Employment status and subjective well-being: the role of the social norm to work

Abstract

This article examines to what extent a social norm to work moderates the relationship between employment status and subjective well-being. It was expected that the detrimental impact of non-employment on subjective well-being would be larger in countries with a stronger social norm. Using a direct measure of the social norm to work and employing data from 45 European countries, this study assessed subjective well-being levels of five employment status groups for men and women separately. Results showed that subjective well-being of unemployed men and women is unaffected by the social norm to work. However, non-working disabled men are worse off in countries with a stronger norm. Living in such a country also decreases the well-being gap between employed and retired men, whereas retired women are worse off in these countries. This effect for retirees disappears when a country’s GDP is taken into account, suggesting that norms matter less than affluence.

Keywords

cross-national comparison employment status multi-level analyses social norm to work subjective well-being

Introduction

A vast body of research shows that unemployment is detrimental to well-being (e.g. Jahoda, 1982; Kassenboehmer and Haisken-DeNew, 2009; Shields and Price, 2005) and that it has an enduring effect even after re-employment (e.g. Clark et al., 2001; Oesch and Lipps, 2012; Winkelmann and Winkelmann, 1998). The relationship between employment status and subjective well-being¹ is primarily explained by differences in individual resources. However, norms in society about how individuals should behave with regard to employment may also be an important factor. In societies with a strong social norm to work, non-working groups may experience lower levels of social approval and acceptance than the employed, which may lower their well-being. This study examines to what extent a social norm to work moderates the relationship between employment status and subjective well-being (defined as life satisfaction), using cross-national survey data.

Social norms are expected to moderate the relation between employment status and subjective well-being through two mechanisms. The first is known as the ‘stigmatization hypothesis’ (Kalmijn and Uunk, 2007). Individuals who deviate from existing social norms experience informal social sanctions (e.g. social exclusion, gossiping, etc.) from others, which create feelings of shame and inferiority and consequently reduce well-being. Complying with a social norm, conversely, leads to approval and acceptance from others and thus to higher well-being. The second mechanism argues that social norms are an important factor in shaping predispositions (e.g. preferences) through internalization of these norms (Etzioni, 2000). Not complying with internalized norms generates guilt and self-sanctioning (McAdams, 1997), which is expected to decrease well-being. In contrast, compliance will lead to intrinsic affirmation (Etzioni, 2000). Both mechanisms predict that non-working individuals will have lower subjective well-being than employed individuals, because they experience internal or external sanctions for not complying with the social norm to work.

Considering the arguments above, it is expected that non-working individuals suffer from more sanctions in countries that hold a strong social norm compared to their counterparts who live in weak-norm countries. As a result, the gaps in well-being between working and non-working individuals (e.g. the unemployed, retirees, homemakers and the disabled) are expected to be larger if the social norm to work is stronger.

So far, research on this topic is scarce. The few studies that are available use indirect measures of social norms, such as unemployment rates (Clark, 2003; Clark et al., 2008; Oesch and Lipps, 2012) or referendum data on unemployment benefit reduction (Stutzer and Lalive, 2004). In this study, the social norm to work is measured directly by countries’ work ethic, which is defined as the conviction that one should work. Work ethic therefore refers to the moral embeddedness of work and is substantively different from personal motives or preferences about work (De Witte, 2000; Niles, 1999). Earlier research shows that countries’ work ethic varies significantly across European countries (Stam et al., 2013).

Alongside indirect measures, all previous studies are limited to comparisons of regions within one or two countries; in contrast, this study will conduct a country-comparative analysis by using data on 45 European countries. Since variation in social norms can be expected to be larger between countries than between regions within a country, this research design is more appropriate to study the moderating role of the social norm to work.

Another important contribution is that this study distinguishes among five employment statuses: employed, unemployed, non-working disabled, retired and (for women only) homemakers. Previous research mainly focuses on the employed and unemployed (Fryer and Payne, 1984). However, considering other employment statuses as well will increase insights into the way that social norms operate and it allows for a more extensive test of the social norm hypothesis. In addition, this research will conduct separate analyses for both men and women, because it is expected that their well-being is affected differently by their employment status. The traditional assumption is that unemployment affects women to a lesser extent than men (Ervasti and Venetoklis, 2010), for instance because women have lower work commitment (Hakim, 1991). Furthermore, normative pressures with regard to work are expected to be smaller for women than for men, because in many (Western) societies the breadwinner role is more strongly assigned to men (Kassenboehmer and Haisken-DeNew, 2009).

Theory

The relationship between employment status and well-being

In the literature, there is a broad consensus about the deteriorating effect of unemployment on individual subjective well-being (Creed and Macintyre, 2001). Both longitudinal and cross-sectional research shows that unemployment leads to a substantial loss of subjective well-being, regardless of which definition of subjective well-being is used (Frijters et al., 2004; Jackson et al., 1983; Shields and Price, 2005). Furthermore, many studies show that the decline in well-being is to a large extent due to the event of becoming unemployed and not to reverse causality (Creed and Evans, 2002; Shields and Price, 2005).

Paid employment is argued to be important for individuals’ well-being because it provides an income and fulfils various psychological needs (e.g. Jahoda, 1982; Nordenmark, 1999; Van der Meer, 2010). During unemployment, a person lacks both these pecuniary and non-pecuniary benefits (often referred to as manifest and latent benefits: e.g. Creed and Watson, 2003; Jahoda, 1982; Paul and Batinic, 2010), a lack which is argued to lead to lower subjective well-being.

The loss of, or drop in, pecuniary benefits (income) during unemployment is argued to be detrimental to well-being because it restricts unemployed individuals in looking forward and planning their future. Moreover, it leads to relative poverty, which is argued to be psychologically corrosive (Fryer, 1998), for instance because it makes it more difficult to participate in society and therefore increases the chance of being socially excluded (Van Echtelt, 2010). The lack of non-pecuniary benefits during unemployment leads to lower well-being because a number of psychological needs cannot be fulfilled (Jahoda, 1982). The literature distinguishes five non-pecuniary benefits: time structure; shared experiences and contacts outside the nuclear family; shared goals; personal status and identity; and enforced activity (Jahoda, 1981). Even though employment may not always be agreeable – for instance because of high job insecurity (e.g. Burchell, 2011; De Witte, 1999) or because of working more or less hours than preferred (being over- or under-worked) e.g. Muffels and Headey (2013), which are all factors that negatively affect subjective well-being – it is argued by other authors that unpleasant ties to reality, such as too rigid time structures or low status, are preferable to none at all (Grün et al., 2010; Jahoda, 1981). Although paid employment is not the only institution that offers these non-pecuniary benefits, it is argued to be the dominant one and the only one that combines them with the necessity of earning a living (Jahoda, 1982).

