepreneurship from self-employment,
of which we abstract here.
the treatment of interest. However, establishing a counterfactual is much harder when the
researcher is not dealing with randomized experimental data but instead observational data.
The problem is that individuals are prone to self-select into their preferred employment
category, which implies that comparing individuals from different employment categories
is prone to a selection bias (this is like “comparing apples with oranges”). Conventional
regression analysis is not suitable to dealing with this kind of selection bias. The approach
we take is to carefully match individuals from the treatment group with individuals from
the control group, to obtain more accurate estimates of the counterfactual. By comparing
the treatment group with the control group, we can thus identify the causal effect of the
self-employment decision on happiness. We here understand “causal effect” as defined by
Rubin (1974), viz. “the causal effect of one treatment, E, over another, C, for a particular
unit and an interval of time from tito t2is the difference between what would have happened
at time t2if the unit had been exposed to E initiated at tiand what would have happened
at t2if the unit had been exposed to C initiated at ti” (p. 689).8
We are therefore interested in comparing the outcome (in our case: life satisfaction) Yi(0)
for an individual ifrom the control group, with the outcome Yi(1) for the same individual
after undergoing the treatment of interest (in our case: going into self-employment):
τi=Yi(1) Yi(0) (1)
However, the drawback is that we can never observe both Yi(0) and Yi(1) for the same
individual (Imbens, 2004). One way of dealing with this problem is to estimate the (Popu-
lation) Average Treatment Effect:
τAT E =E(τ) = E[Y(1) Y(0)] (2)
A drawback of this estimate, however, is that the treatment is not intended for everyone,
and that individuals self-select into the treatment group. It would be better to estimate the
Average Treatment effect for the Treated (ATT), which focuses explicitly on the subsample
of individuals that are the most affected by the treatment (i.e. those individuals that actually
did decide to be self-employed):
τAT T =E(τ|D= 1) = E[Y(1)|D= 1] E[Y(0)|D= 1] (3)
where D is the treatment indicator, taking the value 1 if an individual underwent the treat-
ment (i.e. choosing self-employment) and 0 otherwise. Unbiased estimates for E[Y(1)|D= 1]
can be obtained by taking mean values of the outcome variable for self-employed individuals.
Obtaining unbiased estimates for E[Y(0)|D= 1] will be more difficult, however, because we
cannot observe the case of individuals who chose self-employment but are not self-employed.
Since individuals that self-select into self-employment are expected to be different from in-
dividuals who do not, (i.e. E[Y(0)|D= 1] 6=E[Y(0)|D= 0]), it is not possible to calculate
τAT T by simply comparing the outcomes of the self-employed with those of other individuals
who are not self-employed.
8For a more detailed introduction to matching, see the surveys in Imbens (2004) and Caliendo and Kopeinig
To identify the parameter of interest, τAT T , we need to make two further assumptions.
The first assumption is called the “conditional independence assumption (CIA)”, and is
also known as “selection on observables” or “unconfoundedness”. This assumption means
that the potential outcome (life satisfaction) and participation in the treatment (i.e. choice
to enter self-employment) are independent for individuals with the same set of exogenous
characteristics (Almus and Czarnitzki, 2003). Under this assumption, we have:
Y(0), Y (1)D|X(4)
If this first assumption is correct, we can use the fact that E[Y(0)|D= 1, X =x] =
E[Y(0)|D= 0, X =x] to identify τAT T . Under this CIA assumption, all individual charac-
teristics (X) that influence both the treatment assignment and potential outcomes simulta-
neously must be observed by the econometrician. Unobserved variables are not allowed to
influence treatment assignment and potential outcome.
The second assumption is known as “overlap”, or also as “strong ignorability” or the
“common support condition”, and can be expressed as:
0< P (D= 1|X)<1 (5)
This assumption ensures that those individuals with the same characteristics have a posi-
tive probability of being both participants (i.e. choosing self-employment) or nonparticipants
(not choosing self-employment). If the overlap assumption does not hold, then the resulting
estimates can be heavily biased (Heckman et al., 1996).
The first assumption, CIA, is a strong assumption, and it cannot be verified directly.
There are ways in which the robustness of the matching estimator τAT T can be investigated,
although unfortunately most previous work that applies matching estimators has not verified
the robustness of the estimates in a satisfactory way (Caliendo and Kopeinig, 2008). We
examine the robustness of our estimates in Section 5.3 using the procedure described in
Ichino et al. (2008) and Nannicini (2007), which explores the sensitivity of the matching
estimates to simulated confounding variables.
In contrast to the CIA assumption, the overlap assumption is relatively easy to verify,
and in our dataset it is indeed verified.
We use two different matching procedures in this paper and begin our matching analysis
by using the nearest-neighbour matching estimator outlined in Abadie et al. (2004), which
finds the nearest neighbour from the control group for each of the dimensions of X. If we have
many matching covariates X, however, it becomes prohibitively difficult to find good matches
for individuals in all dimensions simultaneously. On the one hand, it has been argued that
omitting important variables can seriously increase bias in the resulting estimates (Heckman
et al., 1997; Dehejia and Wahba, 1999). On the other hand, however, including too many
variables should also be avoided, because it becomes more difficult to find suitable matches,
and the variance of the estimates increases. Caliendo and Kopeinig (2008, p. 39) write that
“there are both reasons for and against including all of the reasonable covariates available”,
and suggest that the choice of matching covariates be undertaken with reference to theory
and previous empirical findings.
One alternative to nearest neighbour matching, that does not suffer from dimensionality
problems when a large number of matching covariates are considered, is propensity score
matching, which matches individuals by collapsing the vector of individual characteristics into
a scalar propensity score. This synthetic propensity score can then used as the single matching
criterion (Almus and Czarnitzki, 2003). Matching according to a propensity score implies
that there is a (data-driven) tradeoff between the different dimensions — one observation
might be matched to another observation that scores higher in one dimension but this is
compensated for by a lower score in another dimension. These sorts of compensation are not
done in nearest neighbour matching.
Propensity score matching relies on the following corollary to Assumption 1:
Y(0), Y (1)D|P(X) (6)
where P(X) is the propensity score given the observed covariates X. Using both types
of matching analysis helps to ascertain the robustness of the chosen approach.
4. Data
For our analysis, we use a data set that is not primarily concerned with entrepreneurs but
which offers a rich variety of employment status information for a representative sample of
the British populace. The British Household Panel Survey (BHPS) is a longitudinal survey
of private households in Great Britain that contains information on various areas of the
respondents’ lives, ranging from income to household consumption, education, health, but
also social and political values.9
We are using unbalanced panel data from 1996 to 2006 (waves f to p) and have a total of
76,752 observations after cleaning the panel: during the time period, two waves had to be
deleted since not all of our variables have been asked in them, leaving us with a total of 9
waves. We will now discuss the indicators chosen for our analysis as well as characteristics
according to which we later match our individuals. While our main analysis will focus on
the matching methodology described in Section 3, a benchmark will be a set of preliminary
regressions, where we analyze the impact of different job situations on life satisfaction, job
satisfaction and mental well-being.
To examine an individual’s life satisfaction, we use the BHPS’s life satisfaction question.
It covers the response to the question “How dissatisfied or satisfied are you with your life over-
all?” It is effectively tracking an individual’s life satisfaction ordinally on a seven point Likert
scale, ranging from “not satisfied at all” (1) to “completely satisfied” (7). Comparatively
more studies on the BHPS center on the GHQ-12 measure of mental well-being, but recent
work took up using the life satisfaction question too (Binder and Coad, 2010b; Clark and
Georgellis, 2010; Powdthavee, 2009). Nevertheless, to explore the robustness of our findings,
we also use the broader GHQ-12 “mental well-being” variable, which is more encompassing
9The survey is undertaken by the ESRC UK Longitudinal Studies Centre with the Institute for Social and
Economic Research at the University of Essex, UK (BHPS, 2009). Its aim is to track social and economic
change in a representative sample of the British population (for more information on the data set, see Taylor,
2009). The sample comprises about 15,000 individual interviews. Starting in 1991, up to now, there have
been 17 waves of data collected with the aim of tracking the individuals of the first wave over time (there is
a percentage of rotation as some individuals drop out of the sample over time and others are included, but
attrition is quite low, see Taylor, 2009).
(1) (2) (3)
employed self-employed unemployed
mean mean mean
life satisfaction 5.2449 5.3171 4.6258
mental well-being 26.2908 26.5934 24.0307
log(income) 10.1228 9.9603 9.5732
health status 4.0152 4.0555 3.6626
d married 0.5850 0.6840 0.2915
d separated 0.0228 0.0271 0.0390
d widowed 0.0150 0.0123 0.0091
d divorced 0.0834 0.0909 0.1178
d disabled 0.0201 0.0229 0.0580
gender 1.4976 1.2601 1.4158
age 39.4423 45.0684 35.0702
(age-mean age)2183.1217 140.8183 294.1427
education 3.6222 3.4488 2.7861
Observations 40859 5455 2309
Table 1: Summary statistics
as it also relates to mental health. It is an index from the ‘General Health Questionnaire’ of
the BHPS, composed of the answers to 12 questions that assess happiness, mental distress
(such as existence of depression or anguish), and well-being. This subjective assessment is
measured on a Likert scale from 0 to 36, which we have recoded to values of one (lowest
well-being) to 37 (highest scores in mental well-being).
