Relatório Mundial da Felicidade - Geografia (2024)

Prévia do material em texto

<p>WORLD</p><p>HAPPINESS</p><p>REPORT</p><p>Edited by John Helliwell, Richard Layard and Jeffrey Sachs</p><p>TABLE OF CONTENTS</p><p>1. Introduction</p><p>PART I</p><p>2. The State of World Happiness</p><p>3. The Causes of Happiness and Misery</p><p>4. Some Policy Implications</p><p>References to Chapters 1-4</p><p>PART II</p><p>5. Case Study: Bhutan</p><p>6. Case Study: ONS</p><p>7. Case Study: OECD</p><p>WORLD</p><p>HAPPINESS</p><p>REPORT</p><p>Edited by John Helliwell, Richard Layard and Jeffrey Sachs</p><p>65409_Earth_Chapter1v2.indd 1 4/30/12 3:46 PM</p><p>2</p><p>Part I.</p><p>Chapter 1.</p><p>INTRODUCTION</p><p>JEFFREY SACHS</p><p>Jeffrey D. Sachs: Director, The Earth Institute, Columbia University</p><p>65409_Earth_Chapter1v2.indd 2 4/30/12 3:46 PM</p><p>3</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>3</p><p>We live in an age of stark contradictions. The world enjoys technologies of unimaginable sophistication; yet</p><p>has at least one billion people without enough to eat each day. The world economy is propelled to soaring</p><p>new heights of productivity through ongoing technological and organizational advance; yet is relentlessly</p><p>destroying the natural environment in the process. Countries achieve great progress in economic development</p><p>as conventionally measured; yet along the way succumb to new crises of obesity, smoking, diabetes, depression,</p><p>and other ills of modern life. 1</p><p>These contradictions would not come as a shock to the greatest sages of humanity, including Aristotle and</p><p>the Buddha. The sages taught humanity, time and again, that material gain alone will not fulfi ll our deepest</p><p>needs. Material life must be harnessed to meet these human needs, most importantly to promote the end</p><p>of suffering, social justice, and the attainment of happiness. The challenge is real for all parts of the world.</p><p>As one key example, the world’s economic superpower, the United States, has achieved striking economic and</p><p>technological progress over the past half century without gains in the self-reported happiness of the citizenry.</p><p>Instead, uncertainties and anxieties are high, social and economic inequalities have widened considerably,</p><p>social trust is in decline, and confi dence in government is at an all-time low. Perhaps for these reasons, life</p><p>satisfaction has remained nearly constant during decades of rising Gross National Product (GNP) per capita.</p><p>The realities of poverty, anxiety, environmental degradation, and unhappiness in the midst of great plenty</p><p>should not be regarded as mere curiosities. They require our urgent attention, and especially so at this</p><p>juncture in human history. For we have entered a new phase of the world, termed the Anthropocene by the</p><p>world’s Earth system scientists. The Anthropocene is a newly invented term that combines two Greek roots:</p><p>“anthropo,” for human; and “cene,” for new, as in a new geological epoch. The Anthropocene is the</p><p>new epoch in which humanity, through its technological prowess and population of 7 billion, has become</p><p>the major driver of changes of the Earth’s physical systems, including the climate, the carbon cycle, the water</p><p>cycle, the nitrogen cycle, and biodiversity.</p><p>The Anthropocene will necessarily reshape our societies. If we continue mindlessly along the current</p><p>economic trajectory, we risk undermining the Earth’s life support systems – food supplies, clean water, and</p><p>stable climate – necessary for human health and even survival in some places. In years or decades, conditions</p><p>of life may become dire in several fragile regions of the world. We are already experiencing that deterioration</p><p>of life support systems in the drylands of the Horn of Africa and parts of Central Asia.</p><p>On the other hand, if we act wisely, we can protect the Earth while raising quality of life broadly around the</p><p>world. We can do this by adopting lifestyles and technologies that improve happiness (or life satisfaction)</p><p>while reducing human damage to the environment. “Sustainable Development” is the term given to the com-</p><p>bination of human well-being, social inclusion, and environmental sustainability. We can say that the quest</p><p>for happiness is intimately linked to the quest for sustainable development.</p><p>The Search for Happiness</p><p>In an impoverished society, the focused quest for material gain as conventionally measured typically makes</p><p>a lot of sense. Higher household income (or higher Gross National Product per capita) generally signifi es an</p><p>improvement in the life conditions of the poor. The poor suffer from dire deprivations of various kinds: lack</p><p>of adequate food supplies, remunerative jobs, access to health care, safe homes, safe water and sanitation,</p><p>and educational opportunities. As incomes rise from very low levels, human well-being improves. Not</p><p>surprisingly, the poor report a rising satisfaction with their lives as their meager incomes increase. Even</p><p>small gains in a household’s income can result in a child’s survival, the end of hunger pangs, improved nutrition,</p><p>better learning opportunities, safe childbirth, and prospects for ongoing improvements and opportunities in</p><p>schooling, job training, and gainful employment.</p><p>65409_Earth_Chapter1v2.indd 3 4/30/12 3:46 PM</p><p>4</p><p>Now consider the opposite end of the income spectrum. For most individuals in the high-income world, the</p><p>basic deprivations have been vanquished. There is enough food, shelter, basic amenities (such as clean water</p><p>and sanitation), and clothing to meet daily needs. In fact, there is a huge surfeit of amenities above basic</p><p>needs. Poor people would swap with rich people in a heartbeat. Yet all is not well. The conditions of afflu-</p><p>ence have created their own set of traps.</p><p>Most importantly, the lifestyles of the rich imperil the survival of the poor. Human-induced climate change</p><p>is already hitting the poorest regions and claiming lives and livelihoods. It is telling that in much of the rich</p><p>world, affluent populations are so separated from those they are imperiling that there is little recognition,</p><p>practical or moral, of the adverse spillovers (or “externalities”) from their own behavior.</p><p>Yet the problems of affluence also strike close to home. Affluence has created its own set of afflictions and</p><p>addictions. Obesity, adult-onset diabetes, tobacco-related illnesses, eating disorders such as anorexia</p><p>and bulimia, psychosocial disorders, and addictions to shopping, TV, and gambling, are all examples of dis-</p><p>orders of development. So too is the loss of community, the decline of social trust, and the rising anxiety levels</p><p>associated with the vagaries of the modern globalized economy, including the threats of unemployment or</p><p>episodes of illness not covered by health insurance in the United States.</p><p>Higher average incomes do not necessarily improve average well-being, the U.S. being a clear case in point, as</p><p>noted famously by Professor Richard Easterlin, and shown in Figure 3.2. U.S. GNP per capita has risen by a</p><p>factor of three since 1960, while measures of average happiness have remained essentially unchanged over the</p><p>half-century. The increased U.S. output has caused massive environmental damage, notably through green-</p><p>house gas concentrations and human-induced climate change, without doing much at all to raise the well-being</p><p>even of Americans. Thus, we don’t have a “tradeoff” between short-run gains to well-being versus long-run</p><p>costs to the environment; we have a pure loss to the environment without offsetting short-term gains.</p><p>The paradox that Easterlin noted in the U.S. was that at any particular time richer individuals are happier</p><p>than poorer ones, but over time the society did not become happier as it became richer. One reason is that in-</p><p>dividuals compare themselves to others. They are happier when they are higher on the social (or income)</p><p>ladder. Yet when everybody rises together, relative status remains unchanged. A second obvious reason is</p><p>that the gains have not been evenly shared, but have gone disproportionately to those at the top of the income</p><p>and education distribution. A third is that</p><p>At the global level, the genetically based share of life satisfaction</p><p>differences will of course be much smaller, since life circumstances differ much more among people around</p><p>the globe than among people living in the same country.</p><p>Finally, if most inter-personal happiness differences were personality-driven, and if judgments returned to set</p><p>point levels after a period of adaptation, then there could be no sustained trend differences in the relative happiness</p><p>of different groups within larger populations. But data from a series of Canadian General Social Surveys</p><p>spanning almost 25 years reveals that residents of Québec, especially those who are francophone, have had, in</p><p>the decades following Québec’s Quiet Revolution, steadily growing life satisfaction compared to residents of the</p><p>rest of Canada.49 The accumulated trend difference is both large and statistically very significant, equivalent in life</p><p>satisfaction terms to more than a doubling of household incomes. This finding shows that life satisfaction</p><p>captures much more than temporary departures from personality-driven set points, and also that social changes</p><p>can cause sustained trends in well-being far beyond those explicable by conventional economic measures.</p><p>Is happiness serious enough to be taken seriously?</p><p>In most of the social, political, caring and policy sciences, the focus of attention is on eradication of disease,</p><p>crime, poverty and war. In a world where there is still so much hardship left, is it an unearned luxury to be</p><p>concerned with measuring and building happiness? The case for taking happiness seriously, even in a world</p><p>still marked by evils of many types, is based on a belief, increasingly supported by evidence, that it provides a</p><p>broader range of possible ways to build a better world, including more effective solutions for poverty, illness</p><p>and war. Happiness research is sometimes seen as having a “giggle factor,” too frivolous for serious study. It</p><p>has taken a long time to build convincing evidence that the measurement and maintenance of positive states</p><p>of mind can suggest new routes to longer and healthier lives, above and beyond conventional medical care, but</p><p>the case has now been made.50</p><p>Another related issue, with deep philosophical roots, is the contrast between the hedonistic life, spent in the</p><p>pursuit of pleasure, and the eudaimonistic life, aimed at achieving excellence.51 This distinction is captured</p><p>in modern psychology as the difference between hedonic and eudaimonic well-being, where the hedonic</p><p>approach has a focus on positive emotions and the eudaimonic approach emphasizes flourishing, meaning</p><p>and purpose.52 Does this distinction support the skeptical view of happiness as too frivolous? Does happiness</p><p>unduly emphasize current pleasures and ignore the deeper and more fundamental aspects of life? These</p><p>questions hark back to the distinction we have made between emotional reports and life evaluations. Whether</p><p>framed as questions about happiness or life satisfaction, life evaluations appear to take pleasures and purpose</p><p>21</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>both into account, just as Aristotle suggested they should and would. This is somewhat less so for short-term</p><p>emotional reports, including those on happiness. This difference can be illustrated by the first data available</p><p>from the 2011 U.K. ONS well-being surveys. Four questions are asked. One asks about life satisfaction, one</p><p>asks about the respondent’s sense of life purpose (a eudaimonic question), and two relate to emotions yesterday:</p><p>one about happiness and the other about anxiety. The results show that the eudaimonic answers are correlated</p><p>with both emotional measures, but more closely to life satisfaction than to either emotion.53 Even emotional</p><p>reports are likely to depend on more than current pleasures. Life evaluations, whether based on happiness,</p><p>life satisfaction, or the Cantril ladder, are well placed to attach an even greater weight to the deeper features of</p><p>a good life.</p><p>Happiness measures are part of a larger effort to understand well-being</p><p>Although there is always intrinsic interest in finding out how happy people are, such measures will be of little</p><p>help unless they can be combined with sufficient other information to build an understanding of what makes</p><p>for better lives. Thus many national and international efforts to measure and promote happiness have been set</p><p>within broader frameworks involving the measurement and reporting of other variables that have themselves</p><p>been used as indicators or supports for well-being.54 The Bhutanese case study shows how measures of</p><p>happiness are part of a larger Gross National Happiness framework that monitors many variables that have been</p><p>found to contribute to a higher quality of individual and community life. Similarly, the OECD’s recent accounting</p><p>for well-being in OECD countries includes many other variables.55 And in the United Kingdom, although most</p><p>attention has been given to subjective well-being there is also recognition of the need to collect a much broader</p><p>set of information relevant to the understanding and improvement of well-being.</p><p>Within the broader framework of well-being measures, what is special about happiness and other indicators</p><p>of subjective well-being? The distinctive feature of happiness and other subjective well-being measures is that</p><p>they offer people the chance to report on the quality of their own lives, reflecting their own histories, personalities</p><p>and preferences. These are arguably the most democratic of well-being measures, since they reflect not what</p><p>experts or governments think should define a good life, but instead represent a direct personal judgment. Seen</p><p>in this light, the subjectivity of happiness is to be seen as a strength rather than a weakness. The most fundamental</p><p>indicator of your happiness is how happy YOU feel, not whether others see you smiling, your family thinks you</p><p>are happy, or you have all the presumed material advantages of a good life.</p><p>When pulled together for a neighborhood, community or nation, subjective well-being scores can thus be seen</p><p>as directly democratic measures of the quality of individual and community life within that geographic zone.</p><p>Other measures of well-being, and of the presumed supports for happy lives, can then provide the evidence</p><p>required to explain why some lives, and some communities, are happier than others. Chapter 3 provides many</p><p>examples of what can be discovered, and Chapter 4 shows how this information can be put to work to suggest,</p><p>test and evaluate better ways of designing and delivering public and private services.</p><p>Building a strong information base requires that subjective well-being measures be collected widely and</p><p>frequently. Geographic detail is needed to better understand what features of community life are most supportive</p><p>of well-being. To be of most use, assessments of happiness should be made within a wide variety of surveys</p><p>already being conducted for other purposes, since such surveys will thereby automatically provide a range of</p><p>descriptions of the social and economic contexts of people’s lives. These in turn can support a more fine-grained</p><p>assessment of what makes for happier communities.</p><p>Running surveys frequently, and spreading data collection over the whole year, is useful to help pinpoint and</p><p>hence understand the sources of changes in happiness. Average happiness measures generally move slowly,</p><p>and the changes over time are small relative to the range of happiness differences among cities and countries.</p><p>This increases the value of frequent assessments, especially in geographic detail, as the frequency makes it</p><p>easier to spot and understand trends as they develop. We turn now to consider in slightly more detail how dif-</p><p>ferent types of measures can best be collected and used.</p><p>22</p><p>Different measures for different purposes</p><p>What can be learned by measuring and tracking happiness on different time scales? Time use surveys</p><p>involving</p><p>the diary-based daily reconstruction method56 or the pager-based experience sampling methods each have their</p><p>own most appropriate uses.57 Experience sampling and diary methods, despite their differences, can be used in</p><p>complementary ways to track happiness and its correlates in the context of daily life.58</p><p>This chapter has made a distinction between life evaluations (whether ladder, SWL or happiness) and emotional</p><p>reports, including both positive and negative affect. Life evaluations, positive affect, and negative affect are</p><p>ranked in that order in terms of what they tell us about the relative importance of different life circumstances,</p><p>as will be shown in Chapter 3. But for analyzing the fabric of daily life, the priorities are reversed, with the most</p><p>valuable information being provided by momentary and remembered emotions and reactions during the daily</p><p>course of activities and events.59</p><p>What are the implications for future collection and use of happiness data?</p><p>What sorts of happiness data are needed to support better institutions and policy choices? How can the results</p><p>of well-being research be used to design and deliver better policies? We outline a few possibilities here, to set</p><p>the stage for fuller consideration in Chapter 4.</p><p>First, regular large-scale collection of happiness data in the context of a variety of existing surveys will permit</p><p>the establishment of baseline values and trend changes for subjective well-being within and across nations and</p><p>communities. This will permit the well-being consequences of subsequent events and policy changes to be</p><p>better assessed. It is critically important, if happiness data are to be able to support the uses described below,</p><p>that there should be information available also about the key variables likely to support better lives. It is also</p><p>necessary to know where respondents live, to permit measurement and explanation of happiness differences</p><p>among neighborhoods, cities, and demographic groups.</p><p>Second, analysis of these data in their broader economic and social contexts will permit more comprehensive</p><p>estimation of the relative importance of different factors supporting happiness. This will in turn allow conventional</p><p>benefit/cost analysis to be changed to attach specific values to many features of life, and especially the social</p><p>context. Thus it becomes possible to lift important non-market variables from the footnotes to the center of</p><p>benefit/cost analysis.60</p><p>Third, the resulting research can have implications for macroeconomic policies by improving the information</p><p>used to assess the relative importance of different macroeconomic objectives,61 as well as to alter how such</p><p>policies are designed and delivered.62</p><p>Fourth, well-being results can be used to suggest alternative ways of designing and delivering63 public services</p><p>ranging from elder care64 and community services65 to prisons.66</p><p>In all of these cases, subjective well-being measures are needed. First, it is necessary to build the broad base of</p><p>information to establish baseline levels and to enable more solid research. Second, the research gives rise to</p><p>a range of policy possibilities that need to be assessed in experimental and field trial conditions, in each case</p><p>supported by monitoring the well-being consequences for those involved in designing, delivering and receiv-</p><p>ing public and private services.</p><p>23</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>1. In preparing and revising this chapter, we have benefited from the kind advice of Rafael Di Tella, Paul Dolan, Richard Easterlin,</p><p>Nicole Fortin, Bruno Frey, Dan Gilbert, Carol Graham, Stephen Hicks, Haifang Huang, Ron Inglehart, Daniel Kahneman, Richard</p><p>Layard, Richard Lucas, Conal Smith, Neil Thin, Chris Barrington-Leigh, and Dan Weijers. This research was supported by</p><p>R01AG040640 from the US National Institute on Aging.</p><p>2. Kahneman, Diener & Schwarz, eds, (1999).</p><p>3. e.g. Andrews & Withey (1976), Diener (1984).</p><p>4. Cohen et al. (2003), Lyubomirsky et al. (2005), Schimmack (2003), Wiest et al. (2011).</p><p>5. Russell & Carroll (1999).</p><p>6. This is shown by the similarity between the European Social Survey happiness and life satisfaction data in Figures 2.7 and 2.8, and</p><p>between life satisfaction and the Cantril ladder in the Gallup World Poll, in Diener et al., eds. (2010) Table 10.1.</p><p>7. Cantril (1965) introduced the ladder as a “self-anchoring striving scale.”</p><p>8. See, for example, De Neve et al. (2011).</p><p>9. OECD (2011b).</p><p>10. Redelmeier & Kahneman (1996).</p><p>11. Kahneman et al. (1997) took this Benthamite view.</p><p>12. E.g. Helliwell (2008) and Helliwell & Barrington-Leigh (2010).</p><p>13. Wirtz et al. (2003).</p><p>14. Kahneman (2011), p. 410.</p><p>15. Dolan et al. (2011).</p><p>16. Helliwell et al. (2010) Table 10.1.</p><p>17. Helliwell et al. (2010) Table 10.1.</p><p>18. The comparisons with neighboring countries are indirect because the satisfaction with life answers of the Gallup World Poll need to</p><p>be compared to the happiness answers in Bhutan. Both are asked using the same 0 to 10 scale, and comparisons are implemented using</p><p>averages from ESS answers to the same two questions. In the ESS, life satisfaction answers are on average 0.40 points lower than</p><p>those for happiness. The Bhutanese happiness average of 6.05 is thus converted to a 5.65 “SWL-equivalent” value for comparison</p><p>with the Gallup SWL averages of 5.51 for India, 5.32 for Nepal, 5.25 for Bangladesh and 5.24 for China. The Bhutan estimate is statistically</p><p>significantly higher than for Nepal, Bangladesh and China.</p><p>19. Kahneman & Deaton (2010).</p><p>20. Krueger et al. (2009).</p><p>21. See the notes to Figures 2.11 to 2.13 for the exact questions asked.</p><p>22. Easterlin (1974). Empirical welfare functions based on happiness data were being estimated even earlier in Europe, e.g. van Praag</p><p>(1971).</p><p>23. For surveys and examples, see Stiglitz et al. (2009), Diener et al. (2009), Krueger et al. (2009), and Layard (2010).</p><p>24. Bilaliü et al. (2008).</p><p>25. Nickerson (1998).</p><p>26. Diener (2011), Diener et al. (2009).</p><p>27. e.g. r=0.56 in one year, falling to 0.24 over sixteen years, Fujita and Diener (2005).</p><p>28. Maslow (1943).</p><p>29. For the distributions of country coefficients, see Figure 10.3 of Diener et al., eds, (2010). For the difference between OECD and non-</p><p>OECD coefficients, see Figure 1 of Helliwell & Barrington-Leigh (2010).</p><p>30. Gilbert (2006), p. 66.</p><p>31. For example, see Koivumaa-Honkanen et al. (2000).</p><p>32. Lyubomirsky et al. (2005).</p><p>33. Cohen & Pressman (2006).</p><p>34. See Danner et al. (2001), Chida & Steptoe (2008), and Diener & Chan (2011).</p><p>35. Strack, Martin & Schwarz (1988).</p><p>36. Grice (1975).</p><p>37. As argued by Grice (1975).</p><p>38. Helliwell & Wang (2011b).</p><p>24</p><p>39. Schimmack & Oishi (2005).</p><p>40. See Deaton (2011) and Agrawal & Harter (2011).</p><p>41. Mazar et al. (2008), experiment 1.</p><p>42. Oishi (2010).</p><p>43. Henrich et al. (2010).</p><p>44. As shown in Figure 10.3 of Diener et al., eds. (2010).</p><p>45. Brickman et al. (1978).</p><p>46. Lucas (2007).</p><p>47. Haslam et al. (2008).</p><p>48. See for example Lykken (1999) and De Neve et al. (2011). This share includes the role of any environmental factors that may be</p><p>correlated with the genetic differences. De Neve et al. (2011) also identify one of the candidate genes involved (the 5HTT).</p><p>49. Barrington-Leigh (2011).</p><p>50. For representative surveys of these results, see Steptoe et al. (2005) and Diener & Chan (2011).</p><p>51. For more on the philosophical underpinnings of happiness research, see Bok (2010), Kenny & Kenny (2006), Nussbaum & Sen, eds.</p><p>(1999) and Graham (2011, chapter 2).</p><p>52. E.g. Ryan & Deci (2001).</p><p>53. The correlation between a sense of life purpose and satisfaction with life is r=0.67 (n=4200) compared to r=0.54 between purpose</p><p>and happiness yesterday and r=-0.16 between life purpose and anxiety yesterday.</p><p>54. For a survey, see Møller et al., eds. (2008).</p><p>55. OECD (2011a).</p><p>56. Krueger et al. (2009).</p><p>57. Csikszentmihalyi & Larson (1987).</p><p>58. Kahneman (2011), Stone et al. (2002).</p><p>59. Krueger, ed. (2009).</p><p>60. For an</p><p>experimental study showing how much results can differ when the social context is values using subjective well-being analysis,</p><p>see Gyarmati et al. (2008). For a general review of methods for taking subjective well-being into account in benefit/cost analysis, see</p><p>Fujiwara & Campbell (2011).</p><p>61. For examples, see Di Tella et al. (2001) and Di Tella & MacCulloch (2009).</p><p>62. As in the recent South Korean macroeconomic policies described in Helliwell (2011b).</p><p>63. Frey et al. (2004).</p><p>64. Haslam et al. (2010).</p><p>65. Halpern (2010), Bacon et al. (2010).</p><p>66. Leong (2010), Helliwell (2011a).</p><p>25</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>26</p><p>27</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>28</p><p>29</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>30</p><p>31</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>32</p><p>33</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>34</p><p>35</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>36</p><p>37</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>38</p><p>39</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>40</p><p>41</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>42</p><p>43</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>44</p><p>45</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>46</p><p>47</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>48</p><p>49</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>50</p><p>51</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>52</p><p>53</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>54</p><p>55</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>56</p><p>Figure 2.14 - 1: Distribution of Life Satisfaction in Europe</p><p>Figure 2.14 -2: Distribution of Happiness in Europe</p><p>57</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Notes to Figures:</p><p>(1) Notes to Figures 2.1 and 2.2</p><p>The number of people reporting each Cantril Ladder score in</p><p>a specific country is calculated in two steps: 1) Calculating the</p><p>ratio of respondents reporting the level of ladder by dividing the</p><p>weighted number of respondents reporting that ladder score by the</p><p>weighted total number of ladder respondents. In this step, all-wave</p><p>data in GWP 2005-2011 are used. 2) Multiplying the ratio by the</p><p>total national population aged 15+. Only population aged 15+ is</p><p>considered since only this age group is surveyed in GWP (2011).</p><p>Total population aged 15+ is equal to the proportion of population</p><p>aged 15+ (=one minus the proportion of population aged 0-14)</p><p>multiplied by the total population. To simplify the analysis, we use</p><p>population data in 2008 for all the countries/regions if the data</p><p>are available in WDI (2011). Specifically, the total population and</p><p>the proportion of population aged 0-14 are taken from the series</p><p>“Population ages 0-14 (% of total)” and “Population, total” respec-</p><p>tively from WDI (2011). In the cases where the data are not avail-</p><p>able in WDI (2011), such as in Taiwan and Kosovo, other sources of</p><p>data are used. The data in the year closest to 2008 are used if those</p><p>in 2008 are not available. The population in Taiwan is 22,921,000</p><p>in 2008 (Heston et al., 2011) with 16.7% of it aged 0-14 in 2009</p><p>(CIA, 2009). The proportion of population aged 0-14 in Kosovo</p><p>in 2009 is 28% (Statistical Office of Kosovo, 2011). The data</p><p>on age structure in Somaliland region are not available anywhere,</p><p>therefore the region is not included in the calculation of world or</p><p>regional distribution of ladder.</p><p>The world population reporting a specific level of ladder is the sum</p><p>of population reporting that ladder score over all the countries with</p><p>data on ladder and population. The same method is used to cal-</p><p>culate regional population reporting each ladder score.</p><p>(2) Note to Figure 2.5</p><p>The SWL question was only asked for some waves, and not for all</p><p>countries. There were 54 countries asking the SWL question in</p><p>2007, 68 in 2008, 12 in 2009 (A small fraction of German’s wave</p><p>3 survey which was supposed to be done in 2008 was conducted</p><p>in 2009, we then do not count Germany as one of the 12 countries</p><p>having surveys in 2009), and 6 in 2010. Of the 129 countries</p><p>asking the SWL question, only 11 have asked it in two waves,</p><p>explaining why the total of country-wave observations (140) is only</p><p>slightly more than the total of country observations (129).</p><p>(3) Note to Figures 2.6 and 2.9</p><p>To maximize the coverage of countries and focus on recent evalua-</p><p>tions, we use data in WVS waves 4 and 5 and EVS waves 3 and 4 for</p><p>Figure 2.6 and 2.9, except El Salvador which only has WVS wave</p><p>3 data (1999) (EVS, 2011; WVS, 2009). Therefore all the data used</p><p>are from 1999 and later years.</p><p>The WVS Happy Index is used for ranking happiness in Figure</p><p>2.9, as suggested by the Director of the WVS Archive and ASEP/</p><p>JDS, Jaime Díez Medrano (Medrano, 2012). The Happiness Index</p><p>is defined as the weighted (by sampling weights) rate of respon-</p><p>dents reporting “Very happy” or “Quite happy” less the weighted</p><p>rate of respondents reporting “Not very happy” or “Not at all</p><p>happy,” plus 100. The index thus ranges from 0 to 200. This trans-</p><p>formation makes the WVS happiness rankings closer to those for</p><p>other measures. The 4-point happiness scores from WVS/EVS are</p><p>in any event not easily comparable to other series with more</p><p>complete scales.</p><p>(4) Note to Figures 2.11-2.13</p><p>Positive affect is defined as the average of happiness, laughter, and</p><p>enjoyment yesterday for waves 3-5, but as the average of laughter</p><p>and enjoyment for waves 1 and 2 since the happiness question was</p><p>not asked in the first two waves. Negative affect is defined as the</p><p>average of worry, sadness, depression, and anger yesterday except</p><p>that in Mauritius it is defined as the average of worry, sadness,</p><p>and anger, since the depression question was not asked there. Net</p><p>affect is defined as positive affect minus negative affect. For the</p><p>four negative affect items, and for enjoyment and happiness, the</p><p>general question form was “Did you experience the following</p><p>feelings a lot of the day yesterday:..”. The laughter question was</p><p>“Did you smile or laugh a lot yesterday?”</p><p>58</p><p>Part I.</p><p>Chapter 3.</p><p>THE CAUSES OF HAPPINESS</p><p>AND MISERY</p><p>RICHARD LAYARD, ANDREW CLARK AND CLAUDIA SENIK</p><p>Richard Layard: Director, Well-being Programme, Centre for Economic Performance, London School of Economics</p><p>Andrew Clark: CNRS Research Professor, Paris School of Economics</p><p>Claudia Senik: Professor, University Paris-Sorbonne; Professor, Paris School of Economics</p><p>65409_Earth_Chapter3v2.indd 58 4/30/12 3:48 PM</p><p>59</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Part I.</p><p>Chapter 3.</p><p>THE CAUSES OF HAPPINESS</p><p>AND MISERY</p><p>RICHARD LAYARD, ANDREW CLARK AND CLAUDIA SENIK</p><p>If we want to infl uence the levels of happiness and misery, we need to know what causes them. As a result of</p><p>some 30 years of research, we now know a good deal about this. 1</p><p>If we think of each individual, every one of us has her own genetic make-up, but the person she becomes depends</p><p>on the interaction of those genes with the environment she encounters. Together, genes and environment</p><p>determine the main features of a person’s life – both those that are very “personal” and those that are more clearly</p><p>“external.” And these features in turn determine a person’s well-being, as illustrated in Figure 3.1.</p><p>Figure 3.1</p><p>Among the more “external” factors, key determinants of happiness include:</p><p>• income</p><p>• work</p><p>• community and governance, and</p><p>• values and religion</p><p>and, among the more “personal” features, key determinants include:</p><p>• mental health</p><p>• physical health</p><p>• family experience</p><p>• education, and</p><p>• gender and age</p><p>Thus a person’s happiness at a point in time is determined by the whole of her life course. The current external</p><p>features of her life are important, but so are the personal features that have developed over the previous course</p><p>of her life.</p><p>For policy-makers the main issues are the environmental factors affecting happiness, since these are what</p><p>can be changed. Ideally we would study their effects holding the genes constant, but most research so far has</p><p>been unable to do this.2 So in this chapter we look directly at the ways in which external</p><p>and personal factors</p><p>(however they arise) affect a person’s happiness.</p><p>What the policy-maker wants to know is how big an effect each factor has on happiness. So we concentrate</p><p>both on that, and on the share of variance explained by the factor (which depends of course on the sample of</p><p>people being considered).</p><p>We look fi rst at the more external factors (income, work, community, governance, values and religion) and then at</p><p>the more personal ones (mental health, physical health, family, education, gender and age). In many cases there is</p><p>65409_Earth_Chapter3v2.indd 59 4/30/12 3:48 PM</p><p>60</p><p>a two-way interaction between the factor and happiness. For example, education affects a person’s happiness,</p><p>but happiness also affects the ability to learn. Likewise health affects happiness and happiness affects health. That</p><p>is why, in the diagram above, arrows run from personal and external factors towards happiness (one direction of</p><p>causality) but also run from happiness to health, education and so on (reverse causality). In what follows, we discuss</p><p>each individual factor one by one, including (where it exists) the two-way workings of causality.</p><p>Types of Evidence</p><p>On every factor we offer a wide range of evidence. To isolate the causal effect of each factor is not easy. It clearly</p><p>requires us to hold as much else as possible constant while we look at the co-movement of well-being and the</p><p>factor in question. In most cases this cannot be done experimentally. So the next best is to study the same</p><p>individuals (or countries) over time and see how their well-being moves when different factors change. Much</p><p>of the evidence we shall quote is of this longitudinal, time-series form.</p><p>But some insights can also be got from cross-sectional evidence. In this case we are comparing different individuals</p><p>(or countries) at the same point in time. The problem here is that, when we compare individuals or countries, there</p><p>are many ways in which they may differ (for example in personality or values) that cannot easily be measured and</p><p>controlled for when we are examining the effect of those factors that can be measured. But when we have</p><p>longitudinal data on the same person or the same country we can assume that these unmeasured factors are more</p><p>similar at each observation, and may have a better chance of tying down what is causing what.3</p><p>As a background to what follows we include in Appendix A standard individual happiness equations using two well-</p><p>known sets of panel data and the World Values Survey. We give both cross-sectional and longitudinal estimates.</p><p>We can now review the main causes of happiness one by one.4 We begin with the more “external” causes.</p><p>Income</p><p>Does economic growth improve the human lot? In 1974 Richard Easterlin wrote a seminal article on what has</p><p>become known as the Easterlin paradox. He presented evidence of two apparently contradictory phenomena.</p><p>“Fact 1” At a point in time within any society, richer people are on average happier than poorer people</p><p>(a cross-sectional “fact”).</p><p>“Fact 2” Over time within many societies, the population does not on average become happier when</p><p>the country’s income rises (a time-series “fact”).</p><p>To reconcile Facts No 1 and 2 Easterlin proposed the relative income hypothesis. People are comparing themselves</p><p>with other people: it is relative income rather than absolute income that matters. Thus at any particular point</p><p>in time richer people would compare favorably with poorer people (explaining “Fact No 1”). But over time,</p><p>the aggregate of relative income in the population remains constant (thusexplaining “Fact No 2”).</p><p>There is no doubt that Fact No 1 is correct. In multiple regressions, income always emerges as a factor explaining</p><p>the variation in life satisfaction within a country – not the most important factor (see below) but an important</p><p>one. It is also now possible to tie down quite closely the functional form of that relationship. It is well described</p><p>by a logarithmic form where the absolute level of life satisfaction varies linearly with the logarithm of income.5</p><p>This means that, for example, an extra dollar increases the satisfaction of a poor person by 10 times as much as it</p><p>increases the satisfaction of a person who is 10 times richer. For centuries people have intuitively believed in the</p><p>“diminishing marginal utility of income” but happiness research now provides empirical estimates that policy-</p><p>makers can use when considering the distributional impact of their policies.</p><p>65409_Earth_Chapter3v2.indd 60 4/30/12 3:48 PM</p><p>61</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Turning to the time-series Fact No 2, which is what is relevant to the policy debate, the state of research is more</p><p>unsettled. From micro-economic evidence presented below we know that relative income does matter. This</p><p>is also supported by experiments in neuroscience. This establishes a strong prior that the time-series effects</p><p>of higher absolute income would be less than those found in cross-sectional studies when the average level</p><p>of income is held constant. This prior is in turn reinforced by the fi nding that in some important countries,</p><p>especially the United States, average happiness has not risen despite strong economic growth (see Figure 3.2).6</p><p>This was so during the golden period of economic growth in the 1950s and 1960s, and also more recently,</p><p>when even the top quintile of income recipients experienced no growth in happiness, despite huge increases</p><p>in their income.7</p><p>Figure 3.2</p><p>However the experience of particular countries is not enough to support generalized statements about the</p><p>relation between happiness and economic growth. That requires a more exhaustive study of as many countries</p><p>as possible. We shall therefore proceed as follows. First we shall examine the micro-economic evidence about</p><p>how relative and absolute income infl uences the happiness of individuals. This will also include a discussion</p><p>of adaptation. That done, we shall turn to aggregate evidence about the experience of whole nations.</p><p>Individual happiness and income</p><p>Does relative income raise a person’s happiness and does absolute income do likewise? To examine the effect</p><p>of income on happiness, we must eliminate any effect of a person’s underlying happiness upon their income.</p><p>The best way to attempt this is with panel data in which we trace the same individual over many years and</p><p>examine how changes in the person’s income affects their subsequent happiness. Fortunately we have such</p><p>data from Europe’s leading country. In West Germany the German Socio-Economic Panel (GSOEP) has been</p><p>tracking the same individuals each year since 1984. We can use these data to help us understand the movement</p><p>of average life satisfaction in that country, as illustrated in Figure 3.3, using the Eurobarometer series since</p><p>1972 and the GSOEP since 1984.</p><p>The fi rst fi nding is about the effect of own income on life satisfaction: ceteris paribus, differences in income</p><p>explain about 1% of the variance of life-satisfaction in the population. It is a signifi cant effect, though when we</p><p>use the panel feature of the data, it is somewhat reduced.8</p><p>65409_Earth_Chapter3v2.indd 61 4/30/12 3:48 PM</p><p>62</p><p>The next step is to use the panel data to decompose this effect of income into an effect of absolute income and an</p><p>effect of income relative to the appropriate comparator group. For this purpose relative income is measured relative</p><p>to other people of the same sex, age and education in the year in question. When this analysis is performed with</p><p>suitable controls, there is no effect left for absolute income.9 Only relative income matters and this is clearly what</p><p>explains the fact that in Figure 3.3 average life satisfaction has not risen despite rapid economic growth.</p><p>A third result is also of interest. Many people, including some psychologists, use adaptation to explain why</p><p>happiness is not permanently increased by higher income: “whatever your income, you get used to it.” This</p><p>explanation</p><p>clearly has problems since, if it were wholly true, we should not observe Fact No 1.10 And in the</p><p>GSOEP data there is no strong effect on current life satisfaction of current income relative to income over the</p><p>previous three years – no evidence, that is, of a role for adaptation.11</p><p>Figure 3.3</p><p>We can turn now to the dozens of cross-sectional studies which also indicate strong effects of income comparisons.12</p><p>All cross-section studies run the risk of exaggerating the effect of income on happiness by including the reverse</p><p>effects of happiness on income.13 But many of them provide more useful detail, including explicit questions about</p><p>whom people compare themselves with, and how they think their income compares with that of different groups.