The quality of social life and
subjective well-being
By Klaas Kwakkel
1University of Groningen
July, 27, 2010
Abstract:Essential in this thesis, is the paradox that average happiness levels in Western countries have not increased over time despite a tremendous increase in income per capita since the Second World War in these countries. To
explain this paradox, the relationship between the quality of social life and subjective well-being in The Netherlands is investigated. The approach taken is distinct for two reasons. Firstly, the study is macro in nature. Secondly, the quality of social life is modelled by using several variables that relate to the amount and quality of
enduring friendships and the amount and quality of contacts with friends, family, neighbours and other acquaintances. The results show that there exists a negative relationship between on the one hand the divorce
rate, the unemployment rate and the crime by force rate and on the other hand subjective well-being in The Netherlands on the macro level.
Supervisor: Dr. B.P.A. Gales
Keywords: Economic sociology; Subjective well-being; Happiness JEL classification: Z13
1
Contents
Page
Chapter 1: Introduction in happiness research
3
A: Introduction
3
B: Some background in happiness research
5
C: Stylized facts from previous happiness research
11
D: Explanations for the observed patterns
18
Chapter 2: Empirical part
27
A: Theory
27
B: Hypotheses
31
C: Data collection
32
D: Empirical results
35
Chapter 3: Concluding section
66
A: Discussion
66
Appendices
75
1. Introduction in happiness research
A. Introduction
In this thesis, the observation that average happiness levels in most Western countries have not increased over time in spite of large increases in real income per capita is essential. How can we explain this fact? Obviously, real income per capita is not a sufficient indicator for average happiness in a country. There is much more that matters for individual well-being2. For example, the quality of the living environment, the quality and number of social relations, the employment situation, the marital status, the quality of health care, the quality of the educational system, the level of democratic functioning of the government and so on. These factors all affect happiness at the individual level. The GDP as indicator for national well-being ignores the influence of these factors. Thus, there is a need for a more comprehensive indicator for average happiness within a country. Ronald Coleman,
executive director of the Genuine Progress Indicator for Atlantic Canada, puts it as follows: The more we produce, sell, and buy, the more the GDP grows, and the ‘better off’ we are supposed to be – or so the conventional wisdom tells us. Yet vital social and environmental factors remain invisible in these measures. The more trees we cut, the more fossil fuels we burn, and the faster we deplete our natural wealth, the faster the economy grows. This is poor accounting, like a factory owner who sells off his machinery and counts it as profit.3
More recently, initiatives have been employed to supplement or correct the GDP statistic with other aspects of life that are difficult to grasp in monetary terms. A well-known indicator is the Genuine Progress Indicator (GPI). This indicator starts with personal consumption and adjusts for income inequality. It adds the value of housework, the value of parenting, the value of volunteer work, the value of the service from household infrastructure and the value of service from streets and
highways. It finally subtracts the value of time including the cost of lost leisure time, family
breakdown, commuting time and underemployment. It also subtracts the cost of crime, auto accidents, long-term environmental degradation, air pollution, water pollution, ozone depletion, noise pollution, loss of farmland, loss of forests and the loss of wetlands (Anielski, 1999).
One of the most difficult problems with regard to alternative indicators for national well-being is that it is very arbitrary which factors one includes in such a measure and which not. And in the same direction, how much weight do the several factors receive? As a matter of fact, the composer of such a measure ultimately determines what factors people value the most. Consequently, he or she implicitly decides on what ultimately makes people happy. For example, how do being unemployed and being victim of a crime relate to each other and do they affect individual happiness equally strong? Even
2
Subjective well-being is the umbrella-term for things as individual happiness, individual life satisfaction, positive affect, negative affect, and so on. The research on individual happiness or life satisfaction is part of the broader research on subjective well-being.
3
though the measures suggested focus on the aggregate level, it is a very difficult and arbitrary job to decide on the exact factors to include and what weights to assign to the different variables that affect the ultimate value of such a measure for national well-being.
Next to this kind of research that focuses on the development of other indicators that complement or even substitute the traditional measurement of GDP, there exists research based on self-reported happiness by respondents. Based on this kind of research, one found that over time people have not become any happier in spite of large increases in real income per capita. This source of research might be suitable to uncover the determinants of subjective well-being. Some interesting results already have been found in previous research. For example, divorced people appear to be less happy than married people (among others Layard, 2005; Frey, 2008) and married people on average appear to be significantly happier than non-married people (among others Layard, 2005; Stack and Eshleman, 1998). Furthermore, involuntarily unemployed persons tend to be much less happy than employed ones (among others Layard, 2005; Frey, 2008; Winkelmann and Winkelmann, 1995). And finally, it turns out that people active in society are significantly happier than people that are not (among others Argyle, 1999; Thoits and Hewitt, 2001). Thus, happiness research might play a key role in finding out what variables should be included in a measure for national well-being and what
weights should be assigned to these variables.
In this thesis, I will focus on the relationship between the quality of social life and subjective well-being. Central is the question to what extent social contacts and contacts with closed ones (friends, family members, neighbours and other acquaintances) affect subjective well-being. The ultimate goal is to obtain valuable insights concerning the effect of the quality of social life on subjective well-being. And in the same line of thought: how does the quality of social life perform in comparison with other important determinants of subjective well-being.
B. Some background in happiness research
Before we can focus on the actual research on happiness, we first have to say something about the concept of happiness. We first need an adequate definition of ‘happiness’. Next, I will consider some of the issues that arise when measuring happiness. Finally, I spend some time explaining whether or not happiness is an objective or subjective state, and I will ask myself the question whether this creates serious problems with regard to the analysis or not.
The concept of happiness
The concept of happiness is extremely broad and the first thing that we need to do is to come up with an adequate definition. In essence, I define happiness simply as a feeling. When one asks a random individual whether he or she is happy, this person will be able to answer right away and on intuition. One is happy or one is not.
Veenhoven (1984), among others, states that the concept of happiness consists of both a cognitive and an intuitive part. The intuitive side of being happy simply means that the respondent feels happy most of the time. He or she usually experiences positive feelings. This is what Veenhoven calls ‘hedonic level of affect’. On the other hand, a person may consider himself or herself happy as the outcome of a cognitive process. When a person is asked whether he or she is happy, he or she makes an evaluation of his or her life so far. Vital in this evaluation are the goals, aspirations and needs a person has and whether these goals were achieved and these aspirations and needs were met. If this is the case, in the eyes of the respondent, he or she will consider himself or herself as ‘happy’. This makes happiness a subjective matter; the respondent self determines which goals, aspirations and needs are important and to what extent aims and needs were fulfilled. This cognitive evaluation process is what Veenhoven calls ‘contentment’.