These theoretical ideas about non-pecuniary benefits for the employed and unemployed can be extended to other employment statuses in order to predict the relative well-being levels of the non-working disabled, retirees and homemakers alongside those of the employed and unemployed. The predicted rank order is based on the extent to which the groups have opportunities to compensate for the lack of non-pecuniary benefits, which are usually provided for by paid work (the pecuniary benefits are empirically controlled for by relative household income). It is argued that the employed can obtain non-pecuniary benefits most easily compared to other groups (Paul and Batinic, 2010) and therefore they are expected to have the highest well-being, followed by the retired, the non-working disabled and the unemployed. For women, an extra group is distinguished, namely homemakers.² They are expected to have the second highest well-being after the employed.

Homemakers can – compared to the other non-employed groups – most easily compensate for the loss or lack of non-pecuniary benefits within the family sphere. They have a clear and important role: taking care of a household and children (if any). This may be seen as a full-time job, including recurrent activities that have to be undertaken each day. These activities (e.g. grocery shopping) will likely bring them into contact with others outside the household and provide a certain time structure. Homemakers are however expected to rank lower than employed women, because several studies show that they are less well off in terms of mental health (for an overview, see Ferree, 1984).

Among the other non-working groups, retirees are expected to rank the highest on well-being. Retirement is a stable and predictable state, which makes planning ahead easier than it is for instance during unemployment (Strandh, 2000). Consequently, retirees may engage themselves in enduring and demanding activities, such as volunteering and taking care of family members or grandchildren (Van Ingen, 2009), which will also provide non-pecuniary benefits. In addition, society may believe that retirees have ‘earned’ their free time, which gives them status. Although previous literature shows that the transition from paid employment to retirement can in some cases be problematic (depending on the reason for retirement; see Hershey and Henkens, 2013), the above-mentioned arguments predict that retirees have better opportunities to compensate for the loss of non-pecuniary benefits than the unemployed and the non-working disabled. Female retirees are expected to rank after homemakers because, although retirees may have earned their free time, their main role in society might also be played out (e.g. Adelmann, 1993), whereas homemakers may still play an important part in raising children and running a household.

The unemployed and the non-working disabled are expected to have the most difficulties compensating for the lack of non-pecuniary benefits through other activities. First, unemployment is an uncertain and unpredictable state, which makes it more difficult to plan ahead and to engage in other activities. For many unemployed individuals, looking for a new job is the main daily activity, which in the first place may prevent them from having social contacts with others outside the nuclear family; and in the second place, instead of having a collective goal to strive for, they only have their personal goal of finding a job. In addition, their personal status and identity may also be very low; one’s job is an important source for personal validation and achievement. The non-working disabled are likely to have problems compensating, mainly because of their bad physical and/or mental health status (cf. Van Echtelt, 2010). In the models, however, it is possible to control empirically for health, which is why the non-working disabled are expected to rank higher on well-being than the unemployed.

The moderating role of a social norm to work

The key focus of this study is to investigate to what extent the social norm to work moderates the impact of employment status on well-being. Social norms can be defined as ‘informal social regularities that individuals feel obligated to follow because of an internalized sense of duty, because of a fear of non-legal sanctions, or both’ (McAdams, 1997: 340). Not complying with a social norm is argued to have an impact in two ways: individuals sanction themselves because they do not comply with their internalized norms; and secondly, individuals experience external sanctioning, because they do not live up to the expectations of their social environment. The psychological consequence of internal sanctioning is feeling guilty, while external sanctioning generates shame (McAdams, 1997). These consequences are expected to lower individual subjective well-being, because they may lead to, for instance, lower self-esteem and feelings of inferiority. With regard to a social norm to work, this implies that individuals who do not have a paid job experience a higher degree of sanctioning, which leads to lower levels of well-being compared to employed individuals.

From a country-comparative perspective, it is expected that in strong-norm countries, the gaps in well-being between working and non-working groups are larger compared to weak-norm countries. In strong-norm countries, non-working individuals will probably experience greater sanctions for not complying than their counterparts in weak-norm countries. Although all non-employed groups deviate from the social norm to work, some groups can be expected to be sanctioned more than others (e.g. McFayden, 1995) and thus to suffer more in terms of well-being. For instance, research on attitudes regarding welfare expenditures suggests that not all needy groups in society are viewed as equally deserving,³ which may be related to differences in the degree of stigmatization these groups experience. The unemployed are often stigmatized as being irresponsible, lazy and dishonest; and they are blamed for how their lives turn out (Furåker and Blomsterberg, 2003; Maassen and De Goede, 1989). As a consequence, they may experience a higher degree of sanctioning than, for example, the non-working disabled who cannot be blamed for their situation. Van Oorschot (2006) finds that in European welfare states, the elderly are viewed as the most deserving, followed by sick and disabled individuals and finally the unemployed. Thus, it is expected that well-being gaps between employed and non-employed individuals are larger in strong-norm countries compared to weak-norm countries and that groups that are more subject to stigmatization (i.e. the unemployed, the non-working disabled and retirees respectively) suffer most in terms of well-being when the social norm to work is stronger.

Data and measurement

The fourth wave of the European Values Study (EVS, 2010) was used. In 47 countries, a random sample of approximately 1500 respondents per country aged 18 and older completed a standardized questionnaire on basic human values in face to face interviews. The analyses did not include Azerbaijan (data anomalies) and Kosovo (lack of macro-data). Only respondents aged between 25 and 80 were included in the sample, in order to exclude students who had not entered the labour market yet and to increase comparability across countries, since some countries only surveyed respondents up to the age of 80. After this selection and the exclusion of respondents who had missing values on well-being (0.79%), employment status (2.5%), or any of the covariates (4.6%) except income, the analytical sample consisted of 50,934 respondents in 45 countries.

Main individual-level variables

Subjective well-being was operationalized as general life satisfaction and measured by the following question: ‘All things considered, how satisfied are you with your life as a whole these days?’ The response categories of the 10-point scale were: dissatisfied (p. 1) to satisfied (p. 10). This question has often been used to measure subjective well-being (e.g. Frijters et al., 2004; Verbakel, 2012; Winkelmann and Winkelmann, 1998). According to Clark and Lelkes (2005) it has good psychometric properties and has been validated using physiological, statistical and behavioural techniques.⁴ Veenhoven (1984) argued that people use two sources of information in order to rate their lives: affective and cognitive. Their affective experience is used to assess how well they feel and thus how happy they are. Cognitive evaluations are used to compare their life with standards of how it should be; this leads to a certain level of satisfaction with life. The latter was preferred here, since the social norm to work also refers to a standard of how things should be. The average level of life satisfaction was approximately 7 for both genders (see Table 1).

Table 1.

Descriptive statistics on individual characteristics.