Our main focus lies on analysing the effects of self-employment on life-satisfaction, but we
also include a variable for job satisfaction in our preliminary regressions. Our variable for job
satisfaction is based on the question “How dissatisfied or satisfied are you with your job (if in
employment)?” and also ranges from “not satisfied at all” (1) to “completely satisfied” (7).
Pairwise correlation between life satisfaction and job satisfaction in our sample is ρ= 0.4762
overall. It is higher for the self-employed (ρ= 0.5460) than for the employed (ρ= 0.4793).
Our main explanatory variable is the job status of individuals. A variety of job condi-
tions are detailed in the BHPS, the three most important of which are being unemployed,
employed and self-employed. Pooled over all sample years (n= 76,752), 40,859 (53.24%)
individuals have been in employment, 5,455 (7.11%) were self-employed and 2,309 (3.01%)
were unemployed. The rest were either retired (15,278; 19.91%), in some form of schooling or
studies (3,839; 5.00%), had maternity leave (361; 0.47%), were long-term sick (3,095; 4.03%)
or in family care (5,109; 6.66%). 447 (0.58%) fell into none of these categories. Except for
these different employment types, we control for some important individual characteristics
in our analysis. The most prominent of our control variables are detailed in Table 1, where
we have disaggregated them for the three most important employment categories.
Table 1 shows that the self-employed have higher life satisfaction scores, on average, than
those in regular employment. They tend to be in better health, are more likely to be married,
and are generally older than the employed. However, their expected income is lower than
employed individuals. Our summary statistics are broadly similar to those reported elsewhere
(see e.g. Andersson, 2008, Table 2).
life satisfaction mental wb d employed d selfemployed d unemployed log(income) education age gender
life satisfaction 1.0000
mental wb 0.5543∗∗∗ 1.0000
d employed 0.0037 0.0833∗∗∗ 1.0000
(0.3043) (0.0000)
d selfemployed 0.0168∗∗∗ 0.0371∗∗∗ -0.2951∗∗∗ 1.0000
(0.0000) (0.0000) (0.0000)
d unemployed -0.0857∗∗∗ -0.0598∗∗∗ -0.1879∗∗∗ -0.0487∗∗∗ 1.0000
(0.0000) (0.0000) (0.0000) (0.0000)
log(income) 0.0793∗∗∗ 0.0795∗∗∗ 0.2953∗∗∗ 0.0020 -0.1120∗∗∗ 1.0000
(0.0000) (0.0000) (0.0000) (0.5886) (0.0000)
education -0.0165∗∗∗ 0.0593∗∗∗ 0.2870∗∗∗ 0.0470∗∗∗ -0.0367∗∗∗ 0.3110∗∗∗ 1.0000
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
age 0.0898∗∗∗ -0.0347∗∗∗ -0.3616∗∗∗ -0.0066 -0.1028∗∗∗ -0.0317∗∗∗ -0.3184∗∗∗ 1.0000
(0.0000) (0.0000) (0.0000) (0.0690) (0.0000) (0.0000) (0.0000)
gender -0.0037 -0.1314∗∗∗ -0.0691∗∗∗ -0.1496∗∗∗ -0.0403∗∗∗ -0.0616∗∗∗ -0.0636∗∗∗ 0.0196∗∗∗ 1.0000
(0.3033) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
Observations 76752
P-values in parentheses
p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001
Table 2: Correlation matrix
One important control variable is an appropriate measure of income, for which we have
decided to use net equivalised annual household income (in British Pound Sterling), before
housing costs and deflated to price level of 2008, as provided and detailed by Levy and Jenkins
(2008). As equivalence scales, we have opted for applying the widely accepted McClements
scale (McClements, 1977). In accordance with consensus in the happiness literature, we
use the logarithm of the income measure as a matching variable in our analysis, assuming
that a given change in the proportion of income leads to the same proportional change in
well-being (Easterlin, 2001, p. 468). A remark is in order on self-reports of income in the
context of self-employment and entrepreneurship: it is quite well-known that self-reports of
income are quite unreliable in the context of entrepreneurs and self-employed (e.g., Block
and Koellinger, 2009; Blanchflower and Oswald, 1998), leading to biased estimates when
controlling for income in standard regressions. A second problem lies in theoretical concerns
whether one should control for income at all: “if the hypothesis is that the self-employed
have higher job and life satisfaction but at the same time, they receive lower incomes, it is
not certain that we want to control for income. Including income as a control variable in the
fixed-effects models is even more problematic. Since becoming self-employed for the average
individual means a decrease in income, it can be hard to disentangle the effect of becoming
self-employed from the effect of receiving a lower income on the outcomes” (Andersson, 2008,
p. 218).
Using a matching approach, we are immune to both kinds of problems since we are not
regressing income on life satisfaction but using the variable to match individuals who report
similar incomes. Our approach thus allows us to use the information contained in the income
variable without having to fear that the negative effect of lower income on happiness is
entangled with the positive effect of being self-employed.
To measure individuals’ health, we focus on an individual’s subjective assessment of health
(during the last 12 months). This is ordinally scaled on a five point Likert scale, ranging from
“excellent” (five) to “very poor” (one).10 Subjective assessments of health seem to predict
objective health quite well in some cases (e.g., regarding morbidity). Whether objective
health is sufficiently well captured by subjective health assessments is still debated (Johnston
et al., 2009). In order to account for more objective aspects of individual health, we also
included a dummy variable to account for disability of an individual.
Besides income and health, our control variables also comprise the usual set of gender,
age, and age2(we use the squared difference between age and mean-age instead of age2in
order to avoid problems of multicollinearity), as well as some dummies regarding marital
status (e.g., being married, being separated, divorced or widowed). We have also added a
regional control variable, dummies for different ethnicities and years (which we do not report,
however). Also included is an educational control variable, viz. an individual’s highest level
of education. This is measured ordinally, ranging from one (“none of these”) to seven (“higher
degree”), giving intermediate values to the middle education levels.11 Of our sample, 53.00%
were female. The mean age is 45.49 years (s.d. 17.85) with maximum age at 98 years and
minimum age at 15 (younger individuals were not interviewed in the BHPS).
In Table 2, we report pairwise correlations between our variables. A look at these correla-
tions offers first insights. Most of the correlations are highly significant, with the exception of
life satisfaction and gender (not many studies report significant effects of gender on life sat-
isfaction, see Dolan et al., 2008) as well as log(income) and the self-employment dummy. We
find a negative association between life satisfaction and unemployment (ρ=0.0857) and a
less strong positive relationship between life satisfaction and self-employment (ρ= 0.0168).
Education has a somewhat negative association with life satisfaction (ρ=0.0165). A similar
effect has been found by Binder and Coad (2010b) for the upper quantile of the life satis-
faction distribution. Not surprising is that education correlates strongly (ρ= 0.3110) with
log(income). On the other hand, quite interestingly, education also strongly correlates with
being employed (ρ= 0.2870), but only slightly so with being self-employed (ρ= 0.0470). One
could hypothesize that education is of much less importance for the self-employed (probably
due to the majority of them not owning knowledge-intensive high-tech start-ups but low-tech
businesses). Education is moreover slightly negatively associated with being unemployed
(ρ=0.0367). A last impression concerns the association of gender and self-employment,
which is quite strongly negative (ρ=0.1496), suggesting that comparatively fewer females
are in self-employment (of the 5,455 individuals in our sample in self-employment, only 1,419
were female). This simple correlation exercise can but give a first impression of our data set,
since it does not control for any confounding factors.
Figure 1 shows the cumulative density functions (cdf) for log income (left) and life sat-
isfaction (right). Starting with the cdf for log income, we see that the self-employed earn
less than the employed in most cases, while at the upper end of the distribution (starting at
the 80th percentile) a handful of “superstar” self-employed individuals earn more than their
employed counterparts. Put differently, the income distribution for the employed does not
10We have reversed the numerical order of the Likert scale to consistently use higher values for better
outcomes. The original coding in the BHPS codes a value of one to be excellent health and five to be very
poor health.
11For more information see Taylor (2009), App. 2, pp. 18-9.
Figure 1: Cumulative density functions for log income (left) and life satisfaction (right) for the three employ-
ment categories: unemployed (1), employed (2) and self-employed (3).
stochastically dominate the income distribution of the self-employed. This finding is similar
to results for US data in Hamilton (2000), but seem to be at odds with results on Indian data
(Tamvada, 2010), where the distribution of per capita consumption for the employed stochas-
tically dominates the corresponding distribution of the self-employed. It is important to note,
however, that Tamvada (2010) splits the self-employed group into “solo entrepreneurs” and
“employers”, something that we are unable to do in our dataset. These divergent findings
may well be reconciled if we consider that the income distribution of employers stochastically
dominates the income distribution those in paid employment, which in turn stochastically
dominates the income distribution of solo entrepreneurs.