According to Veenhoven, ‘hedonic level of affect’ and ‘contentment’ are the two building blocks of ‘overall happiness’. This implies that one has to look at both the hedonic level of affect and the contentment component before one can judge whether someone is happy or not. At first, this seems quite reasonable. It will be clear that a person who feels good (hedonic level of affect), whose goals were achieved and whose aspirations and needs were met (contentment) will be happy. But it will be more difficult to assess the happiness of an individual who feels good most of the time, although he or she was dissatisfied with the goals achieved and the aspirations and needs fulfilled in life. The same holds the other way around: someone can be very satisfied with the goals and aspirations realised and needs met in life and still feel depressed. How do you judge whether these persons are happy or not?
with the goals and aspirations realised and needs fulfilled; if one still feels depressed this negative feeling will dominate and the person will not be able to feel happy. The same holds the other way around. One can be dissatisfied with the goals and aspirations realised and needs fulfilled; but if this person feels good and enjoys life most of the time, the person will be able to feel happy because these positive feelings dominate. This is for example the case in some poor Third World countries. People in these countries often lack material possessions and they can hardly fulfil their basic needs.
Nevertheless, many people in these countries often consider themselves as happy.
Essentially, my definition of happiness corresponds to the intuitive part of Veenhoven’s definition of happiness, namely ‘the hedonic level of affect’. Layard (2005) comes up with a similar definition. According to him being happy is nothing more than ‘feeling good’. He does not refer to feeling good as the outcome of some kind of peak experience, the kind of feeling that arises as a consequence of, for example, the use of drugs or alcohol. Instead, Layard emphasizes that feeling good is a state and that happy people feel good on average (or most of the time as Veenhoven suggests). In my view, feeling good is indeed the most appropriate definition of happiness. Positive feelings matter most.
However, it might be the case that respondents still make an evaluation of their life in terms of goals achieved and aspirations and needs met, when they are confronted with a question about their individual happiness. They might base their answer on the outcome of such an evaluation of life. This implies that there is a discrepancy between how I define happiness here and how some of the
respondents will define happiness when answering a happiness question in a survey. Things become even more complicated because research about happiness makes use of both happiness questions and life satisfaction questions. I will go into this distinction in the next section.
Happiness versus Life Satisfaction
aspirations and needs and successes and failures. However, when we consider the definition used for happiness in this article, feeling good on average, one has to admit that happiness and life satisfaction are truly different concepts. After all, to assess whether one feels good on average, one does not need to evaluate the individual’s goals, aspirations and needs in life and to what level these had been fulfilled. Most of the time, the question about happiness is answered intuitively. This is an important difference between the concepts ‘life satisfaction’ and ‘happiness’, which is slightly underestimated in the existing happiness literature, one could say. It is important to be aware of the existence of this conceptual difference between happiness and life satisfaction. And in the same line of thought, one should be aware of the fact that true happiness, as defined in this thesis, is modelled by the questions about happiness, thus feeling good, in the various surveys.
The measurement of happiness
How can one prove, in an objective way, that someone is happy? This is a rather difficult question to answer and it might even be the case that it is not possible at all to measure individual happiness in an objective way, although one might claim that it is possible to measure individual happiness objectively based on outcomes of scientific experiments in modern brain research. I will return to this subject later on.
Up to now, nearly all research on subjective well-being is based on respondent’s answers on happiness or life satisfaction questions. These answers are used to analyse to what extent individuals are happy or not. See table 1 for some examples of happiness and life satisfaction questions. These questions are part of broader surveys. These surveys are carried out on a regular basis by order of the European Union (for example the Eurobarometer Surveys), the United States (for example the General Social Survey), other nations or states or independent scientific institutions, such as the World Values
Table 1: some examples of life satisfaction and happiness questions used in surveys
Happiness Life satisfaction
"Taken all things together, how would you say things are these days - would you say you're very happy, fairly happy, or not too happy these days?" (source: Eurobarometer 25)
"On the whole, are you very satisfied, fairly satisfied, not very satisfied or not at all satisfied with the life you lead?" (source: Eurobarometer 25)
1 - very happy 1 - very satisfied
2 - fairly happy 2 - fairly satisfied
3 - not too happy 3 - not very satisfied
4 - not at all satisfied “Taking all things together, would you say you
are…?” (source: World Values Survey)
“All things considered, how satisfied are you with your life as a whole these days?” (source: World Values Survey)
1 - very happy 1 – completely dissatisfied
2 - quite happy 2,3,4,5 - ….
3 - not very happy 6,7,8,9 - ….
Survey Association (World Values Survey). Generally, people are quite capable in reporting their individual happiness (Veenhoven, 1984). More problematic for respondents is how to explain what in particular makes them happy. Most scientific research is based on this method of generating
information on subjective well-being and this holds for this thesis too.
Happiness research by asking questions
The fact that happiness is measured on the basis of large surveys entails the problem of suitability: to what extent is this kind of data suitable for scientific research? How reliable is data gathered this way? In the following section, I will discuss some of the difficulties with this kind of data.
The first serious drawback of this kind of data is that happiness becomes a subjective matter. Every respondent has his or her own ideas about the concept of happiness. In fact, individual A can be very happy, whereas individual B is relative unhappy in an identical situation. This is not a large problem for the scientist. The fact that each respondent has a different idea of the concept happiness is completely irrelevant, as long as the respondent answers the question sincerely. The only task for the scientist is to uncover the factors that affect happiness. Based on the outcomes of such research, scientists might be able to explain differences between subjective well-being on the individual level. Thus, using survey outcomes as basis for the research proposed in this thesis should not be too problematic, even though the concept of happiness is subjective on the individual level (Veenhoven, 1984).
Another difficulty with this kind of data is to what extent individuals are capable of assessing their personal happiness. Do people think about their personal happiness? Do they know what the concept happiness means? Different studies show that individuals are quite capable of answering happiness questions and that individuals know very well what is meant with happiness (among others Veenhoven, 1984, p.41; Easterlin, 2001, p.466; Kahneman and Krueger, 2006, p.6). The percentage of respondents that answers happiness and life satisfaction questions in surveys with ‘no response / don’t know’ is extremely low, indicating that individuals know very well how happy or satisfied with life they are.
Another complexity with regard to the use of this kind of data for scientific research is the claim that answers in a survey depend strongly on contextual factors. It is claimed that the answers in large surveys depend on, at first sight, irrelevant factors, such as the weather on the concerning day, the mood of the respondent, the questions in the survey that precede the question about happiness or life satisfaction and so on. In general, the effects of these contextual factors tend to counterbalance each other, if the sample of respondents is sufficiently large (Kahneman and Krueger, 2006). This implies, for example, that the number of people in a bad mood answering the survey is offset by the number of people in a good mood.
respondents understand the concept of happiness or not. See for a review on common problems with regard to the reliability and validity of happiness questions in surveys among others Veenhoven (1984, 1993) and Diener (1984). Furthermore, Michael W. Fordyce (1988) investigated how surveys about life satisfaction and happiness could be optimally structured. In his research he analyzes all kinds of survey questions about life satisfaction and happiness. He focuses on the structure of the question and the answer categories, the number of answer categories, the exact formulation of the questions and answer possibilities and so forth. In the end, Fordyce concludes that the exact formulation of the questions and answer possibilities and the structure of the question do not matter that much. The results of happiness research are surprisingly consistent, given the huge variety in happiness and life satisfaction questions available (p.375).