</p><p>For example, in a representative sample of rural Chinese, people said they mainly compared themselves with others</p><p>in the same village, and multiple regression results showed that among all possible factors the most important for</p><p>happiness was perceived relative income within the village.14</p><p>In advanced countries the comparators are different. The European Social Survey asked people “How important</p><p>is it for you to compare your income with other people’s incomes?” and those who said income comparisons were</p><p>more important were also on average less satisfi ed with their lives – a common fi nding.15 Respondents were also</p><p>asked “Whose income would you be most likely to compare your own with?” The most important group</p><p>mentioned was “colleagues,” and the same was found in a one-year-only set of questions in the GSOEP.16 In the</p><p>GSOEP study people were then asked to rank their income compared with their colleagues, and also with their</p><p>friends, neighbors, etc. In explaining life satisfaction it was confi rmed that perceived relative income has a large</p><p>effect on life satisfaction. Similar fi ndings have been found in repeated cross-sections in the United States,17 which</p><p>helps to explain the fact that happiness has not increased in the U.S. (see Figure 3.2).</p><p>6</p><p>6.</p><p>5</p><p>7</p><p>7.</p><p>5</p><p>M</p><p>ea</p><p>n</p><p>re</p><p>po</p><p>rte</p><p>d</p><p>lif</p><p>e</p><p>sa</p><p>tis</p><p>fa</p><p>ct</p><p>io</p><p>n</p><p>1970 1980 1990 2000 2010</p><p>Year</p><p>Eurobarometer 1973-2007 GSOEP 1985-2006</p><p>Source: Eurobarometer and German Socio-Economic Panel.</p><p>Mean life satisfaction reported on a 0-10 scale.</p><p>Reported life satisfaction in West Germany</p><p>65409_Earth_Chapter3v2.indd 62 4/30/12 3:48 PM</p><p>63</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>The preceding studies use data on perceived relative income. This perception can of course be influenced by</p><p>the mood of the respondent, but it is reassuring that the results are very similar if we only include actual rela-</p><p>tive income. Many studies only do this, and most but not all of them find significant effects of income relative</p><p>to income in the surrounding area, be it travel-to-work area, county or province.18</p><p>If happiness depends on income relative to the income of a “comparator” that means that when the “compara-</p><p>tor’s” income rises, happiness falls. As we have seen, this is the most common general finding, but it is not</p><p>always the case. In some cases people appear to take the comparator’s income as an indication of what they</p><p>might themselves attain, rather than as an external benchmark they need to compare well with. The compara-</p><p>tor’s income is then like a light at the end of the tunnel.19 In a number of situations this “tunnel effect” has</p><p>been found to dominate the “external norm effect,”20 so that comparator income sometimes has a positive ef-</p><p>fect or in other studies a zero effect.21</p><p>But the more general finding is that comparator’s income reduces happiness and this has been strikingly con-</p><p>firmed in many laboratory experiments. One neuroscience experiment involved the task of guessing the number</p><p>of dots on a screen.22 Good guesses were rewarded by a monetary payment. Each subject was paired with another</p><p>subject, and after each of the 300 trials the subject was told the accuracy of his own guesses and the associated</p><p>income he would receive, as well as the same information for his “pair.” At the same time fMRI scans measured</p><p>the blood oxygenation in the subject’s relevant reward center (the ventral striatum). Blood oxygenation responded</p><p>strongly to both the subject’s own income (positively) and to the pair’s income (negatively). And the negative affect</p><p>of the pair’s income was at least two thirds as large as the positive effect of the subject’s own income.</p><p>Country-level income and happiness: cross-sectional evidence</p><p>We can turn now to country-level data and ask, how far does absolute income affect the happiness of nations?</p><p>We can begin with a cross-sectional analysis of data similar to those presented in Chapter 2, before turning to</p><p>the time-series, which is far more informative about what causes what.</p><p>In Table 3.1, we examine the Gallup World Poll’s measures of well-being reported in Chapter 2. We show first</p><p>a simple regression of average well-being (according to each measure) on the log of GDP per head across the</p><p>sample of countries. There is a very strong relation.23 However material well-being is not the only determinant</p><p>of well-being. In order to get a sensible idea of its effect, we should also include other obvious determinants.</p><p>Beginning with the other indicators in the UNDP’s Human Development Index (HDI), we include:</p><p>Health (we use healthy life expectancy from the WHO)</p><p>Education (we use the HDI average educational level among adults)</p><p>We also include measures of the degree of social support, freedom and corruption that individuals experience</p><p>in their country. These come from the World Gallup Poll’s answers to the following questions:</p><p>Social Support</p><p>If you were in trouble, do you have relatives or friends you can count on to help you whenever you</p><p>need them, or not? (proportion of respondents saying Yes)</p><p>Freedom</p><p>In your country, are you satisfied or dissatisfied with your freedom to choose what you do with</p><p>your life? (proportion of respondents saying Yes)</p><p>Corruption</p><p>Average proportion of respondents saying Yes to the following 2 questions:</p><p>1. Is corruption widespread within business located in your country, or not?</p><p>2. Is corruption widespread throughout the government in your country, or not?</p><p>65409_Earth_Chapter3v2.indd 63 4/30/12 3:48 PM</p><p>64</p><p>Finally we take into account the strength of family life – measured by the proportion of people separated,</p><p>divorced, or widowed.</p><p>In Panel A of Table 3.1 we include only income as an explanatory factor. It has a strong positive impact on life</p><p>evaluation, a smaller impact on positive affect, and an insignificant impact on negative affect. For life satisfaction</p><p>the�ȕ-coefficient on income is high at 0.81; it thus explains 65% (ȕ2) of the variation across countries. However,</p><p>when in Panel B we introduce the social variables discussed above, the positive effect of income falls sharply</p><p>– by more than half. Most of the social variables are highly significant. When it comes to positive and negative</p><p>affect, only the social variables play a significant role.</p><p>A parallel analysis focusing only on European countries shows similar results using the European Social Survey.</p><p>The dependent variable is the average of life satisfaction and happiness these days. When regressed on log</p><p>GDP per head only, ȕ is .84. but when we introduce one additional variable – the average of social trust and</p><p>trust in police -the ȕ-coefficient on trust is .62, and that on GDP falls to .36.</p><p>Table 3.1 Regressions to explain average well-being across countries24 (standardized ȕstatistics)</p><p>Significance Levels: (1 tailed tests)</p><p>* 0.05 ** 0.01 *** 0.001</p><p>The preceding analyses underline the problems of studying the relation of national income and happiness</p><p>without taking into account other variables. This is the main problem with the careful study by Betsey Stevenson</p><p>and Justin Wolfers in which they compare the effect of income on life evaluation at</p><p>the cross-country level</p><p>with its effect at the individual level within a country.25 They argue that there can be no effects of comparator</p><p>income at the individual level if (as they find) the cross-country effects are as high as the within-country</p><p>individual effects. This statement is logically correct, provided other things are held equal. But they are not:</p><p>the “effect of income” at the cross-country level is estimated with nothing else held constant. But, as we have</p><p>shown in Table 3.1, the cross-country effect falls sharply when other variables are included. It is of course possible</p><p>that high income in a country is good for health, social support, freedom and corruption. But to find out about</p><p>the direct effects of comparator income on family well-being, we would definitely have to keep these other things</p><p>constant. Moreover from a public policy point of view it is important to separate out the effects of income from</p><p>those of health, social support, freedom and corruption, and not to roll them all together.</p><p>Dependent Variable</p><p>Independent</p><p>Variables</p><p>Life-evaluation</p><p>Positive affect</p><p>Negative affect</p><p>Panel A</p><p>log GDP per head .81 *** .40 *** -.08</p><p>�2 .65 .15 -.00</p><p>No of countries 153 153 153</p><p>Panel B</p><p>log GDP per head .28 ** -.18 .22</p><p>Health .25 ** .24 .27</p><p>Education -.01 -.18 -.05</p><p>Social support .29 *** .43 *** -.35 ***</p><p>Freedom .15 *** .49 *** -.24 **</p><p>Corruption -.18 *** .00 .23 ***</p><p>Divorce etc. -.43 -.09 -.08</p><p>�2 .80 .52 .20</p><p>No of countries 139 139 139</p><p>Dependent Variable</p><p>Independent</p><p>Variables</p><p>Life-evaluation</p><p>Positive affect</p><p>Negative affect</p><p>Panel A</p><p>log GDP per head .81 *** .40 *** -.08</p><p>�2 .65 .15 -.00</p><p>No of countries 153 153 153</p><p>Panel B</p><p>log GDP per head .28 ** -.18 .22</p><p>Health .25 ** .24 .27</p><p>Education -.01 -.18 -.05</p><p>Social support .29 *** .43 *** -.35 ***</p><p>Freedom .15 *** .49 *** -.24 **</p><p>Corruption -.18 *** .00 .23 ***</p><p>Divorce etc. -.43 -.09 -.08</p><p>�2 .80 .52 .20</p><p>No of countries 139 139 139</p><p>Dependent Variable</p><p>Independent Variables Life evaluation Positive affect Negative affect</p><p>Panel A</p><p>Log GDP per head 0.81 *** 0.40 *** -0.08</p><p>0.65 0.15 -0.00</p><p>No of countries 153 153 153</p><p>Panel B</p><p>Log GDP per head 0.28 ** -0.18 0.22</p><p>Health 0.25 ** 0.24 0.27</p><p>Education -0.01 -0.18 -0.05</p><p>Social support 0.29 *** 0.43 *** -0.35 ***</p><p>Freedom 0.15 *** 0.49 *** -0.24 **</p><p>Corruption -0.18 *** 0.00 0.23 ***</p><p>Divorce etc. -0.43 -0.09 -0.08</p><p>0.80 0.52 0.20</p><p>No of countries 139 139 139</p><p>65409_Earth_Chapter3v2.indd 64 4/30/12 3:48 PM</p><p>65</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>The problem of other things equal maybe relatively less acute when we now turn to changes in GDP over time</p><p>(not holding other things constant).</p><p>Country-level income and happiness: long-term economic growth</p><p>So we are now ready to scrutinize Richard Easterlin’s Fact No 2. Is it true that, if countries grow in income,</p><p>they become no happier? On this matter, one can make no general statement. As we have seen there are some</p><p>countries like the U.S. and West Germany that have grown over long periods of time but have not become</p><p>happier.26 On the other hand there are other countries where income growth has gone hand in hand with</p><p>increases in happiness.</p><p>It would however be helpful if there were some way to summarize the average of this relationship across all</p><p>countries. In a recent paper, Richard Easterlin offered his own summary.27 In order to concentrate on long-term</p><p>economic change, he confined himself to 37 countries with a long enough range of data (21 years for developed</p><p>countries, 15 for developing countries and 12 for transition countries). He found in each group a flat or negative</p><p>relation between changes in life satisfaction and income per head.</p><p>However, this analysis has been powerfully challenged by Stevenson and Wolfers, whose most recent paper</p><p>deals entirely with the issue of long-term growth. It shows first that, both in the countries covered by the World</p><p>Values Survey and in those covered by Eurobarometer, there has been an increase in life satisfaction in the average</p><p>country over recent decades, see Figure 3.4.28 They then investigate whether the intercountry differences in</p><p>changes in life satisfaction are associated with different rates of growth of per capita income. For the World</p><p>Values Survey they find a strong relation between changes in life satisfaction and changes in trend-GDP per</p><p>head – roughly equal in size to the effect found in simple cross-country regressions. They also find a relationship</p><p>among Eurobarometer countries, though the size of the effect is under half of that in the cross-section. One</p><p>reason why their findings differ from Easterlin’s is that they exclude all countries for which the survey did not</p><p>cover the whole country.29</p><p>Figure 3.4 Average subjective well-being in countries covered by Eurobarometer (EB) and World Values Survey (WVS)</p><p>WVS</p><p>EB</p><p>-0.6</p><p>-0.4</p><p>-0.2</p><p>0.0</p><p>0.2</p><p>0.4</p><p>S</p><p>W</p><p>B</p><p>, n</p><p>et</p><p>o</p><p>f c</p><p>ou</p><p>nt</p><p>ry</p><p>F</p><p>E</p><p>s</p><p>(Z</p><p>s</p><p>ca</p><p>le</p><p>)</p><p>1970 1980 1990 2000 2010</p><p>Year</p><p>65409_Earth_Chapter3v2.indd 65 4/30/12 3:48 PM</p><p>66</p><p>There are clearly a number of issues remaining to be resolved in this area.30 But a reasonable interim conclusion</p><p>is as follows:</p><p>1. In a typical country, economic growth improves happiness, other things equal. But other things are</p><p>not necessarily equal, so economic growth does not automatically go with increased happiness. Thus</p><p>policy-makers should balance the argument for more rapid growth against the arguments for supporting</p><p>other sources of happiness. This applies to countries at every level of development.</p><p>2. In developed countries in particular there is strong micro-level evidence of the importance of in-</p><p>come comparisons, which has not been disproved by aggregate data. For this reason policies to raise</p><p>average happiness must target much else besides economic growth.</p><p>Country-level income and happiness: cyclical fluctuations</p><p>There is of course a sharp distinction between long-term economic growth, which may have little effect on the</p><p>level of unemployment, and short-term growth, which is the only way to reduce the high unemployment</p><p>currently prevailing in most parts of the world.</p><p>Everybody agrees on the importance of short-run growth in such a context. Happiness fluctuates over the business</p><p>cycle. It is generally somewhat higher when employment is high relative to trend and when unemployment</p><p>is therefore low.31 But on the other hand happiness is also lower when inflation is high, as often happens in</p><p>upturns. In the overall balance, happiness rises in booms because a one-point decrease in unemployment has</p><p>at least twice as large an effect on happiness as a one-point increase in the inflation rate.</p><p>Economic stability is a crucial goal for any society, due largely to the fact of loss aversion, whereby individuals</p><p>hate to lose x dollars more than they love to gain x dollars.32 But economic stability is a quite different goal from</p><p>long-term economic growth. Long-term growth has much less impact on human happiness than do human</p><p>relationships in all their dimensions – as we shall see.</p><p>Work</p><p>A key relationship comes through work. It provides not only a livelihood but a source of meaning – feeling</p><p>needed and able to contribute. But not everyone can get work, nor if they can, is it always satisfying.</p><p>Unemployment</p><p>When people become unemployed they experience sharp falls in well-being and their well-being remains</p><p>at this lower level until they are re-employed.33 The estimated effect is typically as large as the effect of</p><p>bereavement or separation, and the unemployed share with these other experiences the characteristic of</p><p>ceasing to be needed.</p><p>The Appendix to this chapter documents that unemployment reduces well-being in all the datasets analyzed.</p><p>It</p><p>also shows that the main impact of unemployment on well-being is not through the loss of income, but rather</p><p>through loss of social status, self-esteem, workplace social life, and other factors that matter.</p><p>Psychologists34 have examined these non-pecuniary benefits of work, and they include the preset time structure</p><p>of the working day, regularly shared experiences and contacts with people outside the family, links to goals and</p><p>65409_Earth_Chapter3v2.indd 66 4/30/12 3:48 PM</p><p>67</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>purposes that transcend the individual, personal status and identity, and the enforcement of activity.35 Unemploy-</p><p>ment is destructive due to its negative effect on these functions.</p><p>High unemployment also has spillover effects not only on the families of the unemployed36 but also on those</p><p>in work, who feel less secure in their jobs.37 Thus private sector employees are more affected than public sector</p><p>employees, whose jobs are more secure.38 When we total up all the well-being effects of a rise in the unemploy-</p><p>ment rate, the loss to the rest of the population (which is a large number of people) is twice as large as the loss</p><p>to the unemployed themselves.39</p><p>There is also another spillover effect of the unemployment rate: upon the unemployed themselves. A priori</p><p>one might expect individual unemployed people to be worse off if more other unemployed people were</p><p>competing with them for the available jobs. However, the evidence suggests that greater unemployment may</p><p>actually reduce the stigma associated with one's own unemployment, or be associated with greater social support.40</p><p>British data reveals that others’ unemployment (at the regional, household, and couple level) generally has a</p><p>positive effect on the well-being of the unemployed (at least for men).41 This social norm effect, whereby my</p><p>own unemployment hurts less when more other people are unemployed, parallels data on suicide and</p><p>para-suicide by the unemployed, which is more probable where unemployment is low.42</p><p>But if unemployment is bad, is any reduction in it unambiguously good? In particular is it better to get people</p><p>into bad jobs rather than no jobs at all? This has been a subject of controversy, but a comprehensive study using</p><p>the German Socio-Economic Panel concludes that “Our main result is that we cannot identify a single job</p><p>feature, nor a combination of such features that constitute such low quality jobs that remaining unemployed</p><p>would be the better choice for the individual. On the contrary, the bulk of our evidence shows that even low</p><p>quality jobs are associated with higher life satisfaction, and this effect is statistically significant for most</p><p>specifications of “bad” jobs.43 A parallel study examines the value of the large German workfare program and</p><p>concludes that people’s life satisfaction rises substantially after moving onto the program from being totally</p><p>out of work.44</p><p>Quality of work</p><p>Thus one of the most important aspects of the labor market in terms of well-being is whether individuals are</p><p>able to find a job, given that they want one. However, when in work the quality of life at work is also crucial. The</p><p>view that job quality consists of pay and hours of work has by now largely been superseded.45 In three waves</p><p>of the International Social Survey Programme workers rank eight different job characteristics, on a one to five</p><p>scale from “Not at all important” to “Very Important.” The characteristics are: high income, flexible working</p><p>hours, good opportunities for advancement, job security, interesting job, allows to work independently, allows</p><p>to help other people, and useful to society. The results46 show that only around 20% of respondents in OECD</p><p>countries say that having a high income is very important and the same figure applies to flexible hours and</p><p>promotion opportunities. But around 60% say that job security is very important, with similar figures for</p><p>interesting work and autonomy (50% and 30% respectively).</p><p>Thus it is not surprising that measured satisfaction is shown to be strongly correlated with not only pay at</p><p>work, but also measures of job security, autonomy, workplace trust, independence and so on. Any evaluation</p><p>based solely on income and hours will omit many key characteristics that workers value.</p><p>An important school of thought focuses on the importance of intrinsic motivation at work – and related to that</p><p>the importance of the intrinsic features of the job (rather than pay) as sources of satisfaction:47 eudaimonic</p><p>returns associated with human flourishing. These features include a sense of overall purpose for the job, a</p><p>degree of autonomy in discharging it, and the competence to do the job – a proper fit between worker and job.</p><p>Allied to this people need support and recognition for their efforts. Experimental work has underlined the role</p><p>of purpose and control in determining individual behavior in hypothetical-choice experiments.48</p><p>65409_Earth_Chapter3v2.indd 67 4/30/12 3:48 PM</p><p>68</p><p>This approach has a number of implications. It plays down the role of pay as the prime system of motivation.</p><p>Psychologists have shown many cases where introducing financial incentives reduces performance by</p><p>undermining intrinsic motivation.49 These findings need to be taken seriously by those who design systems of</p><p>performance-related pay.</p><p>One striking finding of happiness research is that the time of day when people are least happy is when they are</p><p>in the presence of their line manager.50 This suggests that too many managers fail to inspire their workers and</p><p>rely too much on mechanical incentives and command.</p><p>Worker well-being matters to firms as well as workers; it is a good predictor of productivity. It is well-known</p><p>that workers who are more satisfied with their jobs are less likely to quit their jobs. They are also less likely</p><p>to reduce firm productivity via absenteeism or via presenteeism - turning up to work, but contributing little.51</p><p>Self-employment</p><p>When it comes to autonomy, some workers can completely control their quality of work, because they are self-</p><p>employed. The self-employed do worse on many job dimensions, including income, hours of work and job</p><p>security, but even so they often report higher levels of overall job satisfaction than do the employed, at least</p><p>in OECD countries. A positive correlation is found in American and European data,52 and in data from Great</p><p>Britain, Germany and Switzerland.53</p><p>This need not necessarily mean that the self-employed experience greater overall satisfaction with their lives, if</p><p>they are sacrificing other dimensions of their lives to their job. Thus it is interesting that in Appendix A, Table</p><p>2, self-employment has no significant effect on overall life satisfaction.</p><p>The great majority of work on well-being and self-employment has used OECD data. But any comparison of</p><p>self-employment and employment will depend critically on the extent to which the former is a choice.54 If</p><p>individuals are free to choose, they will choose self-employment if their happiness there is higher. However,</p><p>when there are insufficient employment opportunities in the formal sector, self-employment may not be a</p><p>choice but a necessity. If formal-sector employment opportunities are positively linked to financial develop-</p><p>ment, this could help to explain a more positive job satisfaction gap between the self-employed and the</p><p>employed in more developed countries. This proposition has been tested on data from the World Value</p><p>Surveys (WVS) over the 1981-2001 period.55 The self-employed do not always report higher job satisfaction</p><p>scores than employees, but do so more often in developed countries. This pattern is not affected by the</p><p>inclusion of income as a control variable, suggesting that the key difference between employees and the self-</p><p>employed is to be found in the non-pecuniary arena.56 The results for non-OECD countries are line with the</p><p>fact that self-employment is associated with lower satisfaction</p><p>in Latin American countries.57</p><p>Retirement</p><p>Eventually many of us stop working: we retire. Do people enjoy life more after they retire? Analyses of panel</p><p>data from the Survey of Health, Ageing and Retirement in Europe (SHARE) show no large overall effect but</p><p>a wide disparity in the effect on different individuals: more educated workers experience rises in well-being</p><p>on retiring; others’ well-being falls.58 So workers in lower-quality jobs are not necessarily those who gain most</p><p>from retirement. The lower-educated report lower levels of well-being when in work, as might be imagined,</p><p>but they also report a greater drop in well-being when they retire. This seeming paradox might be explained</p><p>if education affects not only the value of employment, but also that of retirement: the lower-educated may</p><p>have worse jobs than the better-educated, but also worse retirements. However most work in this area is still</p><p>preliminary.</p><p>65409_Earth_Chapter3v2.indd 68 4/30/12 3:48 PM</p><p>69</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>It might be expected that retiring voluntarily would be associated with a greater rise in well-being than</p><p>involuntary retirement. However, it is not always easy to distinguish the type of retirement in survey data. Also,</p><p>those who retire voluntarily may have a reason for doing so, such as being in poor health, which will potentially</p><p>confound any comparison of voluntary and involuntary retirement if it is not controlled for.</p><p>Social Capital</p><p>While work is one important part of our social world, our relationships go much further, and include relationships</p><p>with family, friends and community. In Maslow’s pyramid of human needs, love and belonging come just after</p><p>basic physiological and safety needs.59 Clearly, the sources of individual happiness include the set of social</p><p>interactions through which individuals are interconnected. The quantity and quality of social relations in a</p><p>community is sometimes referred to as social capital.60</p><p>Why “capital”? Because people’s social networks are accumulated over time (like financial capital) and because</p><p>they yield benefits, such as informal mutual assistance or simply the pleasure of being socially integrated and</p><p>participating in intense social interactions. As a network, social capital also includes a notion of externality, i.e.</p><p>mutually reinforcing benefits for all members.</p><p>We have already seen in Table 3.1 how strongly the happiness of nations is influenced by the extent to which</p><p>the citizens believe they have others they can rely on in times of trouble. We have also seen the powerful effect</p><p>of the levels of perceived corruption in government.</p><p>Trust</p><p>In a well-functioning society there is a high level of trust – above all between citizens, but also in institutions.</p><p>There is a standard question that has been asked in many surveys over many years in many countries. “In</p><p>general, do you think that most people can be trusted, or, alternatively that you can’t be too careful in dealing</p><p>with people?”</p><p>This has been asked in fewer countries than the questions used in Table 3.1. But among those where it has</p><p>been asked, it performs at least as well as the social variables in Table 3.1. One might ask, Do answers to these</p><p>questions correspond to real differences between countries? Their validity is confirmed by the “lost wallet”</p><p>experiment, first conducted by the Reader’s Digest Europe in 1996. This experiment involved dropping 10</p><p>cash-bearing wallets (including a name and address) in each of 20 cities in 14 western European countries,</p><p>and in each of a dozen U.S. cities. Researchers61 later used these data to validate the classic question of trust. It</p><p>turned out that indeed, the actual frequency of return of the wallets was highly correlated with national average</p><p>social trust, as measured in international surveys. Since then, this experiment has often been replicated, and</p><p>the question about the likelihood that a lost wallet, if found by a stranger, or alternatively by a police officer,</p><p>would be returned intact to the owner was used in the Gallup World Poll, as well as certain national surveys</p><p>(e.g. in Canada and the United States) to provide more specific measures of trust.62</p><p>Studies of various types of trust important to Canadian individuals showed large effects for trust in neighbors,</p><p>trust in police, trust in strangers, and especially workplace trust.63 Higher life satisfaction is correlated with</p><p>having a more intense relational life in general, such as socializing frequently with friends and relatives,64</p><p>attending social gatherings and cultural events, participating in sports, performing volunteer work,65 and pro-social</p><p>behavior (donations of time, donations of money, providing help to a stranger). Such correlation may include</p><p>an element of reverse causality with happier people more likely to enter these situations. But several studies</p><p>have documented the stability of trust over generations: the social trust of descendants of immigrants to the</p><p>United States66 or Canada67 is positively linked with the trust level of their ancestors’ home-country. The</p><p>65409_Earth_Chapter3v2.indd 69 4/30/12 3:48 PM</p><p>70</p><p>current level of trust in Europe68 and Africa69 can be traced back to distant past “critical junctures” such as slavery</p><p>or other historical conditions. These studies suggest that trust causes life satisfaction rather than the reverse.</p><p>Levels of trust have fallen substantially over time in some countries (like the U.S. and U.K.) and risen in others</p><p>(such as Denmark and Italy). This may help to explain the fact that life satisfaction has not risen in the U.S.</p><p>and U.K., while it has risen in a number of continental European countries. Indeed for the U.S. it has been well</p><p>argued that the main offsets to the private benefits of economic growth include not only comparator incomes</p><p>but also a decline in the quality of human relationships, as measured by increased solitude, communication</p><p>difficulties, fear, distrust, family infidelity and reduced social engagement.70</p><p>Bonding and bridging capital</p><p>At this point it is important to contrast the relations between people who are similar to each other (bonding</p><p>capital) with the relations between people who are different (bridging capital).71 We want both - not only good</p><p>social capital within communities but also good links between communities.</p><p>The first is the more obvious. Social capital has a local dimension and is most evident within communities, i.e.</p><p>sub-groups of the population who interact directly and frequently share common norms72 and a sense of common</p><p>identity. Staying rooted in the same neighborhood for a longer time is associated with higher levels of all types</p><p>of trust, especially neighborhood trust;73 while respondents who live in districts where the population is highly</p><p>mobile are less likely to trust their neighbors.74 But longer-distance attachments to similar people also often</p><p>matter a lot, even if not as much as local attachments.</p><p>So may more social capital for some mean less for others, because they are in some way excluded? Studies of</p><p>migrants confirm the importance of close communities. New migrants are often found to be less satisfied with</p><p>their life than natives, even when they share an identical socio-professional situation. This is certainly related to</p><p>the fact that migrants have to leave behind them their networks of friends and family.75 It could also be due to</p><p>racial discrimination, although racial tolerance has increased in many Western countries over the last decade,</p><p>leading to an improvement in the happiness of minorities.76 This bridging capital implies that typical communities</p><p>now involve widened circles, so that one can now identify with a wider range of people.77</p><p>Freedom</p><p>Another key feature of a society is the freedom that it provides to its members. No people can be truly happy if</p><p>they do not feel that they are choosing the course of their own life – subject always of course to the inevitable</p><p>constraints of human</p><p>existence. The importance of freedom is confirmed in Table 3.1.</p><p>It is also the fact that the least happy societies documented in 1990 were those in the former Soviet bloc. It is</p><p>not easy to disentangle the effect of the transition, but earlier surveys for Hungary and for Tambov district in</p><p>Russia also show low levels of happiness at that level of GDP per head.78 These contrasts confirm quite clearly</p><p>the importance of freedom for human flourishing.</p><p>Equality</p><p>In a well-functioning society, there is a high degree of mutual respect between its members. Can such a situation</p><p>be achieved if there are massive gaps in income between rich and poor in a society?</p><p>Evidence of the effects of income inequality on the evaluations of happiness is mixed. There is of course the</p><p>basic point that in any society the value of an extra dollar to a poor person is much greater than to a rich person.</p><p>As we have seen, if we compare a poor person with someone who is x times richer, an extra dollar is worth x</p><p>times more (in terms of life satisfaction) to the poor person than to the one who is richer. So in a country with</p><p>65409_Earth_Chapter3v2.indd 70 4/30/12 3:48 PM</p><p>71</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>a given GDP per head, the average life satisfaction must logically (other things equal) be greater if the income</p><p>is more equally distributed.79</p><p>On top of that mechanism (working through individuals), it is often said that inequality damages happiness</p><p>through increased social tensions. For example, Wilkinson and Pickett claim that income inequality is associated</p><p>with lower well-being of various kinds.80 But they also find that inequality damages the rich as well as the poor</p><p>– indicating some kind of environmental effect.</p><p>Nevertheless empirical work on the effects of inequality on life satisfaction has yielded very mixed results.</p><p>Many studies have failed to find any effect.81 The most positive results are in an interesting time-series study</p><p>using both the U.S. General Social Survey and Eurobarometer.82 This finds that in both the U.S. and</p><p>Europe increases in inequality have (other things equal) produced reductions in happiness. The effect has</p><p>been stronger in Europe than in the U.S. This difference probably reflects ideological differences: some 70%</p><p>of Americans believe that the poor have a chance of escaping poverty, compared with only 40% of Europeans.</p><p>Interestingly, the actual facts are actually the other way round: there is more intergenerational social mobility</p><p>in Europe than the U.S. And there is more mobility where there is greater income equality.83 But attitudes have</p><p>an effect on perceptions and thus on happiness.</p><p>People hate inequality much more when they think it is unfair. For example, in some transition countries,</p><p>particularly Poland, income inequality, initially perceived as a positive signal of increased opportunities, started to</p><p>undermine people’s life satisfaction when individuals became skeptical about the legitimacy of the enrichment</p><p>of those who won out in the reform process.84</p><p>The conclusion must be that, other things equal, equality is desirable for two reasons. First the value of extra</p><p>income is greater for the poor than the rich. And second, greater equality can be associated with reduced social</p><p>tensions, especially when inequality is perceived as unfair. But greater equality is unlikely to come about without</p><p>some greater pre-existing ethos of mutual respect and solidarity.</p><p>Values and Religion</p><p>This brings us directly to the issues of values, including those connected with religious belief. Clearly the</p><p>values of a society are crucial to the inhabitants of a society. They are important in two obvious ways: a person’s</p><p>happiness depends on his own values but also on the values of those around him.</p><p>Many people get their values from religion but many do not. The overlap between values and religious belief</p><p>has not yet been studied in the happiness literature, so we shall treat these as two separate issues, beginning</p><p>with religion, before coming on to those values that can also be expressed in purely secular terms.</p><p>Religion</p><p>Some 68% of adults in the world say that “religion is important in their daily lives.”85 Yet our understanding of</p><p>its effects on human happiness is limited. The Gallup World Poll data reported in Chapter 2 provide a starting</p><p>point.86 They show that religious belief and practice is more common in countries where life is harder (less</p><p>income, life expectancy, education and personal safety). They also show that in the U.S. religious belief is</p><p>higher in those states where life is harder. After controlling crudely for those factors, there is no difference in</p><p>life satisfaction between more and less religious countries. There is however a clear difference when com-</p><p>paring the emotional life of more and less religious regions. In particular, in those countries where life is</p><p>tough, there is strikingly more positive emotion and less negative emotion among those people who are more</p><p>religious. Where life is easier, there is no such difference in this study.</p><p>65409_Earth_Chapter3v2.indd 71 4/30/12 3:48 PM</p><p>72</p><p>It is interesting to understand what aspects of religion produce the positive effects on happiness. Clearly religion</p><p>has both social aspects (especially through attending places of worship) but also deeply personal aspects (as</p><p>connected for example with private prayer). In the Gallup World Poll people are asked about the importance of</p><p>religion in their daily lives and also about whether they “have attended a place of worship or religious service</p><p>within the last seven days?” (roughly half of the world’s population had done this). Though these variables are</p><p>not perfectly correlated they both have similar explanatory power.87</p><p>It is therefore natural that, when further questions are examined, they confirm that religion can help in hard</p><p>circumstances both by providing more “relatives or friends you can count on,” and more feelings of being</p><p>respected, and more feeling that “your life has an important purpose or meaning.”88</p><p>Most of the results we have considered above are based on inter-country comparisons. When it comes to</p><p>comparisons between individuals there is always the problem that people who are naturally happier in given</p><p>circumstances are more willing to believe that there is a benevolent deity. However studies of individuals</p><p>do largely agree with the preceding inter-country findings. Meta-analysis concludes that greater religiosity is</p><p>mildly associated with fewer depressive symptoms89 and 75% of studies find at least some positive effect of</p><p>religion on well-being.90 This effect is particularly prevalent in high-loss situations, such as bereavement, and</p><p>weaker in low-loss situations, such as job loss or marital problems. Thus religion can reduce the well-being</p><p>consequences of stressful events, via its stress-buffering role.91</p><p>A recent large study of individuals in the European Social Survey found small but statistically significant</p><p>effects on life satisfaction of “ever attending religious services” and “ever praying.”92 And interestingly the</p><p>religiosity of others in the region was also found to have positive benefits both on those who are religious and</p><p>on those who are not. This confirms findings from cross-country analysis of the Gallup World Poll that weekly</p><p>church attendance has positive spillovers on the well-being of others at the national level.93</p><p>Altruism</p><p>But many of the values taught by the world’s religions are of course universal values that have also been strong-</p><p>ly supported by secular systems of ethics, going back to Stoicism (the most prominent ethical system of the</p><p>Roman Empire) and beyond. The central principle is “do as you would be done by” – behave to others as you</p><p>would wish them to behave to you. This frequently requires that you incur costs for the benefit of others – the</p><p>fundamental definition of altruism.</p><p>Clearly altruistic behavior benefits those at the receiving end. But</p><p>does it also benefit those who give, as well as</p><p>those who receive? There is substantial evidence that it does, and that this is why it is so much more common</p><p>than the crude teachings of elementary economics might predict.</p><p>There is of course plenty of evidence that people who care more about others are typically happier than those</p><p>who care more about themselves.94 But does that mean that altruism increases happiness in a causal sense?</p><p>Evidence on volunteering and on giving money suggests that it does.95</p><p>When East Germany was united with West Germany, many opportunities for volunteering in East Germany</p><p>disappeared. At the same time those who had previously volunteered experienced much larger falls in</p><p>happiness than those who had not been volunteering. This suggests strongly that volunteering had been a</p><p>cause of happiness for those who did it.96 Acts of kindness have a similar effect – in a randomized experiment,</p><p>the treatment group was told to do three extra acts of kindness a day and this significantly raised their</p><p>happiness for some weeks.97</p><p>65409_Earth_Chapter3v2.indd 72 4/30/12 3:48 PM</p><p>73</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>In an experiment on giving, one group was given some money to spend on themselves and another group was</p><p>given equal amounts of money to spend on others. At the end of the day the second group reported themselves</p><p>to be the happier.98 These effects on happiness can also be observed in the brain’s reward centers – when</p><p>people give money they experience a positive reward.99 Moreover altruism can be trained. After two weeks’</p><p>compassion training, a control group gave more money in a laboratory game and showed more neural activity</p><p>in the reward centers of the brain.100</p><p>There is a parallel question of whether happiness in turn increases altruism – a key question if we are wondering</p><p>whether greater individual happiness would also increase happiness in others. There is some evidence that</p><p>happiness is contagious.101 But the specific channel of altruism is best studied through experimentation. A</p><p>number of experiments have confirmed that happier people are more likely to give help to others.102</p><p>Materialism</p><p>Most ethical systems teach not only altruism but also that material wealth should not be pursued beyond the</p><p>point where it compromises other values. Many studies have shown that other things equal, people who care</p><p>more about money are less happy.103 An important study covers a group of students who were freshmen in</p><p>1976.104 Soon after entering college they were asked “the importance to you personally of being well off financially.”