	Minimum	Maximum	Males		Females
			Mean	SD	Mean	SD
Subjective well-being	1	10	7.02	2.28	6.92	2.33
Employment status
Employed	0	1	0.64		0.49
Homemaker	0	1	N.A.		0.17
Retired	0	1	0.25		0.24
Non-working disabled	0	1	0.02		0.02
Unemployed	0	1	0.09		0.09
Age	25	80	49.52	14.81	49.93	14.76
Age squared	625	6400	2671.79	1517.74	2711.02	1523.42
Education	0	6	3.15	1.35	3.04	1.44
Partnership status
Married	0	1	0.69		0.61
Cohabiting	0	1	0.06		0.05
Dating	0	1	0.05		0.04
Single	0	1	0.11		0.08
Divorced	0	1	0.05		0.08
Widowed	0	1	0.04		0.13
Having child(ren)	0	1	0.76		0.84
Relative householdincome (x1000 euros)	−2.80	13.55	0.07	1.05	−0.04	0.94
Health	1	5	3.73	0.93	3.57	0.97
Religious	0	1	0.25		0.33
Work ethic	1	5	3.71	0.75	3.63	0.77

Source: European Values Study (2010)

Notes: males: N= 22,440; females: N=28,494; in 45 countries.

Employment status was measured by self-reported statuses. The employed were respondents who worked full time, part time, or were self-employed. Furthermore, the unemployed, retired and non-working disabled were distinguished and, for women, also homemakers. Respondents who answered they were ‘in military service’, ‘students’ or ‘other’ as well as men who indicated being a homemaker were excluded from the analyses (2.5% for both genders).

Country-level social norm to work

The social norm to work was measured by the aggregated individual work ethic, which is the averaged sum of at least three valid scores on five items:⁵ ‘To fully develop your talents, you need to have a job’; ‘It is humiliating to receive money without having to work for it’; ‘People who don’t work turn lazy’; ‘Work is a duty towards society’; and ‘Work should always come first, even if it means less spare time.’ A number of previous studies used this set of items as an indicator for work ethic (e.g. Roest et al., 2009; Stam et al., 2013, 2014; Van Oorschot, 2006; Wielers and Raven, 2013). The construction of these items is based on the latent deprivation model (De Witte, 2000), which was developed by Jahoda (1982) and consists of the five non-pecuniary benefits.

The theoretical rationale behind this scale is that a number of non-pecuniary benefits have to be obtained through paid employment, not through alternative sources. Central to all these items is the notion that work is a moral duty (De Witte, 2000). Respondents rated the items on a five-point scale. The average for each country was calculated and used as an indicator for the social norm to work, which runs from 1 (weak norm) to 5 (strong norm). Additional factor analyses showed that all items load on one factor with an eigenvalue of at least 1 in all countries (Cronbach’s alpha ranges from 0.58 to 0.79). There was substantial variation in countries’ work ethic; Iceland scored lowest with 2.81 and Turkey scored highest with 4.24 (see Table 2).

Table 2.

Descriptive results for country characteristics (ordered on social norm to work).

Country	Social norm to work (a)	Unemploy-ment (%) (b)	GDP per capita PPP in USD (x1000) (c)	N respon-dents
Iceland	2.81	2.78	36.25	661
Netherlands	3.13	4.30	38.02	1,331
Finland	3.25	7.64	33.60	951
Northern Ireland	3.26	5.16	34.35	383
Sweden	3.29	6.90	35.81	808
Great Britain	3.30	5.16	34.35	1,205
Belgium	3.31	7.94	34.08	1,270
Croatia	3.44	11.00	17.11	1,118
Switzerland	3.44	4.02	40.93	1,066
France	3.47	8.80	31.30	1,269
Ireland	3.48	4.90	41.08	762
Latvia	3.49	8.52	15.42	1,217
Denmark	3.51	4.28	35.83	1,265
Malta	3.52	6.78	22.95	1,281
Poland	3.52	13.52	15.34	1,143
Lithuania	3.55	6.72	16.92	1,168
Spain	3.56	9.64	29.83	1,206
Bosnia	3.58	27.83	6.33	1,118
Estonia	3.58	6.72	18.82	1,273
Russia	3.58	6.92	13.36	1,136
Norway	3.59	3.44	52.25	925
Czech Republic	3.64	6.60	23.20	1,374
Belarus	3.65	1.26	9.81	1,145
Luxembourg	3.65	4.66	76.07	1,100
Serbia	3.69	18.38	9.44	1,203
Slovenia	3.69	5.62	25.50	1,090
Montenegro	3.71	24.76	10.37	1,168
Ukraine	3.72	7.08	6.29	1,213
Germany	3.73	9.66	33.41	1,760
Italy	3.73	7.06	30.33	1,119
Austria	3.74	4.62	36.20	1,200
Greece	3.81	9.06	26.52	1,268
Slovakia	3.83	13.82	18.67	1,273
Macedonia	3.86	35.84	8.79	1,151
Northern Cyprus	3.86	4.56	12.68	331
Armenia	3.88	7.72	4.81	1,111
Hungary	3.89	7.20	18.17	1,189
Moldova	3.89	6.38	2.55	1,187
Romania	3.89	6.94	11.42	1,113
Portugal	3.90	8.42	22.67	1,298
Georgia	3.93	13.96	4.09	1,228
Albania	3.95	13.50	6.59	1,084
Cyprus	4.04	4.56	27.25	840
Bulgaria	4.11	8.74	11.22	1,194
Turkey	4.24	9.08	12.68	1,789
Mean	3.66	8.94	23.39
SD	0.26	6.56	14.70
Minimum	2.81	1.26	4.09
Maximum	4.24	35.84	76.07

Source: European Values Study (2010)

Notes: N=45 countries.

(a): Measured on a scale from 1 to 5: higher scores imply a stronger social norm to work.

(b): Five-year average for the period 2004–8, in percentages of the labour force (UNECE Statistical database, 2013b).

(c): Five-year average for the period 2004–8 (UNECE Statistical database, 2013a).

Control variables

The control variables are: age and age squared; education: the first digit of ISCED, the International Standard Classification of Education, pre-primary or no education (0) to second stage tertiary education (6); current partnership status (married or registered partnership, cohabiting, dating, being single, divorced and widowed, cf. Verbakel, 2012); and having child(ren) – no children (0) and at least one child (1). Furthermore, relative household income⁶ was measured by the monthly household income in euros per month (corrected for purchasing power parity) minus the average monthly household income of the country of residence. Missing cases (17%) were imputed by the country mean and controlled for by a dummy variable. Other control variables are: health – self-reported state of health, 1 (very poor) to 5 (very good); and religiousness – attend church at least once a month (1) or less (0). Finally, controls are added for the individual’s work ethic, in order to rule out effects of the social norm being an artefact of varying population compositions between countries.