Happiness research by looking at behaviour
From the moment the research on subjective well-being started to grow, researchers became increasingly interested in the issue of proving that someone is happy or unhappy. Some interesting facts with regard to this latter issue are worth mentioning.
First of all, Diener (1984) states that people who laugh and smile often are happier than people who laugh and smile less. These findings are supported by research of Kahneman and Krueger (2006), who state that happier people ‘smile more with the eyes (unfakeable smile)’, next to the fact that they laugh more. Furthermore, Kahneman and Krueger (2006) report that happy people are healthier. They find that happier people tend to be less sensitive to an ordinary cold. In addition, wounds of happy people tend to heal faster than wounds of less happy people. Kahneman and Kruger refer in this context to research of Cohen et al. (2003) and Kiecolt-Glaser, McGuire, Robles and Glaser (2002). These findings are confirmed by Layard (2005), who reports that happier people have better functioning immune systems. This was illustrated with a test where people were exposed to an influenza virus. Happy people appeared to be less susceptible to the virus than less happy people. Furthermore, happier persons appeared to have less cortisol, a stress hormone, in their blood. Besides, both the heart rate and the blood pressure tend to be lower for happy persons (Layard, 2005, p.23-24). The results of one investigation in particular are worth mentioning. Nuns that considered themselves happy in 1932 appeared to live significantly longer than nuns that did not explicitly consider
themselves happy in 1932 (Danner et al. 2001).
In addition to the above mentioned results of particularly physiological research, important results have been achieved in neurology. An exciting discovery is the finding that the left frontal part of the brain becomes active when someone experiences positive feelings. When someone experiences negative feelings, the right frontal part of the brain becomes active. This is a revolutionary
individual’s statement is correct. If not, we should doubt the sincerity of the individual with regard to his or her happiness statement. This way, it is possible to show in an objective way whether an individual is happy or not. Although this discovery is a very promising result of modern brain research, we would better not react overenthusiastic. It appears that it is possible to simulate brain activity in either the left or right frontal part of the brain by thinking about positive or negative things during the brain scan. Thus, although modern brain research finds promising results, it is too early to use brain scans to measure individual happiness objectively (Layard, 2005).
Concluding remarks
As previously noted, most of the existing happiness research is based on surveys that contain either life satisfaction or happiness questions. There exist many difficulties with regard to the use of this kind of data for empirical research. However, the answers on life satisfaction and happiness questions tend to be very reliable. Next to that, results from modern brain research imply that the experience of happiness is globally the same among individuals; the left frontal part of the brain is significantly more active for happy people than for less happy people. Furthermore, the exact formulation and structuring of happiness questions does not affect the outcomes severely. Of course, both the formulation and structuring of the questions are important, but they are not decisive. Fordyce (1988) has shown that the results from many different happiness and life satisfaction questions are surprisingly consistent. All in all, one can conclude that data gathered on the basis of happiness and life satisfaction questions in broader surveys are reliable and of relatively good quality. That makes this kind of data suitable for empirical research on happiness.
The benefit of studying happiness for economic science
Why should economists study happiness or at least listen to what happiness researchers have to say? In brief, it boils down to the fact that results of happiness research support and add to what economic scholars have pointed out. Yet on other occasions, results from happiness research contradict economic findings. For example, happiness research has shown that people in richer countries have not become happier, even though real income per capita in these countries has risen tremendously. ‘Classical’ economic science is unable to explain this fact (Frey and Stutzer, 2002). Economic theory predicts that extra income leads to extra utility, whether at a decreasing rate or not4. After all, more money implies that there are more choices open for the individual. Via optimization theory, extra choices lead to higher quality of life (Diener and Seligman, 2004). In fact, utility in the context of economic theory is a quantitative approximation of happiness. The advantage is that utility can be
4
measured objectively once individual income is known. Utility levels can thus be easilier compared between individuals than happiness levels. An important disadvantage is that it is difficult, or even impossible, to analyse the effects of other, less tangible, factors on individual utility. Essentially, these considerations form the basis for the existing discrepancy between the economic concept ‘utility’ and the concept ‘happiness’ in happiness research.
Furthermore, happiness research contains additional information for economic scholars. Results from happiness research show that being employed has a positive effect on subjective well-being. Being unemployed has detrimental effects on individual subjective well-being, even when the effect of the loss of income is controlled for (Frey and Stutzer, 2002). This is an important insight that should be taken into account when making employment policies, next to ‘classical’ economic insights.
Other examples of how happiness research gives additional insights, concern individual preferences about the unemployment and inflation rate. Results from happiness research show how self-reported happiness is affected as a consequence of changes in the inflation and/or unemployment rate (Di Tella, MacCulloch and Oswald, 2001).
In conclusion, happiness research offers economic scientists and policymakers some very relevant insights. The difference between happiness research and economic research arises from a different use of the concept happiness. Economists today generally use a narrow definition for the concept
happiness, namely ‘utility’. In principle, utility is only affected through income and ignores the influence of other factors (it is assumed that other factors do not affect utility). The advantage of using utility is that it can be measured relatively easy and one can make strong claims how to increase or decrease individual utility. In the field of happiness research a much broader definition of the concept happiness is used. The advantage is that the effect of other factors, besides income, on subjective well-being is taken into account. The main disadvantage is, as one would expect, that it is difficult to measure happiness exactly.
C. Stylized facts from previous happiness research
become any happier, even though real income per capita had grown drastically. These results were very remarkable, because a positive relation between income and happiness was conventional wisdom up to that moment. After all, economic science teaches us that increased real income leads to increased utility (whether at a decreasing rate or not), with utility being a quantitative approximation of
happiness. From that moment on, the literature about the relation between subjective well-being and income started to grow strongly. In the following section, I will present some of the stylized facts of the research in this field.
Within countries
Table 2 is copied from an article of Bruno S. Frey and Alois Stutzer, What can economists learn from happiness research, published in 2002. The table shows that on average richer people tend to be happier than poorer people in the United States. This holds for both the periods 1972-1974 and 1994-1996. People in income decile 10 have a happiness rating of 2.36 (2.36) on average, compared to people in income decile 1 who have a happiness rating of 1.94 (1.92) on average in the period 1994-1996 (1972-1974). The average happiness rating is the weighted average of the different answer categories ‘not to happy (1)’, ‘pretty happy (2)’ and ‘very happy (3)’, with the weights given between brackets. Next to that, the table also shows that over time, people have not become happier at all in the United States. After all, the average happiness rating between both periods changed barely, although real income per capita has increased strongly. Later on in this article, I will return to the development of happiness over time.
