The everlasting scarring effects of unemployment

(1)

The everlasting scarring effects of unemployment

Master Thesis HRM

By

Nick Mulder

Rijksuniversiteit Groningen

Faculty of Economics and Business

January 2015

Thesis supervisor: Peter van der Meer

Gelkingestraat 5-9

9711NA Groningen

06-19455094

(2)

ABSTRACT

This study focuses on the differences in subjective well-being between people who retired and have been unemployed in their life and those who have not. Unemployment negatively

influences the level of subjective well-being, and this effect can last permanently. This has been proved by results showing this permanent negative effect of long lasting unemployment.

Hypotheses are tested by the use of a cross-national survey that has been conducted every two years across Europe since 2001, known as the European Social Survey.

The main hypothesis is that the negative effect of long lasting unemployment diminishes when people retire. However, before testing this hypothesis, the effect of retirement on the level of subjective well-being was tested. Results show that retired people have a significant higher level of subjective well-being compared to people in paid work, which was in line with the

expectation. The main result of this research indicates that the permanent negative effect of unemployment on subjective well-being is diminished when people retire. However, still a significant difference in subjective well-being can be found between retired people who have ever been unemployed and retired people who have never been unemployed.

(3)

1. INTRODUCTION ... 1

2. THEORY ... 3

3. DATA & METHODS ... 7

4. RESULTS ... 9

5. DISCUSSION ... 19

(4)

1

1. INTRODUCTION

The growing unemployment rate the past years raises concerns. According to ‘Statistics Netherlands’ (CBS), the number of unemployed people in the Netherlands in April 2014 increased with a million people compared to exactly five years ago. The effect of being

unemployed has a huge impact on a person his life. People do not only lose their income, also the level of subjective well-being of these people decreases. As Cole et al. (2009) explain, the effects of a decrease in well-being do not have to be temporary and can last for years. Effects like lower self-esteem and insecurity can impair their ability to regain employment. As Hetschko et al. (2011) explain, losing a job is making people less optimistic in general and therefore results in a scarring effect of unemployment. In other words, people will face problems to regain

employment again due to their permanent lower level of subjective well-being.

But what happens with the level of subjective well-being when people retire? The baby boom generation nears retirement, resulting in high rates of retirees. Therefore, it is a relevant topic to investigate.

Former research suggests that overall the subjective well-being of people who retire is equally high or higher compared to ten years ago, when they were still working (Kim & Moen, 2001). The reasoning behind this theory is that they do not have the working stress anymore and can use their newfound leisure time for things they really appreciate (Insler, 2014). As Kim & Moen (2001) explain, conflicts and the difficulty of managing overload, related to the job, disappear. The result is an increase in the level of subjective well-being.

However, retirement does not necessarily lead to an increase in the level of subjective well-being for every individual. The subjective well-being of employed people is related to several factors which disappear after retirement As Jahoda (1984) suggest, the level of

(5)

2

of certain employed people is equally high as compared to their subjective well-being when they are retired.

So, previous research suggests that long-term unemployment leaves a scarring effect on the level of subjective well-being. Furthermore, a positive relation between retirement and subjective well-being is suggested. The question remains whether this scarring effect of unemployment lasts into retirement, diminishes or stops. This might shed light on the question why long-term unemployment is scarring.

This paper, therefore, extends the literature by not looking separately at the negative relation that long lasting unemployment has on the subjective well-being and the positive relation that retirement has on the subjective well-being. Instead, they are combined. By doing this a conclusion can be drawn about the research question of this paper, which is: Does the scarring effect of unemployment on well-being still exist when people retire or does this scarring effect disappear?

(6)

3

2. INFLUENCES OF UNEMPLOYMENT AND RETIREMENT ON

SUBJECTIVE WELL-BEING.

2.1 Long lasting unemployment and subjective well-being

Prior research suggests that long-term unemployment has negative effects on well-being (Cole et al. 2009). There are many researchers providing different explanations for this relation. Clark & Oswald (1994) suggest for instance that unemployed people have much lower levels of mental well-being than employed. Therefore they do not have the ability to exercise control over their life compared to employed people, which causes pain and stress. Another reason for this stress is that the unemployed do not receive an income anymore, which negatively contributes to their well-being (Winkelmann & Winkelmann, 1998). Furthermore, an explanation of how the well-being is negatively influenced by unemployment is because the social contact disappears. Haworth & Ducker (1991) showed that social contacts have a positive effect on well-being, and that these disappear when people get unemployed. Not only do they lose the contact with colleagues, but the unemployed also participate in less other social activities and get less social support from close relations and authority figures (Jackson, 1999). The result is a lower level of well-being which is one explanation that the pain of unemployment goes far beyond the loss of income (Schöb, 2013).