</p><p>Nineteen years later they reported their income and their overall satisfaction with life, as well as with family</p><p>life, friendships, and work. At a given level of income, people who cared more about their income were less</p><p>happy with life overall, with their family life, with their friendships and with their job. Of course people who</p><p>care more about money also tend to earn more, and this helps to offset the negative effect of materialism. But</p><p>in this study a person considering high income essential would need twice as much income to be as happy as</p><p>someone considering high income unimportant.</p><p>If materialistic values tend to reduce social life, so does watching TV. Many studies have shown that watching</p><p>TV is associated with lower happiness, other things equal. An early study exploited the fact that one Canadian</p><p>town gained access to TV some years later than other towns.105 The result was a relative fall in social life and</p><p>increased aggression. So TV may cause problems in many ways, including reduced social life and increased</p><p>violence. But the U.S. General Social Survey also shows that heavy TV watchers see so many rich people on</p><p>the screen that they underestimate their own relative income.106 Against all this TV also provides a great deal</p><p>of enjoyment and instruction.</p><p>Environment</p><p>A final issue of values is the environment. There are two quite distinct issues here. One is the future of the</p><p>planet. This mainly affects the happiness of future generations, rather than adults currently alive. The issue of</p><p>greenhouse gases is a classic free-rider problem, and negotiation alone will not solve it. It will require a major</p><p>dose of altruism world-wide.</p><p>The second environmental issue is the effect of our existing environment on adults who are currently alive.</p><p>The environment we live in is extremely complex and only a few of its aspects have so far been examined for</p><p>their effects on human happiness. The method is the same standard method we have used so far – to compare</p><p>the happiness of people living in different environments.107 Dimensions which have been examined so far in-</p><p>clude air quality (sulfur dioxide),108 airport noise,109 and aspects of the climate (sun, heat, humidity and wind).110</p><p>In all cases sensible results have been obtained. Other studies have examined the effects of the natural world on</p><p>human experience. At a very primary level, people assigned to walk from A to B on a tree-lined path alongside</p><p>the Rideau River were systematically happier than those taking the same trip via the Carleton University’s</p><p>underground tunnel system, and the actual gains were much higher than people thought they would be.111</p><p>Students who can see greenery out their classroom windows do better than those who cannot.112 A hospital</p><p>window with a green view similarly sees patients cured faster,113 and there are many other studies linking green</p><p>spaces to better health, performance, and life satisfaction.114</p><p>65409_Earth_Chapter3v2.indd 73 4/30/12 3:48 PM</p><p>74</p><p>Mental Health</p><p>So far we have focused on factors that are very heavily influenced by the society in which you live. We turn now</p><p>to factors that vary more within a country than across countries, but which are still extremely important areas</p><p>for public policy.</p><p>Before starting, there is one important point to make. Happiness depends crucially on personality, and personality</p><p>is strongly affected by your genetic make-up. This has been conclusively established by studying the similarity</p><p>of happiness among identical twins reared apart, and comparing it with the similarity of happiness among</p><p>non-identical twins who grew up in the same family. Identical twins reared apart are remarkably similar in</p><p>their levels of happiness, while non-identical twins brought up together are very different.115</p><p>One important channel through which genes operate is mental health. This can be seen by taking any serious</p><p>mental illness such as bipolar (or manic-depressive) disorder. If one identical twin has this condition, the other</p><p>will also have it in 65% of cases. But, if the twins are non-identical, this falls to 14%.116 For less serious mental</p><p>health conditions the role of the genes is smaller. But for any condition, however serious, environmental</p><p>interventions can also make a huge difference, as we shall see.</p><p>So how important is mental health in explaining the variation of happiness within any particular country? There</p><p>is obviously a danger here of tautology – we explain the variation of happiness by the variation of misery.</p><p>However, we can largely deal with this problem by measuring mental health earlier in life and using it to explain</p><p>current happiness. We can illustrate this approach using the British Cohort Study of people born in 1970. For</p><p>example, we can measure their life satisfaction when they were 34, and then explain this by the standard factors</p><p>discussed in this chapter plus their level of malaise eight years earlier.117 Of all the influences the most powerful</p><p>were malaise eight years earlier (ȕ�= -.23), general health eight years earlier (ȕ�= .10) and current income (ȕ� = .10).</p><p>Even if mental health is measured at age 16 it still exerts nearly as much impact on life satisfaction at age 34 as</p><p>does current income. These facts in themselves have profound implications for policy.</p><p>These are the direct effects of mental health, holding constant a person’s</p><p>other societal factors – insecurity, loss of social trust, a declining</p><p>confidence in government – have counteracted any benefits felt from the higher incomes. A fourth reason</p><p>is adaptation: individuals may experience an initial jump in happiness when their income rises but then at</p><p>least partly return to earlier levels as they adapt to their new higher income.</p><p>These phenomena put a clear limit on the extent to which rich countries can become happier through the</p><p>simple device of economic growth. In fact, there are still other general reasons to doubt the formula of ever-</p><p>rising GNP per person as the route to happiness. While higher income may raise happiness to some extent,</p><p>the quest for higher income may actually reduce one’s happiness. In other words, it may be nice to have</p><p>more money but not so nice to crave it. Psychologists have found repeatedly that individuals who put a high</p><p>premium on higher incomes generally are less happy and more vulnerable to other psychological ills than</p><p>individuals who do not crave higher incomes. Aristotle and the Buddha advised humanity to follow a middle</p><p>path between asceticism on the one side and craving material goods on the other.</p><p>A further huge problem is the persistent creation of new material “wants” through the incessant advertis-</p><p>ing of products using powerful imagery and other means of persuasion. Since the imagery is ubiquitous on</p><p>all of our digital devices, the stream of advertising is more relentless than ever before. Advertising is now</p><p>a business of around $500 billion per year. Its goal is to overcome satiety by creating wants and longings</p><p>where none previously existed. Advertisers and marketers do this in part by preying on psychological weak-</p><p>65409_Earth_Chapter1v2.indd 4 4/30/12 3:46 PM</p><p>5</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>nesses and unconscious urges. Cigarettes, caffeine, sugar, and trans-fats all cause cravings if not outright</p><p>addictions. Fashions are sold through increasingly explicit sexual imagery. Product lines are generally sold</p><p>by associating the products with high social status rather than with real needs.</p><p>And finally, there is one further word of warning to those who expect to become happier by becoming richer.</p><p>Even if gains in well-being can be eked out by further income gains, the evidence is quite overwhelming that</p><p>after a certain point, the gains are very small. The key idea is known as the “diminishing marginal utility of</p><p>income.” Suppose that a poor household at $1,000 income requires an extra $100 to raise its life satisfaction</p><p>(or happiness) by one notch. A rich household at $1,000,000 income (one thousand times as much as the</p><p>poor household) would need one thousand times more money, or $100,000, to raise its well-being by the</p><p>same one notch. Gains in income have to be of equal proportions to household income to have the same</p><p>benefit in units of life satisfaction. This principle means that poor people benefit far more than rich people</p><p>from an added dollar of income. This is a good reason why tax-and-transfer systems among high-income</p><p>OECD countries on balance take in net revenues from high-income households and make net transfers to</p><p>low-income households. Put another way, the inequality of household income is systematically lower net of</p><p>taxes and transfers than before taxes and transfers.2</p><p>Rethinking the Keys to Happiness</p><p>The western economist’s logic of ever higher GNP is built on a vision of humanity completely at variance</p><p>with the wisdom of the sages, the research of psychologists, and the practices of advertisers. The economist</p><p>assumes that individuals are rational decision-makers who know what they want and how to get it, or to get</p><p>as close to it as possible given their budget. Individuals care largely about themselves and derive pleasure</p><p>mainly through their consumption. The individual’s preferences as consumers are a given or change in ways</p><p>actually anticipated in advance by the individuals themselves. Some economists even say that drug addicts</p><p>have acted “rationally,” consciously trading off the early benefits of drug use with the later high toll of addic-</p><p>tion. These economists may say this, but they don’t dare examine such foolishness too closely!</p><p>We increasingly understand that we need a very different model of humanity, one in which we are a</p><p>complicated interplay of emotions and rational thought, unconscious and conscious decision-making, “fast”</p><p>and “slow” thinking. Many of our decisions are led by emotions and instincts, and only later rationalized by</p><p>conscious thought. Our decisions are easily “primed” by associations, imagery, social context, and advertising.</p><p>We are inconsistent or “irrational” in sequential choices, failing to meet basic standards of rational consistency.</p><p>And we are largely unaware of our own mental apparatus, so we easily fall into traps and mistakes. Addicts</p><p>do not anticipate their future pain; we spend now and suffer the consequences of bankruptcy later; we break</p><p>our diets now because we aren’t thinking clearly about the consequences.</p><p>We also understand (again!) that we are social animals through and through. We learn through imitation,</p><p>and gain our happiness through meeting social norms and having a sense of belonging to the community.</p><p>We feel the pain of others, and react viscerally when others are sad or injured. We even have a set of “mirror</p><p>neurons” that enable us to feel things from the point of view of others. All of this gives us a remarkable</p><p>capacity to cooperate even with strangers, and even when there is little chance of reward or reciprocity, and</p><p>to punish non-cooperators, even when imposing punishment on others is costly or puts us at risk ourselves.</p><p>Of course there are limits to such cooperation and fellow feeling. We also cheat, bluff, deceive, break our</p><p>word, and kill members of an out-group. We engage in identity politics, acting as cruel to outsiders as we are</p><p>loving to our own group.</p><p>All these lessons of human nature matter more than ever, more even than when the Buddha taught humanity</p><p>about the illusions of transient pleasures, and the Greeks warned us against the tempting Siren songs that</p><p>could pull us off our life’s course. For today we have more choices than ever before. In the ancient world,</p><p>65409_Earth_Chapter1v2.indd 5 4/30/12 3:46 PM</p><p>6</p><p>the choice facing most of humanity most of the time was little choice indeed: to work hard to secure enough</p><p>to eat, and even then to face the risk of famine and death from bad weather or bad luck.</p><p>Now we face a set of real choices. Should the world pursue GNP to the point of environmental ruin, even</p><p>when incremental gains in GNP are not increasing much (or at all) the happiness of affluent societies?</p><p>Should we crave higher personal incomes at the cost of community and social trust? Should our governments</p><p>spend even a tiny fraction of the $500 billion or so spent on advertising each year to help individuals and</p><p>families to understand better their own motivations, wants, and needs as consumers?</p><p>Should we consider some parts of our society to be “off bounds” to the profit motive, so that we can foster the</p><p>spirit of cooperation, trust, and community? A recent analyst of Finland’s school system, for example, writes</p><p>that Finland’s excellence (ranking near the top of international comparisons in student performance) has</p><p>been achieved by fostering a spirit of community and equality in the schools.3 This is in sharp contrast to</p><p>the education reform strategy at work in the U.S., where the emphasis is put on testing, measurement, and</p><p>teacher pay according to student test performance.</p><p>There are reasons enough to believe that we need to re-think the economic sources of well-being, more so</p><p>even in the rich countries than in the poor ones. High-income countries have largely ended the scourges of</p><p>poverty, hunger, and disease. Poor countries rightly yearn to do so. But after the end of poverty,</p><p>current circumstances. But mental</p><p>health also has an indirect effect, through its effect on those current circumstances. To see this, we have to</p><p>again look back at the effects of previous mental illness. About one half of those who are mentally ill as adults</p><p>were already ill by the age of 15 (one half with primarily emotional problems and one half with primarily</p><p>behavioral problems).118 If we take those people who were mentally ill as adolescents, we can compare their</p><p>adult circumstances with those of the rest of the population. Holding other things constant, they are more</p><p>likely to have experienced low earnings, unemployment, criminal records, teenage pregnancy, physical illness</p><p>and poor educational performance.119 And these factors will in turn reduce their happiness – and frequently</p><p>that of other members of the community as well.</p><p>Even if mental illness is quite narrowly defined, it affects a large number of people. The estimates in population</p><p>surveys vary between countries for reasons that are not well understood.120 But in the typical advanced country</p><p>roughly 15% would be assessed as ill enough to need treatment – with some 1% suffering from psychotic</p><p>conditions (especially schizophrenia) and most of the rest divided equally between depression on the one hand</p><p>and on the other hand anxiety disorders like social phobia, panic attacks, obsessive-compulsive disorder, PTSD</p><p>and crippling general anxiety. In developing countries rates of psychosis are similar, but measured rates of</p><p>depression and anxiety are somewhat lower.</p><p>Mental illness is extremely disabling. For example a recent WHO study estimated that depression was 50%</p><p>more disabling than chronic physical illnesses like angina, asthma, arthritis or diabetes.121 If we combine the</p><p>high prevalence of mental illness with its severity, we find that in WHO estimates it accounts for 43% of</p><p>disability in advanced countries (as measured by the WHO) and 31% of disability world-wide.122 Indeed among</p><p>65409_Earth_Chapter3v2.indd 74 4/30/12 3:48 PM</p><p>75</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>the working-age population in advanced countries, mental illness accounts for as much disability as all the</p><p>other diseases put together. Even if we include all age groups and measure the overall burden of disease so as to</p><p>include not only disability but also premature death, mental illness accounts for 26% of the burden of disease</p><p>in advanced countries, and 13% world-wide.</p><p>Yet mental illness is in very many cases curable – more so than many physical diseases – and in nearly all cases</p><p>it is treatable with significant benefit. For example, when people with anxiety conditions (which have often</p><p>lasted for decades) are treated by cognitive behavioral therapy, one half will experience a complete and</p><p>permanent recovery. Similarly one half of depressed patients will recover and have far less risk of relapse.123</p><p>These treatments are not expensive. Medication is also an effective treatment for severe depression, with recovery</p><p>rates of around one half and reduced risk of relapse if the medication is continued.</p><p>Despite this, in most advanced countries only a quarter of people with mental illness are in treatment, compared</p><p>with well over three quarters for most physical conditions. This is a cause of much unnecessary misery, and the</p><p>situation is even worse in developing countries. Though genes and childhood experience affect our likelihood of</p><p>mental illness, healthcare interventions in childhood and later can reverse the course of our lives.</p><p>And increasingly preventive interventions are also being developed, which can reduce the likelihood of</p><p>developing mental illness in the first place. These include interventions that can be made in childhood or pursued</p><p>by adults. They all fall under the broad heading of “mind-training” and have been shown to affect not only self-</p><p>reported well-being but also immune responses, hippocampal activity and educational performance.124</p><p>Physical Health</p><p>Physical health too is obviously correlated with well-being. There are however a number of issues involved</p><p>in establishing how far health affects happiness. For example the most common measure of health used in</p><p>these surveys is a subjective one. In the BHPS individuals are asked “Please think back over the last 12 months</p><p>about how your health has been. Compared to people of your own age, would you say that your health has on</p><p>the whole been…,” with response categories of Excellent, Good, Fair, Poor, and Very Poor. While these kind of</p><p>measures are easy to apply, and are indeed widely used, it is possible that individual replies are influenced by</p><p>response style (some people give more positive replies, while others are more negative for the same level of</p><p>underlying health), and that the same response bias is at work in each person’s answers to the well-being questions.</p><p>If so, it is unsurprising that health and well-being are correlated, but this need not necessarily reflect any causal</p><p>relationship.</p><p>One way of dealing with this problem is to use panel data, in which we consider changes in health for the</p><p>same individual and relate these to changes in lif- satisfaction. This method of analysis will eliminate any fixed</p><p>individual response style with respect to either health or well-being. The data in the Appendix to this chapter</p><p>enable us to adopt this approach. Table 1 in the Appendix shows that cross-section analysis of BHPS, GSOEP</p><p>and WVS data produces estimated coefficients on self-assessed health that are not only significant, but also</p><p>very sizeable: ceteris paribus, reporting health in the top two categories is associated with life satisfaction</p><p>scores that are two to three points higher than reporting health in the worst category. Table 2 in the Appendix</p><p>then turns to panel analysis for the two datasets in which individuals are repeatedly interviewed (BHPS and</p><p>GSOEP). The estimated coefficients in the panel analysis are indeed somewhat smaller than those in Table</p><p>1, but nonetheless remain significant at all conventional levels. Thus health has a large impact on individual</p><p>life-satisfaction.</p><p>Another approach is to avoid subjective health measures altogether and turn to more objective measures,</p><p>such as doctor visits, nights spent in hospital, or measures of disability. There can be debate over whether such</p><p>65409_Earth_Chapter3v2.indd 75 4/30/12 3:48 PM</p><p>76</p><p>happenings are partly caused by underlying well-being, but these variables are at least more objective than self-</p><p>reported health. There is also self-reported information from the British Household Panel Survey on whether</p><p>the individual is moderately or severely physically disabled, and this can be related to life satisfaction.125 Panel</p><p>data allows us to compare the life-satisfaction of the same individuals before and after they became disabled.</p><p>The impact effect of severe disability is estimated as being 0.6 points on the one to seven life satisfaction</p><p>scale, and that of moderate disability as 0.4 points. There is also evidence of adaptation to disability, such that</p><p>someone who has been disabled for all of the past three years is less affected (in life-satisfaction terms) than</p><p>someone who is recently disabled. This adaptation is estimated at around 50% for moderate disability and 30%</p><p>for severe disability (so that around one-third of the life-satisfaction effect of the latter dissipates over time).</p><p>There is also a reverse relationship: the impact of happiness on health. More happiness predicts better future</p><p>physical health.126 The medical literature has found high correlations between various low well-being scores</p><p>and subsequent coronary heart disease,127 strokes,128 suicide129 and length of life.130 Individuals with higher</p><p>positive affect have better neuroendocrine, inflammatory and cardiovascular activity.131 Those with higher</p><p>positive affect are less likely to catch a cold when exposed to a cold virus, and recover faster if they do.132</p><p>The Family</p><p>Marriage</p><p>Loving and being loved are key conditions for human happiness. Marriage is an institution which is meant to</p><p>promote those experiences. Does it?</p><p>Marriage is one of the unambiguous, universally positive and statistically significant correlates of life satisfac-</p><p>tion. Basic estimates of happiness always reveal that being married rather than single, divorced or widowed,</p><p>is strongly associated with higher self-declared happiness, in all countries that have been under study, e.g.</p><p>the United States and the countries of the European Union,133 Switzerland,134 Latin America, Russia, Eastern</p><p>Europe135 and Asia.136 In most countries married people are also happier with their life than those who cohabit</p><p>with a partner.</p><p>But does marriage make people happy, or are happy people more likely to become and remain married? An</p><p>obvious way to try and disentangle the two directions of causality is to use longitudinal surveys and follow the</p><p>same individuals over time. Accordingly, a retrospective view over 17 years of individuals living in Germany137</p><p>reveals that among single people aged 20, those who would get married later were already happier with their</p><p>life at the age of 20 than those who would remain single. Also, those who would get divorced were already</p><p>less happy when they were single or newly married. Hence individuals who are already happier when they</p><p>are young have a higher probability of becoming and remaining married. Even so, above and beyond these</p><p>selection effects in and out of marriage, getting married still gives an additional boost to happiness, at least for</p><p>some years.</p><p>Life satisfaction peaks in the years before and after marriage. As with many other life events, the happiness</p><p>boost is subject to considerable adaptation after a honeymoon period. But for those who never get divorced,</p><p>happiness remains permanently higher than before they were married.138</p><p>To many readers this marriage “bonus” will not come as a surprise. Apart from love and companionship, there</p><p>are economic advantages of marriage, such as insurance and buffers against adverse life shocks, economies</p><p>of scale and specialization within the family;139 and studies have documented the fact that, compared to single</p><p>people, married people enjoy better physical and psychological health (e.g. less substance abuse and less</p><p>depression) and live longer.140 Because marriage involves a long-term commitment, accompanied with trust</p><p>65409_Earth_Chapter3v2.indd 76 4/30/12 3:48 PM</p><p>77</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>and companionship, it can also be seen as a form of social capital. Research where the data permit has shown</p><p>that the happiness of spouses is interdependent, in Great Britain,141 Germany142 and Australia.143</p><p>However, as for any social group, couples may also suffer from inter-personal comparisons and relative</p><p>deprivation. The leisure and social activities of one’s spouse cause reduced life satisfaction in the other</p><p>spouse.144 Intra-household comparisons of happiness can be a cause of marital instability:145 couples with a</p><p>higher happiness gap are more at risk of divorcing in the future, especially if the wife is the unhappy one.146 A</p><p>related observation is that assortative mating (i.e. marriage between equally educated people) seems to favor</p><p>greater happiness and lower risk of divorce,147 probably because there is more similarity and experience-shar-</p><p>ing between spouses.</p><p>Finally, should people who are unhappy in their marriages expect to be happier after divorce? It is true that</p><p>divorced people are less happy and suffer from more mental strain than people who remain married; but</p><p>people who live in a marriage of (self-assessed) poor quality are less happy than unmarried people.148 As</p><p>British, German and American data show, a person who gets divorced becomes on average more satisfied with</p><p>their life, some years after the separation, than he or she used to be in the three years preceding separation.149</p><p>Of course the break-up of a couple is also likely to affect other people, and children are obvious potential</p><p>collateral victims. Evidence of the impact of parental separation on children’s welfare is abundant; poor school</p><p>performance being one of the most obvious symptoms. However, it is not divorce as such that appears to damage</p><p>children’s welfare and school grades, but parental conflict. Accordingly, children’s school performance</p><p>deteriorates for several years before the official separation of their parents.150 However, the issue of child</p><p>development is too large to be treated here and most of the evidence comes from quite different sources from</p><p>those considered in this report.</p><p>In sum, (a good) marriage is a source of life satisfaction, and conversely, the equality of happiness between</p><p>spouses is a guarantee of marital stability; less happy people are more likely to get divorced, but once they do,</p><p>divorcees reach higher levels of happiness in the long run than they used to experience before divorce.</p><p>Children</p><p>But do children make their parents happy? Surprisingly, the presence of children in the household appears not to</p><p>be associated with higher life satisfaction.151 This is found in the World Values Survey and in panel data for U.K. and</p><p>Germany (see Appendix). Several surveys of the literature acknowledge this surprising absence of a relationship152</p><p>and in particular single parents with more children are less happy than those with fewer children.153</p><p>There are obviously the responsibilities and time pressures of childcare, especially for those facing too many</p><p>claims on their time. Time-use and experience-sampling studies, which investigate the occurrence of positive</p><p>and negative feelings (happy, relaxed, frustrated, depressed, angry, sad, etc.) during typical daily activities,</p><p>reveal that childcare ranks very low in the hierarchy of daily activities, sometimes as low as 16th out of 19</p><p>daily activities, in terms of net positive feelings.155 Everyday experience with children may not be immediately</p><p>rewarding, but “a man who has children lives like a dog, a man without children dies like a dog,” says the proverb.</p><p>However the effect of adult children has not yet been systematically investigated. Moreover even among young</p><p>children, their age matters. Young children under 3 and teenagers are associated to a lower level of parents’</p><p>happiness, whereas children aged 3-12 are associated with higher happiness.155</p><p>Richer people are on average happier with being parents, and parenthood is also less problematic in the social-</p><p>democratic countries of Northern Europe, where there is more child-support from the state.</p><p>65409_Earth_Chapter3v2.indd 77 4/30/12 3:48 PM</p><p>78</p><p>In summary, having children is no guarantee of higher happiness. The pleasure of parenting depends on the</p><p>age of the children, on the quality of the parenting couple and on the social context, including having enough</p><p>time to enjoy family life.</p><p>Education</p><p>On average, the level of education has no clear direct impact on happiness, but education is of course indirectly</p><p>related to happiness through its effect on income: education increases income and income increases happi-</p><p>ness. Studies of the financial returns to education in recent years, mostly based on natural experiments, show</p><p>high returns from additional years of schooling (between 7% and 15% per year).156 Longer years of education</p><p>are also associated with increased employability and job security, and faster promotion,157 all of those being</p><p>factors conducive to higher happiness.</p><p>By contrast, evidence on the direct effect of education is mixed and varies between countries.158 One obvious</p><p>problem is that happy people may be more likely to persist in education and this effect cannot be controlled for</p><p>in panel studies. But there is one type of natural experiment which can help – the raising of the compulsory</p><p>minimum school leaving age. This has been shown to have directly raised the average happiness of those</p><p>affected by the change, though again largely</p><p>through its effect on income.159</p><p>The conclusion is that education may have some non-income benefits to the individuals who get an education,</p><p>especially in poor countries.160 But this is smaller than is often claimed by educationalists. On top of that there</p><p>may be important social effects through an informed electorate and in poor countries through reduced birth-</p><p>rates and mortality.</p><p>Gender</p><p>In most advanced countries women report higher satisfaction and happiness than men.161 In our Appendix,</p><p>women report higher life satisfaction scores than do men in all three of the data sets analyzed. But this finding</p><p>is dominated by advanced countries. Outside the industrial countries the happiness gap infavor of women is</p><p>often found to be smaller or even reversed.162 Moreover in both the U.S. and Europe women are becoming less</p><p>happy relative to men.163</p><p>In the U.S. it is also possible to investigate gender differences using the U-index, which is defined as the</p><p>proportion of time spent in activities for which the highest-rated feeling was negative.164 Data from the Princeton</p><p>Affect and Time Survey (PATS), where the activities of the day previous to the interview are reconstructed,</p><p>show that U.S. women have lower U-index scores than men – and thus less misery.165 It is also found that</p><p>women are relatively happier in countries where gender rights are more equal.166</p><p>Though women report higher life satisfaction than men, ceteris paribus, their rates of mental illness are also</p><p>higher.167 In the BHPS, where both well-being measures are available, women report higher overall life</p><p>satisfaction scores but more psychological stress, as measured by the twelve-item GHQ-12.168</p><p>Age</p><p>The relationship between age and evaluations of happiness is one of the most robust and common findings</p><p>in happiness research. A priori, most people would expect that happiness steadily declines with age, at least</p><p>in adulthood, as do many of our physical and mental faculties. But the pattern of life-evaluation uncovered</p><p>65409_Earth_Chapter3v2.indd 78 4/30/12 3:48 PM</p><p>79</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>by surveys is essentially U-shaped through life: satisfaction declines, reaches a minimum in middle-age</p><p>(between 40 and 50), and then rises again. This two-part age profile of life-evaluation has been observed</p><p>in many countries in many continents,169 although some work finds different patterns in specific countries.170</p><p>In the U.S. the U-shaped pattern has also been found for affect.171</p><p>The shape is even more pronounced when holding income, marital status, health and employment status</p><p>constant. Thus it is not only higher income or greater family stability that explains the rebound in happiness</p><p>after the mid-life low. Explanations could include the wisdom of maturity, or the beneficial effect of reduced</p><p>(or more realistic) aspirations. However between 70 and 80 worsening health begins to take its effect172 and</p><p>average happiness begins to decline once more.173</p><p>Conclusion</p><p>We have covered a lot of ground. We have shown above all that happiness depends on a huge range of</p><p>influences, many of which can be influenced by government policy. In due course we shall have more precise</p><p>estimates of these effects. But a very rough perspective of orders of magnitude can be got from Appendix B,</p><p>where we compare the effects of the main influences looked at in this chapter with the effect of a 30% increase</p><p>in income. The calculations confirm the powerful effect of many variables other than income.</p><p>1 We are extremely grateful to the U.S. National Institute of Aging (R01AG040640) for financial support, to John Berry, Angus Deaton,</p><p>Ed Diener, Bruno Frey, Carol Graham, Richard Easterlin, John Helliwell, Daniel Sachs, Andrew Steptoe, Betsey Stevenson and Alois</p><p>Stutzer for helpful comments and to Harriet Ogborn for expert handling of the preparation of the manuscript.</p><p>2 On the role of genes see Lykken (1999) and Plomin et al. (2001).</p><p>3 However on the other side, estimates of effects based on longitudinal data are more biased towards zero by measurement error. They</p><p>also reflect a narrower range of variation of the variables under study.</p><p>4 For a classic review of the determinants of happiness see Diener & Biswas-Diener (2008). See also Dolan et al. (2008).</p><p>5 Layard et al. (2008). This implies no level of income at which people become satiated. Kahneman & Deaton (2010) have examined this</p><p>question in the U.S. for a given year. Holding constant the incomes of other people, individual life-evaluation has no point of satiation</p><p>but positive affect becomes satiated with income around the level of $75,000.</p><p>6 Layard et al. (2010), Fig 6.2.</p><p>7 Layard et al. (2010), Table 6.1.</p><p>8 See Appendix A. A similar reduction is found using the British Household Panel Survey. This reduction reflects two features: the intro-</p><p>duction of the fixed effect and the greater measurement error bias in panel data. Bound & Krueger (1991) suggest that for income the</p><p>measurement error reduces the coefficient estimate by a factor of around 0.8 in the cross-section and 0.65 in first differences.</p><p>9 Layard et al. (2010), Table 6.4.</p><p>10 Of course in any year richer people have on average more transitory income. But they also have more permanent income.</p><p>11 Di Tella et al. (2010) find strong effects of lagged income but this is because they exclude comparator income from their explanatory</p><p>variables (and include occupational status instead).</p><p>12 Clark et al. (2008b).</p><p>13 For experimental evidence of such effects see Oswald et al. (2011). But there have been good attempts to handle this problem using</p><p>industry wages or twins as instrumental variables. See Pischke (2011) and Li et al. (2011).</p><p>14 Knight et al. (2010). However in South Africa, Kingdon & Knight (2006) find that the comparator incomes that have negative effects on</p><p>happiness are mainly from outside the village.</p><p>65409_Earth_Chapter3v2.indd 79 4/30/12 3:48 PM</p><p>80</p><p>15 Clark & Senik (2010).</p><p>16 Mayraz et al. (2010).</p><p>17 Layard et al. (2010).</p><p>18 For studies which find effects in the U.S. and Canada, see Luttmer (2005), Helliwell & Huang (2010), and Barrington-Leigh & Helliwell (2008).</p><p>19 Hirschman & Rothschild (1973)</p><p>20 Graham & Pettinato (2002).</p><p>21 In individual regressions in the Gallup World Poll, individual income raises life satisfaction as usual but GDP (the comparator) does</p><p>not reduce it (Helliwell et al., 2010).</p><p>22. Dohmen et al. (2011)</p><p>23. See also Deaton (2008).</p><p>24. Source: Gallup World Poll 2005-2011, equations by Helliwell and Wang. The measure of life evaluation is the ladder. Very similar results</p><p>would be obtained using life satisfaction, see Helliwell et al. (2010), Table 1, for a comparison using identical samples.</p><p>25. Stevenson & Wolfers (2008b).</p><p>26. The U.K. is another example.</p><p>27. See Easterlin et al. (2010).</p><p>28. Sacks et al. (2012). Scores for each year indicate the average value in that year for all countries covered. The units are Z-scores net of</p><p>country FEs. The table implies that over the last 40 years average happiness has risen in total by 0.14 standard deviations of happiness</p><p>across the world’s population.</p><p>29. Sacks et al. (2012), Table 3, Lines 5 and 6. Also, in this study changes in happiness are only measured using the same survey source</p><p>for every observation.</p><p>30. One source of controversy relates to the maintained hypothesis. For example, in many estimates the effect of economic growth on life</p><p>satisfaction is not significantly different from zero, nor is it significantly different from the cross-sectional effect across countries or</p><p>individuals.</p><p>31. Di Tella et al. (2003). Helliwell & Huang (2011b) use cross-sectional U.S. evidence to show that well-being is higher in places where</p><p>unemployment is lower.</p><p>32. Kahneman (2011). See also Wolfers (2003) who shows the harmful effects of unemployment variability.</p><p>33. See, for example, Blanchflower & Oswald (2004), Clark & Oswald (1994), and Winkelmann &</p><p>Winkelmann (1998).</p><p>34. Eisenberg & Lazarsfeld (1938).</p><p>35. Jahoda (1981 and 1988).</p><p>36. McKee & Bell (1986).</p><p>37. Green (2011).</p><p>38. Luechinger et al. (2010).</p><p>39. Helliwell & Huang (2011b).</p><p>40. Jackson & Warr (1987).</p><p>41. Clark (2003). This finding is confirmed for large samples of the U.S. data in Helliwell & Huang (2011b).</p><p>42. Platt & Kreitman (1990) and Platt et al. (1992).</p><p>43. Grün et al. (2010), p.287.</p><p>44. Wulfgramm (2011). See also Gyarmati et al. (2008) for a randomized experiment in which the treatment group were allocated to work</p><p>on community projects.</p><p>45. See Warr (1999) for example.</p><p>46. Clark (2010).</p><p>47. Deci & Ryan (1985).</p><p>48. Benjamin et al. (2012).</p><p>49. Ariely et al. (2005), Pink (2009), Frey (1997).</p><p>50. Kahneman et al. (2004).</p><p>51. Robertson & Cooper (2011) and Cooper & Lundberg (2011).</p><p>52. Blanchflower & Oswald (1998), Blanchflower et al. (2001) and OECD (2000).</p><p>65409_Earth_Chapter3v2.indd 80 4/30/12 3:48 PM</p><p>81</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>53. Frey & Benz (2008).</p><p>54. Bianchi (2012).</p><p>55. Ibid.</p><p>56. Frey (2008), Chapter 7.</p><p>57. Graham & Pettinato (2002).</p><p>58. Clark & Fawaz (2009). Many studies find no overall effect of retirement on health (Coe & Lindeboom, 2008), but some find a negative</p><p>effect of retirement on mental health (Dave et al., 2008). In general this literature has not produced an unambiguous set of results.</p><p>59. Maslow (1943).</p><p>60. The OECD (2001, p. 41) defines social capital as “networks together with shared norms, values and understandings that facilitate</p><p>cooperation within or among groups.”</p><p>61. Knack (2001).</p><p>62. Helliwell & Putnam (2004); Helliwell (2008); Helliwell & Wang (2011a).</p><p>63. This dominance of workplace trust applies both for trust in colleagues (Helliwell & Wang 2011a, and Helliwell & Barrington-Leigh</p><p>2011) and trust in management (Helliwell & Huang 2010). These two measures of workplace trust are compared, and their pre-emi-</p><p>nence confirmed, in Helliwell & Huang (2011a).</p><p>64. Powdthavee (2008).</p><p>65. Meier & Stutzer (2008).</p><p>66. Algan & Cahuc (2010), Uslaner (2008).</p><p>67. Soroka et al. (2007), Helliwell & Wang (2011a)</p><p>68. Durante (2009).</p><p>69. Nunn & Wantchekon (2011).</p><p>70. Bartolini (2011), Sarracino (2010).</p><p>71. Putnam (2000).</p><p>72. Bowles & Gintis (2002). Of course, “communities work because they are good at enforcing norms, and whether this is a good thing depends</p><p>on what the norms are” (p.428). Indisputably, examples of harmful norms, such as the culture of honor and the associated violence, are</p><p>legion. See Nisbett & Cohen (1996); Grosjean (2011).</p><p>73. Helliwell & Wang (2011a).</p><p>74. Alesina & La Ferrara (2005). See also Sampson et al. (1997).</p><p>75. See Nguyen & Benet-Martinez (forthcoming) for a meta-analysis of biculturalism.</p><p>76. Stevenson & Wolfers (2008a).</p><p>77. Singer (1981), Pinker (2011).</p><p>78. See also Easterlin (2010).</p><p>79. Stevenson & Wolfers (2010) use cross-country data to examine this mechanism and cannot refute it (nor can they refute zero effect of</p><p>inequality).