Concerning the life satisfaction distribution, the self-employed generally report higher life
satisfaction, although we do not strictly observe that the life satisfaction distribution for
the self-employed stochastically dominates the corresponding distribution for the employed
because of a handful of very dissatisfied self-employed individuals.12
In Table 3 we depict a transition matrix where (pooled) mean changes in life satisfaction
are correlated with changes in job status (for transition between employment “E”, unem-
ployment “UE” and self-employment “SE” from period t1 to t). We can clearly see that
there are no big changes in life satisfaction exhibited if one’s employment status stays con-
stant (the diagonal in the table). However, moving into unemployment from any type of
(self-)employment is associated with a negative change in life satisfaction and leaving un-
employment is also positively associated with an increase in life satisfaction. Interestingly
enough, the effect of moving from employment to self-employment is larger than vice versa.
The signs of change in well-being are comparable to the analysis by Clark (2003). As the
pairwise correlation table, the transition matrix presented here can but offer a simple first
overview over the data.
Mean change in life satisfaction
Et1-0.0087 -0.3962 0.0820
(.0068) ( .0651) ( .0446)
obs 23,439 419 512
UEt10.4286 0.0109 0.4242
(.0591) (.0689) (.1871)
obs 490 458 66
SEt10.0169 -0.3571 -0.0039
(.05334) ( .2151) (.0190)
obs 415 42 2,804
Standard errors in parentheses.
Table 3: Transition matrix: Mean change in life satisfaction and change in job status (“E”, “S” and “U”
denoting employment, self-employment and unemployment respectively) from t1 to t
5. Analysis
We contribute to the literature by making use of recent developments in matching econo-
metrics to create an accurate control group, and thus identify the causal effect of self-
employment on happiness. Our dataset has comprehensive information on individual charac-
teristics, which allows us to find a “perfect twin” for each self-employed individual, and thus
recreate an appropriate control group in our analysis of how satisfied individuals are with
self-employment (Almus and Czarnitzki, 2003).
5.1. Preliminary regressions
As a first orientation, we want to present a standard regression analysis, where we regress
the typical factors on life satisfaction and job satisfaction. In Table 4, we show two pooled
ordered probit regressions (models (1) and (2)), where life satisfaction (1) and job satisfaction
(2) are the dependent variables. While we clearly see a strong effect of being self-employed
on job satisfaction, this effect is much smaller when taking life satisfaction as a dependent
variable. The latter two columns (models (3) and (4)) now repeat this analysis within a fixed-
effects (FE) regression framework, where we are not interested in the between-variance but
the variance within individuals over time, controlling for time-invariant individual-specific
components. Accounting for fixed effects in happiness regressions does substantively alter
regression results, a fact happiness researchers become increasingly more aware of (Ferrer-
i Carbonell and Frijters, 2004). Since happiness is in part determined by genes and stable
personality traits (Lykken and Tellegen, 1996; Diener et al., 1999), accounting for fixed effects
would seem to be the route to choose. Model (3) here depicts the FE-version of model (1)
and model (4) is a robustness test where we use a broader mental well-being variable as
12We observe that 0.58% of the self-employed report a life satisfaction score of 1, compared to 0.50% of
employed individuals.
(1) (2) (3) (4)
life satisfaction job satisfaction life satisfaction mental well-being
Ord. probit Ord. probit FE FE
d unemployed -0.2716∗∗∗ -0.7025∗∗∗ -0.3269∗∗∗ -1.9826∗∗∗
(-10.61) (-8.40) (-10.14) (-12.90)
d selfemployed 0.0588∗∗∗ 0.3171∗∗∗ -0.0020 -0.0354
(4.15) (20.61) (-0.08) (-0.31)
d retired 0.2368∗∗∗ -0.2284∗∗ 0.0292 0.1469
(13.04) (-2.73) (1.06) (1.31)
d studyschool 0.0857∗∗∗ -0.2129∗∗∗ 0.0492 -0.0325
(4.24) (-7.29) (1.59) (-0.21)
d maternityleave 0.3855∗∗∗ -0.1681∗∗ 0.2724∗∗∗ -0.2217
(7.52) (-2.89) (6.42) (-0.95)
d longtermsick -0.2132∗∗∗ -0.4808∗∗∗ -0.3702∗∗∗ -2.0802∗∗∗
(-8.19) (-5.68) (-8.43) (-9.80)
d familycare 0.0094 -0.1284-0.0574-0.4566∗∗∗
(0.53) (-2.09) (-2.16) (-3.58)
d other 0.0563 0.0734 -0.0328 -0.2485
(1.03) (0.91) (-0.54) (-0.81)
log(income) 0.0999∗∗∗ 0.0339∗∗∗ 0.0309∗∗ -0.0028
(14.00) (3.60) (3.14) (-0.06)
health status 0.3826∗∗∗ 0.2342∗∗∗ 0.2130∗∗∗ 1.3505∗∗∗
(75.57) (37.27) (31.33) (39.45)
d married 0.1894∗∗∗ 0.1482∗∗∗ 0.0611-0.3526∗∗
(16.63) (10.83) (2.28) (-2.62)
d separated -0.2556∗∗∗ -0.0002 -0.1225-1.7689∗∗∗
(-8.92) (-0.01) (-2.38) (-6.52)
d widowed -0.0804∗∗∗ 0.0883 -0.2166∗∗ -1.5190∗∗∗
(-3.54) (1.90) (-3.13) (-5.67)
d divorced -0.1033∗∗∗ 0.0064 0.0475 -0.0844
(-5.97) (0.30) (1.05) (-0.38)
d disabled -0.1839∗∗∗ 0.0045 -0.1530∗∗∗ -0.6140∗∗∗
(-9.88) (0.12) (-6.11) (-5.66)
gender 0.0481∗∗∗ 0.1317∗∗∗
(6.05) (13.71)
age 0.0039∗∗∗ 0.0089∗∗∗ -0.0124 -0.0331
(9.67) (16.22) (-0.84) (-0.54)
(age-mean age)20.0004∗∗∗ 0.0006∗∗∗ 0.0000 0.0003
(20.06) (18.93) (0.33) (1.28)
education -0.0447∗∗∗ -0.0235∗∗∗ 0.0126 0.0600
(-18.37) (-7.69) (0.92) (0.91)
Observations 76752 49419 76752 76752
(Pseudo-) R20.0580 0.0215 0.0388 0.1217
tstatistics in parentheses
p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001
Table 4: Ordered probit and fixed effects regressions for life satisfaction and job satisfaction
a dependent variable.13 In line with the above discussed findings in the literature, we can
observe that the small effect of self-employment on happiness disappears when controlling for
individual-specific time-invariant effects in our regressions (see, similarly, Andersson, 2008).
However, by removing time-invariant characteristics, we may be “throwing out the baby with
the bathwater”, and effectively we may be removing some effects of interest (in particular,
slow-changing character traits) by “over-controlling”. Also, fixed-effect regression suffers
from other drawbacks of regression models discussed above (in particular, lack of a common
support for treatment and control groups). In order to come to more reliable estimates of
the effect of being self-employed on life satisfaction, we turn now to our matching estimates.
5.2. Matching estimates
Our main analysis centers around estimating the causal effect that going into self-employment
has on an individual’s life satisfaction. Figure 2 gives a graphical representation of our match-
ing approach. In both cases, we restrict attention to individuals that are similar, along a
number of dimensions, at time t. We then track these individuals over time and observe
differences between the treatment group (those moving into self-employment) and the con-
trol group (their matched counterparts in regular employment). We are interested here in
two different situations, which correspond to the distinction of opportunity versus necessity
entrepreneurship. In the first case (Figure 2, left) all individuals start off in regular employ-
ment in t, and some individuals move into self-employment in t+ 1. We suggest that this
case corresponds to opportunity entrepreneurship. In the second case (Figure 2, left), all
individuals start off as unemployed in t, and some move into self-employment while the con-
trol group moves into regular employment in t+ 1. This case would correspond to necessity
entrepreneurship, where individuals chose to become self-employed to escape unemployment.
Since we are also interested in the dynamics of well-being, we have chosen to examine
whether there is a lagged effect of self-employment on life satisfaction, i.e. the possible impact
of self-employment on life satisfaction at period t+ 2. A robust finding emerging from
the happiness literature is that individuals adapt to changes in their life circumstances.