Table 3 and 4 are based on data from the World Values Survey. Table 3 shows the average happiness rating to income decile for twelve selected countries. Table 4 shows the average life satisfaction rating to income decile for the same group of countries. The number of observations for each income decile is at least 30, yielding reliable happiness or life satisfaction ratings. Based on both tables, one can conclude that, on average, richer people tend to be happier or more satisfied with life
Table 3: average happiness rating to income decile for some selected countries
Decile UK Canada Japan Mexico
South
Africa Australia Norway Romania Colombia New Zealand Mali 1 3.27 3.19 2.90 3.36 2.83 3.11 3.11 2.07 3.21 3.23 2.87 2 3.07 3.26 3.13 3.42 2.84 3.24 3.24 2.21 3.34 3.24 2.85 3 3.36 3.39 3.27 3.41 3.07 3.30 3.21 2.23 3.42 3.21 2.75 4 3.28 3.34 3.22 3.50 3.08 3.22 3.25 2.49 3.39 3.29 3.22 5 3.40 3.37 3.18 3.56 3.22 3.22 3.37 2.48 3.50 3.52 3.19 6 3.52 3.53 3.27 3.60 3.34 3.30 3.31 2.61 3.48 3.42 3.34 7 3.49 3.39 3.18 3.60 3.46 3.27 3.42 2.61 3.35 3.39 3.46 8 3.54 3.50 3.25 3.60 3.56 3.26 3.48 2.84 3.41 3.36 3.49 9 3.47 3.47 3.20 3.56 3.41 3.37 3.51 2.87 3.59 3.37 3.67 10 3.53 3.55 3.33 3.69 3.28 3.46 3.49 2.92 3.62 3.41 3.54 Total 3.43 3.41 3.17 3.51 3.16 3.26 3.34 2.54 3.35 3.36 3.20
Sources: World Values Survey, round 5 (2005-2008), 58 countries
Notes: The number of observations for these countries per decile is at least 30. The average happiness rating is the weighted average of the number of respondents that answered the question Taking all things together, would you say you are…? either with very happy (4), quite happy (3), not very happy (2) or not at all happy (1) and the nominal value of these answers (see value between brackets).
Table 4: average life satisfaction rating to income decile for some selected countries
Decile UK Canada Japan Mexico South
Africa Australia Norway Romania Colombia New Zealand Mali 1 7.33 7.00 6.16 7.88 5.61 6.73 7.46 4.11 8.04 7.45 5.25 2 7.03 7.33 6.88 7.94 6.49 7.24 7.41 4.47 8.36 7.16 4.65 3 7.40 7.50 6.81 8.07 6.56 7.30 7.77 4.81 8.26 7.78 4.75 4 7.20 7.43 6.93 8.20 7.11 6.96 7.85 5.39 8.32 7.78 5.48 5 7.40 7.69 7.25 8.31 7.33 7.36 8.18 5.46 8.58 8.26 6.24 6 7.68 8.02 7.19 8.56 7.39 7.35 8.14 6.20 8.64 7.78 6.41 7 7.61 7.70 7.16 8.65 8.10 7.42 8.23 6.47 8.67 8.05 6.96 8 7.77 7.85 7.42 8.43 8.27 7.43 8.28 6.49 8.24 7.79 7.68 9 8.00 8.13 7.38 8.33 7.84 7.66 8.10 6.71 8.50 8.19 7.47 10 7.97 8.12 7.80 8.47 7.51 7.89 8.53 7.16 8.78 8.11 8.17 Total 7.63 7.73 7.00 8.23 7.04 7.27 8.00 5.72 8.31 7.89 6.16
Sources: World Values Survey, round 5 (2005-2008), 58 countries
Notes: The number of observations for these countries per decile is at least 30. The average life satisfaction rating is the weighted average of the number of respondents that answered the question All things considered, how satisfied are you with your life as a whole these days? on a scale from 1 to 10 where 1 stands for ‘dissatisfied’ and 10 stands for ‘satisfied’ and the ordinal values of 1 till 10.
holds or not. Another interesting observation is that it seems to be the case that the effect of income on subjective well-being is less strong for higher incomes. In table 3, the difference in average happiness ratings between people in income decile 1 and 5 is much larger than the difference between people in income decile 6 and 10. This observation is confirmed in table 2, where average happiness ratings tend to grow relatively fast in lower income deciles and relatively slow in higher income deciles. These observations suggest that there exists a decreasing marginal effect of income on subjective well-being. The effect of one extra unit of money is smaller the more money one already has. This is quite
intuitive and in accordance with the assumption of decreasing marginal utility of income in classical economic theory.
Table 1A and 1B in appendix 1 show the normalised versions of table 3 and 4 on the previous page. The advantage of normalising is that one can compare the relationships between income and happiness on the one hand and income and life satisfaction on the other hand. See appendix 1 for a comprehensive explanation about the normalisation procedure used. Inspecting the tables in appendix 1, one can see that the relation between income and happiness on the one hand and income and life satisfaction on the other hand are pretty similar. As a consequence, one could conclude that both concepts measure the same. In that case, either respondents do not relate happiness and life satisfaction to complete different matters, or it must be the case that happy respondents are also satisfied with life and vice versa.
All in all, we can conclude that within countries richer persons tend to be happier than poorer persons. However, extra income increases subjective well-being at a decreasing rate. Results based on both happiness and life satisfaction questions support these findings. This is stylized fact number one. I will return to this section and the tables presented here when the relation between happiness and income between countries is considered.
Over time
Besides the average happiness ratings of richer and poorer people in the same country, table 2 also shows the development of the average happiness rating for different income groups over time.
Surprisingly, the average happiness rating has been relatively constant over time, although real income per capita increased. This is a remarkable finding that cannot be explained by classical economic theory, which states that increased real income leads to a higher utility level, where utility functions as a quantitative approximation of subjective happiness. Depending on the exact formulation of
rating increased for only two income deciles (decile 1 and 9) between 1972-1974 and 1994-1996. And finally, the overall average happiness rating has decreased, from 2.21 in 1972-1974 to 2.17 in 1994-1996, despite a rise in income per capita.
The idea that there does not exist a positive relationship between subjective well-being and real income per capita is supported by data presented in appendix 2. The figures in this appendix show the relationship between subjective well-being and real income per capita over time for some selected European countries. Careful inspection leads to the conclusion that the relation might even be negative for some countries, such as Belgium or Portugal. For other countries the relation between subjective well-being and real income per capita is not clearly positive or negative. This, for example, is the case for Ireland, The Netherlands and Spain. For other countries, there seems to exist a positive relation between subjective well-being and real income per capita. This is the case for Denmark that appears to be the happiest country of the countries selected in appendix 2.