</p><p>80. Wilkinson & Pickett (2009). They do not include happiness or life satisfaction in their outcomes.</p><p>81. Stevenson & Wolfers (2010), Blanchflower & Oswald (2004) and Helliwell (2003), p.351. But for positive results see Morawetz et al.</p><p>(1977) and Schwarze & Härpfer (2007).</p><p>82. Alesina et al. (2004).</p><p>83. Blanden (2009), Atkinson et al. (1992), Burkhauser & Poupore (1997).</p><p>84. Grosfeld & Senik (2010).</p><p>85. In this context Buddhists normally report themselves as religious, even if others question this use of words.</p><p>86. Diener et al. (2011), see especially their Figure 1. This also shows that there was a large difference in happiness between religious and</p><p>less religious the U.S. states but only a small one for the U.S. individuals.</p><p>87. See op cit, Table 3.</p><p>88. By contrast, recent theU.S. research has found evidence suggesting that all of the life satisfaction benefits of religiosity flow through</p><p>church attendance and the social networks provided by their congregations, and specifically friendships among those sharing a strong</p><p>sense of religious identity, see Lim & Putnam (2010). Another ingenious the U.S. study uses exogenous changes in laws related to</p><p>Sunday shopping to identify a positive effect of church attendance (Cohen-Zada & Sander, (2010)).</p><p>89. Smith et al. (2003).</p><p>90. Pargament (2002).</p><p>65409_Earth_Chapter3v2.indd 81 4/30/12 3:48 PM</p><p>82</p><p>91. Ellison (1991).</p><p>92. Clark & Lelkes (2009).</p><p>93. Helliwell et al. (2010).</p><p>94. Lyubomirsky (2007). See also Singer et al. (2004).</p><p>95. See also Brown et al. (2003) and Schwartz & Sendor (1999).</p><p>96. Meier & Stutzer (2008).</p><p>97. Lyubomirsky (2007), Chapter 5.</p><p>98. Anik et al. (2010). See also Aknin et al. (2010) and Dunn et al. (2008). In the cross-country regressions in Table 3.1, it is possible to add</p><p>a further variable: the proportion of subjects who gave money to a charity in the last month. This is again highly significant.</p><p>99. Harbaugh et al. (2007). See also Zaki & Mitchell (2011) who also show that inequitable behavior causes activity in brain regions as</p><p>sociated with subjective disutility.</p><p>100. Experiment conducted by Richard Davidson.</p><p>101. Fowler & Christakis (2008).</p><p>102. Anik et al. (2010).</p><p>103. Konow & Earley (2008).</p><p>104. Nickerson et al. (2003).</p><p>105. Williams (1986).</p><p>106. Layard (2005), Annex 6.1. See also Bruni & Stanca (2008).</p><p>107. In principle we need to allow for the fact that people have chosen their environments.</p><p>108. Luechinger (2009).</p><p>109. Van Praag & Baarsma (2005).</p><p>110. Frijters & Van Praag (1998), and Brereton et al. (2008), who also examine proximity to the sea, to landfill, and to transport routes.</p><p>111. Nisbet & Zelenski (2011).</p><p>112. Matsuoka (2010).</p><p>113. Ulrich (1984).</p><p>114. Basu et al. (2012).</p><p>115. Lykken (1999).</p><p>116. McGue & Bouchard (1998), Table 1.</p><p>117. Layard (2012). Similar findings come from the earlier National Child Development Study of people born in 1958.</p><p>118. Kim-Cohen et al. (2003).</p><p>119. Richards & Abbott (2009), Layard & Dunn (2009).</p><p>120. Demyttenaere et al. (2004).</p><p>121. Moussavi et al. (2007).</p><p>122. WHO (2008).</p><p>123. Roth & Fonagy (2005).</p><p>124. Durlak et al. (2011); Williams et al. (2007); Davidson et al. (2003) and Hölzel et al. (2011).</p><p>125. Oswald & Powdthavee (2008).</p><p>126. See Diener & Chan (2011), Frey (2011) and Lyubomirsky et al. (2005).</p><p>127. Sales & House (1971).</p><p>128. Huppert (2006).</p><p>129. Koivumaa-Honkanen et al. (2001).</p><p>130. Palmore (1969) and Mroczek & Spiro (2006).</p><p>131. Steptoe et al. (2005).</p><p>132. Cohen et al. (2003).</p><p>133. Di Tella et al. (2003).</p><p>134. Frey & Stutzer (1999).</p><p>65409_Earth_Chapter3v2.indd 82 4/30/12 3:48 PM</p><p>83</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>135. Graham & Pettinato (2002); Senik (2004); Grosjean et al. (2011).</p><p>136. Knight et al. (2010); Knight & Gunatilaka (2010).</p><p>137. Stutzer & Frey (2006).</p><p>138. Stutzer & Frey (2006).</p><p>139. Becker (1974).</p><p>140. Gardner & Oswald (2002); Gove et al. (1983); Hu & Goldman (1990).</p><p>141. Powdthavee (2009).</p><p>142. Lucas & Schimmack (2006).</p><p>143. Guven et al. (2012, forthcoming).</p><p>144. Clark (2011).</p><p>145. Guven et al. (2012, forthcoming).</p><p>146. Ibid.</p><p>147. Frey & Stutzer (2006).</p><p>148. Chapman & Guven (2010).</p><p>149. Gardner & Oswald (2006); Clark et al. (2008a).</p><p>150. Piketty (2003).</p><p>151. Di Tella et al. (2003); Alesina et al. (2004).</p><p>152. Blanchflower (2008); Clark et al. (2008a); Layard (2005); Gilbert (2006). But few studies find a positive association, see Clark &</p><p>Oswald (2002); Frey & Stutzer (2006); Haller & Hadler (2006).</p><p>153. Frey & Stutzer (2006).</p><p>154. Kahneman et al. (2004).</p><p>155. Margolis & Myrskylä (2011). Children affect happiness more favourably at weekends when parents have more time (Helliwell & Wang,</p><p>2011b).</p><p>156. Angrist & Krueger (1991); Acemoglu & Angrist (2001); Card (2001).</p><p>157. Blanchflower & Oswald (2004).</p><p>158. For some positive results, see Frey & Stutzer (2002); Di Tella et al. (2003) and Appendix A for GSOEP and WVS but not BHPS. For</p><p>evidence varying between countries, see Hartog & Oosterbeek</p><p>(1998), p247.</p><p>159. Oreopoulos & Salvanes (2011) relates to U.K., Canada and U.S.</p><p>160. Hartog & Oesterbeek (1998) p. 247.</p><p>161. See Blanchflower (2008) for evidence from repeated cross-sections covering a wide variety of countries, and Helliwell (2008) for</p><p>Gallup data in 27 OECD countries.</p><p>162. See Senik (2004) for instance. But no difference is observed in the Gallup World Poll.</p><p>163. Stevenson & Wolfers (2009).</p><p>164. Kahneman & Krueger (2006).</p><p>165. Krueger et al. (2009).</p><p>166. Graham & Chattopadhyay (2011).</p><p>167. Nolen-Hoeksema & Rusting (1999).</p><p>168. This BHPS measure is used in Clark & Oswald (1994) and Clark (2003).</p><p>169. Blanchflower & Oswald (2004 and 2008); Hayo & Seifert (2003); Helliwell (2003); Helliwell et al. (2009).</p><p>170. For the U.S. Easterlin (2001) has traced the same birth cohorts over their lifetime and failed to find the U-shape. If cross-section and</p><p>cohort results differ, this must mean that the cohorts born some 40-50 years before the study had particularly bad lifetimes.</p><p>171. Stone et al. (2010).</p><p>172. Gerstorf et al. (2008); Gerstorf et al. (2010).</p><p>173. Frijters & Beatton (2008); Wunder et al. (2009). This can also be seen in the GSOEP and the BHPS, using 10-year age groups. (This</p><p>is despite the fact that the BHPS question refers to health relative to your age-group.)</p><p>65409_Earth_Chapter3v2.indd 83 4/30/12 3:48 PM</p><p>84</p><p>Appendix A</p><p>Standard life-satisfaction equations for individuals in three large data sets</p><p>In this Appendix we present standard equations to explain life-satisfaction. We use the German Socio-Economic</p><p>Panel (GSOEP, 1984-2009), the British Household Panel Study (BHPS, 1996-2008) and the World Values</p><p>Survey (WVS, 1981-2008).</p><p>Table 1 shows OLS cross-section regressions, while Table 2 shows OLS equations including a fixed-effect for</p><p>each individual and year, so that the equation estimates the effect of each variable in explaining the different</p><p>levels of happiness which an individual experiences in each different year.</p><p>The sample used is prime-age (30-55), with the top and bottom 5% of income recipients trimmed. Income is</p><p>household income. Regressions by Sarah Flèche.</p><p>Table 1: Cross-sectional regressions to explain life satisfaction</p><p>(Range of life satisfaction: GSOEP 0-10, BHPS 1-7, WVS 1-10)</p><p>GSOEP (West) BHPS WVS</p><p>Log of Income (monthly) 0.60*(0.01) 0.26*(0.02) 0.65*(0.01)</p><p>Female 0.12* (0.00) 0.09*(0.01) 0.14* (0.01)</p><p>Age 0.11*(0.07) 0.11 (0.09) 0.19* (0.11)</p><p>Age^2/1000 -3.55*(1.80) -4.29* (2.18) -5.37* (2.73)</p><p>Age^3/1000 0.03*(0.01) 0.04* (0.01) 0.04* (0.02)</p><p>Single -0.15*(0.01) -0.34*(0.01) -0.32* (0.03)</p><p>Widowed -0.18* (0.04) -0.46*(0.04) -0.25* (0.03)</p><p>Divorced -0.20* (0.01) -0.41*(0.01) -0.40*(0.03)</p><p>Separated -0.48*(0.02) -0.60*(0.03) -0.55* (0.04)</p><p>Unemployed -0.63*(0.01) -0.33*(0.03) -0.43* (0.02)</p><p>Self Employed -0.15*(0.01) 0.07*(0.01) 0.04* (0.02)</p><p>Out of the labor force -0.03* (0.01) -0.13*(0.01) 0.11* (0.01)</p><p>Student -0.12*(0.06) -0.00 (0.06) -0.34* (0.06)</p><p>Education: high 0.00* (0.01) -0.20 (0.08) 0.14*(0.02)</p><p>Education: medium -0.01(0.01) 0.00* (0.04) 0.07*(0.01)</p><p>One child -0.02* (0.01) -0.07*(0.01) -0.03 (0.02)</p><p>Two children -0.03* (0.01) -0.06 (0.01) -0.03 (0.02)</p><p>Three + children -0.09*(0.01) -0.09*(0.02) -0.00 (0.02)</p><p>Health</p><p>Excellent 3.45*(0.03) 1.94*(0.03) 2.69* (0.08)</p><p>Good 2.82*(0.03) 1.59*(0.03) 2.06* (0.08)</p><p>Satisfactory 2.04*(0.03) 1.09*(0.03) 1.44* (0.08)</p><p>Poor 1.26* (0.03) 0.59*(0.03) 0.61* (0.09)</p><p>Fixed effects No No No</p><p>Time Dummies Yes Yes Yes</p><p>Region Dummies Yes Yes Yes</p><p>Observations 100,945 53,615 106,504</p><p>R2 0.25 0.17 0.28</p><p>1RWH��</p><p>�GHQRWHV�VWDWLVWLFDO�VLJQL¿FDQFH�DW����OHYHO</p><p>65409_Earth_Chapter3v2.indd 84 4/30/12 3:48 PM</p><p>85</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Table 2: Fixed effects regressions to explain life satisfaction</p><p>(Range of life satisfaction: GSOEP 0-10, BHPS 1-7)</p><p>GSOEP (West) BHPS</p><p>Log of Income (monthly) 0.39* (0.01) 0.13* (0.03)</p><p>Female -- --</p><p>Age -0.16* (0.07) -0.08 (0.09)</p><p>Age^2 /1000 2.97 (1.76) 0.56 (2.13)</p><p>Age^3/1000 -0.02* (0.01) 0.00 (0.01)</p><p>Single -0.07* (0.03) -0.13*(0.04)</p><p>Widowed -0.44* (0.07) -0.18* (0.07)</p><p>Divorced 0.03 (0.02) -0.14* (0.03)</p><p>Separated -0.25* (0.03) -0.34* (0.03)</p><p>Unemployed -0.49* (0.01) -0.22* (0.03)</p><p>Self Employed -0.01 (0.02) 0.01 (0.02)</p><p>Out of the labor force -0.13* (0.01) -0.12* (0.02)</p><p>Student -0.14*(0.06) 0.07 (0.06)</p><p>Education: High 0.07 (0.05) -0.07 (0.08)</p><p>Education: Medium 0.10* (0.03) 0.13* (0.04)</p><p>One child 0.07* (0.01) 0.01 (0.01)</p><p>Two children 0.04* (0.02) 0.03 (0.02)</p><p>Three + children 0.06* (0.02) 0.06* (0.03)</p><p>Health</p><p>Excellent 2.25* (0.03) 1.04* (0.03)</p><p>Good 1.92* (0.03) 0.90* (0.03)</p><p>Satisfactory 1.51* (0.03) 0.64* (0.03)</p><p>Poor 0.93* (0.03) 0.36* (0.03)</p><p>Fixed effects Yes Yes</p><p>Time Dummies Yes Yes</p><p>Region Dummies Yes Yes</p><p>Observations 100945 53615</p><p>R2 0.20 0.09</p><p>1RWH��</p><p>�GHQRWHV�VWDWLVWLFDO�VLJQL¿FDQFH�DW����OHYHO</p><p>65409_Earth_Chapter3v2.indd 85 4/30/12 3:48 PM</p><p>86</p><p>GSOEP (1984-2009) (West Germany) Summary Statistics</p><p>Variable Mean Std. Dev Min Max</p><p>Life satisfaction 7.05 1.72 0 10</p><p>Female 0.51 0.49 0 1</p><p>Age 42.05 7.27 30 55</p><p>Age^2/1000 1.82 0.61 0.90 3.02</p><p>Age^3/1000 81.10 40.43 27 166.37</p><p>Single 0.13 0.34 0 1</p><p>Widowed 0.01 0.10 0 1</p><p>Divorced 0.08 0.28 0 1</p><p>Separated 0.02 0.16 0 1</p><p>Unemployed 0.06 0.24 0 1</p><p>Self-employed 0.05 0.22 0 1</p><p>Out of the labor force 0.12 0.33 0 1</p><p>Student 0.00 0.07 0 1</p><p>Log of Income 3.33 0.20 2.70 3.74</p><p>Education: high 0.19 0.39 0 1</p><p>Education: medium 0.31 0.46 0 1</p><p>One child 0.23 0.42 0 1</p><p>Two children 0.22 0.41 0 1</p><p>Three + children 0.08 0.27 0 1</p><p>Heath: excellent 0.09 0.29 0 1</p><p>Health: good 0.46 0.49 0 1</p><p>Health: satisfactory 0.30 0.46 0 1</p><p>Health: poor 0.10 0.30 0 1</p><p>65409_Earth_Chapter3v2.indd 86 4/30/12 3:48 PM</p><p>87</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>BHPS (1996-2008) Summary Statistics</p><p>Variable Mean Std. Dev Min Max</p><p>Life satisfaction 5.07 1.24 1 7</p><p>Female 0.54 0.49 0 1</p><p>Age 42.52 7.12 30 55</p><p>Age^2/1000 1.85 0.61 0.90 3.02</p><p>Age^3/1000 83.41 40.35 27 166.37</p><p>Single 0.09 0.29 0 1</p><p>Widowed 0.01 0.10 0 1</p><p>Divorced 0.07 0.26 0 1</p><p>Separated 0.02 0.16 0 1</p><p>Unemployed 0.02 0.16 0 1</p><p>Out of the labor force 0.15 0.36 0 1</p><p>Student 0.00 0.07 0 1</p><p>Self-employed 0.09 0.29 0 1</p><p>Log of Income 3.40 0.22 2.80 3.79</p><p>Education: high 0.00 0.05 0 1</p><p>Education: medium 0.01 0.10 0 1</p><p>One child 0.22 0.41 0 1</p><p>Two children 0.19 0.39 0 1</p><p>Three +children 0.07 0.27 0 1</p><p>Heath: excellent 0.24 0.43 0 1</p><p>Health: good 0.47 0.49 0 1</p><p>Health: satisfactory 0.18 0.39 0 1</p><p>Health: poor 0.07 0.25 0 1</p><p>65409_Earth_Chapter3v2.indd 87 4/30/12 3:48 PM</p><p>88</p><p>WVS (1981-2008) : 86 Countries Summary Statistics</p><p>Variable Mean Std. Dev Min Max</p><p>Life satisfaction 6.33 2.50 1 10</p><p>Female 0.51 0.49 0 1</p><p>Age 41.05 7.27 30 55</p><p>Age^2/1000 1.73 0.61 0.9 3.02</p><p>Age^3/1000 75.83 39.60 27 166.37</p><p>Single 0.09 0.28 0 1</p><p>Widowed 0.02 0.17 0 1</p><p>Divorced 0.04 0.19 0 1</p><p>Separated 0.02 0.14 0 1</p><p>Unemployed 0.07 0.27 0 1</p><p>Out of the labor force 0.20 0.40 0 1</p><p>Self-employed 0.14 0.34 0 1</p><p>Student 0.01 0.10 0 1</p><p>Log of Income 1.38 0.61 0 2.30</p><p>Education: high 0.22 0.41 0 1</p><p>Education: medium 0.42 0.49 0 1</p><p>One child 0.16 0.36 0 1</p><p>Two children 0.32 0.46 0 1</p><p>Three + children 0.38 0.48 0 1</p><p>Heath: excellent 0.21 0.41 0 1</p><p>Health: good 0.44 0.49 0 1</p><p>Health: satisfactory 0.27 0.44 0 1</p><p>Health: poor 0.05 0.23 0 1</p><p>65409_Earth_Chapter3v2.indd 88 4/30/12 3:48 PM</p><p>89</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Appendix B</p><p>Comparing the impact of different influences on happiness</p><p>For policy analysis it is necessary to compare the quantitative impact of different factors upon happiness.</p><p>A natural way to do this is to compare the effect of a change in each other factor with the effect of a 30%</p><p>increase in income.</p><p>At present the calculations are purely illustrative drawing on a variety of sources, which are indicated in the</p><p>table. They do however show clearly how much else matters to people besides income.</p><p>Effects on life evaluation of each factor, as</p><p>a multiple of the effect of a 30% increase in income</p><p>Source</p><p>Individual unemployment (versus employment) -6.00 GSOEP Panel (Appendix A)</p><p>Unemployment rate (10% point increase) -1.30 Helliwell and Huang, 2011b</p><p>Social support (10% point extra saying yes) 4.5.0 Table 3.1, unstandardized</p><p>Freedom (10% point extra saying yes) 2.1.0 �������Ұ�������������������Ұ</p><p>Corruption (10% point extra saying yes) 1.9.0 �������Ұ�������������������Ұ</p><p>Malaise 8 years earlier (1 standard deviation worse) -10.00 British Cohort Study</p><p>Physical health (Poor versus good, self-assessed) -15.00 GSOEP panel (Appendix A)</p><p>Single (versus married) -2.2.0 ���Ұ������������Ұ</p><p>Separated (versus married) -4.0.0 ���Ұ������������Ұ</p><p>Widowed (versus married) -2.9.0 ���Ұ������������Ұ</p><p>Female (versus male) 1.5.0 GSOEP, using panel for effect of income</p><p>Note: The effect of each other variable is compared with that of individual relative income. The exceptions are for social support,</p><p>freedom and corruption, whose effects are compared with the effects of GDP per head.</p><p>65409_Earth_Chapter3v2.indd 89 4/30/12 3:48 PM</p><p>90</p><p>Part I.</p><p>Chapter 4.</p><p>SOME POLICY IMPLICATIONS</p><p>JOHN HELLIWELL, RICHARD LAYARD AND JEFFREY SACHS</p><p>John F. Helliwell: Professor Emeritus of Economics, University of British Columbia and</p><p>Arthur J.E. Child Foundation Fellow of the Canadian Institute for Advanced Research (CIFAR)</p><p>Richard Layard: Director, Well-being Programme, Centre for Economic Performance,</p><p>London School of Economics</p><p>Jeffrey D. Sachs: Director, The Earth Institute, Columbia University</p><p>65409_Earth_Chapter4.indd 90 4/30/12 12:05 PM</p><p>91</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Part I.</p><p>Chapter 4.</p><p>SOME POLICY IMPLICATIONS</p><p>JOHN HELLIWELL, RICHARD LAYARD AND JEFFREY SACHS</p><p>The UN General Assembly has invited Member States to “pursue the elaboration of additional measures</p><p>that better capture the importance of the pursuit of happiness and well-being in development with a view to</p><p>guiding their public policies.”1 This does not mean that the existing goals of governments should be abandoned.</p><p>Indeed, a focus on well-being is already guiding many national and international development policies. But</p><p>it does mean that the relative importance of different goals needs to be reconsidered, especially as absolute</p><p>poverty is progressively eliminated.</p><p>In this chapter we examine this issue in light of the evidence presented in the previous chapters. We do not</p><p>repeat the evidence but draw out some of its implications. We then go on to consider the wider strategic issue</p><p>of how the overall process of policy-making should be changed, if happiness is to become a major goal.2</p><p>New Policy Priorities</p><p>GDP</p><p>No government has ever had GDP per person as its only goal. But in the last 30 years income creation as</p><p>measured by GDP has become an increasing mantra, and we are often told that we cannot afford the “luxury”</p><p>of harmonious social relationships when they stand in its way.</p><p>The fi rst lesson of happiness research is that GDP is a valuable goal but that other things also matter greatly.</p><p>So GDP should not be pursued to the point where:</p><p>• economic stability is imperiled</p><p>• community cohesion is destroyed</p><p>• the weak lose their dignity or place in the economy</p><p>• ethical standards are sacrifi ced, or</p><p>• the environment, including the climate, is put at risk.</p><p>GDP is important but not all that is important. This is especially true in developed countries, where most or</p><p>all of the population has living standards far above basic material needs. Except in the very poorest countries</p><p>happiness varies more with the quality of human relationships than with income.3 And in the richest countries</p><p>it is essential not to subordinate the happiness of the people to the “interests of the economy,” since the</p><p>marginal utility of income is low when income is so high. The economy exists to serve the people, not vice</p><p>versa. Incremental gains in income in a rich country may be much less benefi cial to the population than steps</p><p>to ensure the vibrancy of local communities or better mental health.</p><p>Nor should maximizing GDP be used as an argument against limiting the scale of greenhouse gas emissions</p><p>or other environmental protections. There would be no point in having a high GDP for another decade or two if</p><p>that would be followed by widespread misery and dislocation due to climate change and the loss of biodiversity.</p><p>To be worth having, economic growth must be environmentally and socially sustainable.</p><p>Work</p><p>Among objectives other than long-term growth, high employment is one of the most important. Mass</p><p>unemployment is a major blow to society. It reduces the happiness of those unemployed by as much as</p><p>bereavement or divorce, and it also infects those who do have jobs with the fear of losing them. In conditions of</p><p>high unemployment, active labor market policies and other means to restore employment should be prioritized.</p><p>Of course employment policies must be judged by their effi cacy, not merely by their intention. Yet governments</p><p>should give great weight to policies that reduce involuntary unemployment, including retraining, job matching,</p><p>public employment, low-wage subsidies, education support (to raise long-term skills), and other policies.</p><p>65409_Earth_Chapter4.indd 91 4/30/12 12:05 PM</p><p>92</p><p>A bad job is often better than no job. Job sharing may play a role to spread jobs within an enterprise facing a</p><p>downturn. Long-term unemployment can be devastating, both psychologically and in the deterioration of skills.</p><p>Spurring socially productive community projects can be an important interim step to long-term employment.4</p><p>But of course as many people as possible should be in good jobs, where goodness is defined by their overall</p><p>contribution to well-being.</p><p>A good job is one that provides happiness and satisfaction to the worker, as well as positive spillovers to others.</p><p>Happiness at work really matters. Employers (including those in the public sector) need to give more weight</p><p>to this. It is in their interest to do so, since happy workers are at least as productive as unhappy ones, and</p><p>are more productive if their happiness results from the right kind of management. Good managers inspire</p><p>their workers by: explaining the purpose and goals of the job; leaving enough autonomy about how to do it;</p><p>and always providing support, appreciation and feedback. Managers need to develop the intrinsic motivation</p><p>of their workers, and to rely less on extrinsic incentives (like performance-related pay), which can often be</p><p>dysfunctional and actually undermine intrinsic motivation.</p><p>Community and governance</p><p>Human beings are social animals. We are happier when we are with others and our most rewarding experiences</p><p>are normally connected with human relationships.5 In all societies the most important relationships are with</p><p>loved ones, yet our relationships at work, with friends and in the community are also important.6 A successful</p><p>society is one in which people have a high level of trust in each other – including family members, colleagues,</p><p>friends, strangers, and institutions such as government. Social trust spurs a sense of life satisfaction.</p><p>High social trust is built on strong mutual respect among people, including those of all ages, genders,</p><p>ethnicities and social groupings. Government policy can contribute importantly to this by pursuing inclusive</p><p>policies (including in education) that ensure that all can participate as fully and equally as possible.</p><p>However, there are limits to the proper sphere of government. In a happy society, individuals feel they are</p><p>charting their own courses through life, without excessive constraints. That is why there was such unhappiness</p><p>in the countries of the Soviet bloc before their transition to functioning and stable open societies, and why the</p><p>happiest countries all have very high shares of their populations who feel free. Making happiness an objective</p><p>of governments would not therefore lead to the “servile society,”</p><p>and indeed quite the contrary, if governments</p><p>pay proper regard to the findings of happiness research. Happiness comes from an opportunity to mold one’s</p><p>own future, and thus depends on a robust level of freedom. Moreover, corruption in government is a major</p><p>cause of unhappiness in many countries, and needs to be rooted out.</p><p>Values and religion</p><p>The government is never the main source of values in a healthy community. Values are transmitted mainly by</p><p>parents, educators, faith organizations, the media, writers, and a whole range of organizations of civil society.7 In</p><p>well-functioning societies there is widespread support for the universal value that we should treat others as we would</p><p>like them to treat us. In one form or another, the golden rule ranks high in every system of religious commandments.</p><p>This is not an easy standard to live up to. But there is clear evidence from psychology and neuroscience that people</p><p>who care more about others are on average happier. When people do good deeds for others, they may not always feel</p><p>happier, but in general they do. It is difficult to see how harmonious societies could even exist unless this emotional</p><p>pay-off is there. We are hard-wired for altruism and cooperation, even if we are also able to cheat and violate the trust</p><p>of others. Our two-sided nature – both trust and deception – is at the core of human nature itself.</p><p>But people vary enormously in their standards of behavior and their “fellow feeling” for others. These qualities</p><p>can and should be systematically taught, as is already done in many schools. Unfortunately, the opposite can</p><p>also be taught. Students of game theory in economics courses often end up more likely to “defect” (that is, to</p><p>cheat) vis-à-vis their fellow students in classroom experiments. We need to inculcate the values of cooperation.</p><p>Our happiness and well-being in the community depend on it.</p><p>65409_Earth_Chapter4.indd 92 4/30/12 12:05 PM</p><p>93</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Most of the world’s religions also teach these principles. Among poorer societies, people in more religious regions</p><p>experience more positive emotion. But among richer societies there is no such difference and a good, modern</p><p>society needs to rely also on secular bases for morality. Unfortunately, some recent studies show that the more</p><p>affluent members of the society may have systematically less regard for others than poorer members of society</p><p>do. Wealth can lead to a sense of entitlement, a “right” to violate the trust of others.8 We need to cultivate social</p><p>norms so that the rich and powerful are never given a feeling of impunity vis-à-vis the rest of society.</p><p>Mental health</p><p>We come now to the more personal aspects of happiness and misery. Of these, mental illness is the most</p><p>important, and typically affects a third of all families in one way or another.9 Yet it is often barely mentioned</p><p>because of the feelings of guilt and inadequacy felt by those who suffer, and their relatives, as well as the feeling</p><p>that nothing can be done.</p><p>Mental illness, however, is as responsive to evidence-based treatment as most physical conditions. But in</p><p>most advanced countries only a quarter of mentally ill people are in treatment, and even fewer in developing</p><p>countries This contrasts with over three quarters for most physical illnesses in advanced countries. The</p><p>difference is a simple matter of discrimination.</p><p>In developed countries mental illness accounts for one half of all illness among people of working age. Yet it was</p><p>not covered by the recent UN high-level meeting on non-communicable diseases. A major change of perception</p><p>is clearly needed. The key need is for wider access to modern evidence-based psychological therapies.</p><p>Physical health</p><p>Improved physical health is probably the single most important factor that has improved human happiness</p><p>in recent centuries. This is true in countries at all levels of development. But major health gaps remain in</p><p>poor countries that require urgent increases in public investments, and increased support by donor countries.</p><p>And, as rich countries grow richer, people are clearly willing to contribute an increasing share of their income</p><p>to healthcare. This should be anticipated. There is of course a problem of out-of-control healthcare costs,</p><p>especially in insurance-based systems of care. But happiness research confirms that people greatly value</p><p>healthcare, and it is likely that growing health expenditures often through the public sector will give more</p><p>satisfaction than equivalent increases in private consumption.</p><p>Family and friends</p><p>Of all types of social life, close personal relationships with loved adults explain the greatest variation in</p><p>happiness. Traditionally, external support for family life was provided largely through faith communities.</p><p>But in secular societies all social organizations and institutions, including those managed by the state, have</p><p>important roles to play.</p><p>Many people sacrifice aspects of their family life in order to raise their income, or to achieve personal success.</p><p>These problems of work-life balance can be a major source of stress in families, and are an important source of</p><p>the U-shape in the age-happiness relationship described in Chapter 3. To reduce this stress, it is important that</p><p>workplaces and government policies, including those for childcare and family support, should operate flexibly,</p><p>in ways that moderate these pressures.</p><p>Education</p><p>A decent education for all is essential. It is an important support for building incomes, improving health,</p><p>creating trust, and increasing the efficiency and accountability of government. Indeed, it is primarily through</p><p>these important channels that education supports happier lives. Universal access to education is widely judged</p><p>to be a basic human right, because every aspect of an individual’s dignity, productivity, and ability to thrive is</p><p>enhanced by educational attainment.</p><p>65409_Earth_Chapter4.indd 93 4/30/12 12:05 PM</p><p>94</p><p>The Formation of Policy</p><p>At this stage the science of happiness is in its infancy and its policy implications are inevitably piecemeal</p><p>and tentative. But the happiness agenda is a progressive project that will have a steadily increasing impact on</p><p>society as knowledge and experience accumulate. Over time the explanation of happiness (its causes, direct</p><p>and indirect) should become a central purpose of social science.</p><p>So, how should this increased understanding affect the ways in which governments make policy? There are</p><p>three steps:</p><p>• the measurement of happiness in the population</p><p>• the explanation of how happiness is determined in individuals, communities and whole populations, and</p><p>• the use of this understanding to make policy aimed at greater happiness and less misery.</p><p>Measurement</p><p>The first reason for measuring happiness is to enable citizens and policy-makers to know what their problems</p><p>and opportunities are, and how well difficulties are being solved and doors are being opened. Measurement</p><p>needs to be done in sufficient depth to show how happiness varies among cities and regions, as well as among</p><p>demographic and social groups within those regions.</p><p>The government should systematically survey the subjective well-being of the population. It should also be</p><p>possible to link this information to information on health, education, poverty reduction, and so forth. A</p><p>wide range of surveys will help to link surveys across different subject areas, and to enlarge the range of</p><p>information required to better understand the sources and consequences of happiness.</p><p>It is also highly desirable that happiness be measured by firms, communities, schools, hospitals and even medical</p><p>practitioners. This will permit a more rapid increase in knowledge about the sources and consequences of happiness.</p><p>These measures can also, at the same time, improve the evaluation and performance of all types of organizations.</p><p>A central issue is how governments should measure happiness</p><p>(as proposed in the General Assembly</p><p>resolution). Every organization and nation will have their own ideas of what aspects of life are most important</p><p>to them, and this will dictate what they choose to measure. To make the knowledge they discover of most use</p><p>to others, it is important to have some core questions that are asked in the same way everywhere.</p><p>As described later in this report, the OECD has a major process underway to develop recommendations</p><p>for national statistical offices to consider when expanding the well-being content of their national statistical</p><p>systems. We would not wish to prejudge these recommendations. However, the analysis in Chapters 2 and 3</p><p>found that there would be some advantage in asking two different life evaluation questions, mainly to achieve</p><p>the gains from multiple measures, and partly to reduce the risks that a particular question might not resonate</p><p>in the same way in different languages and cultures. The European Social Survey has for several rounds used</p><p>the following two questions, and our analysis has shown that together they provide a more robust measure of</p><p>well-being than either on its own.</p><p>• Taking all things together, how happy would you say you are?</p><p>(where 0 means extremely unhappy, and 10 means extremely happy)</p><p>• All things considered, how satisfied are you with your life as a whole nowadays?</p><p>(where 0 means extremely dissatisfied and 10 means extremely satisfied)</p><p>Countries may also wish to ask questions about the experience of positive and negative affect, either in general</p><p>surveys or in surveys that track the ups and downs of daily life.</p><p>65409_Earth_Chapter4.indd 94 4/30/12 12:05 PM</p><p>95</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Explanation</p><p>Up to now the science of happiness has focused mainly on the role of a person’s current circumstances in</p><p>determining their happiness. We have reported some of the results in Chapter 3. But, as the diagram at the</p><p>beginning of that chapter shows, these features are in many cases experiences in a person’s life (and their</p><p>genes as well as their environment). A new frontier for happiness research will be a more developmental</p><p>approach, using very long birth-cohort panels to identify factors affecting happiness over the life-course.</p><p>There is also a pressing need for more controlled experiments to find the effects of different interventions.</p><p>From multiple sources, we need quantitative results, which show us how much different factors really matter</p><p>in determining happiness as variously measured.</p><p>Policy-making using happiness as a criterion</p><p>As knowledge increases, societies will have a growing basis for a new type of policy-making aimed at increasing</p><p>happiness and reducing misery. This would involve a change of perspective and a change in the techniques of</p><p>policy analysis.</p><p>At present many countries use a traditional form of cost-effectiveness analysis, in which benefits are measured</p><p>in money units on the basis of what citizens would be willing to pay for those benefits.10 This works quite well</p><p>when the primary benefits are indeed financial or can be readily transferred into monetary equivalents. This</p><p>is often true for policies on industry, transport, education and employment. However expenditure in these</p><p>areas is often no more than a quarter or so of public expenditure. The bulk of public expenditure is on health,</p><p>social care, law and order, the environment, child welfare, and income support. In none of these cases does</p><p>willingness to pay provide adequate guidance to the benefits that arise. Happiness would be an excellent added</p><p>criterion for evaluating these expenditures.</p><p>So we can well envisage a parallel system of evaluation taking shape over time where policies are judged by the</p><p>changes in happiness that they produce per unit of net public expenditure.11 Developing such systems should</p><p>be a goal, at least provisionally. To make them fully operational will of course require more information and</p><p>much more verification, but here the chicken-and-the-egg issue must be confronted. We should get started in</p><p>serious thinking about the links of policies to produce subjective well-being, just as Bhutan is doing with Gross</p><p>National Happiness.</p><p>More knowledge is needed before such methods can be used, but the knowledge is more likely to be produced</p><p>if there is an adequate demand. It is therefore important for governments to foresee their own requirements for</p><p>knowledge about the well-being of the population, and to set in motion the relevant sequence of data collection</p><p>and research to develop that knowledge. If this is done, there will be then an ever-growing understanding of</p><p>what things matter most to people and in which ways. This growing understanding may well provide a new</p><p>basis for policy-making in the age of sustainable development.</p><p>Philosophical issues</p><p>Of course most governments will have many other aims besides increasing the self-reported happiness of their</p><p>population. Leaving aside the desire to be re-elected (which is generally helped if the population are happy),</p><p>governments will surely care about health, freedom, honor, the realization of human potential, social justice, and the</p><p>well-being of future generations. These are all important aims to society. What then is the special role of happiness?</p><p>In Ancient Greece, Aristotle argued that happiness was the only good that was “good in itself.” This argument</p><p>still has relevance.12 If we ask why health matters, we can give reasons: people feel bad when they are sick.</p><p>Similarly people feel bad when they are not free. And so on. But if we ask, “Why does it matter if people feel</p><p>good or bad?” we often end up with the proposition that people’s feelings – their happiness – is the ultimate</p><p>standard for judging the importance of health or some other objective.</p><p>65409_Earth_Chapter4.indd 95 4/30/12 12:05 PM</p><p>96</p><p>People sometimes say that such a focus encourages the selfish pursuit of individual happiness at the expense</p><p>of others. Nothing could be more fallacious. The evidence is very strong that a society cannot be happy unless</p><p>there is a high degree of altruism and trust among its members. That is why Aristotle advocated that happiness</p><p>should be mainly pursued through virtuous acts. The Buddha and countless other sages, as well as many of</p><p>today’s leading psychologists and moral leaders, argue the same.</p><p>As regards social justice, this too is of course vital. It is about the distribution of the benefits of life among</p><p>the members of the community. In the context of happiness it is an issue about the distribution of happiness.</p><p>So social justice is not, strictly speaking, an alternative value to happiness; it is intrinsic to how happiness</p><p>outcomes should be measured and evaluated.</p><p>For a given average level of happiness, most would prefer a more equal to a less equal distribution. But how</p><p>far would we accept a fall in average happiness in order to promote a more equal spread of happiness? Or, to</p><p>put the point another way, what weights would we give to increased happiness accruing to people at different</p><p>points of the happiness scale? These are important ethical issues that have to be decided by the community,</p><p>through politics, culture, and public debate and deliberation.</p><p>Governments will differ on these issues and in the importance that they should attach to happiness in general.</p><p>But the General Assembly’s resolution that happiness should be given greater attention seems to be on the</p><p>mark. Societies want to be happy, and for good and deep reasons. The evidence given in this Report will, we</p><p>hope, give governments some indication of what this means: both the new priorities it implies, and alternative</p><p>modes of policy formation.</p><p>Sustainable Development</p><p>In Chapter 1 we proposed that happiness data and research should underpin the design and attainment of</p><p>the four pillars of sustainable development: ending extreme poverty, environmental</p><p>sustainability, social</p><p>inclusion, and good governance. The happiness consequences of ending poverty, building social inclusion</p><p>and achieving good government have been well documented by the data and research in Chapters 2 and 3.</p><p>There are as yet many fewer established links between happiness and environmental sustainability. We would</p><p>argue that the tools of happiness research have real potential to recast the debate between economic growth</p><p>and environmental protection, and to show new ways to harness human ingenuity to improve all four pillars</p><p>at the same time.</p><p>The environmental debate could be importantly recast by changing the fundamental objectives from economic</p><p>growth to building and sustaining the quality of lives, as assessed by those whose lives they are. This will</p><p>depend crucially on the human capacity for cooperation that we have documented. The assumption that</p><p>individuals are only interested in their own material standards of life has made the possibilities for preserving</p><p>the environment seem unrealistic to many observers. But such pessimism is misplaced. On the contrary people</p><p>gain in happiness by working together for a higher purpose. There can be no higher purpose than promoting</p><p>the Earth’s environmental balance, the well-being of future generations, and the survival and thriving of other</p><p>species as well. Sustainability is an instrumental goal, because without it, our health and prosperity are bound</p><p>to collapse. But environmental sustainability is also an end goal: we care about nature, we care about other</p><p>species, and we care about future generations. Building a wider sense of common identity among all peoples,</p><p>with each other, with other species, and with the future of a threatened planet, will not be easy. But it is the only</p><p>possible way forward for human thriving and even survival.</p><p>65409_Earth_Chapter4.indd 96 4/30/12 12:05 PM</p><p>97</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>1 UN General Assembly Resolution A/65/L.86 (13 July 2011).</p><p>2 For a more extended discussion, see Bok (2010), Diener et al. (2009) and Layard (2011). For a discussion of recent evidence-based</p><p>changes in Britain, see Halpern (2010).</p><p>3 Helliwell & Barrington-Leigh (2010).</p><p>4 Layard et al. (2005), Gyarmati et al. (2008).</p><p>5 Kahneman et al. (2004). On the social context of well-being see Helliwell (2011).</p><p>6 Layard (2011). Annex 5.2.</p><p>7 For example, Action for Happiness is an organization whose members pledge to try to create more happiness and less misery in the</p><p>world around them.</p><p>8 Piff et al. (2012).</p><p>9 On this section, see Layard (2012).</p><p>10 In “cost-benefit analysis” the government’s budget is assumed to adjust to accommodate all justified expenditure. A more practical</p><p>procedure is “cost-effectiveness analysis” where the budget is assumed constant and the aim is that it deliver the maximum benefit.</p><p>11 Policies evaluated this way will of course have to be compared with policies evaluated using willingness to pay. At that stage it is</p><p>important to recognize that, when an individual benefits by some monetary amount that we can convert into units of happiness, society</p><p>may benefit less in total happiness due to a negative effect on those who do not benefit, stemming from the increase in the income of</p><p>their “comparator.”</p><p>12 Layard (2011), Chapter 15.</p><p>65409_Earth_Chapter4.indd 97 4/30/12 12:05 PM</p><p>98</p><p>References</p><p>Acemoglu, D., & Angrist, J. (2001). How large are human capital</p><p>externalities? Evidence from compulsory schooling laws. NBER</p><p>Macroannual, 9–59.