However, these reasons mentioned above do not fully explain why the unemployed perceive their well-being level lower as the employed. As Knabe et al. (2010) explain, a

desirable aspect of life is being employed because it gives our lives meaning. For this reason, the argument of Kahneman et al. (2004), which suggest that positive feelings are strongest during leisure activities and when interacting with family and friends, does not count when people are unemployed. What is striking about the relationship between long lasting unemployment and the level of subjective well-being is that there seems to be a permanent negative decrease in

(7)

4

researchers. Clark (2003) suggested that we humans identify with our work, especially men. It is one’s status that is often defined by one’s job (Jackson, 1982). Being employed is what is

expected from society. As especially men deviate from the social norm to work, this negatively affects their social identity. Goldsmith et al. (1996) explain that current and previous

unemployment lower current self-esteem. The long-term consequences may be feelings of discouragement and hopelessness. Unemployment also damages individuals’ perceptions of self-worth (Goldsmit et al. 1996). They perceive themselves as inherently adequate in their life (Kelvin & Jarret, 1985). It is a scarring effect of unemployment, whereby losing a job is making people less optimistic in general (Hetschko et al. 2004). Another possible explanation for the permanent negative effect of long lasting unemployment on well-being is an explanation given by Knabe & Rätzel (2011). They suggest that unemployed people are afraid of becoming

unemployed again in future, and that this future insecurity is the reason why people have a lower level of subjective well-being in general.

The hypothesis stated below is in line with prior research such as Bell and Blanchflower (2011), but necessary to investigate further for the remaining parts of the model.

Hypothesis 1: Long lasting unemployment permanently lowers subjective well-being.

2.2 The moderating effect of retirement on the relation between unemployment and well-being Originally, people retired at the age of 65. However, the government increased the

retirement age till 66 and will probably keep increasing this in the coming years. Many people do not like this increase in the retirement age. There are many reasons for this. As Insler (2014) for instance suggests, people have the feeling of being free to go wherever they want to when they retire. They do not have the work stress anymore that they used to have. They can use their time for the things they appreciate, like friends and family. When people reach their retirement age, the working life is left behind. It is a new stage in the course of life, which brings several consequences. A transformation in identity, expectations, preferences and meaning takes place when people retire, which have an influence on subjective well-being (Dannefer, 1984).

(8)

5

factors, which disappear when people retire. The subjective well-being of employees is positively influenced by income. Besides, the structure of the days, the work-related activities and the social contacts at work positively influence the subjective well-being of employees. When people retire, the income is reduced, the days are less structured and they lose their work-related activities and contacts. Therefore, not every retiree is positively influenced by retirement. Lucas et al. (2012) describe retirement as a ‘neutral’ life event resulting in a stable level of subjective well-being.

But there are also researchers who provide arguments whereby retirees remain active and busy, allowing them to feel important, and even increase their social life instead of losing it. Retirees can socialize with people whenever they want and with whomever they want. If they do not like something they can stop doing it. The same counts for social contacts, which they might not like. When retired, people are not forced to behave in specific ways as they might had to when they were employed (McPherson, 1990). So, this contradicts the argument of Lucas et al. (2012).

The stated hypothesis is in line with Easterlin (2003), who suggested that overall, the subjective well-being is increased when people retire.

Hypothesis 2: Retirement will positively influences subjective well-being.

Nevertheless, we need to find out if the scarring effect of unemployment on well-being still exists when people retire or if the scarring effect disappears.

(9)

6

retirement, the scarring effect of unemployment which negatively influences the level of subjective well-being, will diminish when retired. The hypothesis is stated as follows:

H3: Retirement will diminish the negative effect of long lasting unemployment on subjective well-being.

However, there might also be an explanation whereby the negative scarring effect of long lasting unemployment on subjective well-being lasts into retirement. As said before, Goldsmit et al. (1996) suggested that unemployment damages individuals’ perception of self-worth. If the scarring effect of unemployment is the result of a permanent loss of self-esteem, retirement will not change this scarring effect.