Overall, we can conclude that there does not exist a clear positive effect between subjective well-being and income per capita over time. This is stylized fact number two. In many countries, income per capita has grown sharply over time, whereas happiness levels stayed more or less constant over time. One might even conclude that in some countries happiness levels have decreased over time. This is a rather remarkable result, in contradiction with what economic theory predicts.
Between countries
If we look back at the previous sections, where we discussed the relation between subjective well-being and real income per capita within countries and over time, one will also find information concerning this relation between countries. Both table 3 and 4 show that Mexico and Colombia are relatively happy, even though income per capita in these countries is relatively low. Japan, on the other hand, turns out to be less happy, despite the fact that income is much larger than Mexico’s or Colombia’s income per capita. The observation that the inhabitants of a country do not necessarily need to be rich to be happy is supported by figure 1 and 2.
Figure 1 is based on happiness data from the World Values Survey, wave 4. These numbers have been gathered in the period 1999-2004. The GDP data are from the World Development
Indicators. The figure shows that, in general, richer countries tend to be happier than poorer countries. However, a country does not necessarily be rich to be happy. For example, Nigeria, Tanzania,
Vietnam, Venezuela, Mexico and Saudi-Arabia all appear to be relatively happy while their real income per capita is relatively low. Surprisingly, the happiest nations appear to be poor nations
Figure 1: average happiness and income per capita at one point in time for some selected countries happiness and income for some selected countries
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 GDP p/c a v g h a p p in e s s r a ti n g
Albania Algeria Argentina Austria
Bangladesh Belgium Bulgaria Belarus
Canada Chile China Croatia
Czech Republic Denmark Estonia Finland
France Germany Greece Hungary
Iceland India Indonesia Iran (Islamic Republic of)
Ireland Israel Italy Japan
Jordan Republic of Korea Kyrgyzstan Latvia
Lithuania Luxembourg Malta Mexico
Republic of Moldova Morocco Netherlands Nigeria
Pakistan Peru Philippines Poland
Portugal Romania Russian Federation Saudi Arabia
Singapore Slovakia Viet Nam Slovenia
South Africa Zimbabwe Spain Sweden
Turkey Uganda Ukraine Macedonia, Republic of
Egypt Tanzania, United Republic Of United States Venezuela
Sources: World Values Survey, round 4 (1999-2004); WDI
Notes: The average happiness rating is the weighted average of the number of respondents that answered the question “Taking all things together, would you say you are…?” either with very happy (4), quite happy (3), not very happy (2) or not at all happy (1) and the nominal value of these answers (see value between brackets). GDP per capita is denominated in constant 2000 international dollars corrected for Price Power Parity.
Figure 2 gives us a comparable picture. Again, the inhabitants of the richer countries appear to be relatively happy or satisfied with life, whereas the inhabitants of some poorer countries are
Figure 2: average life satisfaction and income per capita at one point in time for some selected countries life satisfaction and income for some selected countries
2 3 4 5 6 7 8 9 10 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 GDP p/c a v g l if e s a ti s fa c ti o n r a ti n g
Albania Algeria Argentina Austria
Bangladesh Belgium Bulgaria Belarus
Canada Chile China Croatia
Czech Republic Denmark Estonia Finland
France Germany Greece Hungary
Iceland India Indonesia Iran (Islamic Republic of)
Ireland Israel Italy Japan
Jordan Republic of Korea Kyrgyzstan Latvia
Lithuania Luxembourg Malta Mexico
Republic of Moldova Morocco Netherlands Nigeria
Pakistan Peru Philippines Poland
Portugal Romania Russian Federation Saudi Arabia
Singapore Slovakia Viet Nam Slovenia
South Africa Zimbabwe Spain Sweden
Turkey Uganda Ukraine Macedonia, Republic of
Egypt Great Britain Tanzania, United Republic Of United States
Venezuela
Sources: World Values Survey, round 4 (1999-2004); WDI
Notes: The average life satisfaction rating is the weighted average of the number of respondents that answered the question “All things considered, how satisfied are you with your life as a whole these days?” on a scale from 1 to 10 where 1 stands for “dissatisfied” and 10 stands for “satisfied” and the ordinal values of 1 till 10. GDP per capita is denominated in constant 2000 international dollars corrected for Price Power Parity.
Tables 3 and 4 and figures 1 and 2 show that it is not necessarily true that richer nations are happier than poorer nations. Some poor countries are happier compared to richer countries. Stylized fact number three is that a country does not necessarily need to be rich to be happy.
Stylized facts of happiness research
In summary, it can be concluded that the findings from preceding happiness research can be captured in three stylized facts. Stylized fact number two and three cannot be accurately explained by classical economic theory.
1. Richer people tend to be happier than poorer people within countries, although extra income increases subjective well-being at a decreasing rate.
2. There does not exist a strong positive relation between subjective well-being and real income per capita within countries over time.
3. Richer countries are not automatically happier than poorer countries.
D. Explanations for the observed patterns
Of course, many researchers have attempted to explain the stylized facts mentioned above.
These attempts have led to the development of all kinds of theories that explain one or more of the stylized facts. Basically, these theories can be subdivided in three broad classes of theories, namely ‘basic needs theories’, ‘aspiration level theories’ and ‘cost-of-affluence theories’.
I. Basic needs theory
The basic needs theory states that every human being has fundamental needs. These needs have to be met before one can even start discussing the value of life. Examples of these basic needs include food, drink, shelter and clothing. Sufficient amounts of these goods function as preconditions for life, one could say. Essentially, if these needs cannot be fulfilled, for whatever reason, human life will be worthless (Veenhoven, 1991). The basic needs theory thus suggests that being capable of fulfilling one’s basic needs makes one’s life worthy and leading a worthy life functions as a precondition for a happy life. Without worthiness no happiness.
example 10,000 US dollars (Frey, 2008, p.42), the relation between income and subjective well-being is ambiguous or even disappears at all.
The basic needs theory is able to explain why richer people tend to be happier than poorer people within countries and why richer countries generally tend to be happier than poorer countries. Richer people and richer countries are better able to fulfil basic needs simply because they are richer. However, the basic needs theory cannot explain why several poor countries are relatively happy, despite the fact that their income per capita is rather low. As figure 1 and 2 show, this is the case for numerous countries.
Figure 3: life satisfaction and income for some selected countries
life satisfaction and income for some selected countries
2 3 4 5 6 7 8 9 10 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 GDP p/c a v g l if e s a ti s fa c ti o n r a ti n g
Sources: World Values Survey, round 4 (1999-2004); WDI (same figure as figure 2)
Notes: The average life satisfaction rating is the weighted average of the number of respondents that answered the question All things considered, how satisfied are you with your life as a whole these days? on a scale from 1 to 10 where 1 stands for ‘dissatisfied’ and 10 stands for ‘satisfied’ and the ordinal values of 1 till 10. GDP per capita is denominated in constant 2000 international dollars corrected for Price Power Parity.
