Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland

Abstract

The main aim of this paper is to investigate the impact of becoming unemployed on the well-being of young people. Particularly, we examine to what extent this relationship is moderated by the local unemployment rate. To address this topic, we use a longitudinal dataset: Social Diagnosis: Objective and Subjective Quality of Life in Poland (2005-2015). Using a fixed effect model, we find that young individuals who lost a job feel worse off than their counterparts who remained employed, and that these effects differ by gender, with men being more susceptible to changes in the employment status. Moreover, we show that the impact of unemployment on the well-being of men is the smaller, the higher the unemployment rate is. However, contrary to expectations, this effect is driven by the decreasing well-being of employed men and not by a decline in the distress level of unemployed men when aggregate unemployment grows.

Full text

Magdalena Rokicka, Marta Palczyńska & Małgorzata Kłobuszewska
68
Unemployment and Subjective Well-being Among the Youth:
Evidence from a Longitudinal Study in Poland
Magdalena Rokicka*, Marta Palczyńska & Małgorzata Kłobuszewska
* E-mail address of the corresponding author: m.rokicka@ibe.edu.pl
STSS Vol 10 / Issue 3
Studies of Transition States and Societies
Abstract
The main aim of this paper is to investigate the impact of becoming unemployed on the well-
being of young people. Particularly, we examine to what extent this relationship is moderated
by the local unemployment rate. To address this topic, we use a longitudinal dataset: Social
Diagnosis: Objective and Subjective Quality of Life in Poland (2005-2015).
Using a fixed effect model, we find that young individuals who lost a job feel worse off than
their counterparts who remained employed, and that these effects differ by gender, with men
being more susceptible to changes in the employment status. Moreover, we show that the
impact of unemployment on the well-being of men is the smaller, the higher the unemployment
rate is. However, contrary to expectations, this effect is driven by the decreasing well-being of
employed men and not by a decline in the distress level of unemployed men when aggregate
unemployment grows.
Key words: unemployment, subjective well-being, youth, longitudinal data.
I. Introduction
In this paper, we address the issue of the relationship between both: the individual and aggregated
labour market situation and the subjective well-being of young people. In light of the growing
concern about deteriorating mental health and well-being (European Pact for Mental Health,
2008), especially during the new economic realities (WHO, 2015), it is important to identify the
association between the labour market situation and the subjective well-being of young people
and provide empirical evidences.
This paper builds on previous empirical studies, which analysed the impact of job loss and the
local unemployment rate on subjective well-being. Unexpected life events such as job loss could
depress subjective well-being. Losing a job can mean not only a significant decrease in income but
also a loss of non-monetary benefits associated with work, such as time structure, social contacts
or clearly defined life goals. However, the assessment of individual unemployment also involves
comparing to the labour market situation of the reference group. In a society with strong work
ethics, staying without work can be seen as a breach of the social norm. On the one hand, the
local unemployment rate is considered to proxy the strength of the social norm to work: the more
unemployment there is, the weaker the impact of becoming unemployed on subjective well-being
(e.g. Clark, 2003). On the other hand, there is a competing explanation: a higher unemployment
rate means bleak labour market prospects and poor chances for finding a new job; thus, it might
strengthen the negative impact of individual unemployment on well-being. Therefore, not only
do we analyse the subjective well-being of unemployed versus employed youth, but we also
investigate the moderating role of the regional unemployment rate.
Minimal research attention has been previously directed toward the analysis of these issues
among young people, therefore, our paper fills this gap. The novelty of our paper also lies in
analysing a post-transition country, while the majority of studies look at Western European and

Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland 69
Anglo-Saxon countries. To the best of our knowledge, there is no evidence of the moderating role
of the unemployment rate on the impact of job loss among youth in a post-transition country.
In addition, we analysed the situation of women and men separately. Previous findings indicate
that men are more vulnerable to changes in the labour market situation (e.g. Angelescu, 2008;
Artazcoz et. al., 2004). Men more often assign central importance to employment than women,
who perform alternative roles, which might be substitutes to employment (Cinamon & Rich, 2002).
Also, due to the gender pay gap, the job loss of men translates into more detrimental changes
in the financial situation of households and other dependent household members, which could
reinforce negative feelings of job loss especially in the case of the main breadwinner. Furthermore,
there are different social expectations towards men’s and women’s employment (Van der Meer,
2014), which lead us to conduct analysis separately by gender.
We take advantage of the longitudinal character of data and estimate individual fixed effect models
that solve the problem of unobserved heterogeneity present in cross-sectional studies.
In sum, we address the following research questions: First, what are the effects of individual
unemployment on the well-being of youth? Second, do the effects among youth differ between
men and women? Finally, how does the local unemployment rate moderate these effects?
The paper is organised as follows: Section II vividly reviews the literature on subjective well-
being and individual unemployment, the moderating role of regional unemployment and gender
differences in this respect. In Section III, we describe data and the methodology adopted to address
our research questions. Section IV presents the results, while section V summarises and concludes.
II. Literature and hypotheses
Individual unemployment and subjective well-being
A large volume of literature discusses the theories behind the negative relationship between
unemployment and well-being. Jahoda’s functional theory of employment still remains the most
influential among them. According to Jahoda (1981), besides visible outcomes such as remuneration,
employment also has other latent functions that are beneficial for a person’s well-being (e.g., time
structure, social contact, shared experiences and goals, activity). The lack of employment deprives
people of these beneficial functions and via this channel decreases their well-being (for more
theories linking unemployment and well-being see e.g. Ezzy (1993), McKee-Ryan, Song, Wanberg,
& Kinicki (2005)). According to van der Meer (2014), unemployment influences the individual’s
well-being via three channels. First, unemployment deprives a person of stimulation and comfort
that comes with the job, which are measured by job satisfaction and income. Second, a job and
income define the social status of a person, and people in general prefer to take higher positions
within society, while the social status of an unemployed person is rather low. Last, most people
desire social approval, and having a job means adhering to the social norms to work. There is a
vast amount of empirical literature on the detrimental consequences of job loss for well-being
(e.g. Winkelmann (2009), Clark, Knabe, & Rätzel (2010) for Germany, Theodossiou (1998) for Britain,
Ervasti & Venetoklis (2010) for 21 EU countries). Clark and Oswald (1994) show for Great Britain,
that the negative effect of the unemployment on well-being is larger than the effect of divorce or
separation.
The few existing studies focusing on youth investigate the effects in Western economies. For
example, the influence of short and long-term unemployment on youth well-being and the
mitigating impact of manpower programs were examined by Korpi (1997) for Sweden. His results
indicate that participation in manpower programs improves the well-being of unemployed youth
and makes it indistinguishable from the well-being of employed youth, while unemployment
benefits do not have such a beneficial impact (when comparing the unemployed with and without
benefits). This suggests that employment and activities similar to employment are important