Figure 1: Conceptual Model

(10)

7

3. DATA & METHOD

The data used to test the hypothesis comes from the European Social Survey, held in 2012 (ESS round 6, 2012). The ESS contains data from 29 countries from all over the world. Example countries are for instance Belgium and the Netherlands.

Data is used from 25 countries out of 29 that participated in the sixth round of the survey. The countries which were not included in the sample were Albania, Russia, Ukraine and

Moldavia. The reason why these countries were not included are the highly influential circumstances in which these people are currently living, which has a huge influence on the answers given.

The sample consists of two subsamples. Firstly people in paid work, aged between 25 and 65, without missing answers on the most relevant variables, which are life satisfaction, happiness and past unemployment were selected. Secondly, retired people, aged between 50 and 90,

without missing answers as well on the most relevant variables, which are life satisfaction, happiness and past unemployment were selected. The final sample contains 30338 persons.

To find out if we can accept the first created hypothesis, a comparison is made between all the people who have ever been unemployed for more than 3 months but shorter than 12 months, people who have been unemployed for more than 12 months, and people who have never been unemployed to see if the level of their subjective well-being is significantly influenced by long lasting unemployment.

To test the second hypothesis, a comparison is made between people who have never been unemployed and retired versus people who have never been unemployed and are still doing paid work to see if there is a significant difference between their levels of subjective well-being. Finally, the main hypothesis is tested. A comparison is made between people in paid work who have ever been unemployed more than 12 months and people who retired and have ever been unemployed more than 12 months, followed by a comparison of the subjective well-being between retired and employed people who have and have not been long lasting unemployed. In this way, the conclusion can be drawn whether retirement influences the negative effect of long lasting unemployment on subjective well-being.

(11)

8

The second question reads (F37): ‘Have any of these periods lasted for 12 months or more?’. These questions were asked to both people being in paid work and people who retired. The ESS contains one question about retirement and person being in paid work. The question reads (F17a): ‘What have you been doing for the last 7 days?’, where they had the possibility to select the option retired and being in paid work. Based on the questions about retirement and unemployment, four different variables could be created, namely:

1) Retired – Never been long lasting unemployed

2) Retired – Ever been long lasting unemployed ( ≥ 12 months) 3) In paid work – Never been long lasting unemployed

4) In paid work – Ever been long lasting unemployed ( ≥ 12 months)

To create the subjective well-being variable, two questions can and should be used of the ESS. As Busseri & Sadava (2011), explain, subjective well-being can be divided into two components, namely affective well-being (e.g. feelings of happiness) and cognitive well-being (e.g. overall life satisfaction). Both components of subjective well-being are asked to all people in the ESS. The first question is about overall life satisfaction and reads (B20): ‘All things considered, how satisfied are you with your life as a whole?”. The second question is about the feelings of happiness and reads (C1): ‘Taking all things together, how happy would you say you are?’. Both questions were rated on a 11 point scale from 0 (lowest) to 10 (highest). These questions were combined into one single variable, because together they suggest the level of subjective well-being. It was allowed to do so because of the correlation between the questions (0.72) and the high cronbach’s alpha (0.82).

There are also some control variables included which are known to have an effect on subjective well-being of persons. These are gender, subjective income, marital status, general health and education level. Missing cases of these control variables were replaced by sample average or mode. Using list wise deletion of missing data would lead to a substantial smaller sample.

(12)

9

4. RESULTS

In this section the results will be presented, following the order of the hypotheses. First, some descriptive data is presented for the main variables used in testing each hypothesis.

The permanent effect of long-lasting unemployment on average well-being

In the table below, descriptive data is presented for people who have never been

unemployed, people who have been unemployed between 3 and 12 months and people who have ever been unemployed longer than 12 months.