</p><p>Agrawal, S., & Harter, J. K. (2011). Context effect on life evaluation</p><p>items: Forming new estimates for data prior to January 6, 2009.</p><p>Gallup Technical Report, Washington: Gallup Organization.</p><p>Aknin, L. B., Barrington-Leigh, C. P., Dunn, E. W., Helliwell, J.</p><p>F., Biswas-Diener, R., Kemeza, … Norton, M. I. (2010). Prosocial</p><p>spending and well-being: Cross-cultural evidence for a psychological</p><p>universal. NBER Working Paper No. 16415, Cambridge: National</p><p>Bureau of Economic Research. Retrieved from http://www.nber.</p><p>org/papers/w16415.</p><p>Alesina, A., Di Tella, R., & MacCulloch, R. (2004). Inequality and</p><p>Happiness: Are Europeans and Americans Different? Journal of Pub-</p><p>lic Economics, 88, 2009-2042. doi:10.1016/j.jpubeco.20033.07.006</p><p>Alesina, A., & La Ferrara, E. (2005). Preferences for Redistribution in</p><p>the Land of Opportunities, Journal of Public Economics, 89, 897-931.</p><p>Algan, Y., & Cahuc, P. (2010). Inherited trust and growth. American</p><p>Economic Review, 100, 2060-92.</p><p>Andrews, F. M., & Withey, S. B. (1976). Social indicators of well-being:</p><p>American perceptions of life quality. New York: Plenum Press.</p><p>Angrist, J. D., & Krueger, A. (1991). Does compulsory school</p><p>attendance affect schooling and earnings? Quarterly Journal of</p><p>Economics, 106(4), 979–1014.</p><p>Anik, L., Aknin, L. B., Norton, M. I., & Dunn, E. W. (2010). Feeling</p><p>Good about Giving: The Benefits (and Costs) of Self-Interested</p><p>Charitable Behavior. Harvard Business School Discussion Paper No.</p><p>10-012.</p><p>Ariely, D., Gneezy, U., Lowenstein, G., & Mazar, N. (2005). Large</p><p>Stakes and Big Mistakes. Federal Reserve Bank of Boston Working</p><p>Paper No 05-11, Boston: Federal Reserve Bank of Boston. Retrieved</p><p>from http://www.bos.frb.org/economic/wp/wp2005/wp0511.pdf</p><p>Atkinson, T., Bourguignon, F., & Morrisson, C. (1992). Empirical</p><p>Studies in Earnings Mobility. New York: Harwood.</p><p>Bacon, N., Brophy, M., Mguni, N., Mulgan, G., & Shandro, A.</p><p>(2010). The States of happiness: Can public policy shape peoples</p><p>wellbeing and resilience? London: The Young Foundation. Retrieved</p><p>from http://www.youngfoundation.org/files/images/wellbeing_</p><p>happiness_Final__2_.pdf</p><p>Balliet, D. (2010). Communication and cooperation in social</p><p>dilemmas: A meta-analytic review. Journal of Conflict Resolution,</p><p>54(1), 39-57. doi:10.1177/0022002709352443</p><p>Barrington-Leigh, C.P. (2011). The Québec convergence and</p><p>Canadian life satisfaction 1985-2008. Retrieved from http://</p><p>wellbeing.ihsp.mcgill.ca/publications/Barrington-Leigh-DRAFT-</p><p>Quebec.pdf</p><p>Barrington-Leigh, C.P., & Helliwell, J.F. (2008). Empathy and</p><p>emulation: Life satisfaction and the urban geography of comparison</p><p>groups. NBER Working Paper 14593.</p><p>Bartolini, S. (2011). Sociability predicts happiness in nations:</p><p>Evidence from macro and micro data. In S. Bartolini (Ed.), Policies</p><p>for Happiness. Oxford: Oxford University Press.</p><p>Basu, A., Kaplan, R., & Kaplan, S. (2012). Reasonableness as a</p><p>way to move towards sustainable well-being. In T. Hämäläinen</p><p>& J. Michaelson, (Eds.), New Theories and Policies for Well-Being.</p><p>(forthcoming).</p><p>Becker, G. S. (1974). A theory of marriage: Part II. Journal of Political</p><p>Economy, 82(2), S11-S26.</p><p>Benjamin, D., Heffetz, O., Kimball, M., & Rees-Jones, A. (2012).</p><p>What Do You Think Would Make You Happier? What Do You Think</p><p>You Would Choose? American Economic Review, forthcoming.</p><p>Bianchi, M. (2012). Financial Development, Entrepreneurship</p><p>and Job Satisfaction. The Review of Economics and Statistics, 94(1):</p><p>273–286.</p><p>Bilaliü, M., McLeod, P., & Gobet, F. (2008). Why good</p><p>thoughts block better ones: The mechanism of the pernicious</p><p>Einstellung (set) effect. Cognition, 108(3), 652-661. doi: 10.1016/j.</p><p>cognition.2008.05.005</p><p>Blanchflower, D. G., (2008). International evidence on well-being.</p><p>NBER Working Paper 14318.</p><p>Blanchflower, D. G., & Oswald, A. J. (1998). What makes an entre-</p><p>preneur? Journal of Labor Economics, 16, 26–60.</p><p>Blanchflower, D. G., & Oswald, A. J. (2004). Well-Being over time in</p><p>Britain and the USA. Journal of Public Economics, 88(7–8), 1359-1386.</p><p>Blanchflower, D. G., & Oswald, A. J. (2008). Is Well-being U-Shaped</p><p>over the Life Cycle? Social Science & Medicine, 66(6), 1733-1749.</p><p>Blanchflower, D. G., Oswald, A. J., & Stutzer, A. (2001). Latent</p><p>Entrepreneurship Across Nations. European Economic Review, 45,</p><p>680-691.</p><p>Blanden, J. (2009). How much can we learn from international</p><p>comparisons of intergenerational mobility? CEE Discussion Paper</p><p>No 0111.</p><p>Bok, D. (2010). The</p><p>Politics of Happiness: What Government Can</p><p>Learn from the New Research on Well-Being. Princeton and Oxford:</p><p>Princeton University Press.</p><p>Bok, S. (2010). Exploring happiness from Aristotle to brain science.</p><p>New Haven: Yale University Press.</p><p>65409_Earth_Reference.indd 98 4/30/12 12:08 PM</p><p>99</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>99</p><p>Bound, J. & Krueger, A.B. (1999). The Extent of Measurement</p><p>Error in Longitudinal Earnings Data: Do Two Wrongs Make a</p><p>Right? Journal of Labor Economics, 9(1), 1-24.Bowles, S., & Gintis,</p><p>H. (2002). Social Capital and Community Governance. Economic</p><p>Journal, 112, 419-436.</p><p>Brickman, P., Coates, D., & Janoff-Bulman, R. (1978). Lottery</p><p>winners and accident victims: Is happiness relative? Journal of</p><p>Personality and Social Psychology, 36(8), 917-927. doi:10.1037/0022-</p><p>3514.36.8.917</p><p>Brereton, F., Clinch, J., & Ferreira, S. (2008). Happiness, geogra-</p><p>phy and the environment. Ecological Economics, 65, 386-396.</p><p>Brown, S. L., Nesse, R.M., Vinokur, A.D., & Smith, D.M. (2003).</p><p>Providing social support may be more beneficial than receiving</p><p>it: Results from a prospective study of mortality. Psychological</p><p>Science, 14(4), 320-7.</p><p>Bruni, L., & Stanca, L. (2008). Watching alone: Relational goods,</p><p>television and happiness. Journal of Economic Behavior & Organisation,</p><p>65, 506-528.</p><p>Burkhauser, R. V., & Poupore, J. G. (1997). A Cross-National</p><p>Comparison of Permanent Inequality in the United States and</p><p>Germany. Review of Economics and Statistics, 81, 251-368.</p><p>Cantril, H. (1965). The Pattern of Human Concerns. New</p><p>Brunswick, N.J.: Rutgers University Press.</p><p>Card, D. (2001). Estimating the return to schooling: progress</p><p>on some persistent econometric problems. Econometrica, 69(5),</p><p>1127–1160.</p><p>Card, D., Mas, A., Moretti, E., & Saez, E. (2010). Inequality at</p><p>work: The effect of peer salaries on job satisfaction. NBER Work-</p><p>ing Paper 16396.</p><p>Central Intelligence Agency (CIA) (2009). The world factbook</p><p>2009. US Central Intelligence Agency. Retrieved from http://www.</p><p>geographic.org/wfb2009/taiwan/taiwan_people.html</p><p>Chapman, B. J., & Guven, C. (2010). Marital Status is Misunder-</p><p>stood in Happiness Models. Deakin University Economics Working</p><p>Paper Series No 2010/03.</p><p>Chida, Y., & Steptoe, A. (2008). Positive psychological well-being</p><p>and mortality: A quantitative review of prospective observational</p><p>studies. Psychosomatic Medicine, 70(7), 741-756. doi:10.1097/</p><p>PSY.0b013e31818105ba</p><p>Clark, A. E. (2003). Unemployment as a Social Norm: Psycho-</p><p>logical Evidence from Panel Data. Journal of Labor Economics, 21,</p><p>323-351.</p><p>Clark, A. E. (2010). Work, Jobs and Well-Being across the Millen-</p><p>nium. In E. Diener, J. F. Helliwell, & D. Kahneman (Eds.), Inter-</p><p>national differences in well-being. Oxford: Oxford University Press.</p><p>Clark, A.E. (2011). Happiness, Habits and High Rank: Compari-</p><p>sons in Economic and Social Life. IZA Discussion Paper No 5966.</p><p>Clark, A. E., Diener, E., Georgellis, Y., & Lucas, R. (2008). Lags</p><p>and Leads in Life Satisfaction: A Test of the Baseline Hypothesis.</p><p>Economic Journal, 118, F222–F243.</p><p>Clark, A. E., & Fawaz, Y. (2009). Valuing Jobs via Retirement: Eu-</p><p>ropean Evidence. National Institute Economic Review, 209, 88-103.</p><p>Clark, A.E., Frijters, P., & Shields, M. (2008). Relative Income,</p><p>Happiness and Utility: An Explanation for the Easterlin Paradox</p><p>and other Puzzles. Journal of Economic Literature, 46, 95-144.</p><p>Clark, A. E., & Lelkes, O. (2009). Let us Pray: Religious Interac-</p><p>tions in Life Satisfaction. PSE Working Paper No 2009-01.</p><p>Clark, A. E., & Oswald, A. J. (1994). Unhappiness and Unemploy-</p><p>ment. Economic Journal, 104, 648-659.</p><p>Clark, A.E., & Oswald, A. J. (2002). A Simple Statistical Model</p><p>for Measuring How Life Events Affect Happiness. International</p><p>Journal of Epidemiology, 31, 1139-1144.</p><p>Clark, A.E., & Senik, C. (2010). Who compares to Whom? The</p><p>Anatomy of Income Comparisons in Europe. The Economic Journal,</p><p>120, 573-594. doi:10.1111/j.1468-0297.2010.02359.x.</p><p>Coe, N., & Lindeboom, M. (2008). Does Retirement Kill You?</p><p>Evidence from Early Retirement Windows. IZA Discussion Paper</p><p>No.3817.</p><p>Cohen, S., Doyle, W. J., Turner, R. B., Alper, C. M., & Skoner,</p><p>D. P. (2003). Emotional style and susceptibility to the common</p><p>cold. Psychosomatic Medicine, 65(4), 652-657. doi:10.1097/</p><p>01.PSY.0000077508.57784.DA</p><p>Cohen, S., & Pressman, S. D. (2006). Positive Affect and Health.</p><p>Current Directions in Psychological Science, 15(3), 122-125. doi:</p><p>10.1111/j.0963-7214.2006.00420.x</p><p>Cohen-Zada, D., & Sander, W. (2010). Religious participation</p><p>versus shopping: What makes people happier? IZA Discussion</p><p>Paper No. 5198.</p><p>Cooper, C. L., & Lundberg, U. (2011). The science of occupational</p><p>health: stress, psychobiology and the new world of work. Chichester:</p><p>John Wiley & Sons.</p><p>Csikszentmihalyi, M., & Larson, R. (1987). Validity and reliability</p><p>of the experience sampling method. Journal of Nervous and Mental</p><p>Disease, 175(9), 526-536. Retrieved from http://www.ncbi.nlm.nih.</p><p>gov/pubmed/3655778</p><p>Danner, D. D., Snowdon, D. A., & Friesen, W. D. (2001). Positive</p><p>emotions in early life and longevity: Findings from the nun</p><p>study. Journal of Personality and Social Psychology 80(5), 804-813.</p><p>doi:10.1037/0022-3514.80.5.804</p><p>65409_Earth_Reference.indd 99 4/30/12 12:08 PM</p><p>100</p><p>Davidson, R.J, et al. (2003). Alterations in brain and immune</p><p>function produced by mindfulness meditation. Psychosomatic</p><p>Medicine, 65, 564-70.</p><p>Dave, D., Rashad, I., & Spasojevici, J. (2008). The Effects of Retire-</p><p>ment on Physical and Mental Health Outcomes. Southern Economic</p><p>Journal, 75, 497-523.</p><p>Deaton, A. (2008). Income, Health, and Well-Being around the</p><p>World: Evidence from the Gallup World Poll. Journal of Economic</p><p>Perspectives, 22(2), 53-72.</p><p>Deaton, A. (2012). The financial crisis and the well-being of</p><p>Americans. Oxford Economic Papers, 64(1), 1-26. doi:10.1093/oep/</p><p>gpr051</p><p>Deci, E. L., & Ryan, R. M. (1985). Intrinsic Motivation and Self-Deter-</p><p>mination in Human Behaviour. New York: Plenum Press.</p><p>Demyttenaere, K., et al. (2004). Prevalence, Severity, and Unmet</p><p>Need for Treatment of Mental Disorders in the World Health</p><p>Organisation World Mental Health Surveys. Journal of the American</p><p>Medical Association, 291 (21), 2581-2590.</p><p>De Neve, J.-E., Christakis, N. A., Fowler, J. H., & Frey, B. S. (2011).</p><p>Genes, economics, and happiness. Retrieved from http://jhfowler.</p><p>ucsd.edu/genes_economics_and_happiness.pdf</p><p>Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95(3),</p><p>342-375. doi:10.1037/0033-2909.95.3.542</p><p>Diener, E. (2011). The validity of life satisfaction measures.</p><p>Unpublished manuscript.</p><p>Diener, E. & Biswas-Diener, R. (2008). Happiness: Unlocking the</p><p>mysteries of psychological wealth. Malden, MA: Blackwell Publishing.</p><p>Diener, E., & Chan, M. Y. (2011). Happy people live longer:</p><p>Subjective well-being contributes to health and longevity. Applied</p><p>Psychology: Health and Well-Being, 3(1), 1-43. doi:10.1111/j.1758-</p><p>0854.2010.01045.x</p><p>Diener, E., Helliwell, J. F., & Kahneman, D. (Eds.). (2010).</p><p>International differences in well-being. New York: Oxford University</p><p>Press. doi:10.1093/acprof:oso/9780199732739.001.0001</p><p>Diener, E., Lucas, R., Schimmack, U., & Helliwell, J. F. (2009).</p><p>Well-Being for public policy. New York: Oxford University Press.</p><p>Diener, E., Tay, L., & Myers, D., (2011). The religion paradox: If re-</p><p>ligion makes people happy, why are so many dropping out? Journal</p><p>of Personality and Social Psychology, 101, 1278-1290.</p><p>Di Tella, R., Haisken-De New, J., & MacCulloch, R. (2007). Hap-</p><p>piness adaptation to income and to status in an individual panel.</p><p>NBER Working Paper No 13159.</p><p>Di Tella, R., & MacCulloch, R. (2009). Happiness, contentment</p><p>and other emotions for central bankers. In C. L. Foote, L. Goette,</p><p>& S. Meier (Eds.), Policymaking insights from behavioural economics</p><p>(311-373). Boston: Federal Reserve Bank of Boston. (Proceedings of</p><p>a conference held in September 2007)</p><p>Di Tella, R., MacCulloch, R. & Oswald, A. (2003).</p><p>what comes</p><p>next? What are the pathways to well-being when basic economic needs are no longer the main drivers of</p><p>social change? What will guide humanity in the Anthropocene: advertising, sustainability, community, or</p><p>something else? What is the path to happiness?</p><p>Taking Happiness Seriously</p><p>Most people agree that societies should foster the happiness of their citizens. The U.S. Founding Fathers</p><p>recognized the inalienable right to the pursuit of happiness. British philosophers talked about the greatest</p><p>good for the greatest number. Bhutan has famously adopted the goal of Gross National Happiness (GNH)</p><p>rather than Gross National Product. China champions a harmonious society.</p><p>Yet most people probably believe that happiness is in the eye of the beholder, an individual’s choice, some-</p><p>thing to be pursued individually rather than as a matter of national policy. Happiness seems far too subjec-</p><p>tive, too vague, to serve as a touchstone for a nation’s goals, much less its policy content. That indeed has</p><p>been the traditional view. Yet the evidence is changing this view rapidly.</p><p>A generation of studies by psychologists, economists, pollsters, sociologists, and others has shown that</p><p>happiness, though indeed a subjective experience, can be objectively measured, assessed, correlated with</p><p>observable brain functions, and related to the characteristics of an individual and the society. Asking people</p><p>whether they are happy, or satisfied with their lives, offers important information about the society. It can</p><p>signal underlying crises or hidden strengths. It can suggest the need for change.</p><p>Such is the idea of the emerging scientific study of happiness, whether of individuals and the choices they</p><p>make, or of entire societies and the reports of the citizenry regarding life satisfaction. The chapters ahead</p><p>summarize the fascinating and emerging story of these studies. They report on the two broad measurements</p><p>of happiness: the ups and downs of daily emotions, and an individual’s overall evaluation of life. The former</p><p>is sometimes called “affective happiness,” and the latter “evaluative happiness.”</p><p>65409_Earth_Chapter1v2.indd 6 4/30/12 3:46 PM</p><p>7</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>What is important to know is that both kinds of happiness have predictable causes that reflect various facets</p><p>of our human nature and our social life. Affective happiness captures the day-to-day joys of friendship, time</p><p>with family, and sex, or the downsides of long work commutes and sessions with one’s boss. Evaluative</p><p>happiness measures very different dimensions of life, those that lead to overall satisfaction or frustration</p><p>with one’s place in society. Higher income, better health of mind and body, and a high degree of trust in</p><p>one’s community (“social capital”) all contribute to high life satisfaction; poverty, ill health, and deep divisions</p><p>in the community all contribute to low life satisfaction.</p><p>What we learn in the chapters ahead is that happiness differs systematically across societies and over time,</p><p>for reasons that are identifiable, and even alterable through the ways in which public policies are designed</p><p>and delivered. It makes sense, in other words, to pursue policies to raise the public’s happiness as much as</p><p>it does to raise the public’s national income. Bhutan is on to something path breaking and deeply insightful.</p><p>And the world is increasingly taking notice.</p><p>A household’s income counts for life satisfaction, but only in a limited way. Other things matter more:</p><p>community trust, mental and physical health, and the quality of governance and rule of law. Raising incomes</p><p>can raise happiness, especially in poor societies, but fostering cooperation and community can do even more,</p><p>especially in rich societies that have a low marginal utility of income. It is no accident that the happiest</p><p>countries in the world tend to be high-income countries that also have a high degree of social equality, trust,</p><p>and quality of governance. In recent years, Denmark has been topping the list. And it’s no accident that the</p><p>U.S. has experienced no rise of life satisfaction for half a century, a period in which inequality has soared,</p><p>social trust has declined, and the public has lost faith in its government.</p><p>It is, of course, one thing to identify the correlates of happiness, and quite another to use public policies to</p><p>bring about a society-wide rise in happiness (or life satisfaction). That is the goal of Bhutan’s GNH, and the</p><p>motivation of an increasing number of governments dedicated to measuring happiness and life satisfaction</p><p>in a reliable and systematic way over time. The most basic goal is that by measuring happiness across a</p><p>society and over time, countries can avoid “happiness traps” such as in the U.S. in recent decades, where</p><p>GNP may rise relentlessly while life satisfaction stagnates or even declines.</p><p>The Bhutan case study tells the story of GNH in Bhutan, a story of exploration and progress since the King</p><p>declared in 1972 the goal of happiness over the goal of wealth. Happiness became much more than a</p><p>guidepost or inspiration; it became an organizing principle for governance and policy-making as well. The</p><p>Gross National Happiness Index is the first of its kind in the world, a serious, thoughtful, and sustained</p><p>attempt to measure happiness, and use those measurements to chart the course of public policy. I leave</p><p>description of Bhutan’s wonderful adventure, still unfolding while already inspiring others, to the case study.</p><p>Happiness and the Sustainable Development Goals</p><p>As the world enters the dangerous next decades of the Anthropocene, we must intensify our efforts to achieve</p><p>a new course, one that ensures poor countries have the right to develop, and all countries have the right to</p><p>happiness, while simultaneously curbing the human-induced destruction of the environment. It is too late</p><p>to head off entirely climate change and loss of biodiversity. There is still time, though, to mitigate the damage</p><p>and to build resilience to the changes ahead. The quest for happiness will be carried out in the context of</p><p>growing environmental risks.</p><p>According to the recent recommendations of the UN Secretary-General’s High-level Panel on Global Sustain-</p><p>ability, the Millennium Development Goals, set to end in 2015, should be followed by a new set of Sustainable</p><p>65409_Earth_Chapter1v2.indd 7 4/30/12 3:46 PM</p><p>8</p><p>Development Goals. More succinctly, the MDGs should be followed by the SDGs. It is likely that the concept</p><p>of the SDGs will be adopted by the UN member states at the Rio+20 Summit in Rio de Janeiro in June 2012.</p><p>The Sustainable Development Goals should have four pillars. The first should be to carry on the crucial work</p><p>of the MDGs in order to end extreme poverty by 2030. The developing countries have successfully cut the</p><p>overall poverty rate by half comparing 1990 and 2010, from around 44% to 22%. The biggest gains have</p><p>come in China, while Africa has lagged behind, though Africa too is now on a path of poverty reduction. No</p><p>later than 2030 the remaining extreme poverty and hunger should be eradicated. Happiness in the poorest</p><p>countries would be strongly boosted by such an historic breakthrough.</p><p>The second pillar of the SDGs should be environmental sustainability. Without that, no gains against pov-</p><p>erty, hunger, or disease can endure long. The environmental pillar of the SDGs may be guided by the concept</p><p>of “planetary boundaries,” the notion that humanity must avoid specific thresholds of environmental damage</p><p>to avoid creating irreparable harms to the Earth and to future generations.</p><p>The third pillar should be social inclusion, the commitment of every society that the benefits of technology,</p><p>economic progress, and good governance should be accessible to everybody, women as well as men, minority</p><p>groups as well as the majority. Happiness must not be the preserve of a dominant group. The goal</p><p>The Macro-</p><p>economics of Happiness. Review of Economics and Statistics, 85,</p><p>809–827.</p><p>Di Tella, R., MacCulloch, R., & Oswald, A. (2001). Preferences over</p><p>inflation and unemployment: Evidence from surveys of happiness.</p><p>American Economic Review, 91(1), 335-341. doi:10.1257/aer.91.1.335</p><p>Dohmen, T., Falk, A., Fliessbach, K., Sunde, U., & Weber, B. (2011).</p><p>Relative versus absolute income, joy of winning, and gender: Brain</p><p>imaging evidence. Journal of Public Economics, 95, 279-285.</p><p>Dolan, P., Peasgood, T., & White, M. (2008). Do we really know</p><p>what makes us happy? A review of the economic literature on the</p><p>factors associated with subjective well-being. Journal of Economic</p><p>Psychology, 29(1), 94-122.</p><p>Dolan, P., Layard, R., & Metcalfe, R. (2011). Measuring subjective</p><p>well-being for public policy. London: Office for National Statistics.</p><p>Dunn, E., Aknin, L., & Norton, M. (2008). Spending Money on</p><p>Others Promotes Happiness. Science, 319, 1687-1688.</p><p>Durante, R. (2009). Risk, cooperation and the economic origins of</p><p>social trust: an empirical investigation. Retrieved from http://www.</p><p>cepr.org/meets/wkcn/2/2413/papers/Durante.pdf</p><p>Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., &</p><p>Schellinger, K. B. (2011). The Impact of Enhancing Students’ Social</p><p>and Emotional Learning: A Meta-Analysis of School-Based Univer-</p><p>sal Interventions. Child Development, 82(1), 405-432.</p><p>Easterlin, R. (1974). Does economic growth improve the human</p><p>lot? Some empirical evidence. In P. A. David & M. W. Reder (Eds.),</p><p>Nations and Households in Economic Growth: Essays in Honor of</p><p>Moses Abramovitz (pp. 89-125). New York: Academic Press.</p><p>Easterlin, R.A. (2006). Life Cycle Happiness and Its Sources:</p><p>Intersections of Psychology, Economics, and Demography. Journal of</p><p>Economic Psychology, 27(4), 463-482.</p><p>Easterlin, R.A. (2010). Happiness, growth, and the life cycle. IZA</p><p>Prize in Labor Economics Series. Oxford: Oxford University Press.</p><p>Easterlin, R.A., Angelescu McVey, L., Switek, M., Sawangfa,</p><p>O., & Smith Zweig, J. (2010). The happiness-income paradox</p><p>revisited. PNAS Early Edition, 107(52), 22463-22468. doi:10.1073/</p><p>pnas.1015962107</p><p>Eisenberg, P., & Lazarsfeld, P. (1938). The psychological effects of</p><p>unemployment. Psychological Bulletin, 35, 358-390.</p><p>Ellison, C. G. (1991). Religious involvement and subjective well-</p><p>being. Journal of Health and Social Behavior, 32, 80-99.</p><p>65409_Earth_Reference.indd 100 4/30/12 12:08 PM</p><p>101</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>ESS Round 4: European Social Survey Round 4 Data. (2008).</p><p>Data file edition 4.0. Norwegian Social Science Data Services,</p><p>Norway - Data Archive and distributor of ESS data. Retrieved</p><p>from http://ess.nsd.uib.no/downloadwizard/</p><p>EVS. (2011). European values study 1981-2008, Longitudinal data</p><p>file (ZA4804 Data File Version 1.0.0, 2011-04-30) [Integrated_</p><p>values_surveys_1981-2008.sav]. GESIS Data Archive, Cologne,</p><p>Germany. doi:10.4232/1.1480</p><p>Ferrer-i-Carbonell, P., & Frijters, P. (2004). How important</p><p>is methodology for the estimates of the determinants of</p><p>happiness? Economic Journal, 114(497), 641-659. doi:10.1111/</p><p>j.1468-0297.2004.00235.x</p><p>Fowler, J.H., & Christakis, N. A. (2008) Dynamic spread of hap-</p><p>piness in a large social network: longitudinal analysis over 20</p><p>years in the Framingham Heart Study. British Medical Journal.</p><p>doi:10.1136/bmj.a2338</p><p>Frey, B. S. (1997). Not just for the money: an economic theory of per-</p><p>sonal motivation. Massachusetts and Cheltenham: Edward Elgar.</p><p>Frey, B. S. (2008). Happiness: A revolution in economics. Boston:</p><p>MIT Press.</p><p>Frey, B. S., & Benz, M. (2008). Being Independent is a Great</p><p>Thing: Subjective Evaluations of Self-Employment and Hierar-</p><p>chy. Economica, 75, 362–383.</p><p>Frey, B. S., Benz, M., & Stutzer, A. (2004). Introducing</p><p>procedural utility: Not only what but also how matters. Journal</p><p>of Institutional and Theoretical Economics, 160(3), 377-401.</p><p>Retrieved from http://www.ingentaconnect.com/content/mohr/</p><p>jite/2004/00000160/00000003/art00002</p><p>Frey, B. S. & Stutzer, A. (2002). Happiness and Economics. Princ-</p><p>eton and Oxford: Princeton University Press.</p><p>Frey, B. S., & Stutzer, A. (2002a). The Economics of Happiness,</p><p>World Economics, 3(1), 1-17.</p><p>Frey, B. S., & Stutzer, A. (2002b). What Can We Learn from</p><p>Happiness Research?. Journal of Economic Literature, 40(2), 402-</p><p>435. Retrieved from http://www.jstor.org/stable/2698383</p><p>Frijters, P., & Beatton, T. (2011). The mystery of the U-shaped</p><p>relationship between happiness and age. National Center for</p><p>Econometric Research Working Paper No. 26. Retrieved from</p><p>http://www.ncer.edu.au/papers/documents/NCER_WpNo-</p><p>26Jul08v2.pdf</p><p>Frijters, P., & Van Praag, B.M.S. (1998). The Effects of Climate</p><p>on Welfare and Well-Being in Russia. Climatic Change, 39, 61-81.</p><p>Fujita, F., & Diener, E. (2005). Life satisfaction and set point:</p><p>Stability and change. Journal of Personality and Social Psychology,</p><p>88(1), 158-164. doi:10.1037/0022-3514.88.1.158</p><p>Fujiwara, D., & Campbell, R. (2011). Valuation techniques for</p><p>social cost-benefit analysis: Stated preference, revealed preference and</p><p>subjective well-being approaches, London: Departmental Working</p><p>Paper of HM Treasury and Department of Work and Pensions.</p><p>Retrieved from http://www.hm-treasury.gov.uk/d/green_book_</p><p>valuationtechniques_250711.pdf</p><p>Gardner, J., & Oswald, A. J. (2002). Is it money or marriage that</p><p>keeps people alive? Retrieved from http://www2.warwick.ac.uk/</p><p>fac/soc/economics/staff/academic/oswald/augmortalitygardner-</p><p>os2002final.pdf</p><p>Gardner, J., & Oswald, A. J. (2006). Do divorcing couples</p><p>become happier by breaking up? Journal of the Royal Statistical</p><p>Society A, 169, 319-336.</p><p>Gerstorf, D., Ram, N., Estabrook, R., Schupp, J., Wagner, G., &</p><p>Lindenberger, U. (2008). Terminal Decline in Life Satisfaction</p><p>in Old Age - Longitudinal Evidence from the German Socio-Eco-</p><p>nomic Panel Study. Developmental Psychology, 44(4), 1148-1159.</p><p>Gerstorf, D., Ram, N., Mayraz, G., Hidajat, M., Lindenberger,</p><p>U., Wagner, G., & Schupp, J. (2010). Late-Life Decline in</p><p>Well-Being Across Adulthood in Germany, the UK and the US:</p><p>Something is Seriously Wrong at the End of Life. Psychology and</p><p>Aging, 25(2), 477-485.</p><p>Gilbert, D. (2006). Stumbling on happiness. New York: Random</p><p>House.</p><p>Gove, W., Hughes, M., & Briggs Style, C. (1983). Does marriage</p><p>have positive effects on the psychological well-being of the indi-</p><p>vidual?. Journal of Health and Social Behavior, 24(2), 122–131.</p><p>Graham, C. L. (2011). The pursuit of happiness: An economy of</p><p>well-being. Washington: Brookings Institution.</p><p>Graham, C., & Chattopadhyay, S. (2011). Gender and Well-Being</p><p>around the World: Some Insights from the Economics of Happi-</p><p>ness. Mimeo, The Brookings Institution.</p><p>Graham, C., & Pettinato, S. (2002). Happiness and Hardship.</p><p>Opportunity and Insecurity in New Market Economies. Washington</p><p>D.C.: Brookings Institution Press.</p><p>Green, F. (2011). Unpacking the misery multiplier: how employ-</p><p>ability modifies the impacts of unemployment and job inse-</p><p>curity on life satisfaction and mental health. Journal of Health</p><p>Economics, 30(2), 265-276.</p><p>Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L.</p><p>Morgan (Eds.), Syntax and Semantics, Vol. 3, Speech Acts (pp. 41-</p><p>58), New York: Academic Press.</p><p>Grosfeld I., & Senik, C. (2010). The Emerging Aversion to In-</p><p>equality. Evidence from subjective data. Economics of Transition,</p><p>18(1), 1-26.</p><p>Grosjean, P. (2011). A History of Violence: The Culture of Honor</p><p>as a Determinant of Homicide in the US South. Retrieved from</p><p>http://economics.stanford.edu/files/GrosjeanApril.pdf</p><p>65409_Earth_Reference.indd 101 4/30/12 12:08 PM</p><p>102</p><p>Grosjean, P., Ricka, F., & Senik, C. (2011). Learning, Political At-</p><p>titudes and the Crisis in Transition countries. EBRD Working Pa-</p><p>per. Retrieved from http://www.ebrd.com/downloads/research/</p><p>economics/workingpapers/wp0140.pdf</p><p>Grün, C., Hauser, W., & Rhein, T. (2010). Is any job better than</p><p>no job? Life satisfaction and re-employment. Journal</p><p>of Labour</p><p>Research, 31, 285-306. doi:10.1007/s12122-010-9093-2</p><p>Guven, C., Stichnoth, H., & Senik, C. (2012). You can't be happier</p><p>than your wife. Happiness Gaps and Divorce. Journal of Economic</p><p>Behavior and Organization, 82(1), 110-130. Retrieved from http://</p><p>dx.doi.org/10.1016/j.jebo.2012.01.003</p><p>GWP. (2011). Gallup world poll 2005-2011 [the_gallup_071511.dta].</p><p>Gallup Inc.</p><p>Gyarmati, D., de Raaf, S., Palameta, B., Nicholson, C. & Hui, T.</p><p>(2008). Encouraging work and supporting communities: Final results</p><p>of the community employment innovation project. Ottawa: Social</p><p>Research and Demonstration Corporation. Retrieved from http://</p><p>www.srdc.org/uploads/CEIP_finalrpt_ENG.pdf</p><p>Haller, M., & Hadler, M. (2006). How social relations and struc-</p><p>tures can produce happiness and unhappiness: An international</p><p>comparative analysis. Social Indicators Research, 75, 169-216.</p><p>Halpern, D. (2010). The Hidden Wealth of Nations. Cambridge:</p><p>Polity Press.</p><p>Harbaugh, W.T., Mayr, U., & Burghart, D.R. (2007). Neural</p><p>Responses to Taxation and Voluntary Giving Reveal Motives</p><p>for Charitable Donations. Science, 306, 1622-1625. doi:10.1126/</p><p>science.1140738</p><p>Hartog, J., & Oosterbeek, H. (1998). Health, Wealth and Happi-</p><p>ness: Why Pursue a Higher Education? Economics of Education</p><p>Review, 17(3), 245-256. Retrieved from http://upi-yptk.ac.id/Eko-</p><p>nomi/Hartog_Health.pdf</p><p>Haslam, C., Haslam, S. A., Jetten, J., Bevins, A., Ravenscroft, S.,</p><p>& Tonks, J. (2010). The social treatment: The benefits of group</p><p>interventions in residential care settings. Psychology and Aging,</p><p>25(1), 157-167. doi:10.1037/a0018256</p><p>Haslam, C., Holme, A., Haslam, S. A., Iyer, A., Jetten, J., &</p><p>Williams, W. H. (2008). Maintaining group memberships:</p><p>Social identity continuation predicts well-being after a</p><p>stroke. Neuropsychological Rehabilitation, 18(5-6), 671-691.</p><p>doi:10.1080/09602010701643449</p><p>Hayo, B., & Seifert, W. (2003). Subjective economic well-being in</p><p>Eastern Europe. Journal of Economic Psychology, 24(3), 329-348.</p><p>Helliwell, J.F. (2003). How's Life? Combining Individual and</p><p>National Variables to Explain Subjective Well-Being. Economic</p><p>Modelling, 20, 331-360.</p><p>Helliwell, J. F. (2008). Life satisfaction and quality of</p><p>development. NBER Working Paper No. 14507. Cambridge:</p><p>National Bureau of Economic Research. Retrieved from http://</p><p>www.nber.org/papers/w14507</p><p>Helliwell, J. F. (2011a). How can subjective well-being be</p><p>improved? In F. Gorbet & A. Sharpe (Eds.), New directions for</p><p>intelligent government in Canada (pp. 283-304). Ottawa: Centre for</p><p>the Study of Living Standards. Retrieved from http://www.csls.</p><p>ca/festschrift/Helliwell.pdf</p><p>Helliwell, J. F. (2011b). Institutions as enablers of well-being: The</p><p>Singapore prison case study. International Journal of Wellbeing,</p><p>1(2), 255-65. doi:10.5502/ijw.v1i2.7</p><p>Helliwell, J. F., & Barrington-Leigh, C.P. (2008). Empathy and</p><p>emulation: Life satisfaction and the urban geography of com-</p><p>parison groups. NBER Working Paper No. 14593.Retrieved from</p><p>http://www.nber.org/papers/w14593.pdf</p><p>Helliwell, J. F., & Barrington-Leigh, C. P. (2010). Measuring</p><p>and understanding subjective well-being. Canadian Journal of</p><p>Economics, 43(3), 729-753. doi: 10.1111/j.1540-5982.2010.01592.x</p><p>Helliwell, J. F., & Barrington-Leigh, C. P. (2011). How much is</p><p>social capital worth? In J. Jetten, C. Haslam, & S. A. Haslam</p><p>(Eds.), The social cure (pp. 55-71). London: Psychology Press.</p><p>Helliwell, J. F., Barrington-Leigh, C. P., Harris, A., & Huang, H.</p><p>(2010). International evidence on the social context of well-being.</p><p>In E. Diener, J. F. Helliwell, & D. Kahneman (Eds.), International</p><p>differences in well-being (Chapter 10, pp. 291-350). New York:</p><p>Oxford University Press.</p><p>Helliwell, J. F., & Huang, H. (2008). How’s your government?</p><p>International evidence linking good government and well-being.</p><p>British Journal of Political Science, 38(4), 595-619. doi:10.1017/</p><p>S0007123408000306</p><p>Helliwell, J. F., & Huang, H. (2010). How’s the job? Well-being</p><p>and social capital in the workplace. Industrial and Labor Relations</p><p>Review 63(2), 205-228. Retrieved from http://digitalcommons.ilr.</p><p>cornell.edu/cgi/viewcontent.cgi?article=1513&context=ilrreview</p><p>Helliwell, J. F., & Huang, H. (2011a). New measures of the costs</p><p>of unemployment: Evidence from the subjective well-being of 2.3</p><p>million Americans. NBER Working Paper No. 16829. Cambridge:</p><p>National Bureau of Economic Research. Retrieved from http://</p><p>www.nber.org/papers/w16829</p><p>Helliwell, J. F., & Huang, H. (2011b). Well-being and trust</p><p>in the workplace. Journal of Happiness Studies, 12(5), 747-767.</p><p>doi:10.1007/s10902-010-9225-7</p><p>Helliwell, J. F., & Putnam, R. D. (2004). The Social context</p><p>of well-being. Philosophical Transactions of the Royal Society B,</p><p>359(1449), 1435-1446. doi:10.1098/rstb.2004.1522</p><p>65409_Earth_Reference.indd 102 4/30/12 12:08 PM</p><p>103</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Helliwell, J. F., & Wang, S. (2011a). Trust and well-being.</p><p>International Journal of Wellbeing, 1(1), 42-78. Retrieved from</p><p>www.internationaljournalofwellbeing.org/index.php/ijow/</p><p>article/view/3/85</p><p>Helliwell, J. F. & Wang, S. (2011b). Weekends and subjective</p><p>well-being. NBER Working Paper No. 17180. Cambridge: National</p><p>Bureau of Economic Research. Retrieved from http://www.nber.</p><p>org/papers/w17180.pdf</p><p>Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest</p><p>people in the world? Behavioral and Brain Sciences, 33(2-3), 1-22.</p><p>doi:10.1017/S0140525X0999152X</p><p>Heston, A., Summers, R., and Aten, B. (2011). Penn World Table</p><p>Version 7.0. Center for International Comparisons of Production,</p><p>Income and Prices at the University of Pennsylvania. Retrieved from</p><p>http://pwt.econ.upenn.edu/php_site/pwt70/pwt70_form.php</p><p>Hirschman, A. O., & Rothschild, M. (1973). The Changing Toler-</p><p>ance for Income Inequality in the Course of Economic Develop-</p><p>ment. The Quarterly Journal of Economics, 87(4), 544-566.</p><p>Hölzel, B.K., Carmody, J., Vangel, M., Congleton, C., Yer-</p><p>ramsetti, S.M., Gard, T., & Lazar, S.W. (2011). Mindfulness</p><p>practice leads to increases in regional brain gray matter density,</p><p>Psychiatry Research: Neuroimaging, 191, 36–43.</p><p>Hu, Y., & Goldman, N. (1990). Mortality differentials by marital</p><p>status: An international comparison. Demography, 27 (2),</p><p>233–250.</p><p>Huppert, F. (2006). Positive emotions and cognition: devel-</p><p>opmental, neuroscience and health perspectives. In J. Forgas</p><p>(Ed.), Hearts and minds: Affective influences on social cognition and</p><p>behavior. Philadelphia: Psychology Press.</p><p>Jackson, P., & Warr, P. (1987). Mental health of unemployed</p><p>men in different parts of England and Wales. British Medical</p><p>Journal, 295, 525.</p><p>Jahoda, M. (1981). Work, employment, and unemployment:</p><p>Values, theories, and approaches in social research. American</p><p>Psychologist, 36, 184-191.</p><p>Jahoda, M. (1988). Economic recession and mental health:</p><p>Some conceptual issues. Journal of Social Issues, 44, 13-23.</p><p>Kahneman, D. (1999). Objective happiness. In D. Kahneman,</p><p>E. Diener, & N. Schwarz (Eds.), Well-Being: The foundations of</p><p>hedonic psychology (pp. 3-25). New York: Russell Sage.</p><p>Kahneman, D. (2011). Thinking, Fast and Slow. Toronto:</p><p>Doubleday.</p><p>Kahneman, D., & Deaton, A. (2010). High income improves</p><p>evaluation of life but not emotional well-being. Proceedings of</p><p>the National Academy of Sciences of the United States of America,</p><p>107(38), 16489-16493. doi:10.1073/pnas.1011492107</p><p>Kahneman, D., Diener, E., & Schwarz, N. (1999). Well-being: The</p><p>foundations of hedonic psychology. New York: Russell Sage.</p><p>Kahneman, D., & Krueger, A. B. (2006). Developments in the</p><p>measurement of subjective well-being. Journal of Economic</p><p>Perspectives, 20(1), 3-24. doi:10.1257/089533006776526030</p><p>Kahneman, D., Krueger, A., Schkade, D., Schwarz, N., & Stone,</p><p>A. (2004). A survey method for characterizing daily life experi-</p><p>ence: The day reconstruction method (DRM). Science, 306, 1776-</p><p>1780. doi:10.1126/science.1103572</p><p>Kahneman, D., Wakker, P. P., & Sarin, R. (1997). Back to</p><p>Bentham?</p><p>Exploration of experienced utility. Quarterly Journal of</p><p>Economics, 112(2), 375-406. doi:10.1162/003355397555235</p><p>Kenny, A., & Kenny, C. (2006). Life, liberty and the pursuit of</p><p>utility. Charlottesville: Imprint Academic.</p><p>Kim-Cohen, J., Caspi, A., Moffitt, T. E., Harrington, H., Milne,</p><p>B.J., & Poulton, P.J. (2003). Prior juvenile diagnoses in adults:</p><p>Developmental follow back of a prospective longitudinal cohort.</p><p>Archives of General Psychiatry, 60, 709-717.</p><p>Kingdon, G. & Knight. J. (2006). Subjective well-being poverty</p><p>versus income poverty and capabilities poverty. Journal of Devel-</p><p>opment Studies, 42(7), 1199-225.</p><p>Knack, S. (2001). Trust, associational life and economic</p><p>performance. In J. Helliwell and A. Bonikowska (Eds.), The</p><p>Contribution of Human and Social Capital to Sustained Economic</p><p>Growth and Well-Being. Ottawa: HRDC and OECD.</p><p>Knight, J., & Gunatilaka, R. (2010). Great expectations? The</p><p>subjective well-being of rural-urban migrants to China. World</p><p>Development, 38(1), 113-124.</p><p>Knight, J., Song, L., & Gunatilaka, R. (2010). The determinants</p><p>of subjective well-being in China. China Economic Review, 20(4),</p><p>635-49.</p><p>Koivumaa-Honkanen, H., Honkanen, R., Viinamäki, H.,</p><p>Heikkilä, K., Kaprio J., & Koskenvuo, M. (2000). Self-reported</p><p>life satisfaction and 20-year mortality in healthy Finnish adults.</p><p>American Journal of Epidemiology, 152(10), 983-991. doi:10.1093/</p><p>aje/152.10.983</p><p>Koivumaa-Honkanen, H., Honkanen, R., Viinamaeki, H., Heik-</p><p>kilae, K., Kaprio, J., & Koskenvuo, M. (2001). Life satisfaction</p><p>and suicide: A 20 year follow-up study. American Journal of</p><p>Psychiatry, 158, 433-439.</p><p>Konow, J., & Earley, J. (2008). The Hedonistic Paradox: Is Homo</p><p>Economicus Happier? Journal of Public Economics, 92, 1-33.</p><p>Krueger, A. B. (Ed.). (2009). Measuring the subjective well-being</p><p>of nations: National accounts of time use and well-being. Chicago:</p><p>University of Chicago Press.</p><p>65409_Earth_Reference.indd 103 4/30/12 12:08 PM</p><p>104</p><p>Krueger, A. B., Kahneman, D., Schkade, D., Schwarz, N., &</p><p>Stone, A. A. (2009). National time accounting: The currency of</p><p>life. In A. B. Krueger (Ed.), Measuring the subjective well-being of</p><p>nations: National accounts of time use and well-being (pp. 9-86).</p><p>Cambridge: National Bureau of Economic Research. Retrieved</p><p>from http://www.nber.org/chapters/c5053.pdf</p><p>Layard, R. (2005). Happiness: Lessons from a new science. New</p><p>York: Penguin.</p><p>Layard, R. (2010). Measuring subjective well-being. Science,</p><p>327(5965), 534-535. doi:10.1126/science.1186315</p><p>Layard, R. (2012). Mental Health: The New Frontier for the Welfare</p><p>State. Lecture, 6 March 2012. Retrieved from http://cep.ac.uk/</p><p>layard</p><p>Layard, R., & Dunn, J. (2009). A Good Childhood: Searching for</p><p>Values in a Competitive Age. London: Penguin.</p><p>Layard, R; Mayraz, G., & Nickell, S. (2010). Does Relative Income</p><p>Matter? Are the Critics Right? In E. Diener, J. Helliwell & D.</p><p>Kahneman (Eds.), International Differences in Well-Being. New</p><p>York: Oxford University Press.</p><p>Layard, R., Nickell, S., & Jackman, R. (2005). Unemployment:</p><p>Macroeconomic Performance and the Labour Market. (Second edi-</p><p>tion). Oxford: Oxford University Press.</p><p>Layard, R., Nickell, S., & Mayraz, G. (2008). The marginal utility</p><p>of income. Journal of Public Economics, 92(8–9), 1846–57.</p><p>Leong, L. (2010). Story of the Singapore prison service: From</p><p>custodians of prisoners to captains of life. Singapore: Civil Service</p><p>College.</p><p>Li, H., Pak Wai, L., Ye, M., & Zhang, J. (2011). Does Money Buy</p><p>Happiness? Evidence from Twins in Urban China. Working</p><p>Paper (under review).</p><p>Lim, C., & Putnam, R.D. (2010). Religion, social networks, and</p><p>life satisfaction. American Sociological Review, 75(6), 914-943.