To pose a more stringent test of the effect of the social norm to work, robustness checks were performed by controlling for countries’ unemployment rate and GDP. Previous studies used unemployment rates as an indicator for the social norm to work, arguing that higher unemployment rates indicate a weaker social norm to work. However, the two appear to be positively correlated (r = 0.22; p = 0.000). This implies, first, that unemployment rates are not a good proxy for the social norm to work and, second, that unemployment rates may affect the moderating effect of the social norm to work. Unemployment rate was measured by the five-year (2004–8) average of the percentage of unemployed in the labour force per country.

Previous studies showed that GDP is negatively related to countries’ work ethic (e.g. Stam et al., 2013; the present study: r = −0.48; p = 0.000). Because economic affluence may reduce well-being gaps between employment statuses, neglecting GDP could result in overestimating the moderating effect of the social norm: part of its effect may actually be due to the moderating effect of GDP. GDP (per capita at current prices and purchase power parity) in US dollars was measured by a five-year average (2004–8). Unemployment rates and GDP were retrieved from the UNECE statistical database (2013a, 2013b).

Models

Multilevel regression models were used to analyse the data. These models take into account the nested structure of the data (here: individuals nested in countries) and allow for variance decomposition at the country and individual level. Scores on linear individual and country-level variables are standardized: coefficients refer to the average individual in the average country. Model 1 includes the uncorrected effects of employment status on well-being. Model 2 adds the individual-level control variables and Model 3 the interaction effects between employment status and the social norm to work on both the individual and the country level. The final two models are robustness checks that include countries’ unemployment rates and GDP and their interactions with employment status. Random slopes are included for all employment statuses, allowing the effects of employment status on well-being to vary between countries, except for the non-working disabled (samples were too small).

Results

Model 1 in Table 3 shows that for men, the employed, as expected, have the highest well-being, followed by the retired, unemployed and non-working disabled. For instance, the well-being of non-working disabled men is 1.54 points lower on a 10-point scale, compared to employed men. For women (Table 4), the employed and homemakers score highest on well-being, followed by the retired, unemployed and non-working disabled.⁷ Furthermore, unemployment is more detrimental for men: the well-being of the unemployed compared to the employed is 1.24 points (on a scale of one to 10) lower for men, compared to 0.91 for women.

Table 3.

Multilevel regression analysis on subjective well-being of men.

Males	Model 1			Model 2			Model 3			Model 4			Model 5
	b		se	b		se	b		se	b		se	b		se
Individual level
Employed (ref.)
Retired	−0.302	**	0.077	0.153	*	0.068	0.144	*	0.064	0.150	*	0.066	0.159	*	0.062
Non-working disabled	−1.536	**	0.101	−0.329	**	0.098	−0.356	**	0.104	−0.416	**	0.102	−0.359	**	0.104
Unemployed	−1.237	**	0.114	−0.739	**	0.096	−0.733	**	0.098	−0.770	**	0.095	−0.799	**	0.094
Controls
Age				−0.345	**	0.108	−0.304	**	0.108	−0.336	**	0.108	−0.296	**	0.108
Age squared				0.417	**	0.112	0.359	**	0.113	0.406	**	0.112	0.347	**	0.112
Education				0.076	**	0.015	0.084	**	0.015	0.074	**	0.015	0.085	**	0.015
Married (ref.)
Cohabiting				−0.093		0.061	−0.078		0.061	−0.089		0.061	−0.079		0.061
Dating				−0.150	*	0.072	−0.143	*	0.072	−0.151	*	0.072	−0.140		0.072
Single				−0.448	**	0.058	−0.441	**	0.058	−0.449	**	0.058	−0.440	**	0.058
Divorced				−0.690	**	0.064	−0.681	**	0.064	−0.681	**	0.064	−0.681	**	0.064
Widowed				−0.796	**	0.074	−0.787	**	0.073	−0.787	**	0.073	−0.779	**	0.073
Having child(ren)				0.095	*	0.046	0.081		0.046	0.082		0.046	0.083		0.046
Relative household income				0.108	**	0.014	0.111	**	0.014	0.110	**	0.014	0.111	**	0.014
Health				0.730	**	0.016	0.726	**	0.016	0.726	**	0.016	0.723	**	0.016
Religious				0.104	**	0.034	0.098	**	0.034	0.098	**	0.034	0.100	**	0.034
Work ethic							0.116	**	0.018	0.116	**	0.018	0.116	**	0.018
x Retired							0.033		0.035	0.033		0.035	0.030		0.035
x Non-working disabled							0.135		0.091	0.135		0.091	0.137		0.091
x Unemployed							−0.065		0.048	−0.065		0.048	−0.064		0.048
Country level
Social norm to work							−0.310	**	0.059	−0.278	**	0.061	−0.214	**	0.059
x Retired							−0.156	**	0.052	−0.155	**	0.054	−0.084		0.055
x Non-working disabled							−0.362	**	0.102	−0.276	**	0.102	−0.293	*	0.115
x Unemployed							0.024		0.097	−0.032		0.094	−0.103		0.106
Unemployment rate										0.025		0.063
x Retired										−0.012		0.055
x Non-working disabled										−0.108		0.101
x Unemployed										0.153		0.086
GDP per capita PPP (x1000)													0.210	**	0.062
x Retired													0.161	**	0.055
x Non-working disabled													0.134		0.108
x Unemployed													−0.283	*	0.112
Intercept	7.237	**	0.094	7.089	**	0.087	7.071	**	0.075	7.065	**	0.076	7.076	**	0.070

Variance components
Individual-level variance	4.51		0.04	3.99		0.04	3.97		0.04	3.98		0.04	3.97		0.04
Country-level variance
Intercept	0.38	#	0.09	0.24	#	0.05	0.15	#	0.04	0.15	#	0.04	0.12	#	0.03
Random slope retired	0.21	#	0.06	0.09	#	0.03	0.07	#	0.03	0.08	#	0.03	0.05	#	0.02
Random slope unemployed	0.40	#	0.12	0.25	#	0.08	0.26	#	0.09	0.23	#	0.08	0.22	#	0.08

Notes: N= 22,440 in 45 countries; **p<0.01; *p<0.05; # at least twice the standard error.

Source: European Values Study (2010).

Table 4.

Multilevel regression analysis on subjective well-being of women.