Overall, people who have never been unemployed show on average a higher level of subjective well-being compared to those who have ever been unemployed. Furthermore, the table above shows that on average people who have ever been long-lasting unemployed show on average a lower subjective well-being compared to those who have been unemployed between 3 and 12 months. To test if these differences between different groups are significant, an ANOVA test was conducted. The results, F(2,30335) = 187,081, p < 0.01, indeed show that the average differences in the level of subjective well-being between the three different groups were indeed all significant. These results are in line with the descriptive table. However, further analysis is needed by including the country level effect to find out if these differences are in that case still significantly different. Therefore, multi-level modeling (MLM) is used with a random intercept for the respondent’s country, hereby controlling for general health, subjective income, education, gender and marital status. The main results are presented in Table 4. The empty model without

Table 1

Average Subjective Well-being of people in 25 European Countries

N Mean SD

Never Unemployed 22768 7.28 1.86

Unemployed 3 ≤ months ≤ 12 4553 7.0 1.92

Unemployed ≥ 12 months 3016 6.6 2.13

Total 30338 7.17 1.91

(13)

10

control variables can be found in Table 2 and the empty model with control variables only is presented in Table 3.

The variances of the countries and the residual are approximately .70 and 2.97. With these numbers, an intra-class correlation can be measured:  = .70 / (.70 + 2.97) = .19. This means that 19% of the total variance of subjective well-being is attributed to the fact that people come from different countries, which proves that multilevel modeling is necessary.

The controlling variables used were tested in an empty model to see if the restricted log

likelihood decreased significantly, in other words to find out if there was a significant better fit between the model and the data. Therefore, some control variables were included as a fixed effect whereas others as a factor effect.

Table 2

MLM with a random intercept: Covariance Parameters

Parameter 95% Confidence Interval

Estimate Std. Error Wald Z Lower Bound Upper Bound

Residual 2.97* .024 125.076 2.922 3.016

Intercept Variance .70* .202 3.451 .394 1.228 (Subject = country)

N 30338

Restricted log likelihood 123077.010 Dependent Variable: Subjective well-being * p <0.001

(14)

11

The variances of the countries and the residual are approximately .26 and 2.37. With these numbers, an intra-class correlation can be measured:  = .26 / (.26 + 2.37) = .09. This means that now only 9% of the total variance of subjective well-being is attributed to the fact that people come from different countries.

Furthermore, the restricted log likelihood compared to table 1 decreased with 8200. With a parameter increase of 5, a table with 2 values show that 2 (5) = 15.09, out of which can be concluded that there is a significant better fit between the model and the data by including the control variables.

All the control variables show a positive significant effect on subjective well-being, except education. Instead, the level of education has hardly any effect on subjective well-being, and therefore is not significant.

Table 3

MLM with a random intercept: Covariance Parameters & Control Variables

Parameter 95% Confidence Interval

Estimate Std. Error Wald Z Lower Bound Upper Bound

Residual 2.37* .019 124.421 2.335 2.410

Intercept Variance .26* .075 3.433 .145 .453 (Subject = country)

N 30338

Controlled for: general health, subjective income, marital status, gender and education

(15)

12

The table above shows that both short-lasting unemployment and long-lasting

unemployment have a significant negative effect on subjective well-being compared to those who have never been unemployed. These results, in combination with the descriptive statistics presented earlier, prove that the first hypothesis is accepted: Long-lasting unemployment permanently lowers subjective well-being.

Retirement will positively influence subjective well-being.

The effect on subjective well-being when people retire is the second hypothesis tested. As said before, the group of people who have never been unemployed have been split up in two different groups, which are the employed and retired people. First of all the descriptives are presented of these groups in table 5, followed by the results of an independent t-test in table 6, and finally the MLM results are presented in table 7.

Table 4

MLM with a random intercept & Dummy Groups for Unemployment

Parameter Estimate Std. Error

Intercept 3.05* .12

Unemployed 3 ≤ months ≤ 12 -.24* .02

Unemployed ≥ 12 months -.29* .03

Never Unemployed Reference Category

N 30338

(16)

13

The table above shows that people being in paid work while never been unemployed score higher (7.47) on subjective well-being compared to people being retired while they have never been unemployed in their working life (6.98). To test whether this difference on average subjective well-being is significant, an independent t-test was conducted. The table below indeed shows that on average, people who have never been unemployed in their life score higher when they are employed compared to those who are retired.

However, again further analysis is needed by including the country level effect to find out if these differences between the two groups are still significantly different in this case. Therefore, multi-level modeling (MLM) is used with a random intercept for the respondent’scountry, hereby controlling for general health, subjective income, education, gender and marital status.