II. Aspiration level theory
people are so little happier (if happier at all) than poorer people, given the assumption that basic needs are fulfilled.
In the original article (Does Economic Growth Improve the Human Lot? Some Empirical Evidence, 1974), Easterlin tried to explain the stylized facts with the aid of the relative income theory. Easterlin formulated the basic intuition of the relative income theory as follows: The basic idea was stated quite simply by Karl Marx over a century ago: “A house may be large or small; as long as the surrounding houses are equally small it satisfies all social demands for a dwelling. But if a palace rises beside the little house, the little house shrinks into a hut (Easterlin, 1974, p.11-12).” It is not absolute income that matters to subjective well-being, but relative income (in this case expressed as the quality of one’s house). One’s position on the fictive income ladder is what matters most for subjective well-being. In microeconomic terminology, one might say that interdependent utility functions apply in this case; next to the individual’s income, individual utility also depends on the income of other individuals or more general, the average income of the inhabitants of the country or state. Again, utility in this context is a quantitative approximation of individual happiness. In the context of the aspiration level theory, one assumes that individuals compare their (income) situation with that of their ‘reference’ group. The reference group mainly consists of colleagues, friends, family members and other acquaintances (Frey, 2008). The attentive reader will understand what this
principle implies: increasing the income of everyone does not lead to increased happiness for everyone. If everyone exchanges his or her little house for a palace, subjective well-being of all individuals will stay the same; the relative position of one individual vis-à-vis others has not changed. The relative position of individuals only changes as a result of changes in income inequality in a country. Thus, according to the aspiration level theory, the average happiness level in a country can only change as a result of changes in the income inequality in a country. I will return to the relation between income inequality and happiness in the next section. This mechanism might play a key role in explaining why happiness over time stayed the same in spite of large increases in real income per capita over time within countries. In addition, this mechanism might play an important role in explaining why richer countries are no happier than poorer countries.
Income inequality
predicted to be happier than countries characterized by a less equal income distribution. This reasoning seems to be very sound. Surprisingly, up till now this hypothesis lacks empirical evidence. As noted earlier, this principle might play a vital role in explaining the differences that exist in average
happiness ratings between various rich countries. I will return to this principle in the empirical part of this thesis, where the effect of income inequality on subjective well-being is controlled for.
An interesting article about the relation between income inequality and subjective well-being is from David Morawetz et al. (1977). In this article, two Israeli municipalities are compared to each other. In the first municipality there exist large income differences, whereas in the second municipality the inhabitants hand in their income which is equally divided between all inhabitants of the
municipality. The inhabitants of this second municipality appeared to be 0.8 till 0.9 points happier on a scale from 0 till 10 than the inhabitants of the first municipality. This is a very interesting finding. However, based on these findings, it is impossible to draw any solid conclusions, because the scope of the research is rather limited. Other explanations, besides income inequality, might play a role as well. The difference in subjective well-being between the inhabitants of the first and the second
municipality might be explained by the fact that the social and public character differ significantly between both municipalities. I will come back to this alternative explanation later on in this thesis.
Social comparison
In connection with Easterlin’s work in this field, much research has been carried out about the aspiration level theory. Frey (2008) and Layard (2005) both state that individuals always compare their situation with a certain standard. On the one hand, this standard arises because people tend to compare their situation with that of others, the so-called reference group. This is what is called ‘social comparison’. On the other hand, people tend to compare their situation with situations in the past. This is what is called ‘adaptation’ or ‘habituation’. Experiments show that it is very important to
individuals what position they have on the fictive income ladder. Layard (2005, p.41-42) reports the outcome of one experiment in particular. Suppose one can choose between two imaginary worlds to live in. In world A, you will earn 50,000 US dollars annually, whereas everybody else will earn 25,000 US dollars annually. In world B, you will earn 100,000 US dollars annually, whereas everybody else will earn 250,000 US dollars annually. Prices in both worlds will be the same. It appeared that most of the respondents preferred world A. In conclusion, relative income matters substantially to people.
Habituation
major drop in subjective well-being. Over time however, subjective well-being returned to their initial level. These findings are very interesting and show that habituation effects play an important role in the experience of happiness on the individual level. As a result of external factors, for example a pay raise, a divorce or a car accident, individuals have to cope with a new situation. It takes time to get used to this new situation. Once they get used to the new situation, the level of subjective well-being returns to the initial level. However, it is not certain whether or not human beings are capable of adapting to every external change possible. Veenhoven (1991) suggests that people are unable to get used to structural shortages of basic needs, such as food, drink, shelter and clothing. Other scholars claim that the positive effects of marriage or the negative effects of unemployment are lasting. Subjective well-being is claimed to be structurally larger as a consequence of marriage, according to Easterlin (2004) and Frederick and Loewenstein (1999). In addition, the negative effects of being unemployed are claimed to be permanent, according to Layard (2005, p.67).
The aspiration level model
Figure 4: Happiness as function of aspiration level and income (based on Easterlin 2001)
individual happiness. After a while, individual happiness has returned to its original level, whereas the amount of income and/or assets is at a higher level compared to the original situation. This process is what is called ‘the hedonic treadmill’ in the happiness literature (among others Layard, 2005; Frey, 2008; Easterlin, 2001; Kahneman and Krueger, 2006). To become happier, one needs a positive change every once in a while to remain in front of the process of rising aspirations.
In real life, not every individual will respond equally sensitive to a change in income. One can model this by changing the slope of the adaptation curves in figure 4. The slope represents the
sensitiveness or elasticity between income and subjective well-being. A large slope implies that the individual’s happiness is strongly affected by a change in income. This might be the case for individuals with highly materialistic values. A small slope implies that happiness does not respond very heavily as a result of an income change. This might be the case for individuals who attach less value to money. Horizontal aspiration curves imply that there is no relation between individual aspirations and income. In that case, subjective well-being is not affected by changes in income on the individual level. On the personal level, it depends on both personality and character traits (materialistic or not), to what extent subjective well-being is affected by changes in income. On the macro level, as we saw earlier, a larger income goes together with a higher level of happiness, at least within countries (stylized fact number one).
The perfect explanation?
In all fairness, the aspiration level theory seems to be a very plausible theory and is able to explain some of the stylized facts with regard to happiness research. However, an important disadvantage of the aspiration level theory is that it implies that the actual experience of happiness is in fact unrelated to the actual quality of life. Given that basic needs are fulfilled, the actual experience of happiness only depends on the comparison of the individual’s situation with some kind of standard. By simply lowering one’s aspiration level, one could be significantly happier. This is at least counterintuitive, not to say strange. How can this be explained?