Magdalena Rokicka, Marta Palczyńska & Małgorzata Kłobuszewska
70
factors of youths’ well-being, not necessarily their financial situation. Clark and Oswald (1994)
show that the well-being of young people (under age 30) is less affected by unemployment than
the well-being of older people. A plausible explanation is that young people are more often
unemployed than prime age workers, thus it is easier for them to accept such a situation. On
the contrary, the meta-analysis by Paul and Moser (2009) suggests that the relation between age
and psychological distress connected with unemployment is rather U-shaped, thus youth and
the elderly are more vulnerable than the prime age population, although this effect is not stable
among analysed studies.
Moderating effect of unemployment rate on subjective well-being
Are these negative effects moderated by the local unemployment rate? There are two opposite
theoretical arguments about the direction of the impact. On the one hand, the local unemployment
rate can be considered as a proxy for social norms to work. The lower the unemployment rate,
the stronger the norm to work. Such an approach implies that an individual’s own unemployment
impacts their well-being less when there is a higher level of unemployment among the reference
group/relevant others (Clark, 2003). The underlying mechanism assumes that well-being is a proxy
for utility and that the main sources of unemployed individuals’ utility are benefits and leisure
time. When the norm to work is strong, the unemployed might enjoy leisure time less, as they
can be subject to social sanctions for deviating from the norm (Stutzer & Lalive, 2004). Also, the
concept of adaptive preferences, first introduced by Elster (1982), predicts that an individual’s own
unemployment impacts their well-being less when there is a higher level of unemployment among
the reference group, but with a different causal mechanism. Adaptive preferences assume that
people adjust their aspirations to feasible possibilities (Elster, 1982). When work is more difficult to
find, people downgrade this inaccessible option and, consequently, job loss has a smaller impact
on their subjective well-being. Moreover, when the unemployment rate is high, job loss may be
easier attributed to external factors resulting in less guilt and shame for unemployment, thus
moderating its negative effects.
On the other hand, the higher unemployment rate might strengthen the negative impact of
individual unemployment on well-being, as it indicates poor labour market prospects and lower
chances of finding a new position. Therefore, it is often noted that a higher unemployment rate
lowers the well-being of the whole workforce, also of those in employment (e.g. Turner, 1995;
Clark, 2003). This effect might be due to the increasing subjective job insecurity and the weakening
bargaining power of workers when the surrounding unemployment rate grows.
Studies examining the effects of the regional unemployment rates within a country find a positive
effect of unemployment rates on individual experiences of unemployment, as predicted by social
norm, and adaptive preferences interpretations (Clark, 2003; Clark et al., 2010; Stutzer & Lalive,
2004), no effect at all (Dooley, Catalano, & Rook, 1988; Oesch & Lipps, 2013; Strandh, Novo, &
Hammarström, 2011) or even a negative effect (Turner, 1995). On the other hand, when analysed on
the country level, the unemployment rate either does not interact with personal unemployment
experiences in terms of well-being (Eichhorn, 2013) or exacerbates its negative consequences on
well-being (Calvo, Mair, & Sarkisian, 2015). A meta-analysis by McKee-Ryan et al. (2005) finds no
relation between the unemployment rate at the time of the study and the effects of unemployment
on well-being.
Very few studies have analysed the moderating role of the unemployment rate on the well-
being and health of unemployed youth (see: Scanlan & Bundy, 2009 for an exception). Australian
data show that the unemployed aged 18-24 report poorer mental health during a period of low
unemployment compared to a period of high unemployment. The major drawback of the study
is its methodology - a comparison of means - which does not control for neither observed nor
unobserved heterogeneity.
Most of the literature concentrates on Western Europe and Anglo-Saxon countries, and the research
on post-transition countries is limited, although their socio-economic situation is specific. On the

Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland 71
one hand, these countries experience on average higher unemployment rates, and after transition
unemployment became a relatively common phenomenon, suggesting that the norm to work
may be weaker. On the other hand, there are also cultural differences. The negative effect of
unemployment on well-being could be intensified with very strong social norms inherited from
communism, an era when the non-working people were stigmatised (Selezneva, 2011). During the
times of communism and central planning economy, employment was the obligation of each able-
bodied citizen. For example, in Poland every secondary school or university graduate received a
work order and could not resign from work for 2-3 years (Act of 7 March 1950..., 1950). There was
also massive propaganda against people evading work. The lack of employment without any good
cause indicated a lack of willingness to work and “leading antisocial parasitic way of life” (Porket,
1995, p. 34). Therefore, unemployed faced social ostracism and legal consequences (Porket, 1995).
To the best of our knowledge, there are only two studies on the moderating role of the
unemployment rate on the individual experience of unemployment in post-transition countries.
Using panel data for Russia, Eggers et al. (2006) find that the regional unemployment rate
increases the well-being of both the employed and the unemployed, with the effect stronger for
unemployed. The authors discuss the unusual effect for the employed in terms of the change of
expectations: when individuals’ reference groups fare worse in a troubled economy, people lower
their standards and appreciate more what they have, reporting being better off than in times
of higher unemployment. The methodology of this analysis differs from earlier contributions by
employing split-sample regressions for the employed and unemployed. The analysis of Kazakhstan
(Kalyuzhnova & Kambhampati, 2008) looks at three different years interpreted as different stages
of economic transition. Authors show that the higher regional unemployment rate increases the
well-being of both the unemployed and employed at the beginning of the market reforms and
during their maturity, while in the post-reform period the unemployment rate has a positive impact
only on the well-being of the employed. It would be worthwhile to gather more empirical evidence
to see which effect prevails, as there are different scenarios of transition in these countries.
For Poland, Angelescu (2008) analyses the levels of well-being in 1990 and 1999. She shows that
the well-being of Poles was lower in 1999 compared to 1990, but the difference disappears when
employment status and marital status are accounted for. According to the results of her pooled
analysis, being unemployed has a detrimental effect on subjective well-being. Unfortunately, this
study did not investigate whether the effect of individual unemployment on well-being differs
between 1990 when the unemployment rate was low and 1999 when it was much higher.
Gender differences in response to unemployment
Many studies indicate that men and women respond differently to unemployment. This may be the
result of different social norms and expectations towards genders (Van der Meer, 2014). The World
Bank report on gender equality and development (World Bank, 2011) confirms such conclusions.
The qualitative study shows that having a job is still the basic determinant of a man’s position in
the society. Having a high-quality job and above-average income makes a man a good husband and
increases his self-esteem, while the lack of work disturbs these feelings and leads to frustration. In
many cultures, expectations towards women are different. Women can be successful as mothers
and wives, not necessarily as employees, and their position in their families gives them comfort.
Artazcoz et al. (2004) show that being married and having children are protective factors for
unemployed women’s well-being but not for men’s. In the case of unemployed men, being married
even has an opposite effect.
The fact that negative effects of unemployment are different for women and men is also a result of
cultural determinants, as was shown by Strandh et al. (2013). In more traditional gender regimes,
unemployment has a stronger negative connection with mental health among men than among
women. In countries where gender roles are more equal this difference does not occur. Poland
and other post-transition countries have rather traditional gender roles. In the aforementioned
study on Poland, Angelescu (2008) shows that men are more affected by unemployment than
women in terms of well-being. Similar conclusions for other post-transition countries are derived