</p><p>Lucas, R. E. (2007). Long-term disability is associated with lasting</p><p>changes in subjective well-being: Evidence from two nationally</p><p>representative longitudinal studies. Journal of Personality and</p><p>Social Psychology, 92(4), 717-730. doi:10.1037/0022-3514.92.4.717</p><p>Lucas, R., & Schimmack, U. (2006). Marriage matters: Spousal</p><p>similarity in life satisfaction. Journal of Applied Social Science</p><p>Studies, 127(1), 105–111.</p><p>Luechinger, S. (2009). Valuing Air Quality Using the Life Satis-</p><p>faction Approach. Economic Journal, 119, 482–515.</p><p>Luechinger, S., Meier, S., & Stutzer, A. (2009). Why does unem-</p><p>ployment hurt the employed? The Journal of Human Resources,</p><p>45(4), 998-1045.</p><p>Luttmer, E. (2005). Neighbors as Negatives: Relative Earnings</p><p>and Well-Being. Quarterly Journal of Economics, 120, 963-1002.</p><p>Lykken, D. (1999). Happiness: The nature and nurture of joy and</p><p>contentment. New York: St. Martin’s Griffin.</p><p>Lyubomirsky, S. (2007). The How of Happiness: A practical guide to</p><p>getting the life you want. London: Sphere.</p><p>Lyubomirsky, S., King, L., & Diener, E. (2005). The benefits</p><p>of frequent positive affect: Does happiness lead to success?</p><p>Psychological Bulletin, 131(6), 803-855. doi:10.1037/0033-</p><p>2909.131.6.803</p><p>Margolis, R., & Myrskylä, M. (2011). A Global Perspective on</p><p>Happiness and Fertility. Population and Development Review, 37(1),</p><p>29–56.</p><p>Maslow, A. H. (1943). A theory of human motivation.</p><p>Psychological Review, 50(4), 370-396. doi:10.1037/h0054346</p><p>Matsuoka, R.H. (2010). Student performance and high school</p><p>landscapes: Examining the links. Landscape and Urban Planning,</p><p>97, 273-82.</p><p>Mazar, N., Amir, O., & Ariely, D. (2008). The dishonesty of</p><p>honest people: A theory of self-concept maintenance. Journal of</p><p>Marketing Research, 45(6), 633-644. doi:10.1509/jmkr.45.6.633</p><p>McKee, L., & Bell, C. (1986). His unemployment, her problem:</p><p>the domestic and marital consequences of male unemployment.</p><p>In S. Allen, S. Waton, K. Purcell, & S. Wood (Eds.), The experience</p><p>of unemployment. Basingstoke: Macmillan.</p><p>Medrano, J. D. (2012). Map of happiness. Retrieved from http://</p><p>www.jdsurvey.net/jds/jdsurveyMaps.jsp?Idioma=I&SeccionTexto</p><p>=0404&NOID=103</p><p>Meier, S. & Stutzer, A. (2008). Is volunteering rewarding in</p><p>itself? Economica, 75(1), 39-59.</p><p>Møller, V., Huschka, D., & Michalos, A. C. (Eds.). (2008).</p><p>Barometers of quality of life around the globe: How are we doing?</p><p>(Social Indicators Research Series, Vol. 33). Heidelberg,</p><p>Germany: Springer.</p><p>Moussavi, S., Chatterji, S., Verdes, E., Tandon, A., Patel, V., &</p><p>Ustun, B. (2007). Depression, chronic diseases, and decrements</p><p>in health: results from the World Health Surveys. The Lancet,</p><p>370(9590), 851-858.</p><p>Nguyen, A-M., & Benet-Martinez, V.(2012). Biculturalism and</p><p>Adjustment. A meta-analysis. Journal of Cross-Cultural Psychology</p><p>20(10), 1-38. doi:10.1177/0022022111435097</p><p>Nickerson, R. S. (1998). Confirmation bias: A ubiquitous</p><p>phenomenon in many guises. Review of General Psychology, 2(2),</p><p>175-220. doi:10.1037/1089-2680.2.2.175</p><p>65409_Earth_Reference.indd 104 4/30/12 12:08 PM</p><p>105</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Nisbet, E.K., & Zelenski, J.M. (2011). Underestimating nearby</p><p>nature: Affective forecasting errors obscure the happy path to</p><p>sustainability. Psychological Science, 22(9), 1101-6.</p><p>Nisbett, R., & Cohen, D. (1996). Culture of Honor: The Psychology</p><p>of Violence in the South. Colorado: Westview Press, Inc.</p><p>Nolen-Hoeksema, S., & Rusting, C. L. (1999). Gender differences</p><p>in well-being. In D. Kahneman, E. Diener, & N. Schwarz (Eds.),</p><p>Well-Being: The Foundations of Hedonic Psychology. Russell Sage</p><p>Foundation, New York.</p><p>Nunn, N., & Wantchekon, L. (2011). The Slave Trade and the</p><p>Origins of Mistrust in Africa. American Economic Review, 101(7),</p><p>3221-3252.</p><p>Nussbaum, M., & Sen, A. (Eds.). (1999). The quality of life. Oxford:</p><p>Oxford University Press.</p><p>OECD. (2000). The Partial Renaissance of Self-Employment.</p><p>OECD Employment Outlook. Paris: Organization for Economic</p><p>Cooperation and Development.</p><p>OECD. (2001). The well-being of nations: The role of human and</p><p>social capital. Paris: Organization for Economic Cooperation and</p><p>Development.</p><p>OECD. (2008). Growing Unequal? Income Distribution and Poverty</p><p>in OECD Countries. Paris: Organization for Economic Cooperation</p><p>and Development.</p><p>OECD. (2011a). Divided we stand: Why inequality keeps rising, Paris:</p><p>OECD. doi:10.1787/9789264119536-en</p><p>OECD. (2011b). How’s life? Measuring well-being, Paris: OECD.</p><p>doi:10.1787/9789264121164-en</p><p>Oishi, S. (2010). Culture and well-being: Conceptual and</p><p>methodological issues. In E. Diener, J. F. Helliwell, & D.</p><p>Kahneman (Eds.), International differences in well-being (pp. 34-</p><p>69). New York: Oxford University Press. doi:10.1093/acprof:o</p><p>so/9780199732739.003.0003</p><p>Oishi, S., Diener, E., Lucas, R. E., & Suh, E. M. (2009). Cross-</p><p>cultural variations in predictors of life satisfaction: Perspectives</p><p>from needs and values. In E. Diener (Ed.), Culture and Well-</p><p>Being: The collected works of Ed Diener (Social Indicators Research</p><p>Series 38, pp. 109-127). Dordrecht Heidelberg London New York:</p><p>Springer. doi:10.1007/978-90-481-2352-0_6</p><p>Oreopoulos, P. (2007). Do Dropouts Drop Out Too Soon? Wealth,</p><p>Health, and Happiness from Compulsory Schooling. Journal of</p><p>Public Economics, 91(11-12), 2213-2229.</p><p>Oreopoulos, P., & Salvanes, K. (2011). Priceless: the Nonpecuni-</p><p>ary Benefits of Schooling. Journal of Economic Perspectives, 25(1),</p><p>159-184.</p><p>Oswald, A. J., Proto, E., & Sgroi, D. (2011). Happiness and Pro-</p><p>ductivity. Department of Economics, University of Warwick, mimeo.</p><p>Retrieved from http://www2.warwick.ac.uk/fac/soc/economics/</p><p>staff/academic/proto/workingpapers/protosgroioswalddivorcefi-</p><p>nalmay2011.pdf</p><p>Oswald, A. J., & Powdthavee, N. (2008). Does happiness adapt? A</p><p>longitudinal study of disability with implications for economists</p><p>and judges. Journal of Public Economics, 92, 1061-1077.</p><p>Palmore, E. (1969). Predicting longevity: A follow-up controlling</p><p>for age. Journal of Gerontology, 39, 109-116.</p><p>Pargament, K. I. (2002). The bitter and the sweet: An evaluation of the</p><p>costs and benefits of religiousness. Psychological Inquiry, 13, 168-181.</p><p>Piff, P. K., Stancato, D.M., Côté, S., Mendoza-Denton, R., &</p><p>Keltner, D. (2012). Higher social class predicts increased unethical</p><p>behavior. Proceedings of the National Academy of Sciences, 109(9).</p><p>doi:10.1073</p><p>Piketty, T. (2003). The Impact of divorce on school performance:</p><p>evidence from France, 1968-2002. CEPR discussion paper series</p><p>No 4146. Retrieved from http://piketty.pse.ens.fr/fichiers/Pik-</p><p>etty2003.pdf</p><p>Pink, D.H. (2009). Drive: The surprising truth about what motivates</p><p>us. New York: Riverhead Books.</p><p>Pinker, S. (2011). The Better Angels of Our Nature: The Decline of</p><p>violence in history and its causes. London: Allen Lane.</p><p>Pischke, J-S. (2011). Money and Happiness: Evidence from the</p><p>Industry Wage Structure. CEP Discussion Paper No 1051. Retrieved</p><p>from http://cep.lse.ac.uk/pubs/download/dp1051.pdf</p><p>Platt, S., & Kreitman, N. (1990). Long term trends in parasuicide</p><p>and unemployment in Edinburgh, 1968-87. Social Psychiatry and</p><p>Psychiatric Epidemiology, 25, 56-61.</p><p>Platt, S., Micciolo, R., & Tansella, M. (1992). Suicide and Unem-</p><p>ployment in Italy: Description, Analysis and Interpretation of</p><p>Recent Trends. Social Science & Medicine, 34, 1191-1201.</p><p>Plomin, R., DeFries, J.C., McClearn, G.E.m & McGuffin, P.</p><p>(2001). Behavioral Genetics. Fourth Edition. New York: Worth</p><p>Publishers.</p><p>Powdthavee, N. (2008). Putting a price tag on friends, relatives</p><p>and neighbours: using surveys of life satisfaction to value social</p><p>relationships. Journal of Socio-Economics, 37(4), 1459-1480.</p><p>Powdthavee, N. (2009). I can’t smile without you: Spousal cor-</p><p>relation in life satisfaction. Journal of Economic Psychology, 30(4),</p><p>675–689.</p><p>65409_Earth_Reference.indd 105 4/30/12 12:08 PM</p><p>106</p><p>Putnam, R. D. (2000). Bowling alone: The collapse and revival of</p><p>American community. New York: Simon & Schuster.</p><p>Redelmeier, D., & Kahneman, D. (1996). Patients’ memories</p><p>of painful medical treatments: Real-time and retrospective</p><p>evaluations of two minimally invasive procedures. Pain, 66(1),</p><p>3-8. doi:10.1016/0304-3959(96)02994-6</p><p>Richards, M., & Abbott, R. (2009). Childhood mental health and</p><p>life changes in post-war Britain. Insights from three national birth</p><p>cohort studies. Sainsbury Centre for Mental Health and other</p><p>sponsors. Retrieved from http://www.centreformentalhealth.org.</p><p>uk/pdfs/life_chances_report.pdf</p><p>Rio+20 United Nations Conference on Sustainable Development.</p><p>(2012). The Future We Want - Zero draft of the outcome document.</p><p>Robertson, I., & Cooper, C.L. (2011). Well-being: Productivity and</p><p>Happiness at Work. London: Palgrave Macmillan.</p><p>Roth, A., & Fonagy, P. (Eds.). (2005). What Works for Whom?</p><p>(Second edition). New York: Guilford Press.</p><p>Russell, J. A., & Carroll, J. M. (1999). On the bipolarity of</p><p>positive and negative affect. Psychological Bulletin, 125(1), 3-30.</p><p>doi:10.1037/0033-2909.125.1.3</p><p>Ryan, R. M., & Deci, E. L. (2001). On happiness and human</p><p>potentials: A review of research on hedonic and eudaimonic</p><p>well-being. Annual Review of Psychology, 52, 141-166. doi:10.1146/</p><p>annurev.psych.52.1.141</p><p>Sacks, D., Stevenson, B., & Wolfers, J. (2011). Growth in Income</p><p>and subjective well-being over time. Mimeo. Retrieved from</p><p>http://www.sole-jole.org/12513.pdf</p><p>Sahlberg, P. (2007). Education policies for raising student learn-</p><p>ing: The Finnish approach. Journal of Education Policy, 22(2),</p><p>147-171.</p><p>Sales, S.M., & House, J. (1971). Job dissatisfaction as a possible</p><p>risk factor in Coronary Heart Disease. Journal of Chronic Diseases,</p><p>23, 861-873.</p><p>Sampson, R., Raudenbush, S., & Earls, F. (1997). Neighborhoods</p><p>and violent crime: A multilevel study of collective efficacy.</p><p>Science, 277, 918–24.</p><p>Sarracino, F. (2010). Social capital and subjective well-being</p><p>trends: Comparing 11 western European countries. Journal of</p><p>Socio-Economics, 39, 482–517.</p><p>Schimmack, U. (2003). Affect measurement in experience</p><p>sampling research. Journal of Happiness Studies, 4(1), 79-106.</p><p>doi:10.1023/A:1023661322862</p><p>Schwartz, C. E., & Sendor, M. (1999). Helping others helps</p><p>oneself: Response shift effects in peer support. Social Science &</p><p>Medicine, 48(11), 1563-1575. doi:10.1016/S0277-9536(99)00049-0</p><p>Schwarze, J., & Härpfer, M. (2007). Are People Inequality</p><p>Averse, and Do They Prefer Redistribution by the State? Evidence</p><p>from German Longitudinal Data on Life Satisfaction. Journal of</p><p>Socio-Economics, 36(2), 233-249.</p><p>Senik, C. (2004). When Information Dominates Comparison.</p><p>Learning from Russian Subjective Panel Data. Journal of Public</p><p>Economics, 88, 2099-2133.</p><p>Shiffman, S., Stone, A. A., & Hufford, M. R. (2008). Ecological</p><p>momentary assessment. Annual Reviews of Clinical Psychology, 4,</p><p>1-32. doi:10.1146/annurev.clinpsy.3.022806.091415</p><p>Shultz, S., & Dunbar, R. I. M. (2007). The evolution of the social</p><p>brain: Anthropoid primates contrast with other vertebrates.</p><p>Proceedings of the Royal Society B, 274(1624), 2429-2436.</p><p>doi:10.1098/rspb.2007.0693</p><p>Singer, P. (1981). The Expanding Circle: Ethics and Sociobiology.</p><p>Oxford: Oxford University Press.</p><p>Singer, T., Seymour, B., O’Doherty, J., Kaube, H., Dolan, R., &</p><p>Frith, C. (2004). Empathy for pain involves the affective but not</p><p>sensory components of pain. Science, 303, 1157–62.</p><p>Smith, T. B., McCullough, M. E., & Poll, J. (2003). Religiousness</p><p>and depression: evidence for a main effect and the moderating in-</p><p>fluence of stressful life events. Psychological Bulletin, 129, 614-636.</p><p>Soroka, S., Helliwell, J.F., & Johnston, R. (2007). Measuring and</p><p>modeling interpersonal trust. In F.M. Kay & R. Johnston (Eds.), So-</p><p>cial Capital, Diversity and the Welfare State. Vancouver: UBC Press.</p><p>Statistical Office of Kosovo. (2011). Demographic, social</p><p>and reproductive health survey in Kosovo, November 2009.</p><p>Department of Population Statistics, Statistical Office of</p><p>Kosovo. Retrieved from http://esk.rks-gov.net/ENG/publikimet/</p><p>doc_download/967-demographic-social-and-reproductive-health-</p><p>survey-in-kosovo-</p><p>Steptoe, A., Wardle, J., Marmot, M., & McEwen, B. S. (2005).</p><p>Positive affect and health-related neuroendocrine, cardiovascular,</p><p>and inflammatory processes. Proceedings of the National Academy</p><p>of Sciences of the United States of America, 102(18), 6508-6512.</p><p>doi:10.1073/pnas.0409174102</p><p>Stevenson, B., & Wolfers, J. (2008a). Economic Growth and Sub-</p><p>jective Well-Being: Reassessing the Easterlin Paradox. Brookings</p><p>Papers on Economic Activity, Spring. Retrieved from www.gallup.</p><p>com/poll/File/116605/EasterlinParadox.pdf</p><p>Stevenson, B., & Wolfers, J. (2008b). Happiness Inequality in the</p><p>United States. Journal of Legal Studies, 37(S2): S33-S79.</p><p>Stevenson, B., & Wolfers, J. (2009). The Paradox of Declining</p><p>Female Happiness. American Economic Journal: Economic Policy,</p><p>1, 190–225.</p><p>65409_Earth_Reference.indd 106 4/30/12 12:08 PM</p><p>107</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Stevenson, B., & Wolfers, J. (2010). Inequality and Sub-</p><p>jective Well-Being, mimeo. Retrieved from https://</p><p>editorialexpress.com/cgi-bin/conference/download.cgi?db_</p><p>name=ALEA2010&paper_id=266</p><p>Stiglitz, J., Sen, A., & Fitoussi, J.-P. (2009). Report by the</p><p>commission on the measurement of economic and social progress.</p><p>Paris: OECD. Retrieved from http://www.stiglitz-sen-fitoussi.fr/</p><p>documents/rapport_anglais.pdf</p><p>Stone, A. A., Schiffman, S., Schwartz, J. E., Broderick, J. E.,</p><p>& Hufford, M. R. (2002). Patient non-compliance with paper</p><p>diaries. British Medical Journal, 324(7347), 1193-1194. doi:10.1136/</p><p>bmj.324.7347.1193</p><p>Stone, A. A., Schwartz, J. E., Broderick, J. E., & Deaton, A.</p><p>(2010). A snapshot of the age distribution of psychological well-</p><p>being in the United States. Proceedings of the National Academy</p><p>of Sciences of the United States of America, 107(22), 9985-9990.</p><p>doi:10.1073/pnas.1003744107.</p><p>Strack, F., Martin, L. L., & Schwarz, N. (1988). Priming and</p><p>communication: Social determinants of information use</p><p>in judgments of life satisfaction. European Journal of Social</p><p>Psychology, 18(5), 429-442. doi:10.1002/ejsp.2420180505</p><p>Stutzer, A., & Frey, B.S. (2006). Does Marriage Make People</p><p>Happy, Or Do Happy People Get Married?. Journal of Socio-</p><p>Economics, 35(2), 326-347.</p><p>Ulrich, R.S. (1984). View through a window may influence</p><p>recovery from surgery, Science, 224(4647), 420-1.</p><p>Uslaner, E.M. (2008). Corruption, Inequality, and the Rule of Law.</p><p>Cambridge: Cambridge University Press.</p><p>Van Praag, B.M.S. (1971) The welfare function of income in</p><p>Belgium: an empirical investigation. European Economic Review,</p><p>2(Spring), 337-69.</p><p>Van Praag, B.M.S., & Baarsma, B.E. (2005). Using Happiness</p><p>Surveys to Value Intangibles: the Case of Airport Noise. Eco-</p><p>nomic Journal, 115, 224-246.</p><p>Warr, P.B. (1999). Well-Being and the Workplace. In D. Kahne-</p><p>man, E. Diener, & N. Schwarz (Eds.), Well-Being: The Founda-</p><p>tions of Hedonic Psychology. New York: Russell Sage Foundation.</p><p>WHO. (2008). Global Burden of Disease: 2004 update. World</p><p>Health Organisation.</p><p>Wiest, M., Shüz, S., Webster, N., & Wurm, S. (2011). Subjective</p><p>well-being and mortality revisited: Differential effects of</p><p>cognitive and emotional facets of well-being on mortality.</p><p>Health Psychology, 30(6), 728-735. doi:10.1037/a0023839</p><p>Wilkinson, R., & Pickett, K. (2009). The Spirit Level: Why more</p><p>equal societies almost always do better. London: Allen Lane/</p><p>Penguin.</p><p>Williams, M., Teasdale, J., Segal., Z., & Kabat-Zinn, J. (2007).</p><p>The Mindful Way through Depression: Freeing Yourself from Chronic</p><p>Unhappiness. New York: Guilford.</p><p>Williams, T. M. (Ed.). (1986). A Natural Experiment in Three</p><p>Communities. New York: Academic Press. Summarised in The</p><p>impact of television: A longitudinal Canadian Study. In B.</p><p>Singer (Ed.), Communications in Canadian Society (1991). Scar-</p><p>borough, ON: Nelson Canada.</p><p>Winkelmann, L., & Winkelmann, R. (1998). Why Are the Un-</p><p>employed So Unhappy? Evidence from Panel Data. Economica,</p><p>65, 1-15.</p><p>Wirtz, D., Kruger, J., Scollon, C. N., & Diener, E. (2003). What</p><p>to do on spring break? The role of predicted, on-line, and</p><p>remembered experience in future choice. Psychological Science,</p><p>14(5), 520-524. doi:10.1111/1467-9280.03455</p><p>Wolfers, J. (2003). Is business cycle volatility costly? Evidence from</p><p>surveys of subjective well-being. International Finance, 6(1), 1-26.</p><p>World Development Indicators (WDI). (2011). World development</p><p>indicators [WDI_GDF_Dec102011.xslx]. Washington, World</p><p>Bank. Retrieved from http://databank.worldbank.org/databank/</p><p>download/WDIandGDF_excel.zip</p><p>Wulfgramm, M. (2011). Can activating labour market policy</p><p>offset the detrimental life satisfaction effect of unemployment?</p><p>Socio-Economic Review, 9, 477-501. doi:10.1093/ser/mwr006</p><p>Wunder, C., Wiencierz, A., Schwarze, J., Küchenhoff, H., Kleyer,</p><p>S., & Bleninger, P. (2009). Well-Being over the Life Span:</p><p>Semiparametric Evidence from British and German Longitudi-</p><p>nal Data. IZA Discussion Paper No 4155. Retrieved from http://</p><p>papers.ssrn.com/sol3/papers.cfm?abstract_id=1402502&http://</p><p>www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1</p><p>&cts=1331740279240&ved=0CCUQFjAA&url=http%3A%2F%2</p><p>Fpapers.ssrn.com%2Fsol3%2FDelivery.cfm%3Fabstractid%3D14</p><p>02502&ei=c75gT7m8NqHV0QHK0LCKBw&usg=AFQjCNFe7u</p><p>TlU46NZwCn_ti2xkGMjXjksQ</p><p>WVS. (2009). World value survey 1981-2008 official aggregate</p><p>(v.20090901, 2009) [wvs1981_2008_v20090914.sav]. World</p><p>Values Survey Association, Aggregate File Producer: ASEP/JDS,</p><p>Madrid. Retrieved from www.worldvaluessurvey.org</p><p>Zaki, J., & Mitchell, J.P. (2011). Equitable decision making is as-</p><p>sociated with neural markers of intrinsic value. Proceedings of the</p><p>National Academy of Sciences, 108(49), 19761-19766.</p><p>65409_Earth_Reference.indd 107 4/30/12 12:08 PM</p><p>108</p><p>Part II.</p><p>Case Study: Bhutan</p><p>GROSS NATIONAL HAPPINESS</p><p>AND THE GNH INDEX</p><p>KARMA URA, SABINA ALKIRE AND TSHOKI ZANGMO</p><p>Karma Ura: President, Centre for Bhutan Studies</p><p>Sabina Alkire: Director, Oxford Poverty and Human Development Initiative</p><p>Tshoki Zangmo: Centre for Bhutan Studies</p><p>The Centre for Bhutan Studies would like to thank the United Nations Development Programme, Bhutan; Joint Support</p><p>Programme (JSP) supported by DANIDA and UNDP-UNEP, and Royal Government of Bhutan for fi nancing the second GNH</p><p>Survey in 2010. I would also like to express my gratitude to the International Development Research Centre (IDRC), Canada</p><p>for funding the analysis of the survey data. A longer and more complete version of the 2010 GNH survey report will be</p><p>printed in May 2012 and it too will be funded by International Development Research Centre (IDRC), Canada.</p><p>65409_Earth_Chapter5v2.indd 108 4/30/12 3:57 PM</p><p>109</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Part II.</p><p>Case Study: Bhutan</p><p>GROSS NATIONAL HAPPINESS</p><p>AND THE GNH INDEX</p><p>KARMA URA, SABINA ALKIRE AND TSHOKI ZANGMO</p><p>Summary</p><p>Bhutan’s GNH Index is a multidimensional measure, and is linked with a set of policy and program screening</p><p>tools so that it has practical applications. The GNH Index is built from data drawn from periodic surveys that</p><p>are representative by district, gender, age, rural-urban residence, income, etc. Representative sampling allows its</p><p>results to be decomposed at various sub-national levels, and such disaggregated information can be examined</p><p>and understood more by organizations and citizens for their uses. In the GNH Index, unlike certain concepts of</p><p>happiness in current western literature, happiness is itself multidimensional – not measured only by subjective</p><p>well-being, and not focused narrowly on happiness that begins and ends with oneself and is concerned for and</p><p>with oneself. The pursuit of happiness is collective, though it can be experienced deeply personally. Different</p><p>people can be happy in spite of their disparate circumstances but the options for trade off must be wide.</p><p>The GNH Index is meant to orient the people and the nation towards happiness, primarily by improving the</p><p>conditions of not-yet-happy people. We can break apart the GNH Index to see where unhappiness is arising</p><p>from and for whom. For policy action, the GNH Index enables the government and others to increase GNH in</p><p>two ways. It can either increase the percentage of people who are happy or decrease the insufficient</p><p>conditions</p><p>of people who are not-yet-happy. In the way the GNH Index is constructed, there is a greater incentive for the</p><p>government and others to decrease the insufficiencies of not-yet-happy people. This can be done by mitigating</p><p>the many areas of insufficiencies the not-yet-happy face. Not-yet-happy people in rural Bhutan tend to be those</p><p>who attain less in education, living standards and balanced use of time. In urban Bhutan, not-yet-happy people</p><p>are insufficient in non-material domains such as community vitality and culture and psychological well-being.</p><p>In Thimphu, the capital, for example, the biggest insufficiencies are in community vitality.</p><p>The GNH Index provides an overview of performances across nine domains of GNH (psychological well-</p><p>being, time use, community vitality, cultural diversity, ecological resilience, living standard, health, education,</p><p>good governance). The aggregation method is a version of Alkire Foster method (2007, 2011). The index is</p><p>aggregated out of 33 clustered (grouped) indicators. Each clustered indicator is further composed of several</p><p>variables. When unpacked, the 33 clustered indicators have 124 variables, the basic building blocks of the</p><p>GNH Index. Weights attached to variables differ, with lighter weights attached to highly subjective variables.</p><p>A threshold or sufficiency level is applied to each variable. At the level of domains, all the nine domains are</p><p>equally weighted as they are all considered to be equally valid for happiness.</p><p>Three cutoff points have been established to identify degrees of happiness. Not all people need to be sufficient in</p><p>each of 124 variables to be happy. People are diverse in the ways and the means in which they can have a fulfill-</p><p>ing life. Not all variables need to be present for a person to be happy. People have freedom of choice in the ways</p><p>in which they can make life fulfilling, so not all variables have universal applicability. For such reason, we divide</p><p>the Bhutanese into four groups depending upon their degree of happiness. We use three cutoffs: 50%, 66%, and</p><p>77%. People who have achieved sufficiency in less than 50% are “unhappy,” and they comprise only 10.4% of the</p><p>population. A total of 48.7% of people have sufficiency in 50-65% of domains and are called “narrowly happy.”</p><p>A group of 32.6%, called “extensively happy,” has achieved sufficiency in 66-76% – in between six and seven</p><p>domains. And in the last group, 8.3% of people are identified as “deeply happy” because they enjoy sufficiency in</p><p>77% or more of weighted indicators – which is the equivalent of seven or more of the nine domains.</p><p>In order to have one overall index, the GNH cutoff was set at 66% of the variables, which is the middle cutoff</p><p>used above. People can be considered happy when they have sufficiency in 66% of the (weighted) indicators</p><p>or more – that is, when they were identified as extensively happy or deeply happy. The GNH Index value for</p><p>2010 is 0.737. It shows us that 40.8% of people in Bhutan have achieved such happiness, and the remaining</p><p>59% - who are narrowly happy or unhappy - still enjoy sufficiency in 57% (not 66% as required by the index)</p><p>of the domains on average. On the other hand, if we go by the subjective well-being score, it was 6.06 (SD =</p><p>1.6) for 2010 suggesting a very good level of happiness. The cutoff does make a difference in the GNH Index.</p><p>65409_Earth_Chapter5v2.indd 109 4/30/12 3:57 PM</p><p>110</p><p>The middle cutoff gives a relatively low score of GNH Index as a result of its requirement that a diverse set</p><p>of conditions and states, represented by 124 variables, must be simultaneously prevalent for a person to</p><p>be robustly happy. It is a tougher measure because it is not focused on survival like the poverty measure, but</p><p>rather on flourishing over a wide array of conditions. However the GNH Index, and the four categories of</p><p>people – unhappy, narrowly happy, extensively happy, and deeply happy – will be reported and analyzed when</p><p>the GNH Index is updated over time, as they are in this report. Taken together they will provide a nuanced</p><p>picture of the composition, diversity, and evolution of GNH across Bhutan.</p><p>Introduction</p><p>This guide 1 introduces the 2010 Gross National Happiness (GNH) Index of Bhutan. It explains the origins of</p><p>the concept of GNH, its grounding in Bhutanese culture and history, and describes how the concept is being</p><p>operationalized in the form of the GNH Index in some novel and innovative ways. Any discussion of GNH</p><p>in Bhutan must begin from the understanding that it is distinct from the western literature on “happiness”</p><p>in two ways. First it is multidimensional – not focused only on subjective well-being to the exclusion of other</p><p>dimensions – and second, it internalizes other-regarding motivations. While multidimensional measures of</p><p>the quality of life and well-being are increasingly discussed, Bhutan is innovative in constructing a multidi-</p><p>mensional measure, which is itself relevant for policy and is also directly associated with a linked set of policy</p><p>and program screening tools. This chapter presents the GNH Index, which provides an overview of national</p><p>GNH across nine domains, comprising of 33 clustered indicators, each one of which is composed of several</p><p>variables. When unpacked, the 33 clustered indicators have 124 variables.</p><p>The 2010 GNH survey from which the index is drawn evolved from a 2006 pre-pilot and a 2008 nationally</p><p>representative survey. In its present form it is nationally representative and also representative at the rural and</p><p>urban area and by districts or dzongkhags. In-depth sections on the domains and indicators cover the motiva-</p><p>tion behind the selection of each as well as the weights, cut-offs and results. The GNH Index identifies and ag-</p><p>gregates information on happiness drawing on a special adaptation of the Alkire-Foster method for measuring</p><p>multidimensional concepts such as poverty and well-being. This ensures that the national measure is rigorous,</p><p>and that it is intuitive and can be examined in many policy-relevant ways.</p><p>Overall, in 2010, 10.4% of people are “unhappy” according to the GNH index; 47.8% are “narrowly happy,”</p><p>32.6% are “extensively happy,” and 8.3% are “deeply happy.” These four groups correspond to people who have</p><p>achieved sufficiency in less than half, 50-65%, 66-76%, and more than 77% of domains. The 2010 GNH Index</p><p>uses the middle cutoff. Its value is 0.737, and shows that overall, 41% of Bhutanese are identified as happy</p><p>(meaning they are extensively or deeply happy), and the remaining 59% enjoy sufficiency in 57% of the do-</p><p>mains on average. Recall that 48.7% of these 59% are already narrowly happy, but because we wish to expand</p><p>GNH we consider them not-yet-happy for policy purposes. The low score of GNH is a result of the GNH Index,</p><p>which requires diverse conditions and states, represented by 124 variables, to be prevalent for a person to be</p><p>robustly happy. GNH Indices and their subcomponents are also reported for each of the 20 districts, by gender,</p><p>by rural-urban area, and, for illustrative purposes, by age and certain occupational categories.</p><p>Table 1 below presents the definition of each of the groups used in this analysis. It then gives the percentage</p><p>of the population who belong in each category in the 2010 GNH Index results. The final column provides the</p><p>average percentage of weighted indicators, or domains, in which people in each group, on average, enjoy suf-</p><p>ficiency.</p><p>65409_Earth_Chapter5v2.indd 110 4/30/12 3:57 PM</p><p>111</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>The analysis has two parts: first, the well-being of the people who have been identified as “happy” is examined,</p><p>to show the indicators in which they enjoy satisfaction. The in-depth analysis of who is happy according to</p><p>the 2010 GNH Index includes analysis at the district level, as well as by rural and urban categories, gender,</p><p>occupation, education and income-levels. Some individual examples are presented, to show</p><p>that the “happiest”</p><p>people vary by age, district, occupation, gender, and sufficiency profiles.</p><p>The second part focuses on how to increase happiness. For as well as helping us to understand better the</p><p>diverse kinds of happiness, the GNH Index was primarily devised to provide policy guidance to increase hap-</p><p>piness, particularly by focusing on the not-yet-happy people so that their situation can be improved. Hence a</p><p>second part of the analysis scrutinizes the domains in which not-yet-happy people lack sufficiency. As such the</p><p>“not-yet-happy” and the question “how can GNH be increased?” are key components of the section.</p><p>The GNH Index, like the philosophy of GNH that motivates it, is very much a living experiment, seeking to</p><p>convey more fully the color and texture of people’s lives than does the standard welfare measure of GNI per</p><p>capita, to enrich the dimensions and the methodology well beyond the Human Development Index, and to draw</p><p>together some innovative work from other initiatives seeking to measure human progress on a shared planet.</p><p>Origins of the Concept of GNH</p><p>Although the term “Gross National Happiness” was first coined by the Fourth King of Bhutan the concept has a</p><p>much longer resonance in the Kingdom of Bhutan. The 1729 legal code, which dates from the unification of Bhu-</p><p>tan, declared that “if the Government cannot create happiness (dekid) for its people, there is no purpose for the</p><p>Government to exist” (Ura 2010).2 In 1972, the Fourth King declared Gross National Happiness to be more im-</p><p>portant than Gross National Product (GNP), and from this time onward, the country oriented its national policy</p><p>and development plans towards Gross National Happiness (or GNH). The Constitution of Bhutan (2008, Article</p><p>9) directs the State “to promote those conditions that will enable the pursuit of Gross National Happiness.”</p><p>While there is no single official definition of GNH, the following description is widely used:</p><p>Gross National Happiness (GNH) measures the quality of a country in a more holistic way [than GNP]</p><p>and believes that the beneficial development of human society takes place when material and spiritual</p><p>development occurs side by side to complement and reinforce each other. 3</p><p>Table 1: Overview of GNH domains and breakdown of indicators</p><p>Definition of groups</p><p>~ Sufficiency in:</p><p>Percent of population</p><p>who are:</p><p>Average sufficiency</p><p>of each person across</p><p>domains</p><p>HAPPY 66%-100% 40.9% 72.9%</p><p>Deeply Happy 77%-100% 8.3% 81.5%</p><p>Extensively Happy 66%-76% 32.6% 70.7%</p><p>NOT-YET-HAPPY 0-65% 59.1% 56.6%</p><p>Narrowly Happy 50%-65% 48.7% 59.1%</p><p>Unhappy 0-49% 10.4% 44.7%</p><p>65409_Earth_Chapter5v2.indd 111 4/30/12 3:57 PM</p><p>112</p><p>From the start it is vital to clarify that GNH in Bhutan is distinct from the western literature on “happiness”</p><p>in two ways. First it is multidimensional – not focused only on subjective well-being to the exclusion of other</p><p>dimensions – and second, it internalizes responsibility and other-regarding motivations explicitly. As the first</p><p>elected Prime Minister of Bhutan under the new Constitution of Bhutan adopted in 2008 put it:</p><p>We have now clearly distinguished the ‘happiness’ … in GNH from the fleeting, pleasurable ‘feel good’</p><p>moods so often associated with that term. We know that true abiding happiness cannot exist while others</p><p>suffer, and comes only from serving others, living in harmony with nature, and realizing our innate</p><p>wisdom and the true and brilliant nature of our own minds.4</p><p>It includes harmony with nature (again absent from some Western notions of happiness) and concern for others.</p><p>The brilliant nature he alluded to consists of the various types of extraordinarily sensitive and advanced aware-</p><p>ness with which human beings are endowed and can be realized.</p><p>The nine domains articulate the elements of GNH more fully and form the basis of the GNH Index. The earlier</p><p>four pillars of GNH are included as part of the nine domains.5 The first three domains are very familiar from a</p><p>human development perspective – living standards (such as income, assets, housing), health, and education. The</p><p>next three are a bit newer – the use of time (and time poverty), good governance and ecological resilience. And</p><p>the last are the more innovative – psychological well-being (which includes overall happiness, but also emotions</p><p>and spirituality), community vitality and cultural diversity and resilience.</p><p>The index weights the nine domains equally. Thirty-three cluster indicators are used to identify people as poor and</p><p>create the index. For presentational simplicity they are also combined to produce nine domain-level indicators.</p><p>Each sub-component indicator of the GNH Index is on its own useful for practical purposes of different agencies.</p><p>Purpose of the 2010 GNH Index</p><p>Since the mid-2000s, steps have been taken to build a GNH Index that would draw as fully as possible on the</p><p>holistic and deliberate vision of development as it has evolved in Bhutan. In a 2007 Government Round Table,</p><p>Dasho Karma Ura proposed that a GNH Index would be used in: 1) Setting an alternative framework of devel-</p><p>opment; 2) Providing indicators to sectors to guide development; 3) Allocating resources in accordance with</p><p>Domain Indicators</p><p>1 Psychological well-being 4</p><p>2 Health 4</p><p>3 Time use 2</p><p>4 Education 4</p><p>5 Cultural diversity and resilience 4</p><p>6 Good Governance 4</p><p>7 Community vitality 4</p><p>8 Ecological diversity and resilience 4</p><p>9 Living standards 3</p><p>Total 33</p><p>Table 1: Overview of GNH domains and breakdown of indicators</p><p>65409_Earth_Chapter5v2.indd 112 4/30/12 3:57 PM</p><p>113</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>targets and GNH screening tools; 4) Measuring people’s happiness and well-being; 5) Measuring progress over</p><p>time; and 6) Comparing progress across the country.6 These purposes, each of which has specific implications</p><p>for measurement, are elaborated below.</p><p>1. Setting an alternative framework of development: Bhutan’s GNH vision of development is distinctively</p><p>holistic. The 10th plan explicitly seeks “to address a more meaningful purpose for development than just the</p><p>mere fulfillment of material satisfaction.”7 Hence the nine domains of GNH, taken together, reflect the pur-</p><p>pose of development. If certain dimensions contract, or are being crowded out by material progress, the GNH</p><p>Index must explicitly convey such information as the imbalances enter, in order to catalyze public deliberation</p><p>and if relevant, action.</p><p>2. Providing indicators to sectors to guide development: Certain indicators must either monitor activities by</p><p>the public sector or else change when sector priorities are realized. For example “electricity,” a component of</p><p>the GNH, is a priority in the 10th five year plan. Insofar as the GNH indicators monitor outputs, the GNH In-</p><p>dex provides incentives to ministries to deliver services, because their accomplishments will visibly contribute</p><p>to higher GNH the next time the index is updated. Methodologically this requires an index that can be broken</p><p>down into its component indicators.</p><p>3. Allocating resources in accordance with targets and GNH screening tools: While the composition of the</p><p>GNH is not a sufficient guide for policy, a clear understanding of how the achievements and shortfalls in dif-</p><p>ferent dimensions of GNH vary over time and space and group provides key information for policy design and</p><p>subsequent resource allocation. In terms of targeting, the GNH Index can show which dzongkhags are lacking</p><p>in which indicators, and can also identify and target the “least happy” people and describe them by age, district,</p><p>gender, etc. In terms of screening tools, the GNH indicators can be used as a check list, to convey in concrete</p><p>terms the kinds of activities and achievements that constitute GNH.</p><p>4. Measuring people’s happiness and well-being: The measure and its component indicators aim to capture</p><p>human well-being in a fuller and more profound way than traditional socio-economic</p><p>measures of economic</p><p>development, human development or social progress have done. This also requires the measurement meth-</p><p>odology to be understandable to the general public. Case studies can be provided of differently happy people,</p><p>so that citizens can assess whether the index broadly seems intuitive and has room for their own aspirations</p><p>and values.</p><p>5. Measuring progress over time: The component indicators of the GNH are to be sensitive to changes over</p><p>time. Some indicators must be directly responsive to relevant changes in policy. In this way, the composition</p><p>of well-being, as well as its overall level, can be observed over decades. Similarly, inequalities among groups</p><p>and populations that require special attention can be identified. The GNH Survey hence must be repeated</p><p>regularly, for example every two years.</p><p>6. Comparing progress across the country: The GNH Index should be able to make meaningful comparisons</p><p>across the dzongkhags, which vary widely in terms of climate, culture, access to services, and livelihoods. The</p><p>survey hence must be representative by dzongkhag, and the methodology of measurement must be subgroup</p><p>consistent and decomposable.</p><p>Taken together these six requirements have been used to specify the indicators and composition of the GNH</p><p>Index. It must be policy-sensitive – changing over time in response to public action. In key sectors, the indica-</p><p>tors must reflect public priorities directly. The GNH Index must also reflect the strengthening or deterioration</p><p>of social, cultural, and environmental achievements whether or not at present these are the direct objective of</p><p>policy. The indicators must be assumed to be relevant in future periods as well as at the present time in order</p><p>to measure progress across time. And the GNH Index must be sub-group consistent hence decomposable by</p><p>regions and groups.</p><p>65409_Earth_Chapter5v2.indd 113 4/30/12 3:57 PM</p><p>114</p><p>GNH Survey 2010</p><p>The GNH Index is based on a survey of 7,142 people that was completed in all 20 districts of Bhutan in the</p><p>year 2010 and is representative by rural and urban area and by districts or dzongkhags. The survey itself was</p><p>developed by the Centre for Bhutan Studies (CBS) and builds on previous surveys on GNH. The survey cov-</p><p>ers all nine domains and gives innovative insights into happiness that are not found in most other national</p><p>surveys. Indeed in fielding the GNH surveys, the CBS argues that the quality of the data is unusually high</p><p>and this is because the enumerators working often in remote rural areas took time with the participants to</p><p>explain the purpose of the index, to share the importance of understanding their own insight and perspectives</p><p>and so enabled the respondents to answer fully, completely and reflectively the questions on the survey. The</p><p>survey builds on a 2006 pre-pilot questionnaire and also on the 2008 GNH survey, which was representative</p><p>nationally but not by district. It repeated some of those questions, and learning from those experiences and the</p><p>analysis of that survey, also improved them.</p><p>In order to measure the nine domains of GNH, 33 indicators have been selected according to five different</p><p>criteria. First of all the indicators have to reflect the normative values of GNH, which have been articulated</p><p>in official documents such as the National Development Plan and in statements by His Majesty the King, the</p><p>Prime Minister and other ministers. They also reflect the normative values that are embedded in the culture</p><p>and traditions of Bhutan. The second criterion for the indicators relates to their statistical properties; each</p><p>indicator was analyzed extensively to ensure robustness. Third, the indicators were chosen such that they</p><p>would accurately reflect how happiness is increasing or evolving in different regions over time and among</p><p>different groups accurately. Fourth, the indicators had to be relevant for public action – although government</p><p>policy is by no means the only way of increasing GNH. Many domains of GNH can be facilitated by appropriate</p><p>government policies and by government policies that create incentives for businesses, NGOs and citizens to</p><p>support GNH in its many dimensions. And lastly, the indicators have to be understandable as far as possible</p><p>by ordinary citizens. They have to reflect and relate to people’s own experiences in their own lives, so that the</p><p>GNH Index would not only be a policy tool but would also be something that people could use to imagine the</p><p>many different ways of being happy in the Bhutanese context.