Females	Model 1			Model 2			Model 3			Model 4			Model 5
	b		se	b		se	b		se	b		se	b		se
Individual level
Employed (ref.)
Homemaker	−0.029		0.061	0.086		0.055	0.102		0.056	0.102		0.056	0.106	*	0.055
Retired	−0.538	**	0.067	0.047		0.061	0.029		0.059	0.029		0.059	0.049		0.053
Non-working disabled	−1.445	**	0.103	−0.232	*	0.099	−0.263	*	0.105	−0.244	*	0.107	−0.265	*	0.105
Unemployed	−0.911	**	0.091	−0.581	**	0.078	−0.570	**	0.079	−0.605	**	0.077	−0.588	**	0.080

Controls
Age				−0.361	**	0.094	−0.341	**	0.094	−0.340	**	0.094	−0.333	**	0.094
Age squared				0.453	**	0.098	0.418	**	0.098	0.416	**	0.099	0.407	**	0.098
Education				0.096	**	0.015	0.100	**	0.015	0.100	**	0.016	0.101	**	0.015
Married (ref.)
Cohabiting				−0.168	**	0.058	−0.166	**	0.058	−0.166	**	0.058	−0.164	**	0.058
Dating				−0.327	**	0.068	−0.321	**	0.068	−0.321	**	0.068	−0.322	**	0.068
Single				−0.463	**	0.056	−0.465	**	0.056	−0.463	**	0.056	−0.463	**	0.056
Divorced				−0.735	**	0.047	−0.730	**	0.047	−0.730	**	0.047	−0.729	**	0.047
Widowed				−0.537	**	0.043	−0.530	**	0.043	−0.530	**	0.043	−0.524	**	0.043
Having child(ren)				0.071		0.042	0.066		0.042	0.066		0.042	0.065		0.042
Relative household income				0.067	**	0.013	0.071	**	0.013	0.070	**	0.013	0.071	**	0.013
Health				0.769	**	0.015	0.767	**	0.014	0.767	**	0.015	0.765	**	0.015
Religious				0.141	**	0.029	0.133	**	0.029	0.134	**	0.029	0.133	**	0.029
Work ethic							0.084	**	0.019	0.084	**	0.019	0.084	**	0.019
x Homemaker							−0.020		0.036	−0.020		0.036	−0.020		0.036
x Retired							0.095	**	0.033	0.095	**	0.033	0.095	**	0.033
x Non-working disabled							−0.004		0.091	0.000		0.091	0.004		0.091
x Unemployed							−0.044		0.048	−0.047		0.048	−0.046		0.048

Country level
Social norm to work							−0.366	**	0.057	−0.358	**	0.059	−0.271	**	0.057
x Homemaker							−0.043		0.056	−0.041		0.057	−0.007		0.062
x Retired							−0.150	**	0.053	−0.151	**	0.055	−0.053		0.051
x Non-working disabled							−0.135		0.094	−0.170		0.102	−0.138		0.107
x Unemployed							0.016		0.081	−0.017		0.079	−0.032		0.092
Unemployment rate										−0.033		0.061
x Homemaker										−0.026		0.057
x Retired										0.002		0.055
x Non-working disabled										0.146		0.163
x Unemployed										0.136	*	0.067
GDP per capita PPP (x1000)													0.208	**	0.059
x Homemaker													0.080		0.058
x Retired													0.210	**	0.050
x Non-working disabled													−0.007		0.114
x Unemployed													−0.110		0.100

Intercept	7.216	**	0.106	7.056	**	0.088	7.025	**	0.072	7.025	**	0.072	7.033	**	0.067

Variance components
Individual-level variance	4.7		0.04	4.14		0.03	4.13		0.03	4.13		0.03	4.13		0.03
Country-level variance
Intercept	0.48	#	0.11	0.26	#	0.06	0.13	#	0.03	0.13	#	0.03	0.10	#	0.03
Random slope homemaker	0.08	#	0.03	0.06	#	0.02	0.06	#	0.03	0.06	#	0.03	0.05	#	0.02
Random slope retired	0.16	#	0.05	0.08	#	0.03	0.06	#	0.02	0.07	#	0.03	0.03		0.02
Random slope unemployed	0.21	#	0.07	0.12	#	0.05	0.13	#	0.05	0.11	#	0.05	0.12	#	0.05

Notes: N=28,494 in 45 countries; **p<0.01; *p<0.05; # at least twice the standard error.

Source: European Values Study (2010).

In Model 2, the individual-level controls are added, which changes the ranking of employment statuses’ effects on well-being. This is mainly the result of controlling for health. Male retirees now have the highest well-being (0.15 points higher than the employed), followed by the employed, non-working disabled and unemployed. For women, homemakers, retirees and the employed have the highest well-being followed by the non-working disabled and unemployed. Again, unemployment is more detrimental for men than for women (-0.74 vs -0.58 points), compared to being employed; however the difference has become smaller. Retirement, on the other hand, is clearly more beneficial for men.

The bottom panels in Tables 3 and 4 display the random slopes of the employment statuses, which indicate that the size of the well-being gaps between the employed and all other groups varies between countries for both genders.

Model 3, which is the main model in which all hypotheses are tested simultaneously,⁸ shows that the advantageous well-being position of retired men in countries with an average norm to work vanishes if norms are stronger than average: each unit increase in a country’s social norm to work, decreases the average well-being of retired men with 0.15 points on a 10-point scale, compared to employed men. For female retirees a negative effect of the social norm is found as well. The reduction in the random slope of retirees (Model 3 versus Model 2) signals that variation in countries’ social norm to work explains part of the variation in the size of the well-being gap between employed and retired men and women. The well-being gap between non-working disabled and employed men increases too when the norm to work is stronger. Each unit increase in the social norm lowers well-being by 0.36 points. These results are in line with the hypotheses. Interestingly, the unemployed are not differently affected by the normative climate than the employed. The substantial cross-national differences in well-being gaps between the employed and unemployed (as indicated by the significant random slope for the unemployed) cannot be explained by variation in the social norm to work. This is surprising since the unemployed are expected to be most harmed by the social norm to work. Figure 1 is a graphical representation of the findings showing, at varying levels of the social norm to work, the well-being gaps for all employment categories compared to the employed. Clearly, non-working disabled men (and women, though not significantly) and retired men and women are most affected by the social norm.

Figure 1.

Well-being gaps between several employment statuses and the employed (on well-being scale 1–10) for men and women by the social norm to work.

Models 4 and 5 show the results of the robustness checks.⁹ Model 4 shows that inclusion of the unemployment rate and its interactions does not affect the impact of the social norm. Higher unemployment rates do decrease the well-being gap between employed and unemployed women: the gap decreases by 0.14 points with each unit increase in the unemployment rate. Because the social norm to work is included in this model, the effect of unemployment cannot be interpreted as such, as previous studies claimed;¹⁰ unemployment rates thus are not a good proxy for the social norm to work.

Model 5 shows that differences in GDP explain the moderating impact of the social norm on the well-being gap between employed and retired men and women (Model 3). The fact that the well-being position of retirees relative to employed individuals is worse in strong-norm countries is thus not explained by (self-) sanctioning, but by the lower levels of GDP that are common in these countries. For non-working disabled men, the effect of the social norm does not change.