Hedonic adaptation, the hedonic dulling of repeated or constant affective stimuli (Frederick
and Loewenstein, 1999) is highly domain-specific and varies with the concrete stimulus (for
example, hedonic adaptation to marriage is faster and more complete than hedonic adaptation
to repeated unemployment, see, e.g., Clark et al., 2008b). The yearly structure of our panel
data set suggests to include a second year to check for hedonic adaptation but additional
13The rationale for model (4) lies in some econometric reservations one could have in our using an ordinal
scaled life-satisfaction variable in a fixed effects OLS regression, thus implicitly treating life satisfaction as
a cardinal variable. This is in part motivated by the absence of a commonly agreed-on method to account
for fixed effects in an ordered probit framework. However, econometric research on happiness shows that
there are no substantial differences between both approaches in terms of the results they generate (Ferrer-i
Carbonell and Frijters, 2004) and a cardinal treatment of life satisfaction is common in the psychological
literature on well-being. One reason for the robustness of the life satisfaction measure to being treated as
cardinal could lie in the fact that individuals seem to convert ordinal response labels into similar numerical
values such that these cardinal values equally divide up the response space (Van Praag, 1991; Clark et al.,
2008). Nevertheless, model (4) with a 37-point scale alleviates the possible objection to using life satisfaction
in an FE framework, since treating a 37-point scaled mental well-being construct as a cardinal variable seems
much more uncontroversial.
Figure 2: A graphical depiction of our matching approach. In both cases, we match individuals at time t
and observe these individuals at times t+ 1 and t+ 2, comparing those individuals that have moved into
self-employment with comparable individuals who are in regular employment. In the first case (left) all
individuals start out in regular employment, and the control group remains in regular employment. In the
second case (right) all individuals start out unemployed, and the control group consists of those moving from
unemployment to regular employment.
lags might be added in future work. Our dynamic approach, according to which we match
individuals at time tand observe them again at times t+ 1 and t+ 2, additionally takes
into account the fact that failure to control for lagged outcome can lead to bias in matching
estimators (see, e.g., Gonz´alez and Paz´o, 2008).14
We are carrying out our analysis for two different types of matching, viz. nearest neigh-
bour matching as well as propensity score matching. Both methods differ with respect to
how individuals are matched. Nearest neighbour matching finds a match in many dimensions
simultaneously while propensity score matching collapses all covariates into on composite
variable (the “propensity score”). This difference has the consequence that adding too many
covariates according to which one matches individuals in nearest neighbour matching results
in a dimensionality problem, i.e. one is not likely to find good matches in each of the dimen-
sions simultaneously. Therefore, for the nearest neighbour matching, we matched individuals
according to a smaller number of criteria, namely: previous life satisfaction, log(income),
gender, age, education, subjective health assessment as well as dummies for ethnicity and
being married. Adding more criteria would have made it harder to get good matches in our
For the propensity score matching, we did not have pressing concerns of dimensionality
(since the matching covariates are collapsed into a synthetic propensity score, and matching
is performed with reference to the propensity score only). Therefore with propensity score
matching, we matched individuals according to the above mentioned factors but added also
the following list of covariates: year dummies, regional dummies for the different former
Metropolitan counties and Inner and Outer London, dummies for being separated, divorced
14Cooper and Artz (1995) focus in their analysis on third year business because prospects can fluctuate
and in the beginning, uncertain prospects might lower satisfaction. This initial uncertainty offers thus an
additional reason for taking into account the intertemporal structure of life satisfaction following one’s decision
to go into self-employment.
E to SE vs. E to E
1 lag 0.172 15181
SE 0.062
t-stat 2.76
2 lags 0.241 7896
SE 0.085
t-stat 2.83
U to SE vs. U to E
1 lag -0.130 326
SE 0.210
t-stat -0.62
2 lags -0.031 147
SE 0.317
t-stat -0.10
Table 5: Nearest neighbour matching estimates of the Sample Average Treatment Effect (SATE). 4 matches
are selected for each treatment observation. SATE, Standard Errors and z-stats estimated following Abadie
et al. (2004).
or widowed, a dummy for being disabled and a quadratic age term.15
Table 5 shows the nearest neighbour matching results (“E”, “SE” and “UE” denoting
employment, self-employment and unemployment respectively), while Table 6 shows the es-
timates obtained from a propensity score matching estimator. For both matching estima-
tors, we can see significant positive effects on happiness of switching from employment to
self-employment, compared to a matched sample of those who remain in employment. Our
findings here complement results that have been reported for US data by Hundley (2001),
who find that individuals going from employment into self-employment experience an increase
in job satisfaction. With our matching approach, we are able to show that this increase in
satisfaction is not related only to the job but to satisfaction with life as a whole.
Interestingly enough, both matching estimators show a larger effect at the second lag
compared of the first, which suggests that the positive impact on well-being is not only not
transitory, but it even increases with time. Individuals who move from employment to self-
employment will appreciate this transition even more two years afterwards, once they have
become more accustomed to the self-employed lifestyle. The causal effect self-employment
has on life satisfaction is thus (at least in the first years) not only exempt from hedonic
adaptation, it seems to show the opposite: an increasing antiadaptive effect.
It is important to point out, however, that there is a marked difference between this posi-
tive effect of self-employment on happiness for individuals who switch from employment (i.e.
opportunity entrepreneurs) and those who become self-employed to escape unemployment.
In all these cases, there is no significant difference between the well-being of individuals who
switch from unemployment to self-employment, compared to those to switch from unem-
15We also wanted to match individuals according to industry in which they are employed (or had their last
employment), but since data reporting definitions changed over the sample period we were prevented from
doing this.
ATT Controls Treated
E to SE vs. E to E
1 lag 0.155 23409 512
SE 0.049
t-stat 3.14
2 lags 0.212 23409 512
SE 0.071
t-stat 2.97
U to SE vs. U to E
1 lag -0.225 490 66
SE 0.231
t-stat -0.98
2 lags -0.166 490 66
SE 0.273
t-stat -0.61
Table 6: Propensity score matching estimates of the Average Treatment Effect on the Treated (ATT),
obtained using the kernel option (using the “pscore” command in Stata 11, developed by Sascha Becker
and Andrea Ichino). Analytical SEs cannot be computed; bootstrapped SEs are reported (100 bootstrap
ployment to regular employment. In each case, the estimated effect is actually negatively
signed, but far from significant (perhaps due to the lower number of observations). In the
case of those leaving unemployment (i.e. when it comes to necessity entrepreneurship), self-
employment has no advantage over a regular job in terms of individuals’ life satisfaction.
5.3. Robustness of matching estimates
While such a matching approach offers a robust way of identifying appropriate control
and treatment groups, it can be quite sensitive to identification bias. In particular, problems
might arise if the conditional independence assumption (CIA) is not valid. This aspect is
often ignored in the literature on matching (Caliendo and Kopeinig, 2008). In order to
account for this sensitivity, we conduct various robustness tests.
To begin with, we follow a simulation approach by Nannicini (2007) and Ichino et al.
(2008) that allows us to identify the robustness of our estimation strategy with respect to
simulated confounders that recreate violations of the CIA. The sensitivity analysis is reported
in Table 7. As recommended by Nannicini (2007, p. 6), “the results of this simulation-based
sensitivity analysis should be judged more on the basis of the distance between point esti-
mates associated to different pij , rather than the significance level of the simulated ATTs.”
Our robustness analysis reveals that our results are generally robust with respect to simulated
confounders such as gender and marriage status variables, but not to “strong confounders”.
However, it has been observed that the “strong confounder” configuration taken here (fol-
lowing Nannicini, 2007) is especially stringent, and although our results are not robust to
the presence of a “strong confounder” they are nonetheless fairly robust.
A second robustness test we conducted was to repeat our analysis with mental well-being
UtoSE vs UtoE Fraction U=1 by t/o Outcome effect Γ Selection effect Λ ATT SE
p11 p10 p01 p00
E to SE: 1 lag
No confounder 0.084 0.074
Neutral confounder 0.5 0.5 0.5 0.5 0.996 0.983 0.089 0.096
Strong confounder 0.8 0.8 0.6 0.3 3.489 5.22 -0.021 0.094
d female 0.33 0.34 0.5 0.48 1.048 0.541 0.097 0.092
d married 0.67 0.62 0.63 0.57 1.289 1.235 0.083 0.092
UE to SE: 1 lag
No confounder . .