Table 5

N Mean SD

Employed & Never been unemployed 14003 7.47 1.69 Retired & Never been unemployed 8766 6.98 2.08

N 22769

N=Number; SD=Standard Deviation Source: ESS (2012).

Table 6

Independent t-test for significant difference in average subjective well-being between employed & retired, both never been unemployed

t-test for equality of measures

95% Confidence Interval Sig. Mean Std. Error of the difference

t df (2-tailed) difference difference Lower Upper

Subjective well-being 19.66 22766 0.000 0.50 0.25 0.45 0.54

(17)

14

The results of the multilevel analyses show a significant difference between the two different groups. However, after controlling for the random effect and by introducing the control variables, employed people which have never been unemployed show a significant lower level of subjective well-being (-.27) compared to retired people which have never been unemployed. Therefore, the suggested relation by the t-test and descriptives is incorrect.

The stated hypothesis that retirement will positively influences subjective well-being can therefore be accepted.

Retirement will diminish the negative effect of long lasting unemployment on subjective well-being.

The third and main hypothesis is to find out if the negative effect of long lasting unemployment on subjective well-being can still be found when people retire. Therefore, the group of people who have ever been long lasting unemployed are split up in two different groups: currently employed people versus currently retired people. In combination with the results in the first two hypotheses, this final hypothesis can be tested. In table 8 below, the descriptives can be found, followed by the results of an independent t-test in table 9, and finally the results of the MLM analysis in tables 10 and 11.

Table 7

MLM with a random intercept & Dummy groups 'Current activity and never been unemployed'

Intercept 3.60* .12

Employed & Never been unemployed -.27* .02

Retired & Never been unemployed Reference Category

N 22769

(18)

15

The employed people who have ever been long lasting unemployed score a little higher on subjective well-being (6.64) compared to those who have ever been long lasting unemployed and currently retired (6.54). To test whether this average difference in subjective well-being is significant between both groups, an independent t-test was conducted. The result of the test in the table below shows that on average, the difference in subjective well-being between both groups is not significant.

However, again further analysis is needed by including the country level effect to find out if these differences between the two groups are still not significantly different in this situation. Therefore, multi-level modeling (MLM) is used with a random intercept for the respondent’s country, hereby controlling for general health, subjective income, education, gender and marital status.

Table 8

N Mean SD

Employed & Ever been long lasting unemployed 2237 6.64 2.08 Retired & Ever been long lasting unemployed 779 6.54 2.26

N 3016

N=Number; SD=Standard Deviation

Source: ESS (2012).

Table 9

Independent t-test for significant difference in average subjective

well-being between employed & retired, both ever been long lasting unemployed

t-test for equality of measures

95% Confidence Interval Sig. Mean Std. Error of the difference

t df (2-tailed) difference difference Lower Upper

Subjective well-being 1.06 3014 0.287 0.09 0.88 -.08 0.27

(19)

16

The table above shows that the relation as predicted by the t-test and descriptives is incorrect. Instead, people who retired and have ever been long lasting unemployed show a significant higher subjective well-being compared to employed people who have ever been unemployed in life. These results are in line with the second hypothesis, whereby retired people show a higher subjective well-being. However, another test needed to be conducted to find out if the negative scarring effect of unemployment on subjective well-being showed in the first hypothesis diminishes when people retire, as suggested in the third hypothesis. In the table below, the four different dummy groups are compared.

Table 10

MLM with a random intercept & Dummy groups 'Current activity and ever been unemployed'

Intercept 2.96* .20

Employed & Ever been unemployed -.28* .08

Retired & Ever been unemployed Reference Category

N 3016

(20)

17

The table above shows that retired indeed score higher than employed. Moreover, ever unemployed people score lower than never unemployed. Additionally, the effect of ever been unemployed on subjective well-being increases with (0.15) when people retire compared to employed. To test if the effect of ever unemployed between employed and retired people is significant, the model was tested once again. A dummy variable for ever unemployed was created, and compared with the retired and employed people, both never been unemployed. A difference in log likelihood of 34 was found. With a parameter increase of 1, a table with 2

values show that 2

(1) = 6.63, out of which can beconcluded that there is a significant difference between both groups.

Therefore, the third hypothesis can be accepted as well. In other words retirement will indeed diminish the negative effect of long lasting unemployment on subjective well-being. However, there is still a scarring effect of unemployment present when people retire.