Magdalena Rokicka, Marta Palczyńska & Małgorzata Kłobuszewska
72
from Selezneva’s review (Selezneva, 2011). Also, in declarations work is less important for women
than for men in Poland. According to the European Values Study (2008), 58.9% of men attach high
importance to work while for women it is 53.6%. This gender gap (5.3 pp.) is above the average for
EU countries (4.1 pp.). There are also more women than men in Poland who do not consider work
as a meaningful part of their life.1
A similar conclusion about gender differences could be drawn from Jahoda's theory (Jahoda, 1981).
The author emphasises that work can take different forms than just employment, in the sense that
other forms of engagement that are not regular work, such a household tasks, can also provide the
beneficial functions of employment.
Based on the above arguments, we formulate the following hypotheses:
Hypothesis 1: Becoming unemployed depresses the individual subjective well-being of youth.
Hypothesis 2: Becoming unemployed affects men more severely than women.
Hypothesis 3: A high local unemployment rate mitigates the impact of an individual’s own unemployment
on well-being.
Hypothesis 4: The mitigating effect of the local unemployment rate on own unemployment impact on
well-being is stronger for men.
III. Research design
Data
We draw on the most comprehensive dataset on well-being in Poland – The Social Diagnosis:
Objective and Subjective Quality of Life in Poland (Social Diagnosis ... ,2015). This biennial household
panel collects information about the economic and financial situation of households, as well as
its members. Individuals aged 16 and above are interviewed individually, providing information
about their labour market experiences, health, psychological well-being, lifestyle, engagement
in arts and cultural events. Each round of the study also consists of special thematic, ad hoc
modules. In order to avoid seasonal effects, the field work is conducted in the spring, and the
panel is refreshed in each wave with new households sampled. Our analysis uses data from six
waves collected in 2005-2015. The sample is restricted to individuals aged 18-35, who in the first
observance already finished their education and were employed or searching for work. The age
limit for all waves lies between 18 and 35 years old. This means that all individuals in the first wave
(2005) are of a minimum age of 18 and a maximum age of 25 years old if they participate in all six
waves, or of the maximum age of 33 if they participate in two waves. The adoption of such an
age limit is related to the transition from education into the labour market, which predominantly
happens after completing secondary or vocational school at the age of 18/19 or after finishing
tertiary education at the age of 24. By applying such a broad age limit, we also reflect trends
in postponing and delaying the transition to adulthood observed in Poland, with young people
delaying procreation decisions and family formation till the age of 30. This trend is also taken into
consideration by policy makers in several EU countries, where the age of 35 is used as the limit in
the definition of youth. As stated in the comparative report of national legislation toward youth
in Europe, “Difficult inclusion of young people into labour market is usually seen as the main
reason to prolonging youth age. Consequently, it may be expected that more and more European
countries would keep up with this trend, moving the upper limit further towards 35 or even 40
years for some policy areas.” (Perovic, 2015, p.4).
1 Calculation based of European Values Study 2008, accessed on the 15/02/2018 from http://zacat.gesis.org/webview/index.
jsp?object=http://zacat.gesis.org/obj/fCatalog/Catalog5

Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland 73
Measures
Our dependent variable is individual self-reported subjective well-being, based on the question,
‘Considering all, how would you assess your current life - would you say you are…’ with answers
ranging from 1 (unhappy) to 4 (very happy). The way the question is constructed differs from the
standard, widely used definitions. One of its weaknesses is that while the question itself refers
mostly to the concept of life satisfaction, answers are related to life happiness. However, both
conform to the general definition suggested by Diener, Suh, Lucas, & Smith (1999), according to
which subjective well-being is a very wide concept. Consistently with this approach, our question
measures the pleasant effect of well-being – individuals’ instant evaluation of life events (e.g., joy,
elation, pride, affection, happiness). The distribution of this variable is presented in Figure 1. The
majority of respondents declare that they are rather happy.
Figure 1: Distribution of the dependent variable “Considering all, how would you assess your current
life - would you say you are”
Note: Pooled sample, n=8035.
Source: Own calculation based on The Social Diagnosis. Integrated database.www.diagnoza.com
[download in July 2017].
Our key independent variable is binary and defined on the basis of the current labour market
status: 1 for those who are unemployed and 0 for those employed. The analysis focuses on the
individuals active in the labour force. We decided to exclude the inactive population due to the
huge heterogeneity of this group. Among the inactive group, we have those who are in full-time
education, so unable to work, people with serious health issues, so as well excluded from the
labour market, people who care for others (children, other family members), and those who
are not willing to work. Many diversified motivations are behind the decisions of those groups
to be away from the labour market, and the impact of their employment status on well-being
constitutes a separate topic for further analysis, not addressed in this paper.
In a pooled person-wave sample, there are 19% of the unemployed, and this level varies by year,
with the highest share of the unemployed in 2005 (29%) and the lowest in 2015 (15%). There
are also visible gender differences: there are 23% of unemployed women compared to 16% of
unemployed men (more details in Figure A. 1 in the appendix). It is important to mention here that
unemployed young people have the same access to unemployment benefits as prime age workers,