Neutral confounder 0.5 0.5 0.5 0.5 1.031 1.172 -0.212 0.304
Strong confounder 0.8 0.8 0.6 0.3 10.19 2.07E+16 -0.463 0.483
d female 0.13 0.22 0.39 0.38 1.062 0.308 -0.231 0.313
d married 0.35 0.35 0.3 0.3 1.082 1.543 -0.211 0.293
E to SE: 2 lags
No confounder 0.125 0.093
Neutral confounder 0.5 0.5 0.5 0.5 1.007 1.014 0.14 0.123
Strong confounder 0.8 0.8 0.6 0.3 3.542 3.962 0.012 0.135
d female 0.34 0.33 0.49 0.48 1.055 0.541 0.166 0.125
d married 0.65 0.62 0.6 0.58 1.089 1.267 0.157 0.13
UE to SE: 2 lags
No confounder
Neutral confounder 0.5 0.5 0.5 0.5 1.085 1.107 -0.166 0.353
Strong confounder 0.8 0.8 0.6 0.3 1.50E+09 1.03E+07 -0.746 0.539
d female 0.19 0.15 0.43 0.32 2.231 0.281 -0.106 0.373
d married 0.35 0.35 0.34 0.18 3.772 1.327 -0.132 0.36
Table 7: Robustness of treatment effect estimates. Sensitivity analysis investigating the effect of calibrated
confounders using the simulated approach presented in Nannicini (2007).
as the dependent variable instead of life satisfaction. Mental well-being is a broader concept
of well-being that includes more affective and mental health related aspects of human life (the
correlation between life satisfaction and mental well-being is ρ= 0.5543). We thus repeated
the matching analysis with mental well-being and obtained similar results, for both mod-
els, and for both lags, and for both matching techniques (nearest neighbour matching, and
propensity score matching).16 In other words, using mental well-being instead of life satisfac-
tion we observed that individuals switching from regular employment were happier (at t+ 1
and t+ 2) than those who remained in regular employment. On the other hand, individuals
who moved from unemployment into self-employment were not significantly different from
those who moved from unemployment into regular employment.17 Taken together, we are
reasonably convinced of the soundness and robustness of our matching estimation strategy.
6. Conclusion
In previous studies, the self-employed were found to be more satisfied with their jobs
than employed control groups. This finding has proven robust even though the self-employed
often earn less and work more hours than individuals in regular employment. Explanations
suggest that the autonomy enjoyed by “being one’s own boss” more than compensates en-
trepreneurs for the hardships otherwise associated with self-employment. The present paper
has investigated to which extend higher job satisfaction of the self-employed also translates
into a more global assessment of well-being, namely their satisfaction with life in general.
Few studies so far were able to present empirical evidence on higher life satisfaction of the
16The authors will provide the detailed results of this exercise on request.
17The coefficient was also always negative but not significant.
self-employed, either due to methodological difficulties since the self-employed are a very het-
erogeneous group, or due to a lack of a causal connection between the variables. Following
the latter line of reasoning, self-employed individuals might not enjoy higher life satisfaction
than the employed because their high job satisfaction could result in the self-employed fo-
cussing so strongly on their work that they crowd out other activities that contribute to high
life satisfaction, such as social relations or health.
To account for these methodological difficulties and to explore the above-mentioned the-
oretical intuition, we have applied a matching methodology in order to better identify treat-
ment and control groups and thus being able to examine the causal effect that a transition
into self-employment has on life satisfaction. Since individuals go into self-employment for
quite divergent reasons, we have broadly distinguished two motivations in our regressions:
individuals going into self-employment to escape unemployment (necessity self-employment)
differ from individuals who go into self-employment to exploit a business opportunity (op-
portunity self-employment). In our analysis we found that individuals moving from regular
employment into self-employment (the case of “opportunity entrepreneurship”) experience
a positive and significant increase in life satisfaction, that actually increases from the first
year of self-employment to the second. However, we also observed that individuals moving
from unemployment to self-employment were not better off than those moving from unem-
ployment to regular employment (the case of “necessity entrepreneurship”). Those moving
from unemployment to self-employment actually had lower life satisfaction scores than the
control group, but these differences were not statistically significant.
Further research might fruitfully centre on extending our findings from the British House-
hold Panel Survey (BHPS) data set to other countries as well as extending the analysis to
cover longer horizons in order to explore the longer term causal effects of self-employment
on life satisfaction. This might be a worthwhile undertaking, considering that it has been
observed that the self-employed not only have lower pay, but also that their pay increases at
a lower rate over time (Hamilton, 2000).
Date: December 21, 2010; ca. 7,400 words
Abadie, A., Drukker, D., Herr, J., and Imbens, G. (2004). Implementing matching estimators
for average treatment effects in Stata. Stata Journal, 4:290–311.
Almus, M. and Czarnitzki, D. (2003). The Effects of Public R&D Subsidies on Firms’
Innovation Activities. Journal of Business and Economic Statistics, 21(2):226–236.
Andersson, P. (2008). Happiness and health: Well-being among the self-employed. Journal
of Socio-Economics, 37(1):213–236.
Benz, M. and Frey, B. (2008a). Being independent is a great thing: Subjective evaluations
of self-employment and hierarchy. Economica, 75(298):362–383.
Benz, M. and Frey, B. (2008b). The value of doing what you like: Evidence from the self-
employed in 23 countries. Journal of Economic Behavior & Organization, 68(3-4):445–455.
BHPS (2009). British Household Panel Survey.
Binder, M. and Coad, A. (2010a). An examination of the dynamics of well-being and life
events using vector autoregressions. Journal of Economic Behavior and Organization,
Binder, M. and Coad, A. (2010b). Going Beyond Average Joe’s Happiness: Using Quantile
Regressions to Analyze the Full Subjective Well-Being Distribution. Papers on Economics
& Evolution #1010.
Blanchflower, D. (2000). Self-employment in OECD countries. Labour Economics, 7(5):471–
Blanchflower, D. (2004). Self-employment: more may not be better. Swedish Economic Policy
Review, 11(2):15–73.
Blanchflower, D. and Oswald, A. (1998). What makes an entrepreneur? Journal of labor
Economics, 16(1):26–60.
Blanchflower, D., Oswald, A., and Stutzer, A. (2001). Latent entrepreneurship across nations.
European Economic Review, 45(4-6):680–691.
Block, J. and Koellinger, P. (2009). I can’t get no satisfaction — Necessity entrepreneurship
and procedural utility. Kyklos, 62(2):191–209.
Bradley, D. and Roberts, J. (2004). Self-employment and job satisfaction: Investigating the
role of self-efficacy, depression, and seniority. Journal of Small Business Management,
Brockhaus, R. (1980). The effect of job dissatisfaction on the decision to start a business.
Journal of Small Business Management, 18(1):37–43.
Caliendo, M. and Kopeinig, S. (2008). Some practical guidance for the implementation of
propensity score matching. Journal of Economic Surveys, 22(1):31–72.
Clark, A. E. (2003). Unemployment as a social norm: Psychological evidence from panel
data. Journal of Labor Economics, 21(2):323–351.
Clark, A. E. and Georgellis, Y. (2010). Back to baseline in Britain: Adaptation in the BHPS.
Clark, A. E. and Oswald, A. J. (1994). Unhappiness and unemployment. The Economic
Journal, 104(424):648–659.
Clark, A. E., Frijters, P., Shields, M. A. (2008a). Relative income, happiness, and utility: An
explanation for the Easterlin paradox and other puzzles. Journal of Economic Literature
46, 95–144.
Clark, A. E., Diener, E., Georgellis, Y. and Lucas, R. E. (2008b). Lags and Leads in Life
Satisfaction: A Test of the Baseline Hypothesis. The Economic Journal, 118:F222-F243.
Cooper, A. and Artz, K. (1995). Determinants of satisfaction for entrepreneurs. Journal of
Business Venturing, 10(6):439–457.
Craig, J. B. and Schaper, M. and Dibrell, C. (2007). Life in Small Business in Australia:
Evidence from the HILDA Survey. Mimeo.
Cromie, S. and Hayes, J. (1991). Business Ownership as a Means of Overcoming Job. Per-
sonnel Review, 20(1):19–24.
Dehejia, R. and Wahba, S. (1999). Causal Effects in Nonexperimental Studies: Reevaluating
the Evaluation of Training Programs. Journal of the American Statistical Association,
Diener, E., Suh, E., Lucas, R. E., Smith, H. L., 1999. Subjective well-being: Three decades
of progress. Psychological Bulletin 125, 276–302.
Di Tella, R., MacCulloch, R. J., and Oswald, A. J. (2001). Preferences over inflation and
unemployment: Evidence from surveys of happiness. The American Economic Review,
Dolan, P., Peasgood, T., and White, M. (2008). Do we really know what makes us happy?
A review of the economic literature on the factors associated with subjective well-being.
Journal of Economic Psychology, 29(1):94–122.
Easterlin, R. A. (2001). Income and happiness: Towards a unified theory. The Economic
Journal, 111:465–484.
Ferrer-i Carbonell, A., Frijters, P., 2004. How important is methodology for the estimates of
the determinants of happiness? The Economic Journal 114, 641–659.
Frederick, S., Loewenstein, G. F., 1999. Hedonic adaptation. In: Kahneman, D., Diener, E.,
Schwarz, N. (Eds.). Well-Being: The Foundations of Hedonic Psychology. Russell Sage
Foundation, New York, 302–329.
Fuchs-Schundeln, N. (2009). On preferences for being self-employed. Journal of Economic
Behavior & Organization, 71(2):162–171.
Gonz´alez, X. and Paz´o, C. (2008). Do public subsidies stimulate private R&D spending?.
Research Policy, 37(3):371–389.