Conclusion and discussion

This article focuses on the question to what extent a social norm to work moderates the relationship between employment status and subjective well-being. By using a direct measure for the social norm to work and employing a large-scale cross-national dataset, this study assessed the well-being levels of five employment status groups for men and women separately.

The results show variation in well-being between the employed and non-employed, but also within the latter group. In average-norm countries, retired men had a relatively high well-being compared to employed men, whereas both unemployed men and women had the lowest well-being; even lower than the non-working disabled. Retired women and homemakers displayed the same level of well-being as employed women. In addition, it is observed that men suffered more from unemployment than women. Furthermore, men benefited from retirement, whereas retired women had only as much well-being as employed women (in average work-norm countries).

The relatively poor position of the unemployed has been found in previous studies as well (e.g. Bjørnskov et al., 2008; Winkelmann and Winkelmann, 1998). Presumably, unemployed individuals have the most difficulties compensating for the lack of non-pecuniary benefits (cf. Paul and Batinic, 2010). Moreover, the results suggest that there are also other sources beside paid employment that provide non-pecuniary benefits (cf. Jahoda, 1982), since the well-being of retirees and homemakers is as high or even higher compared to that of the employed. One may argue that these groups may have more freely chosen their status, compared to the non-working disabled and the unemployed, which might explain their relatively high levels of well-being. Note that substantial differences in well-being also exist between individuals within a certain employment status group (for instance retirees, see Hershey and Henkens, 2013),¹¹ a finding which this study has not addressed. Future research is needed on this issue.

With regard to the main interest of this study, it is found that a strong social norm to work reduces the well-being of retired men and women and non-working disabled men relative to their employed counterparts. The moderating effect of the social norm for retirees appears to be statistically confounded by countries’ GDP; retirees in strong-norm countries do not have lower levels of well-being because they are stigmatized, but because they live in less prosperous countries with, presumably, fewer facilities for retirees.

Interestingly, no effect of the social norm was found for the unemployed, whereas the well-being gap with the employed varies substantially between countries. Nor can countries’ unemployment rate or GDP account for this variation. Note that multilevel regression models allow for including only a limited number of country-level predictors, because the number of degrees of freedom is limited. In the analyses, up to eight country-level predictors are included simultaneously, which might cause statistical power issues (Snijders and Bosker, 1999) and might render significant results non-significant. Future research might be able to include more countries, perhaps at more points in time, to tackle this problem. Analysing longitudinal country data could also tackle the problem of causation, which is inherent in cross-sectional research.

As alternative country-level explanations for the variation in the well-being gaps between the employed and unemployed, Eichhorn (2014) suggests that differences in the age dependency ratio are important; unemployment causes well-being to deteriorate more in societies with a greater proportion of the elderly, because falling out of a relatively small and homogeneous working age group results in greater perceived differences from one’s reference group, than when this group is larger. Price inflation also matters: the higher the inflation, the more corrosive is unemployment for well-being, because unemployment already induces economic insecurity and inflation makes, for instance, personal savings worth even less (Eichhorn, 2014). Inflation might also play a role in this study, as the data were collected at the beginning of the worldwide economic crisis; arguably there were substantial differences in inflation between countries.

Another limitation of this study is related to the way the employment status groups are defined; they are based on self-categorization. McFayden (1995) suggests that the risk of being stigmatized affects how respondents categorize themselves. For instance, unemployed individuals who are at risk of stigmatization might categorize themselves not as unemployed, but in alternative socially acceptable groups (e.g. homemaker, retired), in order to avoid stigma (McFayden, 1995). If this is the case, the effect of the social norm could be biased for certain non-working groups. Another factor which might cause biased estimates is the small sample of the non-working disabled. In order to obtain more reliable estimates for this group, future studies should try to increase the number of respondents in this specific group.

Alongside the main results, the analyses showed that unemployment rates are not a good proxy for the social norm to work. First, because they correlate positively to the social norm, instead of negatively as was argued in previous studies; and second, because the moderating effect of the unemployment rate exists, even when the effect of the social norm to work is controlled for. Future studies are encouraged to explore alternative explanations for this finding, for example by studying social comparison mechanisms in a qualitative research design. Such a design might also uncover to what extent individuals subjectively experience normative pressures and social sanctions when it comes to work.

Footnotes

Appendix 1

Table A1.

Average (uncorrected) subjective well-being scores for men and women by work ethic, country and employment status.

Average well-being
Men						Women
Work ethic	Country	Employed	Retired	Disabled	Unemployed	Employed	Homemaker	Retired	Disabled	Unemployed
Weak	Iceland	8.0	8.7	5.8*	6.6	8.3	8.4	7.5	7.1	7.9
	Netherlands	8.0	7.8	7.3	6.9	8.1	8.1	7.8	7.2	7.3*
	Finland	7.7	8.0	4.7*	6.6	8.0	7.6	7.9	7.2*	5.8
	Northern Ireland	7.8	8.1	7.7	7.6	7.8	8.2	8.0	7.4*	7.6
	Sweden	7.7	8.2	6.6	6.0	8.0	9.8*	8.3	5.6	6.7
	Great Britain	7.5	7.8	5.7	5.7	7.8	7.5	8.0	6.0	6.7
	Belgium	7.8	8.0	6.8	6.5	7.7	7.9	7.5	6.2	6.1
	Croatia	7.3	7.0	6.5*	6.8	7.4	6.9	6.3	5.4*	6.9
	Switzerland	8.1	8.2	5.0*	4.6*	8.1	8.4	8.0	5.6	5.3*
	France	7.1	7.2	6.3	5.9	7.2	7.0	6.9	6.3	6.2
	Ireland	7.7	8.2	6.1*	7.0	8.1	7.7	7.6	5.3*	7.1
	Latvia	6.3	5.8	5.5	4.7	6.6	6.7	6.1	5.8	5.1
	Denmark	8.4	8.1	6.5*	7.0*	8.6	8.0*	8.4	N.A.	7.9
	Malta	8.0	8.0	6.7*	7.6	7.8	7.9	7.4	9.0*	6.7
	Poland	7.5	6.7	9.0*	6.4	7.6	7.3	6.5	N.A.	7.5
	Lithuania	6.7	5.7	5.2	4.0	6.4	6.8	5.5	4.5	5.2
Average		7.6	7.6	6.4	6.3	7.7	7.7	7.4	6.3	6.6