Veenhoven (1991) points at the different happiness concepts and definitions used in the literature. As stated earlier, Veenhoven’s concept of happiness exists of both an intuitive part (‘hedonic level of affect’) and a cognitive part (contentment). The aspiration level theory makes pre-eminently use of the cognitive concept of happiness. By evaluating the cognitive part of happiness, the individual assesses to what extent certain goals in life were achieved and to what extent certain
aspiration level is only able to explain the stylized facts to a certain extent, because overall happiness is associated with the fulfilment of basic needs. The aspiration level theory ignores this latter aspect.
III. The cost-of-affluence theory
As opposed to the basic needs theory (absolute income) and the aspiration level theory (relative income), explained in the previous sections, the cost-of-affluence theory does not focus on income. As the name implies, this theory focuses on the adversary effects of increased economic welfare.
Consequently, this theory is mainly applicable to richer countries. In his original article, Easterlin (1974) already noted the following: Other possible interpretations of these data come to mind [next to higher aspirations]. For example, emphasis might be placed on external diseconomies of production. At a given time, it might be argued, the rich are better able to avoid these sources of ‘ill-fare’ and hence are happier. But over time and across societies increases in income are largely or wholly offset by a corresponding growth in pollution, congestion, and so forth (p.113). Today, pollution and
congestion are important problems in many Western countries. Next to that, other problems have come to the surface in many rich countries. One might even argue that these additional problems are caused by the increased material welfare that characterizes richer countries.
The last decennia for example, stress, overweight, depression and other psychic problems have increased dramatically in these countries (among others Diener and Seligman, 2004). Some researchers even argue that these problems are part of a more comprehensive problem, namely the decrease of the quality of social life during the last fifty years (among others De Vos, 2003; Layard, 2005). One might thus consider that the decreased quality of social life is an adverse effect of increased economic welfare.
As noted earlier, the basic needs theory focuses on the steeper part of the curve (in figure 3), whereas the aspiration level theory focuses on the flatter part. The cost-of-affluence theory focuses on the flatter part of the curve as well, without redefining the concept of happiness. After all, the
aspiration level theory makes use of a happiness concept in terms of goals achieved and aspirations and needs met as a starting point. Implicitly, this leads to a disconnection between the actual experience of happiness and the value of life. This is not the case with the cost-of-affluence theory. The adverse effects of economic welfare do affect the value of life. Even though basic needs are fulfilled, environmental pollution, congestion, nuisance, stress, psychic problems and so on do affect the value of life on the individual level. This is an important distinction compared to the aspiration level theory.
The cost of affluence in summary
therefore, inhabitants of these countries have not become any happier during the last fifty years, despite significant increases in income per capita. However, the literature is not completely clear about the exact description of these unpleasant effects of increased material welfare and how these effects exactly affect subjective well-being.
The sociological literature emphasizes the decrease of the quality of social life. Theoretically, the quality of social life might have a large influence on individual happiness, because every human being feels the need to be part of a community with interpersonal contact (Diener and Seligman, 2004). However, ‘decrease of the quality of social life’ is a very broad and to some extent vague description, which is difficult to specify. In addition, one might question whether this is truly a negative effect of increased material welfare or not.
Economists, on the other hand, traditionally focus on the external effects of economic growth. These ‘externalities’ (as they are called in economic jargon) are for example environmental pollution, congestion, nuisance and so on. In the standard production process, these costs are not taken into account by firms5. Although these factors affect happiness negatively, firms generally do not compensate for these adversary effects. It is, however, not certain to what extent these externalities influence individual happiness.
Furthermore, it might be the case that human values have changed over time. An interesting example is how individuals look at the existence of God. During the second half of the twentieth century, the percentage of people that believes in God has deteriorated strongly in most Western countries. Research shows that people who believe in God tend to be happier than people who do not (among others Layard, 2005, p.72). This example shows that a change of personal values (whether one believes in God or not) affects the average happiness rating in a country. However, it is not sure to what extent changing values are caused by an increase in material welfare.
All in all, the cost-of-affluence theory implies that the growth of economic welfare has some negative effects for subjective well-being. However, the literature is not unanimous about the exact description of these negative effects. Sociologists primarily focus on the decreased quality of social life, whereas economists focus more on the externalities of economic growth. Next to that, it might be the case that people’s values have changed over time as a consequence of increased material welfare.
5
2. Empirical part
In the remainder of this thesis, I will focus on stylized fact number two: There does not exist a strong positive relation between subjective well-being and real income per capita within countries over time. I will try to explain this finding with the aid of the cost-of-affluence theory, with a focus on the quality of social life within countries. As we already concluded, there does not exist a clear relation between income per capita and subjective well-being over time. In some cases, there seems to exist a positive relation (Denmark). For some other countries, there seems to exist a negative relation (Belgium, Portugal). To explain these findings, I will not use the basic needs theory. I will primarily focus on countries that have been able to fulfil their basic needs the last fifty years. In terms of figure 1 (page 16), these countries are located on the flatter part of the curve. Hence, the basic needs theory is not suitable. Neither will I use the aspiration level theory. This theory is not suited to play a role in explaining the paradox that happiness has not risen over time despite increases of real income per capita. The main disadvantage of the aspiration level theory is that it is based upon a happiness concept that differs significantly from the one I use throughout this thesis. I will use the cost-of-affluence theory and in essence, I will focus on sociological developments during the second half of the twentieth century in the Netherlands. I will restrict the empirical analysis to The Netherlands, because there is very little sociological data available that can be compared between several
(European) countries. I will focus on the explanation that the quality of social life has decreased over time and this has affected individual (and aggregate) subjective well-being negatively.
A. Theory
As previously noted, some scholars suggest that problems, such as stress, overweight, alcohol abuse, drugs abuse, depressions and other psychic problems are the result of a decrease in the quality of social life. De Vos (2003), for example, stresses that people do not only look at their financial situation when asked to their individual happiness. People also evaluate ‘the rest’, by which they mean the quality of social life (p.288). Thus both the respondent’s financial situation and the quality of the respondents’ social life determine to what extent the respondent feels good. During the second half of the twentieth century, the financial situation, or the quality of material life, has increased drastically in most Western countries. Real income per capita has grown strongly in these countries. In addition, products and services have been developed that please our lives, such as vacuum cleaners,
microwaves, dishwashers, televisions, solar banks, saunas, contraception, mobile phones, the internet, amusement parks, foreign holidays and so forth. Many of these products and services are available to a much larger group of people than before (De Vos, 2003, p.287).
of divorces has increased dramatically during the last fifty years (among others Putnam, 2000; Layard, 2005; De Vos, 2003) and it is more difficult to maintain enduring friendships (Layard, 2005).
Furthermore, interpersonal contact is more diverse and more volatile nowadays. Finally, membership of organized clubs, associations and other communities has decreased steadily during the last fifty years (Layard, 2005; Putnam, 2000). These developments show that the quality of social life in The Netherlands and in other Western countries has decreased, according to De Vos (2003). This explains why people have not become happier over time.