Hamilton, B. (2000). Does Entrepreneurship Pay? An Empirical Analysis of the Returns to
Self-Employment. Journal of Political Economy, 108(3):604–631.
Heckman, J.J. and Ichimura, H. and Smith, J. and Todd, P. (1996). Sources of selection bias
in evaluating social programs: An interpretation of conventional measures and evidence on
the effectiveness of matching as a program evaluation method. Proceedings of the National
Academy of Sciences, 93(23):13416–13420.
Heckman, J., Ichimura, H., and Todd, P. (1997). Matching as an Econometric Evaluation
Estimator: Evidence from Evaluating a Job Training Programme. Review of Economic
Studies, 64(4):605–654.
Helliwell, J. F. (2003). How’s life? combining individual and national variables to explain
subjective well-being. Economic Modelling, 20:331–360.
Hundley, G. (2001). Why and when are the self-employed more satisfied with their work?
Industrial Relations: A Journal of Economy and Society, 40(2):293–316.
Hussinger, K. (2008). R&D and subsidies at the firm level: an application of parametric and
semiparametric two-step selection models. Journal of Applied Econometrics, 23(6):729–
Hyytinen, A. and Ruuskanen, O. (2007). Time use of the self-employed. Kyklos, 60(1):105–
Ichino, A., Mealli, F., and Nannicini, T. (2008). From temporary help jobs to permanent
employment: What can we learn from matching estimators and their sensitivity? Journal
of Applied Econometrics, 23(3):305–327.
Idson, T. L. (1990). Establishment size, job satisfaction and the structure of work. Applied
economics, 22(8):1007–1018.
Imbens, G. (2004). Nonparametric estimation of average treatment effects under exogeneity:
A review. Review of Economics and Statistics, 86(1):4–29.
Johnston, D. W., Propper, C., and Shields, M. A. (2009). Comparing subjective and objective
measures of health: Evidence from hypertension for the income/health gradient. Journal
of Health Economics, 28(3):504–552.
Kawaguchi, D. (2008). Self-employment Rents: Evidence from Job Satisfaction Scores.’.
Hitotsubashi Journal of Economics, 49(1):35–45.
Levy, H. and Jenkins, S. P. (2008). Documentation for derived current and annual net
household income variables, BHPS waves 1-16. Institute for Social and Economic Research,
University of Essex, Colchester.
Lucas, R. E., Clark, A. E., Georgellis, Y., and Diener, E. (2004). Unemployment alters the
set point for life satisfaction. Psychological Science, 15(1):8–13.
Lykken, D., Tellegen, A., 1996. Happiness is a stochastic phenomenon. Psychological Science
7, 186–189.
McClements, L. D. (1977). Equivalence scales for children. Journal of Public Economics,
8(2):191 – 210.
Nannicini, T. (2007). Simulation-based sensitivity analysis for matching estimators. Stata
Journal, 7(3):334–350.
Noorderhaven, N. and Thurik, R. and Wennekers, S. and Van Stel, A. (2004). The role of
dissatisfaction and per capita income in explaining self-employment across 15 European
countries. Entrepreneurship Theory and Practice, 28(5):447–466.
Oswald, A. J. and Powdthavee, N. (2008). Does happiness adapt? A longitudinal study
of disability with implications for economists and judges. Journal of Public Economics,
Powdthavee, N. (2009). I can’t smile without you: Spousal correlation in life satisfaction.
Journal of Economic Psychology, 30(4):675–689.
Reynolds, P., Bosma, N., Autio, E., Hunt, S., De Bono, N., Servais, I., Lopez-Garcia, P., and
Chin, N. (2005). Global entrepreneurship monitor: data collection design and implemen-
tation 1998-2003. Small Business Economics, 24(3):205–231.
Rubin, D. (1974). Estimating Causal Effects of Treatments in Randomized and Non-
randomized Studies. Journal of Educational Psychology, 66:688–701.
Santarelli, E. and Vivarelli, M. (2007). Entrepreneurship and the process of firms’ entry,
survival and growth. Industrial and Corporate Change, 16(3):455–488.
Schjoedt, L. and Shaver, K. (2007). Deciding on an Entrepreneurial Career: A Test of
the Pull and Push Hypotheses Using the Panel Study of Entrepreneurial Dynamics Data.
Entrepreneurship Theory and Practice, 31(5):733–752.
Storey, D. J. (1994). Understanding the small business sector. Thomson: London, UK
Tamvada, J. (2010). Entrepreneurship and welfare. Small Business Economics, 34(1):65–79.
Taylor, M. F. E. (2009). British Household Panel Survey User Manual Volume A: Introduc-
tion, Technical Report and Appendices. edited with John Brice, Nick Buck and Elaine
Prentice-Lane. Colchester: University of Essex.
Van Praag, B. S. v., 1991. Ordinal and cardinal utility: An integration of the two dimensions
of the welfare concept. Journal of Econometrics 50, 69–89.
Van Praag, M. C. v. and Versloot, P. H. (2007). What is the value of entrepreneurship? A
review of recent research. Small Business Economics, 29:351–382.
Vivarelli, M. (1991). The birth of new enterprises. Small Business Economics, 3(3):215–223.
Winkelmann, L. and Winkelmann, R. (1998). Why are the unemployed so unhappy? evidence
from panel data. Economica, 65(257):1–15.
... Analyse du burnout en entrepreneuriat : étude empirique sur les dirigeants de PME (Doctoral dissertation, Montpellier 1). Binder, M., et Coad, A. (2013). Life satisfaction and self-employment: a matching approach. ...
... Cette recherche permet, au travers de la quatrième contribution, de combler un manque dans la littérature concernant l'influence de la motivation entrepreneuriale sur la santé mentale des dirigeants. La littérature montre en effet, que la motivation entrepreneuriale va influencer la satisfaction au travail (Block et Koellinger, 2009 ;Kautonen et Palmroos, 2010 ;Binder et Coad, 2013) mais aussi la santé perçue mais aucune étude ne semblait s'intéresser plus précisément à l'impact de cette motivation entrepreneuriale sur la santé mentale des dirigeants et plus précisément sur le risque de burnout et sur le risque suicidaire. ...
Cette thèse s’inscrit au croisement de l’entrepreneuriat de la psychologie cognitive, ainsi que de la psychologie de la santé. Elle contribue au récent courant de la santé des dirigeants de PME qui vise à mieux comprendre l’impact de la fonction entrepreneuriale sur la santé des indépendants. Cette recherche s'intéresse plus particulièrement à la question de leur santé mentale au travers du risque de burnout et du risque suicidaire. Pour cela, elle mobilise une approche cognitive afin de mieux appréhender l’impact du stress entrepreneurial sur la santé mentale des dirigeants de PME, et de voir pourquoi certains entrepreneurs s’épuisent au travail, voire en viennent à se suicider. L’objectif de ce travail est donc de contribuer à une meilleure compréhension de la souffrance psychique patronale. Pour ce faire, cette thèse réalisée sur travaux s’articule autour de quatre contributions académiques et s’appuie sur une méthodologie principalement quantitative. Une étude qualitative est néanmoins menée de manière exploratoire sur la question du suicide, un sujet encore tabou et très peu étudié dans le monde patronal. La recherche est construite en trois étapes, la première s’intéresse à l’étude des spectres émotionnels de deux stresseurs de la fonction entrepreneuriale. La seconde s’intéresse à deux risques en santé mentale subséquents aux deux facteurs de stress de la fonction entrepreneuriale (risques de burnout et suicidaire). Enfin, la troisième et dernière étape vise quant à elle, à étudier la motivation entrepreneuriale comme facteur modérateur des deux risques en santé mentale ciblés. Les résultats montrent une ambiguïté émotionnelle dans la surcharge de travail des dirigeants de PME. La perception de l’événement accompagnant la surcharge va ainsi avoir un rôle déterminant dans l’impact de ce stresseur sur la santé. Aussi, au quotidien, les dirigeants de PME ne sont pas à l’abri d’un risque de burnout. Et selon les contextes, celui-ci peut être modéré par la motivation entrepreneuriale. Enfin, les résultats mettent également en évidence le caractère tragique que peut prendre l’échec entrepreneurial ainsi que l’omniprésence de l’endettement dans le risque suicidaire patronal.
... This scrap value could include, for example, resources that individuals can retain and potentially apply to a subsequent idea. Further, many of the rewards from venturing are not quantitative/ financial (self-employment, autonomy, intrinsic satisfaction, and more; Binder & Coad, 2013;Fitzsimmons & Douglas, 2011;Scheaf et al., 2020). As expected loss stresses retained value, these nonquantitative benefits enhance the retained value of the venture, which may be appealing to some individuals. ...
... Both options have the same expected value of 16% and are hence equivalent. We used "professional objectives" instead of straight financial rewards because the literature on entrepreneurship consistently reports that financial gain is only one component of entrepreneurs' motivation (Binder & Coad, 2013;Kuratko et al., 1997). ...