Magdalena Rokicka, Marta Palczyńska & Małgorzata Kłobuszewska
74
however, in Poland the basic condition for receiving unemployment benefits is at least one year of
work with at least minimum remuneration, during which social contributions were paid. These
conditions might be difficult to meet for younger workers, who more often have temporary jobs,
internships, short-time job placement, or are entering the job market for the first time.
Table 1: Descriptive statistics
Total Men Women
Mean sd Mean sd Mean sd
Subjective well-being 3.00 0.54 2.99 0.54 3.00 0.54
Unemployed 0.19 0.40 0.16 0.37 0.23 0.42
NUTS2 unemployment
share 8.16 2.14 7.42 1.91 8.88 2.10
Nr of friends 6.77 6.77 7.38 7.61 6.17 5.76
Age 28.48 4.49 28.21 4.57 28.75 4.39
Age squared 831.13 251.33 816.39 254.85 845.57 247.01
Married 0.44 0.50 0.37 0.48 0.50 0.50
Lost a close person 0.16 0.37 0.15 0.36 0.17 0.37
Seriously ill 0.04 0.20 0.03 0.18 0.05 0.22
Civic engagement 0.10 0.30 0.11 0.31 0.09 0.29
Religious activity 2.10 2.15 1.86 2.14 2.34 2.15
Number of observations 8035 3976 4059
Note: Pooled sample, N=8063.
Source: Own calculation based on The Social Diagnosis. Integrated database.www.diagnoza.com
[download in July 2017].
Another important independent variable is the regional unemployment rate by gender. We proxy
it by a gender specific regional share of unemployed in the working age population, provided
by the Polish Statistical Office2, at NUTS2 level (województwo), as data on the unemployment
rate is not available by region. This share is slightly lower than the unemployment rate in our
sample, as it covers the whole working age population, not only youth, as in our dataset, and it
uses a much larger denominator – working age population instead of just economically active
population. However, the yearly and gender trends in both indicators are the same (Figure A.1).
Correspondingly to our sample, the highest share of unemployed was observed in 2005, with
higher unemployment among women. The regional unemployment share varies from 7% to 14%
for women, and from 6% to 11% for men.
We also control for several time-varying, personal characteristics and events, which could affect
subjective well-being in the given year of the study, such as: age, marital status, number of close
friends, a loss of a close person in the current year, and serious illness in the current year (Table
1). Marital status is a binary measure defined 0 for single and 1 for married individuals. As some
changes in life circumstances are related to distressful events such as a death in the network of
close family or friends or serious health problems are confirmed to be negatively associated with
subjective well-being (Suh, Diener, & Fujita, 1996), and they are included in all specifications as
binary indicators. Based on previous empirical evidence, which confirmed the positive relationship
between community involvement and subjective well-being (Helliwell & Putnam, 2004), we also
control for membership in nongovernmental organisations, as well as for engagement in religious
activities, which is also positively related to subjective well-being (Dolan, Peasgood, & White,
2008). Time-invariant controls include education and gender (in the full sample specification only).
However, due to the application of an interaction effect only in the fixed term model, they are not
directly controlled for in the main specification.
2 Variable: Share of the registered unemployed persons in the population in the working age by sex, accessed on the
02/08/2017 from https://bdl.stat.gov.pl.

Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland 75
Method
To address the first and the second hypotheses, the following specification is used:
To test the third and the fourth hypotheses, we include interaction terms between own employment
status and the regional unemployment rate:
, where:
- represents the well-being of an individual i at time t,
- represents the individual fixed effect,
- is a binary indicator of being unemployed (versus employed),
- is a measure of regional unemployment share,
- is a set of individual control variables,
- is a wave indicator,
- represents an error term
Both specifications are estimated using hybrid models (Allison, 2009) with additionally clustered
standard errors on a regional level. This is one of the methods that takes advantage of longitudinal
aspects of data and bears both the characteristics of fixed effects and random effects models,
measuring the same individuals over a few years. This approach has an advantage over simple
cross-sectional regressions. Most of all, it allows for the estimation of the real effect of change of
variable of interest on the change of dependent variable in the give time span, not only association,
as in the case of cross-sectional data.
This approach allows us to check if the individual change in labour market status translates into
changes in individual subjective well-being, and how the changes in regional unemployment rates
moderate the effect of job loss on well-being. Second, by adding time-invariant variables such
as education we can also assess their relation to changes in subjective well-being. As argued by
Allison (2009), the hybrid method offers considerable advantages over simple fixed effect estimator,
allowing for estimation of both time-variant and time-invariant covariates. This approach allows
for decomposing the time-varying predictors into two components: the first one describing the
fixed effects estimates within-person variation and the second one capturing the random effects
between persons estimates variation (which are equal to random intercept model coefficients,
based on the weighted average between and within estimates), both used as predictors in the
model (Schunck et al., 2013). The within-person estimator depicts the relationship between a
predictor and an outcome. Its interpretation is straightforward: if a predictor varies across time,
the outcome decreases or increases by one unit (for a continuous variable). In contrast, the
between estimator represents the effect of the variation of the predictor on the outcome across
respondents.
We treat our dependent variable as a cardinal one, which allows us to apply linear methods. This
is a common practice in life satisfaction and well-being estimations (Clark et al., 2010; Oesch &
Lipps, 2013).

Magdalena Rokicka, Marta Palczyńska & Małgorzata Kłobuszewska
76
IV. Results
Before moving to the multivariate analysis in a longitudinal setting, we first checked if the
distribution of the subjective well-being varies by employment status. In accordance with our
expectations, Figure 2 indicates that unemployed young people more often report being unhappy
and rather unhappy than their working counterparts. The opposite can be said about positive well-
being, which is higher among the employed than the unemployed youth.
We then use the advantage of observations of the same individual over several years to calculate
the mean subjective well-being score, conditional on a labour market status change between two
consecutive waves (which in our data means a 2-year period).
Figure 2: Subjective well-being by employment status
Note: Pooled sample, employed n=6969, unemployed n=1634.
Source: Own calculation based on The Social Diagnosis. Integrated database.www.diagnoza.com
[download in July 2017].
Using longitudinal data, Table 2 compares the change in the subjective well-being of two groups:
those who stayed in employment and those who lost their job. Those who remained employed in
the two consecutive waves had on average higher subjective well-being in the initial year, and it
increased in the consecutive wave. The opposite can be said about those who become unemployed:
not only did they have lower subjective well-being while still in employment, but their subjective
well-being also decreased with job loss. We observe the same pattern for both genders. These
results indicate that a simple comparison of employed versus unemployed individuals, as in a
cross-sectional setting, can be biased by the presence of certain selection.
Table 2: Average level of subjective well-being by changes of labour market status and gender
TOTAL MEN WOMEN
T T+2 DIFF T T+2 DIFF T T+2 DIFF
EMP-->EMP 3,034 3,057 0,023 3,032 3,055 0,023 3,038 3,059 0,021
EMP-->UNEMP 2,920 2,750 -0,170 2,929 2,728 -0,201 2,913 2,774 -0,139
Source: Own calculation based on The Social Diagnosis. Integrated database.www.diagnoza.com
[download in July 2017].

Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland 77
This descriptive finding is in line with our expectations based on theory and previous research.
However, it is a raw effect, without accounting for other personal time variant characteristics and
conditions of local labour markets. Therefore, we moved forward to econometrical multivariate
analysis. We estimated six regressions, presented in Table 3. Model 1 refers to the first specification,
and Model 2 refers to the second specification with the interaction effect. Both specifications
are estimated on the total population and on men and women separately. We first present the
coefficients from within estimation (fixed effect) and then from between individuals’ estimation
(random effect).
Results from within estimation in model 1 indicate that when they become unemployed young
individuals have on average 0.14 (which is 1/4 of standard deviation) lower well-being, compared
to the moment when they work, all others constant. Thus, hypothesis 1 can be confirmed. To
give an idea of the effect’s size, the detrimental effect of unemployment on well-being outweighs
the beneficial impact of getting married (0.10) and is similar to that of becoming seriously ill. As
discussed in the theoretical part, there are several channels via which job loss affects subjective
well-being. One of them is income, which is argued to have a positive although diminishing impact
on well-being (Clark, Frijters, & Shields, 2006). The loss of personal earnings due to unemployment
also translates into a loss of relative income, which is strongly associated with subjective well-being.
Our model does not account for income, as our goal is to estimate the whole effect of becoming
unemployed on well-being, also partly driven by income loss. When income is controlled for, the
coefficient of change of labour market status is treated as if the income remains ceteris paribus, so
assuming that job loss does not involve an income change. As a sensitivity check, we have estimated
models with a proxy for household income3 (results are reported in Table A.1 in the Appendix,).
According to our expectations, when income proxy is taken into account the negative impact of
becoming unemployed on subjective well-being diminishes slightly but remains significant (the
coefficient changes from 0.136 to 0.126). Also, the effect of local unemployment share on individual
subjective well-being is significant and does not differ qualitatively from the main specification. As
even after controlling for family income the detrimental effect of unemployment is persistent, it
suggests that a part of the observed effect is also related to the non-financial aspects of job loss,
as predicted by the functional theory of Jahoda (1981), described in the previous part of this paper.
3 Proxy of household income is a dummy variable indicating those households in which the income is sufficient to cover
all basic needs.

Magdalena Rokicka, Marta Palczyńska & Małgorzata Kłobuszewska
78
Table 3: Impact of unemployment and unemployment share on subjective well-being: mixed
effects coefficients
(1)
Total
(2)
Total
(3)
Men
(4)
Men
(5)
Women4
(6)
Women
Within individual estimates
Unemployed -0.136*** -0.333*** -0.180*** -0.404*** -0.100* -0.189
(0.025) (0.073) (0.031) (0.096) (0.046) (0.152)
NUTS2 unemployment -0.014* -0.018** -0.022** -0.026*** -0.005 -0.007
(0.007) (0.007) (0.007) (0.007) (0.012) (0.012)
Unempl.# NUTS2 unempl. 0.023** 0.029* 0.010
(0.008) (0.013) (0.015)
Nr of friends 0.003 0.003 0.004 0.004 0.002 0.002
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
Age 0.060*0.059*0.038 0.037 0.074*0.074*
(0.029) (0.029) (0.049) (0.049) (0.034) (0.034)
Age squared -0.001 -0.001 -0.001 -0.001 -0.001*-0.001*
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Married 0.096** 0.095** 0.133** 0.133** 0.065 0.064
(0.032) (0.032) (0.048) (0.048) (0.049) (0.049)
Lost a close person -0.028*-0.028*-0.076*** -0.076*** 0.013 0.013
(0.012) (0.012) (0.022) (0.021) (0.022) (0.022)
Seriously ill -0.143** -0.140** -0.187*-0.181*-0.111** -0.111**
(0.049) (0.049) (0.087) (0.087) (0.037) (0.036)
Civic engagement 0.018 0.017 0.010 0.009 0.024 0.024
(0.023) (0.023) (0.045) (0.045) (0.032) (0.032)
Religious activity 0.014** 0.014** 0.018** 0.018** 0.011 0.011
(0.005) (0.005) (0.006) (0.006) (0.007) (0.007)
Between individual estimates (only time invariant)
Men -0.011 -0.011
(0.011) (0.011)
Basic vocational and lower -0.055** -0.055** -0.046*-0.047*-0.069*-0.069*
(0.020) (0.020) (0.024) (0.023) (0.027) (0.027)
Post-secondary and
tertiary 0.099*** 0.099*** 0.094** 0.094** 0.107*** 0.107***
(0.019) (0.020) (0.029) (0.029) (0.028) (0.028)
Constant 3.842*** 3.842*** 4.195*** 4.171*** 3.573*** 3.580***
(0.254) (0.240) (0.419) (0.415) (0.561) (0.548)
Observations 8035 8035 3976 3976 4059 4059
Number of individuals 3844 3844 1951 1951 1893 1893
Standard errors in parentheses, * p<0.05, ** p<0.01, *** p<0.001. Reference Base category: for
unemployed: employed; for married: single, divorced, or widowed; for men: women; for education:
general secondary education.
Source: Own calculation based on The Social Diagnosis. Integrated database.www.diagnoza.com
[download in July 2017].
In line with hypothesis 2, the detrimental effect of becoming unemployed on subjective well-being
is larger among men, while it is smaller and statistically insignificant for women (see Models 3
and 5). Gender differences in attitudes, beliefs, aspiration and societal roles could be the potential
explanation of this result. This result seems to confirm the norm that men in Poland are seen
as the main breadwinners and income providers, so their labour market situation influences
4 There are visible gender differences in the statistical significance of estimated coefficients. In the regression done on
the subsample of women, only four explanatory variables are statistically significant, however the overall F-test for the model
indicates that our variables are jointly significant at 0.0003 level.

Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland 79
their perception of self-worth and well-being more than women. In fact, there are differences in
attitudes toward employment in our sample, where 53% of young men considered work to be the
most important condition of a successful, happy life, while only 40% of young women agreed with
this statement.
We also find evidence that subjective well-being is affected by changes in the local labour market.
An increase by 1 p.p. in the regional unemployment share in working age population diminishes
well-being by 0.014 for the overall population (Model 1) and by 0.22 for men (Model 3), with no
statistically significant evidence for women (Model 5).
To test hypothesis 3 and hypothesis 4, we estimated models with an interaction term between
the level of regional unemployment share and individual employment status (specification 2). Our
results (Models 2, 4 and 6) reveal that this interaction effect is positive and statistically significant
for the total population and men. Looking at the specific results for men (Model 4), subjective
well-being of employed decreases by 0.026 when regional unemployment share grows by 1 p.p.,
while the well-being of the unemployed is not affected by the regional unemployment change (β2
+ β3 = 0.003). Figure 3 illustrates the predicted values of subjective well-being for unemployed and
employed men at different NUTS-2 unemployment levels. These findings contradict hypotheses 3
and 4 and suggest that the impact of aggregated unemployment on the well-being of unemployed
men is smaller the higher the unemployment rate is, but this effect is driven by the decreasing
well-being of employed men and not by a decline in the distress level of unemployed men. There
are also differences in the response to regional unemployment rates between women and men,
but only among employed individuals.
Figure 3: Linear prediction of subjective well-being by employment status and the regional
unemployment share for men
Source: Own calculation based on The Social Diagnosis. Integrated database.www.diagnoza.com
[download in July 2017].
For employed men, the higher the local unemployment the lower their subjective well-being is.
Most likely they perceive it as a threat to their current job. As shown by Clark (2010), the impact of
local unemployment on employees’ well-being is heterogeneous and depends on the assessment
of their labour market security.

Magdalena Rokicka, Marta Palczyńska & Małgorzata Kłobuszewska
80
Other controls used in the specification confirm our expectations based on the theoretical
background and previous empirical findings. Age, marital status, health, and religious activities
are among important determinants of changes in subjective well-being. Moreover, the analysis
of the time-invariant variables indicates that the higher the educational level of an individual
the more satisfied with his/her life he/she is. The coefficient of gender (Models 1,2) is statistically
insignificant due to the adoption of local unemployment share by gender, so the effect of gender is
already captured by this variable. Since they led to different conclusions, models done separately
for men and women confirmed gender differences in the subjective well-being and the gender
variation in the scope and strength of influence of the labour market situation.
V. Summary and Conclusions
This article complements previous research on the effects of unemployment on subjective well-
being and the moderating role of the local unemployment rate (e.g. Clark, 2003; Oesch & Lipps,
2013; Strandh et al., 2011) focusing on young people.
Using the most recent waves of The Social Diagnosis dataset (2005-2015), we find that becoming
unemployed depresses the individual’s subjective well-being of youth in Poland. This result indicates
that the young generation in Poland still values work highly and attaches a high importance to
employment, similarly to that reported for the working age population in Western countries.
Our results are also gender specific, indicating that young men are more affected by a job loss
than young women (the effect is smaller by almost a half for women), which is in line with most
previous empirical evidence and confirms the existence of traditional gender roles in Poland.
Moreover, we show that the impact of becoming unemployed on men's well-being decreases when
the regional unemployment rate increases. However, unlike expectations, this effect is driven by
the decrease in the well-being of employed men and not by a decline in the distress level of
unemployed men. The results of the negative impact of the regional unemployment rate on those
in employment are in line with previous findings from Western economies (Novo, Hammarström,
& Janlert, 2001). Novo et al. (2001) claim that the health effect of high unemployment may be
associated with growing pessimism about the future and inferior financial prospects. Brenner &
Mooney (1983) argue that an increase in the unemployment rate reduces job security and increases
stress related to the potential risk of a job loss. Others argue that during high unemployment,
more people accept undesirable job offers with lower remuneration, which leads to distress and
dissatisfaction (Catalano & Dooley, 1983). An additional perspective is provided by Clark et al.
(2008), who claimed that the effect of aggregate unemployment on subjective well-being differs
not only by the employment status, but rather by the employment opportunities and job prospects.
All these expectations are in line with the theoretical approach suggested by Blanchard (Blanchard,
2016; Blanchard & Giavazzi, 2001). According to this approach, the individual and collective
bargaining power of workers is influenced by the aggregate unemployment rate. When the
unemployment rate is high or is growing, it is harder for workers to change a job, while it is easier
for an employer to find a new worker. Such a situation translates into the weaker bargaining power
of workers, with all adverse consequences: lower wages, less favourable working conditions, less
opportunities for promotion, and, as a result, lower job satisfaction and a decline in the overall
well-being.
The negative impact of the regional unemployment rate on the well-being of those in employment
contradicts, however, the evidence from post-transition countries where an increase of regional
unemployment rates translated into a higher well-being of employed (Eggers et al., 2006;
Kalyuzhnova & Kambhampati, 2008). One possible explanation for this disparity between the
results for Russia and Kazakhstan and our results for Poland might be that there are country
disparities with regard to the stage of transition and economic situation. While in the former
studies the investigated economies were currently in transition, our study uses more recent data
from 2005-2015, when the Polish economy was much more stable.

Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland 81
In sum, a job lost causes a decline in the subjective well-being of both young men and young
women in Poland. Moreover, aggregate unemployment decreases the well-being of employed
men, with no effect on women. Those results have important policy implications, showing that
unemployment affects the psychological health not only of the person directly affected but also
of those who are in employment. Therefore, when assessing the societal costs of unemployment,
the effects of aggregate unemployment should be accounted for. For that reason, policies aimed
at bringing young people back to work should provide more incentives and help in searching
more intensively for a new job, while also providing interventions aimed at strengthening the
bargaining power of those in employment.
References:
Act of 7 March 1950 on preventing the liquidity of human resources in occupations or specializations
particularly important for the socialized economy. (1950). Retrieved from http://prawo.
sejm.gov.pl/isap.nsf/DocDetails.xsp?id=WDU19500100107
Allison, P. D. (2009). Fixed Effects Regression Models (Vol. 160). Thousand Oaks, CA: SAGE publications.
Angelescu, L. (2008). The effects of transition on life satisfaction in Poland. Center for the Study of
Democracy Working Paper.
Artazcoz, L., Benach, J., Borrell, C., & Cortès, I. (2004). Unemployment and mental health:
understanding the interactions among gender, family roles, and social class. American
Journal of Public Health, 94(1), 82-88.
Blanchard, O. (2016). Macroeconomics (7th ed.). Retrieved from https://www.pearsonhighered.com/
program/Blanchard-Macroeconomics-7th-Edition/PGM333935.html
Blanchard, O., & Giavazzi, F. (2001). Macroeconomic effects of regulation and deregulation in goods
and labor markets (Working Paper No. 8120). National Bureau of Economic Research.
Retrieved from http://www.nber.org/papers/w8120
Brenner, M. H., & Mooney, A. (1983). Unemployment and health in the context of economic change.
Social Science & Medicine, 17(16), 1125-1138.
Calvo, E., Mair, C. A., & Sarkisian, N. (2015). Individual troubles, shared troubles: The multiplicative
effect of individual and country-level unemployment on life satisfaction in 95 nations
(1981-2009). Social Forces, 93(4), 1625-1653.
Catalano, R., & Dooley, D. (1983). The health effects of economic instability: A test of the economic
stress hypothesis. In J. John, D. Schwefel & H. Zöllner (Eds.), Influence of Economic
Instability on Health (pp. 225-262). New York: Springer.
Cinamon, R. G., & Rich, Y. (2002). Gender differences in the importance of work and family roles:
implications for work–family conflict. Sex roles, 47(11-12), 531-541.
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., Frijters, P., & Shields, M. A. (2006). Income and happiness: Evidence, explanations and
economic implications. Retrieved from https://hal.archives-ouvertes.fr/halshs-00590436/
Clark, A. E., & Oswald, A. J. (1994). Unhappiness and unemployment. The Economic Journal, 104(424),
648-659. https://doi.org/10.2307/2234639
Clark, A., Knabe, A., & Rätzel, S. (2010). Boon or bane? Others’ unemployment, well-being and job
insecurity. Labour Economics, 17(1), 52-61. https://doi.org/10.1016/j.labeco.2009.05.007
Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three decades of
progress. Psychological Bulletin, 125(2), 276-302.
Dolan, P., Peasgood, T., & 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.
Dooley, D., Catalano, R., & Rook, K. S. (1988). Personal and aggregate unemployment and psychological
symptoms. Journal of Social Issues, 44(4), 107-123. https://doi.org/10.1111/j.1540-4560.1988.

Magdalena Rokicka, Marta Palczyńska & Małgorzata Kłobuszewska
82
tb02094.x
Eggers, A., Gaddy, C., & Graham, C. (2006). Well-being and unemployment in Russia in the 1990s:
Can society’s suffering be individuals’ solace? The Journal of Socio-Economics, 35(2), 209-
242.
Eichhorn, J. (2013). The (non-)effect of unemployment benefits: Variations in the effect of
unemployment on life-satisfaction between EU countries. Social Indicators Research,
119(1), 389-404. https://doi.org/10.1007/s11205-013-0474-9
Elster, J. (1982). Sour grapes – utilitarianism and the genesis of wants. In A. Sen & B. Williams (Eds.),
Utilitarianism and Beyond. Cambridge: Cambridge University Press.
European Pact for Mental Health and Wellbeing (2008). Brussels. https://www.researchgate.net/
publication/236157279_European_Pact_for_Mental_Health_and_Wellbeing
Ervasti, H., & Venetoklis, T. (2010). Unemployment and subjective well-being: An empirical test of
deprivation theory, incentive paradigm and financial strain approach. Acta Sociologica,
53(2), 119-139. https://doi.org/10.1177/0001699310365624
Ezzy, D. (1993). Unemployment and mental health: a critical review. Social Science & Medicine, 37(1),
41-52.
Jahoda, M. (1981). Work, employment, and unemployment: Values, theories, and approaches in
social research. American Psychologist, 36(2), 184-191.
Kalyuzhnova, Y., & Kambhampati, U. (2008). The determinants of individual happiness in
Kazakhstan. Economic Systems, 32(3), 285-299.
Korpi, T. (1997). Is utility related to employment status? Employment, unemployment, labor market
policies and subjective well-being among Swedish youth. Labour Economics, 4(2), 125-147.
McKee-Ryan, F., Song, Z., Wanberg, C. R., & Kinicki, A. J. (2005). Psychological and physical well-
being during unemployment: a meta-analytic study. The Journal of Applied Psychology,
90(1), 53-76. https://doi.org/10.1037/0021-9010.90.1.53
Novo, M., Hammarström, A., & Janlert, U. (2001). Do high levels of unemployment influence the
health of those who are not unemployed? A gendered comparison of young men and
women during boom and recession. Social Science & Medicine, 53(3), 293-303.
Oesch, D., & Lipps, O. (2013). Does unemployment hurt less if there is more of it around? A panel
analysis of life satisfaction in Germany and Switzerland. European Sociological Review,
29(5), 955-967.
Paul, K. I., & Moser, K. (2009). Unemployment impairs mental health: Meta-analyses. Journal of
Vocational Behavior, 74(3), 264-282. https://doi.org/10.1016/j.jvb.2009.01.001
Porket, J. L. (1995). Unemployment in Capitalist, Communist and Post-Communist Economies. New
York: Springer.
Perovic, B. (2015). Defining Youth in Contemporary National Legal and Policy Frameworks across
Europe. Council of Europe.
Scanlan, J. N., & Bundy, A. C. (2009). Is the health of young unemployed Australians worse in times
of low unemployment? Australian and New Zealand Journal of Public Health, 33(1), 79-82.
Schunck, R., & others. (2013). Within and between estimates in random-effects models: Advantages
and drawbacks of correlated random effects and hybrid models. Stata Journal, 13(1), 65-76.
Selezneva, E. (2011). Surveying transitional experience and subjective well-being: Income, work,
family. Economic Systems, 35(2), 139-157.
Social Diagnosis - Objective and Subjective Quality of Life in Poland. (2015). Retrieved 6 March 2018,
from http://www.diagnoza.com/index-en.html
Strandh, M., Hammarström, A., Nilsson, K., Nordenmark, M., & Russel, H. (2013). Unemployment,
gender and mental health: the role of the gender regime. Sociology of Health & Illness,
35(5), 649-665. https://doi.org/10.1111/j.1467-9566.2012.01517.x
Strandh, M., Novo, M., & Hammarström, A. (2011). Mental health among the unemployed and the
unemployment rate in the municipality. The European Journal of Public Health, 21(6), 799-805.

Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland 83
Stutzer, A., & Lalive, R. (2004). The role of social work norms in job searching and subjective well-
being. Journal of the European Economic Association, 2(4), 696-719.
Suh, E., Diener, E., & Fujita, F. (1996). Events and subjective well-being: only recent events matter.
Journal of Personality and Social Psychology, 70(5), 1091.
Theodossiou, I. (1998). The effects of low-pay and unemployment on psychological well-being:
A logistic regression approach. Journal of Health Economics, 17(1), 85-104. https://doi.
org/10.1016/S0167-6296(97)00018-0
Turner, J. B. (1995). Economic context and the health effects of unemployment. Journal of Health
and Social Behavior, 36(3), 213-229.
Van der Meer, P. H. (2014). Gender, unemployment and subjective well-being: Why being
unemployed is worse for men than for women. Social Indicators Research, 115(1), 23-44.
https://doi.org/10.1007/s11205-012-0207-5
WHO-Europe-Mental-Health-Acion-Plan-2013-2020.pdf. (n.d.). Retrieved from http://www.
euro.who.int/__data/assets/pdf_file/0020/280604/WHO-Europe-Mental-Health-Acion-
Plan-2013-2020.pdf
Winkelmann, R. (2009). Unemployment, social capital, and subjective well-being. Journal of
Happiness Studies, 10(4), 421-430. https://doi.org/10.1007/s10902-008-9097-2
WHO (2015). Europe Mental Health Action Plan 2013-2020. (2015). http://www.euro.who.int/__data/
assets/pdf_file/0020/280604/WHO-Europe-Mental-Health-Acion-Plan-2013-2020.pdf
World Bank (2011). World development report 2012: gender equality and development. (World Bank
Publications).
Magdalena Rokicka - economist (PhD, University of Essex), assistant professor at the Educational
Research Institute. Her research interests include: applied micro-econometric analysis of
education, labour market, migration, and gender related issues. Currently she coordinates research
conducted by the Educational Research Institute within the project ‘Social Exclusion of Youth in
Europe: Cumulative Disadvantage, Coping Strategies, Effective Policies and Transfer (EXCEPT)’.
Małgorzata Kłobuszewska - economist (PhD student, University of Warsaw), researcher at the
Educational Research Institute (IBE). She specialises in quantitative analysis, mainly applied micro-
econometric analysis in the field of economics of education and labour economics. Her research
interests include: parental expenditure on education, investment in human capital, vocational
education and youth on the labour market.
Marta Palczyńska - economist (PhD student, Warsaw School of Economics), an economist at
the Institute for Structural Research (IBS). Her general research interests include economics of
education and labour economics. Her current work focuses on the evaluation of the youth labour
market policies, the role of skills and personality in the labour market and the gender wage gap.
Acknowledgements
This project has received funding from the European Union’s Horizon 2020 research and innovation
programme under the grant agreement No 649496.
The authors acknowledge the contributions made by reviewers and participants of the 6nd EXCEPT
Workshop in Thessaloniki, 28-30 August 2017. Special thanks are extended to Vanessa di Paola for
constructive comments and suggestions.

Magdalena Rokicka, Marta Palczyńska & Małgorzata Kłobuszewska
84
Appendix
Figure A.1: Regional unemployment share among aged 16-64 (NUTS2 unemployment) and
unemployment rate in our sample (18-35 years old) by gender and year
Source: Own calculations based on: Diagnoza społeczna. Integrated database.www.diagnoza.com
[download at July 2017] and data accessed on the 02/08/2017 from https://bdl.stat.gov.pl [Share of
the registered unemployed persons in the population in the working age by sex].

Unemployment and Subjective Well-being Among the Youth: Evidence from a Longitudinal Study in Poland 85
Table A.1: Impact of unemployment and unemployment share on subjective well-being: mixed
effects coefficients (income accounted for)
(1)
Total
(2)
Total
(3)
Men
(4)
Men
(5)
Women
(6)
Women
Within individual estimates
Unemployed -0.126*** -0.335*** -0.167*** -0.403*** -0.091*-0.201
(0.024) (0.070) (0.029) (0.100) (0.043) (0.138)
NUTS2 unemployment -0.011 -0.015*-0.019** -0.024*** -0.001 -0.004
(0.007) (0.007) (0.007) (0.007) (0.012) (0.012)
Unempl.# NUTS2 unempl. 0.025** 0.031*0.012
(0.008) (0.013) (0.013)
Nr of friends 0.003 0.003 0.004 0.004 0.003 0.003
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
Age 0.050 0.049 0.033 0.032 0.060 0.060
(0.031) (0.031) (0.051) (0.051) (0.037) (0.037)
Age squared -0.001 -0.001 -0.000 -0.000 -0.001 -0.001
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Married 0.099** 0.097** 0.136** 0.135** 0.067 0.067
(0.030) (0.030) (0.048) (0.049) (0.047) (0.047)
Lost a close person -0.031*-0.031*-0.080*** -0.079*** 0.011 0.011
(0.013) (0.013) (0.022) (0.022) (0.024) (0.024)
Seriously ill -0.142** -0.139** -0.185*-0.178*-0.111** -0.110**
(0.048) (0.048) (0.088) (0.088) (0.034) (0.034)
Civic engagement 0.012 0.012 0.000 -0.001 0.022 0.022
(0.023) (0.023) (0.044) (0.043) (0.031) (0.031)
Religious activity 0.014** 0.014** 0.019** 0.019** 0.011 0.011
(0.005) (0.005) (0.006) (0.006) (0.006) (0.006)
Sufficient income 0.098*** 0.100*** 0.103*** 0.105*** 0.099*** 0.101***
(0.022) (0.022) (0.028) (0.028) (0.025) (0.025)
Between individual estimates (only time invariant)
Men -0.006 -0.006
(0.011) (0.011)
Basic vocational and lower -0.035 -0.035 -0.028 -0.028 -0.047 -0.046
(0.020) (0.020) (0.024) (0.024) (0.027) (0.027)
Post-secondary and tertiary 0.075*** 0.075*** 0.072*0.072*0.082** 0.082**
(0.018) (0.018) (0.028) (0.028) (0.027) (0.027)
Constant 3.698*** 3.700*** 4.008*** 3.988*** 3.470*** 3.480***
(0.228) (0.215) (0.426) (0.423) (0.553) (0.543)
Observations 7965 7965 3943 3943 4022 4022
Number of individuals 3833 3833 1947 1947 1886 1886
Standard errors in parentheses, * p<0.05, ** p<0.01, *** p<0.001. Reference Base category: for
unemployed: employed; for married: single, divorced, or widowed; for men: women; for education:
general secondary education.
Source: Own calculations based on: Diagnoza społeczna. Integrated database.www.diagnoza.com
[download at July 2017] and data accessed on the 02/08/2017 from https://bdl.stat.gov.pl [Share of
the registered unemployed persons in the population in the working age by sex].