Medium	Spain	7.4	7.3	8.0*	6.8	7.3	7.2	7.1	5.5*	7.3
	Bosnia	7.4	7.1	6.3	6.5	7.0	6.7	6.8	5.5*	6.6
	Belarus	6.3	5.7	4.1	5.4	6.1	6.7	5.7	5.5*	5.3*
	Estonia	6.7	6.4	5.1	5.0	6.9	7.1	6.2	4.7	5.0
	Russia	7.0	5.2	5.3	5.9	6.7	7.2	5.7	5.3	5.4
	Norway	8.1	7.8	6.5	7.0	8.3	8.5	8.5	6.6	7.7*
	Czech Republic	7.3	6.6	6.1	5.3	7.2	7.8	7.0	6.1	6.9
	Luxembourg	8.0	8.3	7.4*	4.9	7.6	8.1	7.9	6.5*	7.1
	Serbia	7.1	7.0	5.1*	6.3	7.3	7.2	6.2	5.0*	7.1
	Slovenia	7.7	7.2	7.0*	6.6	7.8	7.5	7.1	8.0*	6.6
	Montenegro	7.7	7.5	6.0*	7.5	7.5	7.7	6.7	7.3*	7.1
	Ukraine	6.4	5.6	5.1*	4.7	6.3	6.3	5.2	3.2*	5.5
	Germany	7.2	6.9	5.1	4.1	7.2	7.0	6.9	4.9	4.9
	Italy	7.2	7.9	5.5*	5.7	7.4	7.4	7.0	N.A.	6.7
Average		7.3	6.9	5.9	5.8	7.2	7.3	6.7	5.7	6.4
Strong	Armenia	6.0	5.2	3.8*	5.1	5.5	5.8	4.9	4.3*	5.1
	Georgia	6.0	4.7	2.3*	5.4	5.9	5.5	4.7	4.8*	5.1
	Turkey	6.5	6.5	3.0*	5.5	6.3	6.6	6.5	6.5*	6.3
	Hungary	6.6	6.2	4.9	4.5	6.7	6.7	5.8	5.9	5.5
	Portugal	7.0	6.4	3.8*	6.0	6.7	6.9	5.9	4.8	5.8
	Austria	7.6	7.7	1.5*	6.4	7.5	7.9	7.8	7.6*	5.8
	Greece	7.1	6.6	4.0*	4.8	7.2	6.4	6.4	5.0*	6.5
	Moldova	7.0	5.8	5.4	6.0	6.7	7.3	5.9	6.1	5.5
	Slovakia	7.5	6.7	5.7	6.4	7.5	7.0	6.8	6.4	6.2
	Romania	6.9	6.8	5.2*	7.0	7.3	6.4	6.4	5.4*	6.3
	Northern Cyprus	6.6	6.9	5.2*	5.7	6.6	6.1	4.1	3.5*	6.1
	Albania	6.6	6.2	6.3*	5.8	6.5	5.9	5.3	4.3	5.9
	Cyprus	7.7	7.5	8.0*	7.3*	7.2	7.6	6.6	4.3*	6.5
	Bulgaria	6.2	5.0	2.5*	4.6	6.1	6.8	5.0	4.0*	4.9
	Macedonia	7.4	6.7	5.0	6.7	6.9	6.9	6.5	8.0*	6.5
Average		6.8	6.3	4.4	5.8	6.7	6.7	5.9	5.4	5.9

Source: European Values Study (2010)

Scores based on less than 10 respondents.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

Kirsten Stam is a PhD candidate in the Department of Sociology at Tilburg University. In her dissertation she focuses on predictors and consequences of work ethic values from both a country-comparative and longitudinal perspective.

Inge Sieben is an assistant professor in the Department of Sociology at Tilburg University. She is interested in cross-national research on family, religious and moral values and in social stratification research and has published articles in journals such as European Sociological Review, the British Journal of Sociology and Work, employment and society.

Ellen Verbakel is an assistant professor in the Department of Sociology / ICS at Radboud University Nijmegen. Her research interests include partner relations, partner effects, work and well-being.

Paul M de Graaf is Professor of Sociology at Tilburg University. His research interests include the sociology of education, social stratification, demographic and social aspects of the life-course and comparative research on values.

References

Adelmann

(1993) Psychological well-being and homemaker vs. retiree identity among older women. Sex Roles 29(3/4): 195–212.

Bjørnskov

Dreher

Fischer

JAV

(2008) Cross-country determinants of life satisfaction: exploring different determinants across groups in society. Social Choice and Welfare 30(1): 119–73.

Burchell

(2011) A temporal comparison of the effects of unemployment and job insecurity on wellbeing. Sociological Research Online 16(1): 9–21.

Clark

(2003) Unemployment as a social norm. Journal of Labor Economics 21(2): 323–51.

Clark

Lelkes

(2005) Deliver Us from Evil: Religion as Insurance. Paris: PSE.

Clark

Georgellis

Sanfey

(2001) Scarring: the psychological impact of past unemployment. Economica 68(270): 221–41.

Clark

Knabe

Rätzel

(2008) Unemployment as a Social Norm in Germany. SOEP papers on Multidisciplinary Panel Data Research 132, Deutsches Institut für Wirtschaftsforschung, Berlin, Germany.

Creed

Evans

(2002) Personality, well-being and deprivation theory. Personality and Individual Differences 33(7): 1045–54.

Creed

Macintyre

(2001) The relative effects of deprivation of the latent and manifest benefits of employment on the well-being of unemployed people. Journal of Occupational Health Psychology 6(4): 324–31.

10.

Creed

Watson

(2003) Age, gender, psychological well-being and the impact of losing the latent and manifest benefits of employment in unemployed people. Australian Journal of Psychology 55(2): 95–103.

11.

De Witte

(1999) Job insecurity and psychological well-being: review of the literature and exploration of some unresolved issues. European Journal of Work and Organizational Psychology 8(2): 155–77.

12.

De Witte

(2000) Houdingen tegenover arbeid in België op de drempel van de eenentwintigste eeuw. In: Dobbelaere

. (eds) Verloren zekerheid: De Belgen en hun waarden, overtuigingen en houdingen. Tielt: Lannoo, 107–23.

13.

Eichhorn

(2014) 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.

14.

Ervasti

Venetoklis

(2010) Unemployment and subjective well-being: an empirical test of deprivation theory, incentive paradigm and financial strain approach. Acta Sociologica 53(2): 119–38.

15.

Etzioni

(2000) Social norms: internalization, persuasion and history. Law and Society Review 34(1): 157–78.

16.

EVS (European Values Study) (2010) European Values Study 2008, 4th Wave, Integrated Dataset. GESIS Data Archive, Cologne, ZA4800 Data File Version 2.0.0 (2010-11-30) DOI: 10.4232/1.10188; DOI: 10.4232/1.10059.

17.

Ferree

(1984) Class, housework and happiness: women’s work and life satisfaction. Sex Roles 11(11/12): 1057–74.