We propose and test a process where potential entrepreneurs (PEs) prioritize a venture idea consideration set using preference-based decision-specific heuristics to assess idea feasibility and desirability. We test our hypotheses through two studies with PEs. The first experiment shows that prioritization occurs, with 113 of 122 PEs voluntarily changing a randomized list of their ideated ventures into a rank-ordered priority list of potential opportunities. Second, we employ a novel “equivocal forced-choice” conjoint design with 250 PEs. We find empirical support that PEs prioritize via relative preferences for experience-based knowledge, strong social ties, and low risk/low reward venture ideas. We contribute to the entrepreneurship literature by theorizing and providing evidence of a prioritization stage for multiple idea sets before evaluation. Further, we demonstrate the influence of individual and social network factors on prioritization and expand our understanding of how PEs conceptualize risk in venturing.
... However, entrepreneurs are not a homogeneous group and one of the ways that they differ from each other is whether they became entrepreneurs by choice or by necessity. Those that are entrepreneurs because they have no other option are labelled necessity entrepreneurs and those that have become entrepreneurs to take advantage of an opportunity are called opportunity entrepreneurs (Binder & Coad, 2013. The factors that lead an individual to entrepreneurship can influence many factors at the individual and at the country levels. ...
... Entrepreneurial activities include elements of stressful factors that are generally antagonistic to subjective wellbeing, such as emotional demand, failure risks, protracted work hours, intense work efforts (Nikolaev et al., 2020;Wiklund et al., 2019). Paradoxically, the literature suggests that entrepreneurs most often report positive state of wellbeing including satisfaction (Binder & Coad, 2013;Nikolaev et al., 2020;Stephan, 2018). This paradox is explained by Lazarus and Folkman (1984) transactional theory of stress, which posits that individuals ponder sressful circumstances as either a threatening hindrance or a promoting challenge to their future gains, goal achievements, and personal growth (LePine et al., 2005). ...
The investigation of personal and environmental factors affecting entrepreneurship is of great importance in terms of how countries encourage entrepreneurs and how they create the necessary environment for them to be more successful. In this study, the factors that encourage or hinder entrepreneurship were examined using the data of the Global Entrepreneurship Monitor. In order to develop the entrepreneurship ecosystem, countries should facilitate entrepreneurs to start businesses more easily, expand the network between entrepreneurs and entrepreneurial candidates and investors, and strive to increase entrepreneurial skills.
... We focus on two distinctive forms of self-employment-ownaccount workers and employers with hired employees (Craig et al., 2012;Demurger & Xu, 2011). From the perspective of the motivation to engage in self-employed activities, own-account workers generally refer to those who have no other job options or are dissatisfied with their current employment and are thus forced to work for themselves (Binder & Coad, 2013;Seva et al., 2016;Wei et al., 2019). Employers with hired employees are those who strive to seek out new business opportunities for a higher income or nonpecuniary benefits (Hamilton, 2000;Larsson & Thulin, 2019). ...
... These two types of selfemployed individuals are very likely to have different levels of living conditions and social integration into their city of residence, which generates varying degrees of citizenship perceptions. This paper thus also fills the gap created by previous works that fail to analyze different types of self-employment separately when examining the consequences of self-employed activities (Binder & Coad, 2013;Levine & Rubinstein, 2017). We use the survey question "What is your current type of employment status?" to create three dichotomous variables: salaried workers (reference group), own-account workers, and employers. ...
Full-text available
Self-employment is an essential driver of economic growth and societal well-being. While previous studies have examined various consequences of self-employed activities, little attention has been paid to how self-employment affects citizenship perceptions among urban migrant workers. This paper seeks to fill this gap by distinguishing between own-account workers and employers with hired employees. Using data from the 2017 China Migrant Dynamic Survey, we find that migrants who are employers or work on their own account are more likely to perceive themselves as citizens of their city of residence than wage earners. The size of the effect is much larger for employers than for own-account workers. We also find that self-employment can be linked to migrants' perceptions of citizenship via three pathways-individual income, type of housing (owned or rented accommodation) and social security. Our results are robust to employing the conditional mixed process estimator and nonparametric matching method to address endogeneity concerns. Our findings highlight the importance of the role that migrant entrepreneurship plays in reforming the current urban household registration system in China. The paper also has implications for other countries worldwide to cope with the group of marginalized immigrants by emphasizing the form of entrepreneurship.
... Prior studies about the satisfaction of entrepreneurs mainly focused on a single country, especially developed countries [25,35]. However, pointed out that life satisfaction at the national level showed notable differences. ...
... First, this paper enriches the research on entrepreneurial satisfaction. Prior research in this field mainly focuses on the individual level [25,35], which suggests that, compared to employees, entrepreneurs have more autonomy, gain more social support, and balance their role pressures better, and thus obtain higher satisfaction. Also, the extant literature explored some factors from an individual level that affect entrepreneurs' satisfaction, such as gender [105][106][107]. ...
Full-text available
Entrepreneurs are known to be more satisfied than employees, with their life satisfaction being built on their satisfaction with their job and work–family balance. We argue that effects differ among societies, drawing on theories about self-determination and culture. Representative samples of 1276 entrepreneurs and 3821 employees in traditional China and modern Finland and Sweden were surveyed by the Global Entrepreneurship Monitor (GEM), which is amenable to multivariate analyses. The effects of occupation upon satisfaction were found to differ among the societies, consistent with their cultural differences. These findings contribute to contextualizing theories about satisfaction being embedded in society and culture.
... Another study found that being in arrear was a significant problem for European selfemployed people during the pandemic and that many were afraid to lose their homes due to being unable to pay the rent [9]. Research comparing the health and well-being of the self-employed with waged workers has yielded mixed results, with some studies reporting poorer well-being in the self-employed [10,11], whereas others state the opposite [12][13][14]. Having high levels of autonomy, flexibility, job satisfaction and many more advantages compared to waged workers [15,16] are commonly used as an explanation for better well-being in the self-employed. ...
Full-text available
Background: The self-employed are at increased risk of negative well-being outcomes when facing adversity such as the COVID-19 pandemic. Studies that examine socio-ecological factors that may protect their well-being are warranted. Methods: Data were drawn from a cross-sectional survey of European self-employed people (n = 1665). The WHO-5 Well-being Index was used to examine the impact on well-being of factors at four socio-ecological levels. Independent sample t-tests, Pearson correlations and linear regression were applied to analyse differences between groups of self-employed and interactions between variables using SPSS. Results: Well-being and the socio-ecological factors of resilience, social support, useful work and finding the rules clear were positively correlated with well-being. For self-employed who reported that it was challenging to run their business during the pandemic, social support and finding rules clear were of significantly greater importance to their well-being. Conclusions: The findings highlight that the socio-ecological factors of resilience, social support, doing useful work and finding the rules clear affect well-being. The results also indicate that it is vital to consider factors at multiple socio-ecological levels to improve the well-being of the self-employed during adversity.
... Vegetti & Adăscăliţei (2017) investigates the effects of the economic crisis and entrepreneurship activity within the European Union and concludes that pushing unemployed individuals into entrepreneurship creates mostly unmotivated, dissatisfied entrepreneurs with limited potential for success. In particular, ventures created out of necessity rarely improve an individual's overall wellbeing (Binder & Coad, 2013;Larsson & Thulin, 2019). Overall, it seems the entrepreneurship emerging typifies a precarious form of work because it provides no guarantee of success or sustainable income. ...
Full-text available
Entrepreneurship has the potential to drive economic development and social advancement. The European Commission implemented entrepreneurship policy as a pragmatic response to its economic and social challenges, especially after the 2008 financial crisis. Institutional changes to promote entrepreneurship and enable individuals to directly contribute to economic growth, job creation and society were introduced to create an entrepreneurial Europe. This study undertakes a systematic review to examine the implications of entrepreneurship policy within Europe. Examining and understanding the impacts of entrepreneurship policy and institutional changes are particularly relevant because of the billions of Euros invested and the impacts on the working lives of European citizens. By examining a broad range of existing literature, the study finds that the entrepreneurial Europe envisioned by policymakers has not been fully realised. Instead, entrepreneurship activity has skewed towards poor quality, necessity entrepreneurship. The European institutional context has also shifted away from the social model on which it was founded, increasing the exposure of European workers to social risks. To promote sustainable growth, wellbeing and well-functioning labour markets, researchers and policymakers are reconsidering the role of social protection. Social protection also has the potential to promote quality entrepreneurship. Based on the review of literature, seven testable propositions about how social protection can promote quality entrepreneurship have been developed for future empirical testing. This study advances knowledge in entrepreneurship research and contributes to debates in policymaking and practice. It also provides a sound basis for subsequent empirical research.
... On the contrary, those who cannot find a formal job and passively become self-employed (necessity self-employed) are often forced to work in lower-end jobs and work longer hours due to the pressure of life and being limited by their skills [52], making them experience poor health [53] compared with opportunity self-employed. Furthermore, the opportunity self-employed start businesses mainly to increase (rather than maintain) their income or to become independent [54,55], which also helps improve their ability to invest in health. ...