Table 11

MLM with a random intercept & Dummy groups 'Current activity and never/ever been unemployed'

Intercept 3.25* .12

Retired & Ever been unemployed -.03 .04

Retired & Never been unemployed .29* .02

Employed & Ever been unemployed -.18* .02

Employed & Never been unemployed Reference Category

N 30338

(21)

18 Summary of results

Overall, support was found for all the three hypotheses. Long lasting unemployment indeed significantly lowers subjective well-being, where additionally a significant negative effect is found between people who have been unemployed for a short period and people who have been long lasting unemployed. These results are in line with previous research. Furthermore, the predicted effect of the second hypothesis was found. Retired people score higher on their level of subjective well-being compared to employed people. Interestingly, evidence was also found for the third hypothesis, showing that indeed the negative scarring effect of long lasting

unemployment diminishes when people retire.

(22)

19

5. DISCUSSION

Unemployment and retirement have been two widely discussed subjects over the past years in research. Many things have been written about how they influence the level of subjective well-being of people.

The main goal with the present research paper was to contribute to the existing

knowledge by trying to dig deeper in what the effects of unemployment and retirement are on the level of subjective well-being of people. By specifically concentrating on the moderating factor retirement, additional information has been provided on how retirement influences the negative effect of long lasting unemployment on the level of subjective well-being.

Many people became unemployed due to the economic breakdown. Moreover, government increased the age upon retirement. Furthermore, as estimated by the European Commission, the elderly population of, for example the Netherlands (65 years and over) is growing with 11.9% in the coming fifty years.

This means that simultaneously the retired population is growing. This makes the effect of unemployment on the level of subjective well-being very relevant, since this level is

influenced by retirement. But it also sheds some light of what causes unemployment to be scarring.

The relevance of this effect and the possible causes for the scarring effect are the strengths of this paper. Especially, because the level of subjective well-being is the main

predictor of health (Graham, 2008). And the health costs are again one of the main concerns for the government these days.

The results show, in accordance with previous research, that unemployment indeed leaves a scarring effect on the level of subjective well-being. The fact that the relationship holds, although controlled for subjective income, indeed confirms that the pain of unemployment goes far beyond the loss of income as is suggested by Schöb (2013) as well.

(23)

20

provide more evidence for the suggested relations in previous research. Therefore, they contribute to the discussion in such a way that it can be confirmed that unemployment indeed leaves a scarring effect on subjective well-being. Additionally, retirement positively influences subjective well-being.

The most important finding of this paper and at the same time the main contribution to science is that the negative effect of long lasting unemployment on subjective well-being diminishes when people retire. However, the effect still significantly differs from the level of subjective well-being of retired people who have never been unemployed in their life.

There might be a few explanations which can explain this relationship, which are all related to self-esteem. Apparently, the scaring effect of becoming unemployed again and the belonging decrease in subjective well-being indeed diminishes when people retire.

However, there is still a significant difference in subjective well-being which might be explained by Goldsmith et al. (1996). They explained that the individuals’ perception of self-worth is negatively influenced by unemployment, resulting in a lower level of subjective well-being. As the results show, the lower level of subjective well-being is not totally recovered by retirement. This might suggest that long lasting unemployment to some extent results in a permanent loss of self-esteem, which negatively influences subjective well-being and is not changed by retirement. Moreover, an explanation by Clark (2003) can be the reason for the significant difference. They explained that people who have been unemployed a long time before their retirement show less distress when retired, resulting in a higher subjective well-being. Unfortunately, such data was not available. Many retirees were too old and therefore it was not possible to find out if they had been unemployed a short or long time ago. Future research could take a closer look at this. In line with the previous explanation, another direction for future research is the employment status before retirement. As Hetschko (2011) explain, retiring from unemployment significantly influences subjective well-being. Furthermore, it was unfortunately not possible to find out if people were voluntary or involuntary unemployed in their lives, which does have a significant influence on subjective well-being as well (Winkelmann & Winkelmann, 1998).

Naturally, there are some limitations to this research as mentioned above. These

(24)

21

effect of unemployment originates and is influenced by retirement. A further limitation of this research is that it is impossible to make claims about causality, because it was impossible to test the hypothesis with longitudinal data.