According to De Vos (2003), a good social life is a prerequisite for a happy life. De Vos (2003) links the decreased quality of social life in The Netherlands during the second half of the twentieth century to the demise of community life. Every human being is a social being that feels the fundamental need to interact with other people and to be part of a larger community of human beings. Human beings flourish in the company of other human beings (among others Diener and Seligman, 2004). Since community life has decreased in the second half of the twentieth century and is still decreasing, human beings spend less time in company of other human beings. In the end, this is detrimental for every human being and for society at large. De Vos (2003) defines a community as a group of people with personal, long lasting relations, where people do not only actually help each other, but also give each other the feeling that one can count on one another, whenever necessary (p.290). Communities are characterized by reciprocal relationships between individual community members. The intrinsic value of these relations in itself is important to people. Within communities, there exists a sense of security for all its members. Individuals feel safe and respected. Research shows that children who grow up in a safer place are more explorative and curious. That predicts how well these children will perform at a later age at school (Bowlby, 1982). The question is, whether people deeply rooted in communities are significantly happier or not. Research shows that people with more social support (more reciprocal relations, deeper rooted in communities) experience less anxiety, less depressions, less additional health problems and have higher life expectations. This relates to effects on (the functioning of) the immune system, (the functioning of) the endocrine system and (the
functioning of) heart and blood vessels (...) (quoted from De Vos, 2003, p.291). Thus, people with more social support tend to be healthier, both physically and mentally. As I described earlier, people who report themselves as happy tend to be both physically and mentally healthier than average. It seems to be logical that people with more social support tend to be happier than people with less social support.
In his research to the ultimate determinants of happiness, Layard (2005) draws a similar conclusion. He distinguishes seven essential determinants of subjective well-being at the personal level.
The second important determinant is ‘family relationships’, according to Layard (2005). Layard shows that divorced people tend to be less happy than married ones (-5 on a scale from 10 till 100). The same holds for people who recently split up (-8), widowed people (-4), singles (-4.5) and couples who recently moved in together (-2) when compared with married people (Layard, 2005, p.64-66). According to Layard, this can be explained because married couples share almost everything with each other. They offer each other loyalty, fidelity and love. In marriage, your life partner accepts you the way you are. In addition, research shows that married people are physically healthier; they have better functioning endocrine systems and are simply healthier (p.66).
As the third important aspect that adds to a happy life, Layard mentions the respondent’s ‘work situation’ (p.66-67). Next to having a life partner, people also feel the need to be part of wider society and to contribute to it. This usually occurs through jobs people have. However, people may be involuntarily unemployed and they sometimes feel quite useless. One does not contribute to wider society and sometimes this feels like one has no purpose in life. In extreme cases, this might even result in psychic problems and clearly this affects one’s happiness negatively. People with jobs tend to be happier because they are able to put energy and creativity in their work. Even when the job does not fit the individual completely, one still feels that he or she contributes to the wider society. Research shows that the negative effect of being unemployed is larger than the negative effect of the income
Figure 6: Through what channels work situation affects individual happiness
loss alone of having no job (among others, Frey, 2008; Layard, 2005; Helliwell, 2003). Therefore, one might conclude that the negative effect of being unemployed on subjective well-being can be
disentangled into an income effect and a social participation effect. This is shown in figure 6. The work situation thus affects individual well-being via income and via social contacts on the work floor.
according to Layard (2005). At first sight, however, it may seem possible to have as many enduring friendships in a ‘bad’ society as in a good society. However, the former society is less safe than a good society and hence, people will tend to distrust one another quicker. One will feel less at ease, which makes it more difficult to build up contacts. Thus indeed, it might be more difficult to build enduring friendships and find a life partner in societies based upon distrust.
The fifth important determinant of subjective well-being is health, according to Layard (2005, p.69). In particular, psychical problems and chronic pain affect subjective well-being negatively. In many circumstances people are able to adapt to physical pain. Frey (2008, p.42) claims that the inhabitants of richer countries tend to be healthier because the level of health care is higher than in poor countries. One might question this claim, because the amount of psychical problems, such as stress, depression, drugs abuse and alcohol abuse, has risen dramatically during the last decennia in most Western countries (among others Diener and Seligman, 2004). As explained earlier, this might be a consequence of changes in the way of life in these countries. Therefore, it is not sure to what extent psychical health is an independent determinant of subjective well-being or that the variation in subjective well-being is better explained by changes in the quality of social life of people (which includes the effect of changes in physical health).
Layard’s (2005) sixth determinant of happiness is personal freedom. In many Western
countries freedom of speech applies. This might seem self-evident, but it is not. Democratic principles are not guaranteed in most countries, with the former Soviet Union countries as sad example. This changes very gradually, but for approximately fifty years it has not been possible to say what one thought without endangering one’s own life in the Soviet Union. This explains why inhabitants of former Soviet Union countries are significantly less happy than people in other countries over the world, according to Helliwell (2003) in his study about the factors that affect subjective well-being (How’s life? Combining individual and national variables to explain subjective well-being). Next to that, there are places in the world characterised by war and terror. It does not need any additional clarification that in places where one fears one’s own life one cannot be happy.
personal values, depending on the exact definition of these personal values. For example, materialistic people are claimed to be less happy than less materialistic people (among others: Kasser and Ahuvia, 2002). Next to that, it is a robust finding that people who believe in God are significantly happier than people who do not (among others, Layard, 2005, p.72).
B. Hypotheses
The central hypothesis I am going to investigate in the empirical part is formulated as follows:
Hypothesis 1: The quality of social life affects subjective well-being negatively in The Netherlands.
But how can one define ‘social life’? In the following sub-hypotheses, I will attempt to specify ‘social life’, making use of the theoretical considerations of especially De Vos (2003) and Layard (2005), which I discussed briefly above.
Hypothesis 1a: The number of divorces affects subjective well-being negatively in The Netherlands.
Hypothesis 1b: The unemployment rate affects subjective well-being negatively in The Netherlands.
In this context, the unemployment rate functions as an approximation for the amount of social contact in the workplace. On the macro level, a larger unemployment rate implies that the total amount of workplace contacts is lower, whereas a lower unemployment rate implies that the level of social contact is higher. There are indications that being employed increases subjective well-being through a higher quality of social life, next to the effect of higher incomes (among others, Frey, 2008; Layard, 2005; Helliwell, 2003). This is graphically shown in figure 6 on page 29.
Hypothesis 1c: The number of enduring friendships affects subjective well-being positively in The Netherlands.
Hypothesis 1d: The quality of enduring friendships affects subjective well-being positively in The Netherlands.
In this respect, it is important to know that the quality of enduring friendships refers to the intensity of the relationship with friends. A better quality implies that the friendship is more enduring.