18.

Frijters

Haisken-DeNew

Shields

(2004) Money does matter! Evidence from increasing real income and life satisfaction in East Germany following reunification. The American Economic Review 94(3): 730–40.

19.

Fryer

(1998) Labour market disadvantage, deprivation and mental health. In: Drenth

PJD

. (eds) A Handbook of Work and Organizational Psychology. Volume 2: Work Psychology. Hove: Psychology Press Ltd, 215–28.

20.

Fryer

Payne

(1984) Proactive behaviour in unemployment: findings and implications. Leisure Studies 3(3): 273–95.

21.

Furåker

Blomsterberg

(2003) Attitudes towards the unemployed: an analysis of Swedish survey data. International Journal of Social Welfare 12(3): 193–203.

22.

Grün

Hauser

Rhein

(2010) Is any job better than no job? Life satisfaction and re-employment. Journal of Labor Research 31(3): 285–306.

23.

Hakim

(1991) Grateful slaves and self-made women: fact and fantasy in women’s work orientations. European Sociological Review 7(2): 101–21.

24.

Hershey

Henkens

(2013) Impact of different types of retirement transitions on perceived satisfaction with life. The Gerontologist. Available (consulted 5 March 2014) at: http://gerontologist.oxfordjournals.org/content/early/2013/02/18/geront.gnt006.full.pdf+html

25.

Jackson

Stafford

Banks

Warr

(1983) Unemployment and psychological distress in young people: the moderating role of employment commitment. Journal of Applied Psychology 68(3): 525–35.

26.

Jahoda

(1981) Work, employment, and unemployment. Values, theories, and approaches in social research. American Psychologist 36(2): 184–91.

27.

Jahoda

(1982) Employment and Unemployment: A Social-Psychological Analysis. Cambridge: Cambridge University Press.

28.

Jeene

Van Oorschot

Uunk

(2013) Popular criteria for the welfare deservingness of disability pensioners: the influence of structural and cultural factors. Social Indicators Research 110(3): 1103–17.

29.

Kahneman

Krueger

(2006) Developments in the measurement of subjective well-being. Journal of Economic Perspectives 20(1): 3–24.

30.

Kalmijn

Uunk

(2007) Regional value differences in Europe and the social consequences of divorce: a test of the stigmatization hypothesis. Social Science Research 36(2): 447–68.

31.

Kassenboehmer

Haisken-DeNew

(2009) You’re fired! The causal negative effect of entry unemployment on life satisfaction. The Economic Journal 119(536): 448–62.

32.

Maassen

De Goede

(1989) Public opinion about unemployed people in the period 1975–1985: the case of the Netherlands. The Netherlands’ Journal of Social Sciences 25(2): 97–113.

33.

McAdams

(1997) The origin, development, and regulation of norms. Michigan Law Review 96(2): 338–433.

34.

McFayden

(1995) Coping with threatened identities: unemployed people’s self-categorizations. Current Psychology 14(3): 233–57.

35.

Möhring

(2012) The Fixed Effect as an Alternative to Multilevel Analysis for Cross-National Analyses. GK SOCLIFE Working Paper Series 16, University of Cologne, Köln, Germany.

36.

Muffels

Headey

(2013) Capabilities and choices: do they make Sen’se for understanding objective and subjective well-being? An empirical test of Sen’s capability framework on German and British panel data. Social Indicators Research 110(3): 1159–85.

37.

Niles

(1999) Toward a cross-cultural understanding of work-related beliefs. Human Relations 52(7): 855–67.

38.

Nordenmark

(1999) Employment commitment and psychological well-being among unemployed men and women. Acta Sociologica 42(2): 135–46.

39.

Oesch

Lipps

(2012) 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): 955–67.

40.

Paul

Batinic

(2010) The need for work: Jahoda’s latent functions of employment in a representative sample of the German population. Journal of Organizational Behavior 31(1): 45–64.

41.

Roest

AMC

Dubas

Gerris

JRM

(2009) Value transmissions between fathers, mothers, and adolescent and emerging adult children: the role of the family climate. Journal of Family Psychology 23(2): 146–55.

42.

Shields

Price

(2005) Exploring the economic and social determinants of psychological well-being and perceived social support in England. Journal of the Royal Statistical Society Series A: Statistics in Society 168(3): 513–37.

43.

Snijders

Bosker

(1999) Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling. London: SAGE.

44.

Stam

Verbakel

De Graaf

(2013) Explaining variation in work ethic in Europe: religious heritage rather than modernization, the welfare state and communism. European Societies 15(2): 268–89.

45.

Stam

Verbakel

De Graaf

(2014) Do values matter? The impact of work ethic and traditional gender role values on female labour market supply. Social Indicators Research 116(2): 593–610.

46.

Strandh

(2000) Different exit routes from unemployment and their impact on mental well-being: the role of the economic situation and the predictability of the life course. Work, Employment and Society 14(3): 459–79.

47.

Stutzer

Lalive

(2004) The role of social work norms in job searching and subjective well-being. Journal of the European Economic Association 2(4): 696–719.

48.

UNECE (United Nations Economic Commission for Europe) Statistics Database (2013a) GDP per Capita, in International Comparable Prices by Country/Region, Expenditure and Year. Available (consulted 28 March 2013) at: http://w3.unece.org/pxweb/database/STAT/20-ME/2-MENA/?lang=1

49.

UNECE (United Nations Economic Commission for Europe) Statistics Database (2013b) Unemployment by Country and Year. Available (consulted 28 March 2013) at: http://w3.unece.org/pxweb/database/STAT/20-ME/3-MELF/?lang=1

50.

Van der Meer

(2010) Gender, unemployment and subjective well-being: why being unemployed is worse for men than for women. Social Indicators Research 115(1) 23–44.

51.

Van Echtelt

(2010) Een baanloos bestaan. De betekenis van werk voor werklozen, arbeidsongeschikten en werkenden. Den Haag: Sociaal en Cultureel Planbureau.

52.

Van Ingen

(2009) Let’s Come Together and Unite: Studies of the Changing Character of Voluntary Association Participation. Tilburg: Ridderprint.

53.

Van Oorschot

(2006) Making the difference in social Europe: deservingness perceptions among citizens of European welfare states. Journal of European Social Policy 16(1): 16–23.

54.

Veenhoven

(1984) Conditions of Happiness. Dordrecht: Kluwer.

55.

Verbakel

(2012) Subjective well-being by partnership status and its dependence on the normative climate. European Journal of Population 28(2): 205–32.

56.

Wielers

Raven

(2013) Part-time work and work norms in the Netherlands. European Sociological Review 29(1): 105–13.

57.

Winkelmann

(1998) Why are the unemployed so unhappy? Evidence from panel data. Economica 65(257): 1–15.