(25)

22

REFERENCES

Bell, D.N.F., & Blanchflower, D.G. 2011. Young people and the Great Recession. Oxford Review of Economics Policy, 27(2): 241-267.

Busseri, M.A., & Sadava, S.W. 2011. A review of the tripartite structure of subjective well- being: Implications for conceptualization; operationalization, analysis, and synthesis. Personality and Social Psychology Review, 36: 917-927.

Clark, A. 2003. Unemployment as a social norm: psychological evidence from panel data. Journal of Labour Economics, 21(2): 323-351.

Clark, A.E., & Oswald, A.J. 1994. Unhappiness and unemployment. The Economic Journal, 104(424): 648-659.

Cole, K., Daly, A., & Mak, A. 2009. Good for the soul: The relationship between work, wellbeing and psychological capital. Journal of Socio-Economics, 38(3): 464-474.

Dannefer, D. 1984. Adult development and social theory: A paradigmatic reappraisal. American Sociological Review, 49: 100-116.

Easterlin, R.A. 2003. Happiness of women and men in later life: Nature, determinants and prospects. Advantages in Quality-of-Life Theory and Research Social Indicators Research Series, 20: 13-25.

ESS Round 6: European Social Survey (2012) ESS-6 2012.

(26)

23

European Commission, Economic and Financial Affairs 2011, The 2012 Ageing Report: Underlying Assumptions and Projection Methodologies, viewed 01-06-2014,

http://ec.europa.eu/economy_finance/publications/european_economy/2011/pdf/ee-2011-4_en.pdf

Graham, C. 2008. Happiness and Health. Health Affairs, 27(1): 72-87.

Goldsmith, A.H., Veum, J.R., & William Darity Jr. 1996. The Psychological impact of unemployment and joblessness. The Journal of Socio-Economics, 25(3): 333-358

Haworth, J.T., & Ducker, J. 1991. Psychological well-being and access to categories of experience in unemployed young adults. Leisure Studies, 10: 265-274.

Hetschko, C. 2011. Changing identity: Retiring from unemployment. SOEP – The german Socio-Economical Panel Study at DIW Berlin, 399: 1-31.

Insler, M. 2014. The health consequences of Retirement. The Journal of Human Resources, 49(1): 195-233.

Jackson, T. 1999. Differences in psychological experiences of employed, unemployed, and student samples of young adults. Journal of Psychology, 133: 49-60.

Jackson, P., Warr, P., & Banks, M. 1982. Duration of unemployment and psychological well- being in young men and women, Current Psychological Research, 2(1-3): 207-214.

(27)

24

Kahneman, D., Krueger, A.B., Schkade, D.A., Schwarz, N., & Stone, A.A. 2004. A survey method for characterizing daily life experience: the day reconstruction method. Science, 306(5702): 1776–80.

Kelvin, P., & Jarrett, J. E. 1985. Unemployment: Its social psychological effects. Cambridge, England: Cambridge University Press.

Kim, J.E., & Moen, P. 2001. Is retirement good or bad for subjective well-being? Journal of the Association for Psychological Science. 10(3): 83-86.

Knabe, A., Clark, A., & Rätzel, S. 2010. Boon or bane? Others’ unemployment, well-being and job insecurity. Labour Economics, 17(1): 53-61.

Knabe, A., & Rätzel, S. 2011. Scarring or scaring? The psychological impact of past unemployment and future unemployment risk. Economica, 78(310): 283-293.

Lucas, R.E., Eid, M., Hofmann, W., & Luhmann, M. 2012. Subjective well-being and adaption to life events: A meta-analysis. Journal of Personality and Social Psychology, 102(3): 592-615.

McPherson, B.D. 1990. Aging as a social process: An introduction to individual and population aging. Canada: Butterworth & Co. (Canada) Ltd, 136-151.

Schöb, R. 2013. Unemployment and Identity. Oxford Journals CESinfo Economic studies,

Snijders, T., & Bosker, R. 1999. Multilevel Analysis: An introduction to Basic and Advanced Multilevel modeling, Sage Publications, London.

(28)

25

Statistics Netherlands 2009, Unemployment significantly increased, viewed 01-06-2014,

http://www.cbs.nl/nl-NL/menu/themas/dossiers/conjunctuur/publicaties/artikelen/archief/2009/2009-051-pb.htm

The everlasting scarring effects of unemployment