</p><p>There are four indicators in every domain, except time use, which has two (sleep and work), and living standards,</p><p>which has three. Because the object of inquiry is happiness people will think the key questions are “How happy</p><p>am I? How can I be happier?” but actually these hedonic questions are not present in the index although they</p><p>were present in the survey and have been analyzed. The following section presents the indicators that have been</p><p>included in the index.</p><p>Domains and indicators</p><p>This section explains each of the nine domains and 33 indicators of the 2010 GNH Index 2010, how they have</p><p>been constructed as well as the cutoffs that have been set. The GNH Index uses two kinds of thresholds: suf-</p><p>ficiency thresholds, and one happiness threshold. Sufficiency thresholds show how much a person needs in</p><p>order to enjoy sufficiency in each of the 33 indicators. The overall happiness threshold meanwhile answers the</p><p>question “how many domains or in what percentage of the indicators must a person achieve sufficiency in</p><p>order to be understood as happy?” The happiness threshold will be presented later in this paper.</p><p>65409_Earth_Chapter5v2.indd 114 4/30/12 3:57 PM</p><p>115</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Psychological Well-being</p><p>Psychological well-being is an intrinsically valuable and desired state of being. Diener, et al (1997) categorize</p><p>indicators of psychological well-being according to refl ective or affective elements, while the Sarkozy Report8(Stiglitz,</p><p>Sen and Fitoussi, 2009a, p. 44) emphasizes the importance of using diverse well-being indicators. It states, “...</p><p>different aspects (cognitive evaluations of one’s life, happiness, satisfaction, positive emotions such as joy and</p><p>pride, and negative emotions such as pain and worry)…should be measured separately to derive a more com-</p><p>prehensive appreciation of people’s lives.” Besides the refl ective life evaluations and hedonic experiences,</p><p>an additional aspect of spirituality has also been included in the domain.</p><p>Life satisfaction</p><p>This indicator combines individuals’ subjective assessments of their contentment levels with respect to health,</p><p>occupation, family, standard of living and work-life balance.9 The respondents were asked to say how satisfi ed</p><p>or dissatisfi ed they were in these fi ve areas on a fi ve-point Likert scale (1= very satisfi ed, 5=very dissatisfi ed).</p><p>The life satisfaction indicator sums their responses across the fi ve areas. It could have a score as low as 5 (low</p><p>satisfaction) or as high as 25 (high satisfaction). The suffi ciency threshold for the life satisfaction score is set at</p><p>19 and 83% of people enjoy suffi ciency in life satisfaction.</p><p>33 GNH</p><p>Indicators</p><p>Psychological</p><p>Well-being</p><p>.�Life satisfaction</p><p>.�Positive emotions</p><p>.�Negative emotions</p><p>.�Spirituality</p><p>Health</p><p>.�Mental health</p><p>.�Self reported health</p><p>.�Healthy days</p><p>.�Disability</p><p>Time Use</p><p>.� Work</p><p>.� Sleep</p><p>Education</p><p>.�Literacy</p><p>.�Educational Level</p><p>.�Knowledge</p><p>.�Values</p><p>Cultural Diversity</p><p>and Resilience</p><p>.�Native Language</p><p>.�Cultural Participation</p><p>.�Artisan Skills</p><p>.� Conduct</p><p>Good Governance</p><p>.�Gov�t performance</p><p>.�Fundamental rights</p><p>.�Services</p><p>.�Political Participation</p><p>Community</p><p>Vitality</p><p>.�Donations (time &</p><p>money)</p><p>.�Community relationship</p><p>.�Family</p><p>.�Safety</p><p>Ecological</p><p>Diversity and</p><p>Resilience</p><p>.�Ecological Issues</p><p>.�Responsibility towards</p><p>environment</p><p>.�Wildlife damage (Rural)</p><p>.�Urbanization</p><p>issues</p><p>Living Standards</p><p>.�Assets</p><p>.�Housing</p><p>.�Household per capita</p><p>income</p><p>Figure 1: The nine domains and 33 indicators of the GNH</p><p>65409_Earth_Chapter5v2.indd 115 4/30/12 3:57 PM</p><p>116</p><p>Emotional balance (positive and negative emotions)</p><p>Ten self-reported emotional items were selected for this indicator. Positive emotions, or non-disturbing emotions,</p><p>such as compassion, generosity, forgiveness, contentment and calmness were included while selfishness, jealousy,</p><p>anger, fear and worry were used to represent negative emotions. In Buddhist perspective, the negative emotions</p><p>may be more accurately called disturbing emotions during which people cannot experience with much clarity</p><p>and that might lead often to formation of poor intentions. For both sets of emotions the respondents were asked</p><p>to rate the extent to which they had experienced them during the past few weeks with reference to a five-point scale. 10</p><p>The scale ranges are: 1 “never,” 2 “rarely,” 3 “sometimes,” 4 “never” and 5 “very much.”</p><p>Both the positive and negative emotion indicator scores run from 5 to 20 (from low to high incidence of positive</p><p>or negative emotions). For positive emotions, a sufficiency threshold of 15 was set which identifies 58.8%</p><p>as being adequate at positive emotions. The negative emotion indicator consists of two components of</p><p>sub-indices. The emotions included are selfishness and jealousy in one sub-index, and anger, fear and worry</p><p>in the other sub-index. A sufficiency threshold of 12 was applied for negative emotions, with about 64.6% of</p><p>the respondents deemed as not suffering from disturbing or negative emotions.</p><p>Spirituality</p><p>The spirituality indicator is based on four questions. They cover the person’s self-reported spirituality level,</p><p>the frequency with which they consider karma,11 engage in prayer recitation, and meditate. Self-reported</p><p>spirituality level describes the person’s judgment on his or her own position on the spirituality continuum. The</p><p>question of the consideration of karma asked people to what extent they take into account their own volitional</p><p>impulses and actions as having moral consequences in future just as they did in the present. Measures of social</p><p>engagements are dealt in both community vitality and time use domains. Here, indicators of sacred activities were</p><p>limited to praying and meditation as two separate events although these activities are not mutually exclusive.</p><p>All the four indicators run on a four-point scale of “regularly” to “not at all” except for the spirituality level,</p><p>which ranges from “very spiritual” to “not at all.”</p><p>The indicator sums the scores across the four questions. Scores range from 4 to 16 with 16 indicating a greater</p><p>degree of spirituality. The threshold has been set at 12, which implies that at least three of the four indicators</p><p>must be rated “regularly” or “occasionally” for individuals to be defined as happy. The indicator identifies 53%</p><p>of people as adequate in terms of spirituality level.</p><p>Health</p><p>In the indigenous healing science practiced as a branch of the official health system in Bhutan, health has</p><p>always been associated with both physical health and mental health. Health is the outcome of the relational</p><p>balance between mind and body, between persons and the environment. Typically, an individual is said to be</p><p>well only if both heat-pain is absent from the body and sorrow is absent from the mind. The social and material</p><p>conditions for creating good health such as clean air or water or nurturing family relationships or community</p><p>relationships have been incorporated in other domains. Similarly, emotional balance and spirituality have also</p><p>been included in the psychological well-being domain.</p><p>Self-reported health status</p><p>Questions persist about how accurately this simple self-reported indicator proxies objective health and nutrition</p><p>states, and the extent to which it is affected by “adaptive preferences” (Easterlin 2003). The self-reported health</p><p>indicator is used here as a proxy measure and to complement other health indicators (healthy days and disability)</p><p>and is consequently given only one-tenth of the total weight for health, and only one-third as much weight</p><p>as any of the other three indicators. The ratings range on a five-point scale from having “excellent” health to</p><p>“poor” health.</p><p>65409_Earth_Chapter5v2.indd 116 4/30/12 3:57 PM</p><p>117</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>For a person to be sufficient in self-reported health status, he or she must have a rating of “excellent” or “very</p><p>good.” A large majority (73.8%) has met the sufficiency condition in self-reported health.</p><p>Healthy days</p><p>This indicator reports the number of “healthy days” a respondent enjoyed within the last month. The mean</p><p>number of healthy days for Bhutan is 26 days (SD=7.7) and the median is 30 days.</p><p>To allow for normal illness and for elderly respondents, the threshold has been set at 26 days and 76.2% meet</p><p>the sufficiency threshold.</p><p>Long-term disability</p><p>This indicator examines an individual’s ability to perform functional activities of daily living without any</p><p>restriction (U.S. Department of Health and Human Services 2000). Participants were asked whether they had</p><p>any longstanding illness that had lasted over six months. If the answer was “yes,” they were then asked, using</p><p>a five-point scale, whether the disability restricted their daily activities. The scale ranged from “never” to “all</p><p>the time.” However, no further information on the intensity of disabilities was elicited.</p><p>The threshold is set such that those individuals who are disabled but are “rarely” or “never” restricted from doing</p><p>their daily chores are classified as sufficient. Conversely, individuals with a disability whose daily activities are</p><p>restricted “sometimes” are classified as deprived. With this threshold, about 89.5% achieve sufficiency.</p><p>Mental health</p><p>This indicator uses a version of the General Health Questionnaire (specifically GHQ-12) developed by</p><p>Goldberg. It consists of 12 questions that provide a possible indication of depression and anxiety, as well as</p><p>confidence and concentration levels. It is calculated and interpreted using the Likert scale with lowest score at</p><p>0 and highest possible score at 36. Each item has a four-point scale, but there are two types of scales depending</p><p>on the structure of statements. Some questions range from “not at all” to “much more than usual” and some</p><p>from “more than usual” to “much less than usual.”</p><p>Since the GHQ-12 satisfied similar reliability and validity tests in Bhutan as in other places, the 12 questions</p><p>were computed using the standard procedure. The threshold was set at normal well-being (15) and 85.8%</p><p>achieve sufficiency.</p><p>Education</p><p>GNH highlights the importance of a holistic educational approach that ensures Bhutanese citizens gain a</p><p>deep foundation in traditional knowledge, common values and skills. In addition to studying reading, writing,</p><p>math, science and technology, students are also encouraged to engage in creative learning and expression. A</p><p>holistic education extends beyond a conventional formal education framework to reflect and respond more</p><p>directly to the task of creating good human beings. It is important for Bhutan that an education indicator in-</p><p>cludes the cultivation and transmission of values (Ura 2009).</p><p>Literacy</p><p>A person is said to be literate if he or she is able to read and write in any one language, English or Dzongkha or Nepali.</p><p>Most Bhutanese who have achieved six years of schooling are also literate, and this measure therefore</p><p>recognizes their educational achievements. In literacy, 48.6% have attained sufficiency. Schooling on a</p><p>universally accessible basis grew from the 1970s onwards. The backlog of older generations who did not go</p><p>to school shows up as low literacy rate.</p><p>65409_Earth_Chapter5v2.indd 117 4/30/12 3:57 PM</p><p>118</p><p>Educational qualification</p><p>The education system in Bhutan has two major components: formal education and non-secular</p><p>institutions</p><p>such as monastic schools, plus non-formal education (NFE). This educational indicator includes formal schooling,</p><p>education imparted by monastic schools and NFE.</p><p>The threshold for education was set such that persons have insufficient education if they have not completed</p><p>six years of schooling from any source, including government, non-formal, or monastic schools. With this</p><p>threshold, only 37.3% have attained six years of schooling, again due to the fact that schooling and non-formal</p><p>education began relatively recently in Bhutan.</p><p>Knowledge</p><p>This indicator attempts to capture learning that could have occurred either inside or outside formal institutions.</p><p>Five knowledge variables were chosen: knowledge of local legends and folk stories, knowledge of local</p><p>festivals (tshechus), knowledge of traditional songs, knowledge of HIV-AIDS transmission, and knowledge of</p><p>the Constitution. The first three kinds of knowledge capture certain forms of local traditions, especially oral</p><p>and performance-based ones. The responses for each question follow a five-point scale that ranges from “very</p><p>good knowledge” to “very poor knowledge.” Responses are aggregated to create a maximum score of 25, which</p><p>indicates “very good” knowledge in all areas, while the minimum score of 5 indicates “very poor” knowledge.</p><p>The threshold is set to 19, which implies that Bhutanese should have an average of “good” knowledge across</p><p>the five variables. When the threshold is applied, only 7.5% have sufficiency in knowledge. Sufficiency in</p><p>knowledge is low compared to other indices; only 3% rated “good” or “very good” in all five knowledge indicators.</p><p>It suggests a divergence between rising literacy and declining knowledge about respective locality.</p><p>Values</p><p>This indicator asked respondents whether they considered five destructive actions to be justifiable: killing, stealing,</p><p>lying, creating disharmony in relationships and sexual misconduct. In a society influenced by good values, e.g.,</p><p>by Buddhism, individuals are expected to tame themselves with respect to five destructive actions. Moral</p><p>consequences of virtues and non-virtues are typically revealed through speech, body and mind and in the case</p><p>of disinformation, the agency of speech is emphasized. The variables have a three-point response scale ranging</p><p>from “always justifiable” to “never justifiable” along with an option of “don’t know.”12 The values have been</p><p>combined into a composite indicator in a particular manner. For killing, stealing and sexual misconduct, a value</p><p>of 1 is assigned if the person reports “never justifiable” while for creating disharmony and lying, responses of</p><p>either “never justifiable” or “sometimes justifiable” are assigned 1. The composite indicator takes the values 1 to 5.</p><p>The threshold is set at four, which implies that a person can consider at least one of the values to be justifiable</p><p>and 97.1% achieve sufficiency in value. The 2010 GNH indicator of values used will be improved in future</p><p>GNH surveys but the present finding provides some preliminary insight into these issues.</p><p>Culture</p><p>The distinctive culture of Bhutan facilitates sovereignty of the country and provides identity to the people.</p><p>Hence the preservation and promotion of culture has been accorded a high priority both by the government</p><p>and the people. Culture is not only viewed as a resource for establishing identity but also for cushioning</p><p>Bhutan from some of the negative impacts of modernization and thereby enriching Bhutan spiritually.</p><p>The diversity of the culture is manifested in forms of language, traditional arts and crafts, festivals, events,</p><p>ceremonies, drama, music, dress and etiquette and more importantly the spiritual values that people share. To</p><p>assess the strength of various aspects of culture, four indicators have been considered: language, artisan skills,</p><p>cultural participation and Driglam Namzha (the Way of Harmony).</p><p>65409_Earth_Chapter5v2.indd 118 4/30/12 3:57 PM</p><p>119</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Language</p><p>The language indicator is measured by a self-reported fluency level in one’s mother tongue on a four-point</p><p>scale. It should be clarified that mother tongue is defined as natal tongue, which is a dialect. There are over</p><p>a dozen dialects. Only in Western parts of the country does the mother tongue coincide with the national</p><p>language, Dzongkha. The ratings vary from “very well” to “not at all.”</p><p>Since almost everyone seems to be fluent in their mother tongue, a high threshold is necessary to maintain</p><p>standards. And for this reason, the threshold is set to “very well.” With this threshold, at present an impressive</p><p>95.2% of respondents are classified as sufficient.</p><p>Artisan skills</p><p>This indicator assesses people’s interest and knowledge in 13 arts and crafts, collectively known as Zorig</p><p>Chusum, and reports on the number of skills possessed by a respondent. These skills and vocations are the</p><p>basis of historical material culture of Bhutan from when it was trading far less. The 13 arts and crafts include</p><p>1) weaving (Thagzo) 2) embroidery (Tshemzo) 3) painting (Lhazo) 4) carpentry (Shingzo) 5) carving (Parzo) 6)</p><p>sculpture (Jinzo) 7) casting (Lugzo) 8) blacksmithing (Garzo) 9) bamboo works (Tszharzo) 10) goldsmithing</p><p>and silversmithing (Serzo and Nguelzo) 11) masonry (Dozo) 12) leather works (Kozo) and 13) papermaking</p><p>(Dezo). For the indicator, people were asked if they possessed any of the above 13 arts and crafts skills. The</p><p>mean was 1.01 with a SD of 1.15.</p><p>A sufficiency threshold has been set at one, which implies that a person must possess at least one skill to be</p><p>identified as sufficient. About 62% of the respondents are categorized as having achieved sufficiency. The</p><p>dominant or commonly shared skills today are masonry, carpentry and textile weaving.</p><p>Socio-cultural participation</p><p>In order to assess people’s participation in socio-cultural activities the average number of days within the past 12</p><p>months is recorded from each respondent. The days are grouped on five-point scale ranging from “none,” and “1</p><p>to 5 days” to “+20 days.” The median is 1 to 5 days and mean is 6 to 12 days. About 15% spent more than 13 days</p><p>attending socio-cultural events in the past year and 1% reported “don’t know” (these respondents were dropped).</p><p>The threshold was set at 6 to 12 days per year.13 It identifies 33.2% that have achieved sufficiency.</p><p>Driglam Namzha</p><p>Driglam Namzha (the Way of Harmony) is expected behavior (of consuming, clothing, moving) especially in</p><p>formal occasions and in formal spaces. It arose fundamentally from the conventions of communal living and</p><p>working in fortress-monasteries. Certain elements of Driglam Namzha are commonly practiced amongst</p><p>Bhutanese when they interact with each other in formal spaces. A minimal part of it is also taught for a few</p><p>days in educational institutions. Respondents were asked to rate its importance on a three-point scale of being</p><p>very important to not important. In addition, respondents were also asked if there were any perceived changes</p><p>in the practice of this particular form of etiquette over the years.</p><p>For Driglam Namzha, two indicators were developed: perceived importance of Driglam Namzha and the</p><p>perceived change in practice and observance during the last few years. The questions run on a three-point</p><p>scale: perceived importance ranges from “not important” to “very important” and perceived change from</p><p>“getting weaker” to “getting stronger.” Both have values of “don’t know” which have been classified as</p><p>insufficient since it is considered vital to have knowledge about etiquette.</p><p>The thresholds have been set at “important” for perceived importance and at “getting stronger” for perceived</p><p>change. Both indicators need to be fulfilled for an individual to be identified as sufficient in Driglam Namzha.</p><p>After applying the thresholds, 59.7% of people enjoy sufficiency.</p><p>65409_Earth_Chapter5v2.indd 119 4/30/12 3:57 PM</p><p>120</p><p>Time Use</p><p>The balance</p><p>should</p><p>be happiness for all.</p><p>The fourth pillar should be good governance, the ability of society to act collectively through truly participa-</p><p>tory political institutions. Good governance is not only a means to an end, but also an end in itself, since</p><p>good governance signifies the ability of people to help shape their own lives and to reap the happiness that</p><p>comes with political participation and freedom.</p><p>Yet how shall we measure success, to know that our society is on track? Here is where new metrics of happiness</p><p>can play a crucial role. To assess the four pillars of sustainable development, we need a new set of indica-</p><p>tors that extend well beyond the traditional GNP. The UN conferees have anticipated this need in the draft</p><p>outcome document for Rio+20:</p><p>Paragraph 111. We also recognize the limitations of GDP as a measure of well-being. We agree to</p><p>further develop and strengthen indicators complementing GDP that integrate economic, social and</p><p>environmental dimensions in a balanced manner. We request the Secretary-General to establish a</p><p>process in consultation with the UN system and other relevant organizations.4</p><p>These are the kinds of indicators – economic, social, and environmental – now being collected by Bhutan’s</p><p>Gross National Happiness Commission in order to create Bhutan’s GNH Index.</p><p>In addition to specific measures of economic, social, and environmental performance, governments should</p><p>begin the systematic measurement of happiness itself, in both its affective and evaluative dimensions. The</p><p>SDGs should include a specific commitment to measure happiness, so that the world as a whole, and each</p><p>individual country, can monitor progress in sustainable development and can make comparisons with the</p><p>achievements elsewhere. This massive effort of data collection has already begun. As this report discusses,</p><p>survey data on happiness are now being collected in various means: the World Values Survey, covering up</p><p>to 65 countries; the Gallup World Poll covering 155 countries; and several other national and international</p><p>surveys mentioned in Chapters 2 and 3. The OECD is now developing important proposals for internationally</p><p>standard measures explained in its case study.</p><p>65409_Earth_Chapter1v2.indd 8 4/30/12 3:46 PM</p><p>9</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Summary of this Report</p><p>When thinking about increasing happiness, one of the most important aspects is measurement. Is there</p><p>a way to accurately measure people’s happiness, both within and across societies? Chapter 2 discusses the</p><p>happiness measures currently in use across countries, specifically the Gallup World Poll (GWP), the World</p><p>Values Survey (WVS), and the European Social Survey (ESS), and asks whether or not these measures can</p><p>provide valid information about quality of life that can be used to guide policy-making. It considers the</p><p>questions of the reliability and validity of well-being measures; how happiness can be compared; whether</p><p>or not there is a happiness set point; and if happiness is “serious” enough to be taken seriously. The chapter</p><p>argues that regular large-scale collection of happiness data will enable analysis of the impacts of policies on</p><p>well-being. It concludes that regular large-scale collection of happiness data will improve macroeconomic</p><p>policy-making, and can inform service delivery.</p><p>In order to both measure and improve happiness levels, we must understand what influences these levels.</p><p>Chapter 3 discusses the causes of happiness and misery, based on 30 years of research on the topic. Both</p><p>external and personal features determine well-being. Some of the important external factors include income,</p><p>work, community and governance, and values and religion. More “personal” factors include mental and</p><p>physical health, family experience, education, gender, and age. Many of these factors have a two-way interaction</p><p>with happiness – physical health may improve happiness, while happiness improves physical health. An</p><p>analysis of all these factors strikingly shows that while absolute income is important in poor countries, in</p><p>richer countries comparative income is probably the most important. Many other variables have a more powerful</p><p>effect on happiness, including social trust, quality of work, and freedom of choice and political participation.</p><p>Chapter 4 discusses some of the policy implications of these findings. GNP is a valuable goal, but should</p><p>not be pursued to the point where economic stability is jeopardized, community cohesion is destroyed, the</p><p>vulnerable are not supported, ethical standards are sacrificed, or the world’s climate is put at risk. While</p><p>basic living standards are essential for happiness, after the baseline has been met happiness varies more</p><p>with quality of human relationships than income. Other policy goals should include high employment and</p><p>high-quality work; a strong community with high levels of trust and respect, which government can influence</p><p>through inclusive participatory policies; improved physical and mental health; support of family life; and a</p><p>decent education for all. Four steps to improve policy-making are the measurement of happiness, explanation</p><p>of happiness, putting happiness at the center of analysis, and translation of well-being research into design</p><p>and delivery of services.</p><p>1 Editorial assistance provided by Claire Bulger.</p><p>2 On average across OECD countries, cash transfers and income taxes reduce inequality by one third. Poverty is around 60% lower than</p><p>it would be without taxes and benefits. Even among the working-age population, government redistribution reduces poverty by about</p><p>50%. See OECD (2008).</p><p>3 Sahlberg, P (2007).</p><p>4 Rio+20 United Nations Conference on Sustainable Development. (2012).</p><p>65409_Earth_Chapter1v2.indd 9 4/30/12 3:46 PM</p><p>10</p><p>Part I.</p><p>Chapter 2.</p><p>THE STATE OF WORLD HAPPINESS</p><p>JOHN F. HELLIWELL AND SHUN WANG</p><p>John F. Helliwell: Professor Emeritus of Economics, University of British Columbia and Arthur J.E. Child</p><p>Foundation Fellow of the Canadian Institute for Advanced Research (CIFAR)</p><p>Shun Wang: Assistant Professor, KDI School of Public Policy and Management</p><p>11</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Introduction</p><p>This chapter presents and explains a range of happiness measures currently available in a comparable format</p><p>for many countries. Some survey data cover almost all countries, and hence can be used to develop an overall</p><p>picture of the state of world happiness in the first decade of the 21st century.1 This accounting makes use of</p><p>measures of subjective well-being, since they capture best how people rate the quality of their lives. “Subjective</p><p>well-being” is the general expression used to cover a range of individual self-reports of moods and life assessments.</p><p>The word “happiness” is often used in an equally general way, as in the title of this report. It does help to focus</p><p>thinking, and attracts attention more quickly than does “subjective well-being.” But there is a risk of confusion.</p><p>A bit of advance explanation may help to keep things clear.</p><p>Among various measures of subjective well-being, the primary distinction to be made is between cognitive life</p><p>evaluations (represented by questions asking how happy or satisfied people are with their lives as a whole), and</p><p>emotional reports.2 Early modern attempts to classify different types of subjective well-being in psychology</p><p>have also made a distinction between two types of emotional reports: positive affect (a range of positive</p><p>emotions) and negative affect (a range of negative emotions).3 The primary distinction between life evaluations</p><p>and emotional reports continues to be accepted today. It is also accepted,4 although less generally,5 that positive</p><p>and negative affect carry different information, and need to be separately measured and analyzed. In this report</p><p>we shall present all three types of measure.</p><p>How does happiness come into this classification? For better or worse, it enters in three ways. It is sometimes</p><p>between paid work, unpaid work and leisure are important for one’s well-being. Similarly, a flexible</p><p>working life is vital for the well-being of individual workers and their families and communities. Since the</p><p>1970s, there has been a growing awareness of how unpaid work both at home and in communities is obscured</p><p>in national accounts and so efforts have been made to include these activities, which are equally fundamental</p><p>to well-being.</p><p>In the GNH survey, a simple time diary was administered. Information on how people use their time was</p><p>collected by asking respondents to recall their activities during the previous day. Survey respondents reported</p><p>activities that they did from the time they woke up until the time they slept on the previous day of the interview.</p><p>For each activity the respondents were asked how long the activity lasted. The activities were then later</p><p>regrouped into 60 different categories of different kinds of activities such as work, leisure, sleep, personal care</p><p>and so on.</p><p>Time use data can yield a range of important information that provides insight into lifestyles and occupations</p><p>of the people. It can also reveal the gap between GDP and non-GDP activities, which reflects the gap between</p><p>market and household economy sectors. Such data are helpful in accounting for a more comprehensive output</p><p>of goods and services that SNA omits (Ironmonger 1999). Time use data on 24 hours in the life of Bhutanese</p><p>people can be broken down into various useful sub-categories. The distribution involves the following</p><p>disaggregation: 20 districts, seven income slabs, 11 age groups, 60 activities, and gender (Ura 2012).14 How-</p><p>ever, the GNH Index incorporates only two broad aggregated time use: work hours and sleep. The definition of</p><p>work15 hours in GNH is not completely congruent with definitions used elsewhere and shows unusually long</p><p>work duration in Bhutan. Some activities not usually defined as work elsewhere are included as part of work.</p><p>Working hours</p><p>The GNH definition includes even unpaid work such as childcare, woola (labor contribution to community</p><p>works), and voluntary works and informal helps etc. In this indicator, all the following categories are classified</p><p>as work: Crop farming and kitchen gardening (agric), Business, trade and services, Care of children and sick</p><p>members of household, Construction and repairs, Craft related activities, Forestry and horticultural activities,</p><p>Household maintenance, Livestock related activities, Processing of food and drinks, and Quarrying work.</p><p>Eight hours is also the legal limit, applied to the formal sector, set by the Ministry of Labour and Human</p><p>Resources of Bhutan for a standard work day. Since a main objective of the indicator is to assess people who</p><p>are overworked, those who work for more than eight hours are identified as time deprived. 45.4% achieve</p><p>sufficiency when this threshold is applied. Those who do not achieve this sufficiency are mainly women</p><p>irrespective of whether they live in towns or villages, and more generally the people in the Eastern districts.</p><p>People in Eastern Bhutan have longer workdays compared to the rest.</p><p>Sleeping hours</p><p>Sleep is clearly beneficial for a person’s health and impacts nearly every area of daily life. In general most healthy</p><p>adults need an average of seven to eight hours of sleep for proper functioning (Kleitman 1963; Doran, Dongen</p><p>and Dinges 2001; Smith, Robinson and Segal 2011). But sleep requirements can vary substantially and some</p><p>people, such as nuns and monks, would prefer and find it much healthier to devote more time to meditation and</p><p>other spiritual practices than sleeping. Indeed, survey confirms that they sleep comparatively less.</p><p>Eight hours is considered the amount necessary for a well-functioning body for everyone. Both the mean and</p><p>median fall at around eight hours for the respondents. With this threshold, about 66.7% achieve sufficiency.</p><p>65409_Earth_Chapter5v2.indd 120 4/30/12 3:57 PM</p><p>121</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Good Governance</p><p>Four measures were developed to signify effective and efficient governance. These include fundamental rights,</p><p>trust in institutions, performance of governmental institutions and political participation. These indicators may</p><p>be adjusted in future surveys. The governance indicators are quite innovative in combining political activities</p><p>with access to government services. These are understood as part of governance and a part of the public services</p><p>to be provided by the government. They also include fundamental rights to vote, freedom of speech, to join a</p><p>political party, to be free of discrimination, and a perceptual indicator on government performance.</p><p>Political participation</p><p>The measure of political participation is based on two components: the possibility of voting in the next election</p><p>and the frequency of attendance in zomdue (community meetings). The respondents are asked if they will vote</p><p>in the next general election and the response categories are simply “yes” or “no” or “don’t know.”</p><p>An individual has to report “yes” in the voting criteria and has to attend at least one meeting in a year to be classified</p><p>as sufficient in political participation. About 92% have expressed an intention to vote in the next general election,</p><p>4.7% declined and 2% don’t know. For voting, the threshold is straightforward because it is agreed by everyone</p><p>that developing true democratic processes requires the active participation of citizens – minimally, by voting. In</p><p>terms of attendance in meetings the threshold has been set to one time. About 60.2% attended at least one meeting.</p><p>Fixing the threshold as such classifies 43.6% as deprived in political participation.</p><p>Political freedom</p><p>These indicators attempt to assess people’s perceptions about the functioning of human rights in the country</p><p>as enshrined in the Constitution of Bhutan, which has an entire article (Article 7, Fundamental Rights) dedicated</p><p>to it. The seven questions related to political freedom ask people if they feel they have: freedom of speech and</p><p>opinion, the right to vote, the right to join a political party of their choice, the right to form tshogpa (association)</p><p>or to be a member of tshogpa, the right to equal access and the opportunity to join public service, the right to</p><p>equal pay for work of equal value, and freedom from discrimination based on race, sex, etc. All have three possible</p><p>responses from 1 to 3: “yes,” “no” and “don’t know.”</p><p>The thresholds for all rights were set to “yes.” So, a person has a sufficient condition in the indicator if he or</p><p>she has all seven rights fulfilled. Of the respondents, 61.7% were identified as sufficient.</p><p>Service delivery</p><p>The indicator comprises four indicators: distance from the nearest health care center, waste disposal method,</p><p>access to electricity and water supply and quality. The goal is to evaluate access to such basic services, which in</p><p>Bhutan are usually provided by the state.</p><p>In health services, people less than an hour’s walk to the nearest health center are considered to have sufficient</p><p>access. In cities, access is attained but crowding can lead to waiting. If households report disposing of trash by</p><p>either “composting,” “burning” or “municipal garbage pickup” they are non-deprived. On the other hand, if the</p><p>response is “dump in forests/open land/rivers and streams” then they are deprived. As access to electricity is at</p><p>the forefront of Bhutan’s objectives, respondents who answer “yes” to the question of whether their house has</p><p>access to electricity are considered non-deprived. The improved water supply indicator combines information</p><p>on access to safe drinking water with information on the perceived quality of drinking water. An improved</p><p>facility would include piped water into a dwelling, piped water outside of a house, a public outdoor tap or a</p><p>protected well. For the perceived quality of water, the threshold has been set to “good” or “very good.” Both</p><p>conditions</p><p>need to be fulfilled in order to be sufficient in water.</p><p>65409_Earth_Chapter5v2.indd 121 4/30/12 3:57 PM</p><p>122</p><p>Overall, people are is classified as having achieved sufficiency in service delivery if they enjoy sufficiency in</p><p>each of the four elements. About 41% have achieved that condition.</p><p>Government performance</p><p>The indicator pertains to people’s subjective assessment of the government’s efficiency in various areas. To test</p><p>people’s perceptions of overall service delivery in the country, respondents are asked to rate the performance</p><p>of the government in the past 12 months on seven major objectives of good governance: employment,</p><p>equality, education, health, anti-corruption, environment and culture. These outcome-based questions enable</p><p>respondents to rank the services on a five-point scale from “very good” to “very poor.” 16,17 The overall indicator</p><p>has a maximum value of 35 and minimum value of 7.</p><p>A threshold of 28 was adopted, which means that a person has to perceive that public services are “very good”</p><p>or “good” in at least five of the seven objectives. With this threshold, about 78.8% are considered to have</p><p>achieved sufficiency.</p><p>Community vitality</p><p>The concept of GNH includes the social capital of the country, which is sustained through co-operative rela-</p><p>tionships and social networks within the community. A vital community can be described as a group of people</p><p>who support and interact positively with each other. The concept outlined here reflects also GNH values and</p><p>Bhutanese moral beliefs.</p><p>From a GNH standpoint, a community must possess strong relationships amongst the community members</p><p>and within families, must hold socially constructive values, must volunteer and donate time and/or money,</p><p>and lastly must be safe from violence and crime. It is vital that volunteering and donations of time and money</p><p>be recognized as fundamental parts of any community development. The values can act as tools through which</p><p>activities can be implemented for positive change in communities. The indicators in this domain cover</p><p>four major aspects of community: 1) social support, which depicts the civic contributions made 2) commu-</p><p>nity relationships, which refers to social bonding and a sense of community 3) family relationships and 4)</p><p>perceived safety.18</p><p>Social support</p><p>These indicators assess the level of social support in a community and the trends across time. They capture</p><p>the giving of time and money (other goods in previous olden days) – volunteering and donating – a traditional</p><p>practice in Bhutanese societies. To capture the rate of volunteering, respondents were asked for the number</p><p>of days they volunteered and for the amount they donated. Donation is expressed in the total amount of financial</p><p>resources donated in the past 12 months and volunteering is measured by the days donated in the past 12 months.</p><p>For donation, giving 10% of household income is considered sufficient, and for volunteering, three days per</p><p>year is considered sufficient. These thresholds have been derived from normative criteria. Overall, if persons</p><p>donate 20% of their income, then even if they do not volunteer it is considered sufficient and if they volunteer</p><p>more than six days, but do not donate 10% of their income, it is also considered sufficient. With these condi-</p><p>tions applied, overall, 46% are sufficient.</p><p>Community relationships</p><p>The two components of this indicator are “a sense of belonging,” which ranges from “very strong” to “weak,”</p><p>and “trust in neighbors,” which ranges from “trust most of them” to “trust none of them.” Both indicators have</p><p>options of “don’t know.” 71% have a very strong sense of belonging, 46% trust most of their neighbors, and</p><p>65409_Earth_Chapter5v2.indd 122 4/30/12 3:57 PM</p><p>123</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>85% trust most or some of their neighbors. The trust indicator may reveal the trustworthiness of the neighbors.</p><p>The thresholds here are based on normative reasons for sustaining and promoting a sense of community. The</p><p>threshold for sense of belonging has been set at “very strong” and for levels of trust “some of them” and “most</p><p>of them” have been selected. For a person to have achieved sufficiency, both conditions have to be satisfied and</p><p>62.5% of people are sufficient in both.</p><p>Family</p><p>For this indicator, six questions on a three-point scale of “agree,” “neutral” and “disagree” have been asked of</p><p>the respondents. They are added together to form an indicator with 18 as the maximum score (high family</p><p>relationships) and 6 as the minimum score (low family relationships).</p><p>A threshold of 16 is applied in order to allow “neutral” responses in any two statements. 92% are satisfied in</p><p>the family indicator.</p><p>Victim of crime</p><p>To assess safety in the community, respondents are asked if they have been a victim of crime in the past 12</p><p>months. The crime indicator has a simple two-point scale of “yes” and “no.”</p><p>The threshold is set at “no.” The crime statistics are low with only about 4% being described as victims. Self-</p><p>reported victimization, however, slightly underestimates victimization when it concerns sexual offenses. In the</p><p>next survey, other safety indicators might be incorporated to improve evaluation.