Full-text available
Rural-to-urban migrant workers are at high risk of health inequalities in cities. Since labor is a central social determinant of health, this paper provided evidence on the health consequences of self-employment among mobile populations in developing countries. The cross-sectional data from the 2017 data of the China Migrants Dynamic Survey (CMDS) and the IV-Oprobit model are used to examine the effects of self-employment on health. The results showed that: (1) Self-employment was positively related to health; (2) among the self-employed, the health effects of opportunity self-employed are larger than those of necessity self-employed; (3) in the subsample analysis, the health effect of self-employment was greater for male and Han nationality migrant workers; (4) self-employment promotes health primarily through reducing manual labor, increasing flexibility time, job stability, financial rewards, and social integration directly or indirectly. Thus, focusing on improving the social security system, granting entrepreneurial subsidies, and optimizing the business environment mean boosting the positive effect of self-employment on economic development.
This study was designed to test if satisfaction with health and personal financial wellbeing mediates the relationship between prosocial motivations and exit intentions among social entrepreneurs. Using a sample of 317 social entrepreneurs, the partial least square structural equation modeling (PLS-SEM) revealed that prosocial motivation decreased the financial satisfaction of entrepreneurs, which increased their exit intentions. However, health satisfaction did not have a mediating effect on the relationship between prosocial motivation and exit intention. Moreover, adopting the multi-group analysis (MGA) technique, we found that the negative impact of prosocial motivation on financial satisfaction was stronger for males than for females, suggesting male entrepreneurs were more likely to experience lower financial satisfaction caused by prosocial motivation than female entrepreneurs. There was no evidence that gender moderated the relationship between prosocial motivation and health satisfaction.
Purpose The ideal-typical entrepreneur presents him/herself in the neoliberal iconography as an autonomous and pro-active individual who is highly engaged with his/her vocation. Nevertheless, empirical research on the actual work engagement of the self-employed is scarce. In addition, phenomena like “necessity self-employment” and “economically dependent self-employment” raise concerns about the potential eudaimonic well-being outcomes of these self-employed. In this study, it was therefore investigated to what extent necessity self-employment and economically dependent self-employment are associated to work engagement and whether this relation is mediated by intrinsic job resources. Design/methodology/approach The authors used data from the 2015 European Working Conditions Survey (EWCS) involving 5,463 solo self-employed participants. For analyzing the data, structural equation modeling (SEM) with the Lavaan package was used. Findings Both necessity self-employment and economically dependent self-employment were linked to poor work engagement, however, intrinsic job resources mediated both effects. Originality/value While previous studies have shown differences in hedonic well-being between opportunity/necessity entrepreneurs, and economically (in)dependent entrepreneurs, this study considers their distinct profiles regarding eudaimonic well-being. Eudaimonic well-being was deemed particularly relevant because of its implications for other outcomes such as life satisfaction, psychological well-being, ill-health, business performance and persistence in self-employment.
In observational studies, investigators have no control over the treatment assignment. The treated and non-treated (that is, control) groups may have large differences on their observed covariates, and these differences can lead to biased estimates of treatment effects. Even traditional covariance analysis adjustments may be inadequate to eliminate this bias. The propensity score, defined as the conditional probability of being treated given the covariates, can be used to balance the covariates in the two groups, and therefore reduce this bias. In order to estimate the propensity score, one must model the distribution of the treatment indicator variable given the observed covariates. Once estimated the propensity score can be used to reduce bias through matching, stratification (subclassification), regression adjustment, or some combination of all three. In this tutorial we discuss the uses of propensity score methods for bias reduction, give references to the literature and illustrate the uses through applied examples. © 1998 John Wiley & Sons, Ltd.
The validity of self‐report measures of subjective well‐being (SWB) was examined and compared with non‐self‐report measures using a sample of 136 college students studied over the course of a semester. A principal axis factor analysis of self‐ and non‐self‐report SWB measures revealed a single unitary construct underlying the measures. Conventional single‐item and multi‐item self‐report measures correlated highly with alternative measures, with theoretical correlates of SWB, and with a principal axis factor underlying five non‐self‐report measures of well‐being. Comparisons of family versus friend informant reports demonstrated the considerable cross‐situational consistency and temporal stability of SWB. Evidence of the discriminant validity of the measures was provided by low correlations of the various SWB measures with constructs theoretically unrelated to well‐being. It was concluded that conventional self‐report instruments validly measure the SWB construct, and that alternative, non‐serf‐report measures are useful for providing a comprehensive theoretical account of happiness and life satisfaction.
In observational studies, investigators have no control over the treatment assignment. The treated and non-treated (that is, control) groups may have large differences on their observed covariates, and these differences can lead to biased estimates of treatment effects. Even traditional covariance analysis adjustments may be inadequate to eliminate this bias. The propensity score, defined as the conditional probability of being treated given the covariates, can be used to balance the covariates in the two groups, and therefore reduce this bias. In order to estimate the propensity score, one must model the distribution of the treatment indicator variable given the observed covariates. Once estimated the propensity score can be used to reduce bias through matching, stratification (subclassification), regression adjustment, or some combination of all three. In this tutorial we discuss the uses of propensity score methods for bias reduction, give references to the literature and illustrate the uses through applied examples.
INTRODUCTION Every epidemiologist knows that observational studies are troublesome because of the potential for confounding, a condition that implies improper comparisons and potentially biased effect estimates. While standardization is still utilized, covariance adjustment via regression has long been the principal tool through which investigators try to recover “proper” comparisons and unbiased estimates. Regression techniques are now remarkably easy to implement due to the availability of powerful yet inexpensive computers and user-friendly statistical software. But with such ease comes the potential for misuse or misunderstanding, be it intentional or accidental (Berk 2004). Too many contemporary analysts, it seems to us, fail to appreciate the assumptions inherent in regression methods to say nothing of the hypothetical experiment their observational study surely aims to mimic. A key concern is that investigators alter their (often implicit) causal regression models based not on theory but on indicators of sampling variability (e.g., p-values) or other aspects of the relationship between dependent and independent variables. Models end up capitalizing on chance, being overfit, and otherwise misleading with respect to causal inference. The problem is not with regression technology itself, but with its application. Though clearly no panacea, we believe that propensity score methods may allow social epidemiologists to (i) better appreciate and more closely mimic experimental study designs, (ii) minimize approaches to model specification that rely on testing, and (iii) increase the transparency of inference. Accordingly, we believe propensity score methods are important and worthy of both study and use. Our goal here is to motivate, explain, and offer an example of how social epidemiologists might use the propensity score approach. Methodologically, we offer nothing especially new or ground-breaking. Instead we aim to synthesize and make accessible existing research and show that the combined use of a counterfactual framework for causal inference, an explicit causal contrast study design, and propensity score matching methods is a useful alternative approach to regression models. “Accessibility” is key since the relevant literature is both vast and often quite technical, if not impenetrable for non-statisticians. Even some of the published tutorials (e.g. D'Agostino 1998; Joffe and Rosenbaum 1999; Little and Rubin 2000) can present challenges, and none are tailored for social epidemiologists. Since social epidemiology is clearly interested in estimating the effect of neighborhood contexts on health outcomes (i.e., neighborhood effects) we use such an investigation as an example and departure point for discussion. Accordingly, we divide this chapter into four sections: (1) the counterfactual framework, (2) propensity score matching methods, (3) example, and (4) conclusion.
Subjects were exposed to two aversive experiences: in the short trial, they immersed one hand in water at 14 °C for 60 s; in the long trial, they immersed the other hand at 14 °C for 60 s, then kept the hand in the water 30 s longer as the temperature of the water was gradually raised to 15 °C, still painful but distinctly less so for most subjects. Subjects were later given a choice of which trial to repeat. A significant majority chose to repeat the long trial, apparently preferring more pain over less. The results add to other evidence suggesting that duration plays a small role in retrospective evaluations of aversive experiences; such evaluations are often dominated by the discomfort at the worst and at the final moments of episodes.
Pleasures of the mind are different from pleasures of the body. There are two types of pleasures of the body: tonic pleasures and relief pleasures. Pleasures of the body are given by the contact senses and by the distance senses (seeing and hearing). The distance senses provide a special category of pleasure. Pleasures of the mind are not emotions; they are collections of emotions distributed over time. Some distributions of emotions over time are particularly pleasurable, such as episodes in which the peak emotion is strong and the final emotion is positive. The idea that all pleasurable stimuli share some general characteristic should be supplanted by the idea that humans have evolved domain-specific responses of attraction to stimuli. The emotions that characterize pleasures of the mind arise when expectations are violated, causing autonomic nervous system arousal and thereby triggering a search for an interpretation. Thus pleasures of the mind occur when an individual has a definite set of expectations (usually tacit) and the wherewithal to interpret the violation (usually by placing it in a narrative framework). Pleasures of the mind differ in the objects of the emotions they comprise. There is probably a