</p><p>Ecological Diversity and Resilience</p><p>Bhutan has always recognized the central role environmental factors play in human development. Pursuant to</p><p>Article 5 (Environment) of the Constitution of Bhutan, every Bhutanese citizen shall “…contribute to the pro-</p><p>tection of the natural environment, conservation of the rich biodiversity of Bhutan and prevention of all forms</p><p>of ecological degradation including noise, visual and physical pollution…”</p><p>The environmental domain includes three subjective indicators related to perceptions regarding environmental</p><p>challenges, urban issues and responsibilities, and one more objective question, related to wildlife damage to</p><p>crops. Like other subjective indicators, the interpretation of these indicators is clouded by different and possibly</p><p>shifting frames of reference, so they are given a light weight of 10% of the environmental domain each.</p><p>Indicators in this domain in particular may be reconsidered for future GNH surveys to better capture the full</p><p>complexity of the ecological system.</p><p>Pollution</p><p>In order to test people’s environmental awareness, a series of questions were developed to test the perceived</p><p>intensity of environmental problems. Seven environmental issues of concern were shared with respondents,</p><p>and their responses follow a four-point scale from “major concern” to “minor concern.”</p><p>They are not added into a single number but rather a conditional threshold is applied whereby an individual is</p><p>insufficient if he or she has rated “major concern” or “some concern” in at least five of the seven environmental</p><p>issues. Their reference frame is within the past 12 months; however, as with many subjective indicators, there</p><p>might be errors with the reference frame and so it is not very practical to give more weight to perceptive data</p><p>by fixing high thresholds. Hence, with the proposed threshold, 69% are sufficient in the pollution indicator.</p><p>65409_Earth_Chapter5v2.indd 123 4/30/12 3:57 PM</p><p>124</p><p>Environmental responsibility</p><p>The indicator attempts to measure the feelings of personal responsibility towards the environment. It is crucial</p><p>to reinforce attitudes that will encourage people to adopt eco-friendly approaches and also to identify any</p><p>deterioration in the current very environmentally aware views of citizens. The responses run on a four-point</p><p>scale ranging from “highly responsible” to “not at all responsible.” When the threshold is set at “highly</p><p>responsible,” 84.4% are sufficient.</p><p>Wildlife</p><p>The wildlife indicator here incorporates information on damage to crops. There has been a growing concern</p><p>about wildlife damage to crops in Bhutan (Choden and Namgay 1996; Wang, Curtis and Lassoie 2006). Wildlife</p><p>damage can have catastrophic economic consequences for farmers, especially</p><p>vulnerable households; it also</p><p>disrupts sleep patterns and may create anxiety and insecurity. A simple self-reported estimate is used as a proxy</p><p>for quantitative assessment. Two simple questions on the presence and absence of damage and the severity of</p><p>damage are applied to determine the impact of wildlife damage on agriculture.</p><p>The first question deals with whether respondents consider damage as a constraint to farming. Responses are</p><p>given on a four-point scale ranging from “major constraint” to “not a constraint.” The threshold has been set</p><p>at “minor constraint.” The second indicator pertains to the severity of damage, i.e. crop loss. Respondents are</p><p>asked to provide an average perceived amount of crop loss, if the crop had been damaged by wildlife. It ranges</p><p>from “a lot” to “not at all.” For both the indicators the reference frame is the past 12 months.</p><p>The threshold is fixed such that respondents are deprived if they report either “some constraint” or “major</p><p>constraint” and account for a crop loss of “a lot” or “some.” The lack of actual numeric amounts or percentages</p><p>of actual crop loss may give rise to errors so both conditions have to be fulfilled. With this threshold, 57.9% of</p><p>the respondents attain the sufficiency condition.</p><p>The wildlife indicator is rural-specific since it pertains to farmers. Individuals from other occupational back-</p><p>grounds such as civil servants or corporate workers are classified as non-deprived. The rural-specific indicator</p><p>is later offset by the urban issue indicator, which in turn applies to urban dwellers only.</p><p>Urban use</p><p>Bhutan is undergoing a rapid urbanization resulting in the growth of city and town populations. Since this has</p><p>both positive impacts on human well-being (such as improvement in energy, health care, infrastructure) and</p><p>negative effects (congestion, inadequate green spaces, polluted ambience) these adverse impacts on well-being</p><p>have been incorporated into the GNH index. Respondents are asked to report their worries about four urban</p><p>issues: traffic congestion, inadequate green spaces, lack of pedestrian streets and urban sprawl.</p><p>The threshold is set such that a person can report any one of the issues as a major threat or worry to be sufficient.</p><p>About 84.4% achieve sufficiency; this is in part because people who live in rural areas have been automatically</p><p>classified as sufficient, to offset the wildlife damage indicator introduced above. This indicator mainly acts as a</p><p>proxy for sustainable urban development, which is one of the major objectives of the government.</p><p>Living Standards</p><p>The living standards domain refers to the material well-being of the Bhutanese people. It ensures the fulfill-</p><p>ment of basic material needs for a comfortable living. Over the years, the material standard of living has risen</p><p>steadily due to advances in development. However, about 23.2% (Royal Government of Bhutan 2007) of Bhu-</p><p>tanese still live in income poverty; some lack assets such as land or adequate housing.</p><p>65409_Earth_Chapter5v2.indd 124 4/30/12 3:57 PM</p><p>125</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>There are a wide range of indicators used in the literature to assess standards of living. For individual-level</p><p>analysis, the actual consumption of goods and services is often argued to be the most accurate. Income and</p><p>expenditure levels are often used if consumption is difficult to detail. Here, we use three indicators to assess</p><p>people’s standards of living: household per capita income, assets and housing conditions. Assets include live-</p><p>stock, land and appliances, while housing conditions pertain to room ratio, roofing and sanitation. These are</p><p>included so that there are enough complementary measures for self-reported household income.</p><p>Household income</p><p>Household income includes income earned by all the individuals in a household from varied sources within or</p><p>outside of the country. The household income here has been adjusted for in-kind payments received.</p><p>In the literature, two types of thresholds are generally used, either a fixed threshold like a poverty line or rela-</p><p>tive thresholds such as mean or median income. The poverty line for Bhutan is Nu. 1,096.94 per person per</p><p>month in the Poverty Analysis Report (Royal Government of Bhutan 2007).19 The mean household per capita</p><p>was generated by dividing the household income by household size, without equivalence scales. In Bhutan</p><p>Living Standards Survey (BLSS) (2007) it was Nu. 31,834.3. When a poverty line threshold (Nu. 1, 096.94) was</p><p>used on individual income, the headcount estimation made by the Poverty Analysis Report (Royal Government</p><p>of Bhutan 2007) was 23.2%.</p><p>For the GNH Index, it would not be sensible to use the poverty line as a threshold because the threshold should</p><p>reflect sufficient income. The GNH living standards domain refers to higher conditions for well-being than</p><p>poverty lines. One option would be to use a relative income threshold for the sufficiency threshold, as is com-</p><p>monly done in European countries. Thresholds like 60% of the median or 50% of mean income are often used</p><p>to identify poverty. 20</p><p>Yet for the GNH indicator an absolute sufficiency threshold was chosen, since the GNH values and encourages</p><p>people to achieve happiness through their accomplishments, and discourages a relative approach in which one</p><p>is satisfied only if one has relatively more income (or other achievements) than one’s peers. In this regard, a</p><p>threshold is computed from a GNH data-adjusted poverty line21 by the multiplying the national poverty line by</p><p>1.5. It would have amounted to Nu. 14,200 per person per year in the BLSS 2007 data.22 The income threshold</p><p>classifies 54% of people as sufficient.</p><p>Assets</p><p>An asset indicator has been used as an indicator of living standards in many studies (Montgomery et al 2000;</p><p>Morris et al 2000; Filmer and Pritchett 2001; Case et al 2004). 23 The indicator uses data on selected household</p><p>assets, such as durable and semi-durable goods of everyday use, to describe household welfare. The concept is</p><p>based on evidence that income/expenditure measures are incomplete measures of the material well-being of</p><p>households especially in developing countries where such data may have higher measurement errors. 24 However,</p><p>it is necessary to note that the items of the indicator are taken from a generic list of goods, the uses of which</p><p>may not be the same across all household members, and quality aspects of the goods owned were not included.</p><p>Commonly, asset indicators are defined by appliances such as a mobile phone, radio, TV or bicycle; however,</p><p>because of the socio-cultural context, livestock and land ownership were also considered assets. Livestock is</p><p>understood as an integral component in agricultural and rural economies in Bhutan. Most farming is still</p><p>subsistence farming, and the difficult terrain makes it challenging to use modern equipment. Thus, the work</p><p>must be done by animals and humans. Moreover, animals provide households with transport, fertilizers and</p><p>foods, and also employment. So, it is a critical asset especially for poor households. Similarly, land ownership</p><p>is particularly relevant for rural agricultural-based economies. In some of the focus group participants’ percep-</p><p>tions, a decent living standard always included livestock and land ownership.25</p><p>The asset indicator consists of three major components: 1) appliances (mobile phone, fixed-line telephone,</p><p>personal computer, refrigerator, color television and washing machine) 2) livestock ownership and 3) land</p><p>ownership.</p><p>65409_Earth_Chapter5v2.indd 125 4/30/12 3:57 PM</p><p>126</p><p>The thresholds are applied at two levels: they are set initially on each of the three indicators and then later, an</p><p>overall threshold is applied to classify insufficiency in the asset indicator.</p><p>For a measure of appliances, a series of household items that could be considered amenities for the family was</p><p>developed. Principal component analysis has been used to determine the</p><p>selection of appliances. The first</p><p>factor explained 80% of the variance and contained six appliances – mobile phone, fixed-line phone, personal</p><p>computer, refrigerator, washing machine and color television. The mobile phone could be dropped from the</p><p>list of appliances since, in general sense, the utility is marginal and limited to the one who owns it. For the</p><p>other appliances, the scope of functional utility is much wider and other members of the household might</p><p>have access. However, in rural areas if a household owned a mobile phone then that would imply that every</p><p>household member had some access to it. Moreover, fixed-line phones are being replaced by mobile phones</p><p>even in urban areas; only 21% of urban households now have fixed-line phones. So, in the end, all six items</p><p>loaded in the first factor were considered for the asset indicator. The sufficiency threshold was set to three and</p><p>31% are sufficient in appliances.</p><p>It is widely known that livestock constitute an important source of income, especially in rural areas and no-</p><p>madic areas of the country. They contribute to a household’s livelihood by providing cash income or in-kind</p><p>income through the sale of animal products or animals themselves and thereby act as savings for future</p><p>security. Although the importance of including livestock as an asset is generally agreed upon, setting a thresh-</p><p>old becomes challenging because of the difference in the capital and maintenance costs of different species,</p><p>which are usually higher for larger ruminants. Larger ruminants require more fodder while smaller domestic</p><p>animals, such as chickens, can survive on a lesser amount. And so, based on the rates of an average domestic</p><p>purchase, a threshold is defined. It was observed that an average price of 40 chickens would be equivalent to</p><p>the average rate of others. Ownership of chickens has been reclassified accordingly. In terms of thresholds,</p><p>Bhutan’s national MPI (2010) sets it at three, but for the GNH Index it has to be set higher. And so, livestock</p><p>has been set to five normatively. About 41.3% of the respondents are sufficient in livestock.</p><p>The data on land were collected in the categories of dry land and (un-terraced); wetland (irrigated and ter-</p><p>raced); panzhing, which is a type of land use where land is cultivated after leaving it fallow to improve soil fer-</p><p>tility; orchards; kitchen gardens; and tseri, which refers to shifting cultivation. Although the Land Act of 2007</p><p>banned tseri cultivation, the survey shows about 14.4% of the respondents still practice it. The average land</p><p>holding is 2.9 acres per household (SD =3.6). The average rural land holding is 3.39 acres per rural household,</p><p>and for urban areas it is 0.86 acres per household.</p><p>In setting the sufficiency cutoff for land, there are numerous factors that need to be taken into consideration</p><p>such as quality of land, household size, area and type of farming practices and sources of other income. The</p><p>household size plays a role as smaller families might require smaller land holdings and larger families might</p><p>need more land. The region of location is also a huge determinant since an agriculture-based economy usually</p><p>requires more land holdings. Lastly, the type of farming must also be considered, for instance whether the</p><p>land is being used for crops or orchards or just as pasture for animals and also whether the particular house-</p><p>hold has other sources of income. Given the wide range of factors that require equal attention, it is challeng-</p><p>ing to set a threshold that fulfils all these conditions.</p><p>The focus group discussions carried out in some districts concluded that five acres was the threshold for a</p><p>rural farming household with an average family size of five. It was decided that for farming-related activities</p><p>an average of five acres would be sufficient to grow crops or fruits or for livestock management. The land asset</p><p>is included to reflect assets for rural areas, and so understanding land ownership in rural areas is pertinent</p><p>for setting the threshold. In rural areas, only 26% of households have five or more acres of land, while about</p><p>44% have three or more acres of land. For the MPI Bhutan 2010, the threshold was set to one acre, but the</p><p>GNH Index is not a poverty measure and so a minimum threshold cannot be applied. The average household</p><p>size in rural areas is 4.7, and the sufficiency threshold for an average land amount was normatively set to</p><p>five acres. About 22% are sufficient; however, note that the GNH Index also includes urban dwellers whose</p><p>income comes mostly from employment, so they would be regarded as deprived in this sub-indicator (but not</p><p>necessarily overall as we see below).</p><p>65409_Earth_Chapter5v2.indd 126 4/30/12 3:57 PM</p><p>127</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>The final threshold across the three assets is applied so that if a household possesses sufficiency in appliances</p><p>or livestock or land then the household is classified as being sufficient in assets overall. This implies that</p><p>any one condition of the three can be satisfied to be in order to be labeled non-deprived. This threshold was</p><p>selected based on its flexibility to incorporate individuals from diverse occupational backgrounds, as well as</p><p>from varied areas of residence. For example, livestock and farm land may not be very relevant to a person who</p><p>is employed in a service occupation but may be particularly valid in remote areas. It must be understood that</p><p>the objective of an asset indicator is to supplement information income with some crude indicator of wealth.</p><p>Asset indices may move more slowly than income and expenditure. This gives rise to data reliability issues</p><p>for GNH Index analysis attempting to capture trends in well-being over time. This requires not only that we</p><p>interpret results with due caution but that we also keep in mind the complexities of combining the three as-</p><p>sets together. However given the issues with the income data mentioned above, both indicators were included</p><p>to improve accuracy. Application of the overall conditional threshold identifies 74.1% of Bhutanese to have</p><p>achieved sufficiency.</p><p>Housing quality</p><p>The domain is incomplete without including an indicator of housing conditions. The benefits of good housing</p><p>can be observed from both an individual as well as from a community perspective. On the individual level, hav-</p><p>ing one’s personal space is considered fundamental for one’s biological, psychological and social needs since</p><p>it is a place where most spend a significant part of their everyday lives. 26 Studies show the critical impacts that</p><p>poor quality, overcrowded and temporary accommodation can have on an individual’s physical and mental</p><p>health. 27 From a community standpoint, aspects such as combating social exclusion and discrimination and</p><p>strengthening social cohesion cannot be achieved unless there are proper living spaces and a decent standard</p><p>of accommodation. Studies show strong associations between the likelihood of criminality and educational at-</p><p>tainment (Lupton and Power 2005; Fagan and Davies 2007; Friedman 2010). Overcrowded accommodation,</p><p>which is based on the number of rooms and number of household members, can lead to family disintegration</p><p>and a weakening community ties, and is considered to give rise to a variety of social ills. Therefore, insufficient</p><p>housing conditions can pose a threat to not only the well-being of individuals but also the community at large.</p><p>The quality of housing is composed of three indicators: the type of roofing, type of toilet and room ratio. The</p><p>thresholds have been set based on the Millennium Development Goals such as corrugated galvanized iron</p><p>(CGI) or concrete brick or stone for roofing, pit latrine with septic tank for toilet and two persons per room for</p><p>overcrowding, and all three conditions must be met. So, overall an individual is sufficient in housing if he or</p><p>she lives in a house that has a good roofing structure (CGI or concrete brick or stone),</p><p>a pit latrine with a septic</p><p>tank, and uncrowded rooms. In reality, having a higher quality roof may by far outweigh toilet condition as</p><p>far as housing quality is considered. With the stated threshold, about 46.2% are sufficient in housing quality.</p><p>Weighting</p><p>The nine domains of GNH are equally weighted. This is because they are of equal importance, so none can be</p><p>permanently ranked as more important than others but each might be particularly important to some person</p><p>or some institution at a given point in time. The 33 indicators are roughly equally weighted but the subjective</p><p>and self-reported indicators have lighter weights and the indicators that are anticipated to be more objective</p><p>and/or more reliable have relatively higher weights when the domains mix subjective and objective indicators. There</p><p>are equal weights among all indicators in three dimensions: psychological well-being, time use and living standards.</p><p>In three domains, health, good governance, and ecological diversity, subjective indicators receive only 10% of</p><p>the weight of the dimensions and the other indicators within those dimensions are equally weighted. The five</p><p>indicators which receive 10% weight of their respective dimension each, because they are subjective, are as fol-</p><p>lows: in the domain of health – self reported health status; in the domain of governance – governance per-</p><p>formance and fundamental rights; in the domain of ecological diversity and resilience – responsibility towards</p><p>the environment and perceptions of ecological issues.</p><p>65409_Earth_Chapter5v2.indd 127 4/30/12 3:57 PM</p><p>128</p><p>In the last three domains, education, culture and community, self-reported indicators are weighted at 20% each</p><p>and the other indicators are weighted at 30%. In education, the two self-report based indicators are knowledge</p><p>and values. In cultural diversity and resilience, the two self-report based indicators are speaking a native lan-</p><p>guage and Driglam Namzha. And in community vitality the two self-report based indicators are community</p><p>relationships and family relationships.</p><p>In this way the weighting on the indicators tries to both preserve accuracy and also to prevent future GNH</p><p>indices being too affected by changes in the frame of reference or changes in the aspirations of people that</p><p>might affect their subjective or self-reported indicators. However these are diffi cult decisions to make. Many</p><p>indicators in the GNH survey could be argued to be self-report based. Indeed to some extent all could be</p><p>self-report based indicators. However we have tested the GNH Index robustness to changes in these weights</p><p>and those results, which are presented later, show that it is relatively robust for policy purposes for small</p><p>changes in the weighting structure.</p><p>Thresholds</p><p>The GNH Index uses two kinds of thresholds: suffi ciency thresholds or cutoffs, and one happiness threshold.</p><p>Suffi ciency thresholds show how much a person needs in order to enjoy suffi ciency in each of the 33 cluster</p><p>indicators. It asks how much is enough to be happy. Each of the 33 cluster indicators has a suffi ciency threshold</p><p>and each person in the survey is identifi ed as enjoying suffi ciency or not in each indicator. How are these suf-</p><p>fi ciency thresholds set? Who decided?</p><p>There were different inputs to calibrate these decisions. Some use relevant and appropriate international standards,</p><p>e.g. for hours of work, and overcrowding in a house. Some use national standards, e.g. a suffi ciency income</p><p>is equivalent to 1.5 times the income poverty line for Bhutan. For other indicators there wasn’t a literature or</p><p>precedent in Bhutan or internationally to set suffi ciency thresholds. For this reason, some rely on normative</p><p>judgments. This is because GNH is innovative and there are no international or national standards for these</p><p>indicators, e.g. for positive emotions. In this case, the GNH thresholds are based on normative judgments that</p><p>Table 2: Weights on the 33 Indicators</p><p>65409_Earth_Chapter5v2.indd 128 4/30/12 3:57 PM</p><p>129</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>have been shared and discussed in consultative sessions. The final and important inputs were participatory</p><p>meetings. The Centre for Bhutan Studies held consultative conversations with different institutions and leaders</p><p>in government, and focus group discussions with communities in different rural areas and sought their input,</p><p>checking with them the thresholds on test or trial GNH indices while the GNH Index was still being finalized.</p><p>Their insights proved very useful but also drew attention to the fact that no one set of thresholds will be accurate</p><p>across all people in Bhutan. And that is why it is very important to have a second cut-off, of a sufficient happiness</p><p>threshold that allows for a lot of variation among people, based on their own personalities and aspirations</p><p>as well as on their material, community and climactic circumstances. All of the indicators with their cut-offs</p><p>will not be equally meaningful or relevant in the many varied contexts of Bhutan – but they need not be. The</p><p>second threshold permits diversity.</p><p>In reporting the GNH, we divide the population into four sub-groups by applying three cutoffs, which refer</p><p>to people who have achieved sufficiency in 50%, 66%, and 77% of the weighted indicators. This enables us</p><p>to identify the unhappy, narrowly happy, extensively happy, and deeply happy. We can and do analyze each</p><p>of these groups’ achievements separately. For each person, we have their personal profile of achievements</p><p>across all 33 cluster indicators, and these profiles provide a rich basis for analyses of these four different</p><p>GNH Groups – the indicators and dimensions in which they lack sufficiency, and how these change by gen-</p><p>der, region, age, and occupation.</p><p>To calculate the GNH Index, we choose one threshold or cutoff. We could choose the lowest cutoff in which</p><p>case we would find that only 10% of Bhutanese were unhappy. However this would restrict the policy focus</p><p>to a small set of the population, leaving the rest unsupported. So instead, we choose the middle happiness</p><p>cutoff of 66%. Thus the not-yet-happy group includes both those who are unhappy and those who are nar-</p><p>rowly happy – a total of 41% of people. Our analysis of how to “increase GNH” focuses on increasing the</p><p>sufficiency of these groups.</p><p>This second cutoff is referred to as the happiness threshold. It is set across the nine domains and the 33</p><p>cluster indicators. The question that it asks is “how many domains or in what percentage of the indicators</p><p>must a person achieve sufficiency in order to be understood as happy?” Here it is important to acknowledge</p><p>that this approach is an experiment. Happiness is a very deeply personal experience and any measure of it is</p><p>necessarily imperfect. The index is offered to the people of Bhutan for understanding, discussion and debate</p><p>to see if it frames and captures their understandings and how this might change or be improved.</p><p>The happiness threshold was set based on three criteria. The first is diversity, as not all of the indicators have</p><p>universal applicability. It may not be necessary to have sufficiency in all of the indicators to be happy; e.g. a</p><p>person who is very old might not need sufficiency in education indicators in order to be happy. They might</p><p>have other members of their family who can read for them or explain things that require a formal education</p><p>and their wisdom and skills may suffice for their own happiness. Some people, such as atheists for example,</p><p>may not participate in prayer recitation or meditation.</p><p>The second is measurement error. Responses might not be completely accurate about people’s values in</p><p>different cultures – for example, people may be hesitant to say what exactly their beliefs or practices are for</p><p>fear of seeming proud or ostentatious. Because of the difficulty of allowing for these differences (as it is done</p><p>in poverty measures) it seemed reasonable not to require sufficiency</p><p>in every domain.</p><p>The third and last criterion is freedom of choice. Many people are fully happy without achieving sufficiency</p><p>in every single indicator. Maybe they are not healthy but they have achieved a kind of flourishing, fulfillment</p><p>and richness of life that is important. Maybe they are illiterate or have material challenges but that need not</p><p>necessarily be decisive for their happiness. Thus to allow some freedom of choice we have set the happiness</p><p>threshold at 66%.</p><p>65409_Earth_Chapter5v2.indd 129 4/30/12 3:57 PM</p><p>130</p><p>Methodology</p><p>The GNH itself is constructed using the Alkire Foster method (2007, 2011) for measuring multidimensional</p><p>concepts such as poverty, well-being or inequality (see the Appendix for the formal methodology). It is a robust</p><p>method that identifi es a group – in this case those people who are not-yet-happy (vs. those who are happy) by</p><p>considering the “suffi ciencies” they enjoy. It is a fl exible method that has been fully tailored to the needs and</p><p>context in Bhutan. This includes identifying the happiness gradient – the four population subgroups accord-</p><p>ing to the percentage of weighted indicators in which they have suffi ciency.</p><p>Like other measures in the Alkire Foster family, the GNH Index is created from two numbers:</p><p>Headcount ratio: Percent of people who are happy</p><p>Breadth: Percent of domains in which people who are not-yet-happy enjoy suffi ciency (this is similar to “intensity”</p><p>in poverty measures using the Alkire Foster method)</p><p>To construct the GNH Index using this methodology six steps are followed:</p><p>1. Choose indicators</p><p>2. Apply suffi ciency thresholds (who has enough)?</p><p>3. Apply weights for each indicator</p><p>4. Apply the happiness threshold</p><p>5. Identify two groups:</p><p>1. Happy people (extensively and deeply happy)</p><p>2. Not-yet-happy people (policy priority) ) (unhappy and narrowly happy)</p><p>6. Identify among the not-yet-happy people, in what percentage of domains they lack suffi ciency,</p><p>and in what percentage they enjoy suffi ciency.</p><p>Figure 2: Identifying who is happy according to the GNHFigure 2: Identifying who is happy according to the GNH</p><p>65409_Earth_Chapter5v2.indd 130 4/30/12 3:57 PM</p><p>131</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Figure 2 uses an illustrative sample of seven people with nine domains to show how step fi ve works in practice.28</p><p>The people at the top have suffi ciency in the fewest domains, while those at the bottom have the most.</p><p>How do we move from this picture to the GNH? Here four out of seven people are not yet happy – 4/7 = 57%,</p><p>while 3 out of 7 people are happy – 3/7 = 43%. Once we have this fi gure, to compute the GNH Index, we only</p><p>need to know one more thing: Among the not-yet-happy people, in what percentage of domains do they enjoy</p><p>suffi ciency?</p><p>Figure 3 shows how we arrive at this fi gure. The not-yet-happy lack suffi ciency in 48.9% of domains, and enjoy</p><p>it in 51.1% of domains in this example.</p><p>To calculate the GNH, the data of the population are aggregated into a decomposable “Adjusted Headcount M0”</p><p>measure that is sensitive to the “breadth” of achievements (Alkire and Foster 2007, 2011). M0 is constructed</p><p>by multiplying HnAn, where Hn represents the percentage of people who have not achieved suffi ciency in six</p><p>domains thus are identifi ed as not-yet-happy, and An is the average proportion of dimensions in which those</p><p>not-yet-happy people lack suffi ciency.</p><p>The Adjusted Headcount ranges in value from 0 to 1, with larger numbers signifying greater insuffi ciencies</p><p>and less happiness. In order to create the GNH Index in which a higher number refl ects greater happiness, the</p><p>Adjusted Headcount is subtracted from 1 to obtain the GNH. GNH = 1- HnAn.</p><p>The GNH Index formulae can also be written GNH = Hh (Hn x As), where Hh is the percentage of happy people</p><p>[Hh = (1=H)] and As</p><p>is the percentage of dimensions in which the average not-yet-happy person enjoys suf-</p><p>fi ciency [As</p><p>= 1-An].</p><p>29 This way of presenting the same results focuses on happiness and suffi ciency; the fi rst</p><p>presentation focuses on the not-yet-happy people and their insuffi ciencies. Both formulae create the same</p><p>number, and both are useful in explaining the GNH Index. The GNH Index can be decomposed by population</p><p>sub-groups and broken down by indicators.30</p><p>Figure 3: Calculating the percent of domains in which not-yet-happy people lack suffi ciency</p><p>65409_Earth_Chapter5v2.indd 131 4/30/12 3:57 PM</p><p>132</p><p>So returning to our example, we take the following three numbers:</p><p>The percentage of happy people we call Hh</p><p>This is 43% in the example</p><p>The percentage of not-yet-happy people Hn</p><p>This is 57% in the example</p><p>The percentage of domains in which not-yet-happy people enjoy suffi ciency we call An</p><p>This is 54% in the example</p><p>They are then combined into a fi nal GNH formula as follows: GNH=(Hh+HnAs )= 57% + (43% x 48.9%) = 0.780</p><p>Now, to identify the happiness gradient, apply the two additional cutoffs – 50% and 77%. These enable the</p><p>identifi cation of the two additional groups.</p><p>As Figure 4 shows, when we apply the 50% cutoff we fi nd that only one person, Thinley, is unhappy. Looking</p><p>between 50-65% we fi nd three people are narrowly happy: Dorji, Jampel and Tashi. Two people have suffi cien-</p><p>cy in 66-76% of domains: Tshering and Chhimi. And fi nally, one person, Sangay, is deeply happy with achieve-</p><p>ments in over 77% of domains. We can compute the average suffi ciency for each group also: for example, in</p><p>the case of the narrowly happy people, the average suffi ciency is [(4.6/9 + 5/9 + 5/9)/3] = 54%. We could also</p><p>look at their composition (see Figure 21).</p><p>Figure 4: Happiness Gradient</p><p>65409_Earth_Chapter5v2.indd 132 4/30/12 3:57 PM</p><p>133</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>What does the GNH Index show us?</p><p>The index provides an overall picture of how GNH is distributed in Bhutan and can also be used to zoom in</p><p>to look at who is happy and those who are “not yet happy,” and to zoom further to look the unhappy, narrowly</p><p>happy, extensively happy, and deeply happy. The GNH can also be unpacked in different ways to tell different</p><p>stories. It can be decomposed by subgroups like Dzonkhags, age groups, gender, or some occupations. It can</p><p>also be analyzed by each dimension and indicator. All of these functions make it a useful tool for policymakers</p><p>as they seek to address the question of “how can GNH be increased?”</p><p>Overall, most Bhutanese enjoy suffi ciency in value, safety, native language, family, mental health, urbanization</p><p>issues, responsibility towards environment, satisfaction in life, government performance, healthy days and</p><p>assets. Between 50-60% of Bhutanese enjoy suffi ciency in ecological issues, negative emotions, community</p><p>relationship, artisan skills and Driglam Namzha. Less than half of Bhutanese enjoy suffi ciency in literacy, hous-</p><p>ing, donations, work, services, schooling, cultural participation and knowledge.</p><p>Each of the GNH indices is also reported for each of the 20 districts, by gender, by rural-urban area, and, for</p><p>illustrative purposes, by age and certain occupational categories. Standard errors are presented, as are robustness</p><p>tests for weights and cutoffs, measured with respect to group rankings and also, for the fi rst time, with respect</p><p>to the percentage contribution of each indicator.</p><p>Understanding happiness and who is happy</p><p>The GNH value is 0.737. It shows us that 40.8% of people in Bhutan have achieved happiness, even after the</p><p>structure of the GNH Index requires a wide array of conditions to be met. Those who are happy enjoy it in</p><p>56.6% of the domains, i.e. have suffi ciency in 56.6% of the 124 weighted conditions. Happiness according to</p><p>the GNH is reached when people reach suffi ciency in roughly four out of the six domains or the equivalent</p><p>proportion of conditions. How do the lives of happy people look? We fi rst look at all happy people, and then</p><p>briefl y examine the “deeply happy” subset of them.</p><p>Domains</p><p>Figure 5 shows in which</p><p>domains happy people enjoy suffi ciency. We can see that all nine dimensions contribute to</p><p>GNH and no domain is unimportant. Happy people live relatively balanced lives.</p><p>Figure 5: In which domains do happy people enjoy suffi ciency?</p><p>65409_Earth_Chapter5v2.indd 133 4/30/12 3:57 PM</p><p>134</p><p>Good health (14%), community (12%), ecology (12%), and psychological well-being (12%) contributed most to</p><p>GNH of happy people in 2010. Happy Bhutanese did not necessarily have high education (9%). Nor did they</p><p>score equally high in good governance (9%).</p><p>Indicators</p><p>Bhutanese enjoy highest suffi ciency in value, safety, native language, family, mental health, etc.</p><p>Figure 6: Indicators in which happy people enjoy suffi ciency</p><p>Figure 7: Indicators in which happy people lack suffi ciency</p><p>65409_Earth_Chapter5v2.indd 134 4/30/12 3:57 PM</p><p>135</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>The indicators in which happy people still often lack suffi ciency were knowledge, participation in festivals,</p><p>donations, having more than six years of schooling, enjoying government services, participating politically, and</p><p>believing in the practice of Driglam Namzha.</p><p>Dzongkhag (district)</p><p>The GNH reveals a large amount of equality between the regions and the range between regions is very small.</p><p>One district is probably the unhappiest – Samdrup Jongkhar.</p><p>Figure 8: GNH index by dzongkhag (district)</p><p>65409_Earth_Chapter5v2.indd 135 4/30/12 3:57 PM</p><p>136</p><p>GNH ranks districts differently than does per capita income. Thimphu (the capital) is not ranked highest in</p><p>GNH terms yet it has the highest per capita income of any district of Bhutan. Dagana and Zhemgang do much</p><p>better in GNH than on income criteria.</p><p>The composition of happiness changes somewhat across Dzongkhags. Thimphu does better in terms of education</p><p>and living standards, but worse in community vitality. Thimphu and Chukha are also home to the highest</p><p>number of happy people – and the highest numbers of not-yet-happy people (they are the biggest two Dzong-</p><p>khags in terms of population) in absolute terms.</p><p>Figure 9: GNH compared with per capita income</p><p>Figure 10: How the nine domains contribute to happiness by Dzongkhag</p><p>65409_Earth_Chapter5v2.indd 136 4/30/12 3:57 PM</p><p>137</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Rural and urban populations</p><p>In general, rural people are less happy than urban people but it is rather balanced. 50% in urban areas are happy</p><p>on GNH criteria and 37% in rural areas. The composition of happiness also differs; in rural areas, community</p><p>vitality, cultural diversity and good governance contribute more to happiness. In contrast, living standards, edu-</p><p>cation and health contribute more to happiness in urban areas. Urban people have insuffi ciency in governance,</p><p>time use and culture, while in rural areas insuffi ciency is worst in education and living standards.</p><p>Gender</p><p>When we decompose the GNH Index by gender we see that men are happier than women. 49% of men are</p><p>happy, while only one-third of women are happy, a result that is both striking and statistically signifi cant.</p><p>Women do better in living standards and ecology. Men do better in education, community vitality and psychological</p><p>well-being. Men and women are about the same in health, time use, governance, and culture.</p><p>Figure 11: Contribution of domains to happiness by region</p><p>Figure 12: GNH index by gender</p><p>65409_Earth_Chapter5v2.indd 137 4/30/12 3:57 PM</p><p>138</p><p>Age groups</p><p>The psychological happiness variable asks people to say, on a scale of 0 to 10, whether they consider themselves:</p><p>0 (Not a very happy person) - 10 (Very happy person). The young are generally the happiest group in Bhutan.</p><p>Figure 13: Percentage of Bhutanese having suffi ciency according to gender</p><p>Figure 14: Self-reported happiness level by age group</p><p>65409_Earth_Chapter5v2.indd 138 4/30/12 3:57 PM</p><p>139</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Educational level</p><p>People who have been identifi ed as happy by the GNH Index don’t necessarily have good education. Those</p><p>who are educated to post-graduate level are a little bit higher, though a lack of formal education clearly goes</p><p>with lower happiness. We can also see that as education increases, the contribution of living standards and edu-</p><p>cation to happiness increases; the contribution of governance and culture decrease.</p><p>Occupation</p><p>Figure 15: GNH Index and percentage of happy people by educational level</p><p>Figure 16: GNH Index and percentage of happy people by occupational status</p><p>65409_Earth_Chapter5v2.indd 139 4/30/12 3:57 PM</p><p>140</p><p>The sample is not fully representative and these are not robust rankings. Members of the national work force</p><p>are clearly and strongly the unhappiest group – they are often poorly paid migrants doing manual labor such</p><p>as taking care of roads. Clearly, it is the worst group followed by farmers, the biggest group in the survey.</p><p>The deeply happy</p><p>Any analysis of the “happy” people would be incomplete without a brief exploration of the subset of happy</p><p>people who are identified as “deeply happy.” These comprise 8.3% of the population. Two-thirds of these are</p><p>male, and one-third are female. 69% of the deeply happy people live in rural areas, and 31% in urban areas.</p><p>The ages are spread from less than 20 years old to more than 65, with 59% of the deeply happy people being</p><p>less than or equal to 40 years old. Deeply happy people live in every single district of Bhutan, with the highest</p><p>numbers living in Thimphu, Samtse and Chukha. Still, only 12% of the deeply happy people live in Thimphu.</p><p>84% of the deeply happy people are married and 12% are never married; the rest are divorced, separated or</p><p>widowed. 26% of deeply happy people have no formal education; 28% have completed primary school; and</p><p>some deeply happy people pertain to the remaining categories of education. Finally, deeply happy people per-</p><p>tain to every occupational category except the national workforce. The highest share of deeply happy people is</p><p>farmers (34%) followed by civil servants (18%). This small snapshot of happiness across Bhutan shows that it</p><p>is accessible to people of different ages, occupational categories, regions, and educational backgrounds. The</p><p>fact that two-thirds of deeply happy people are men is of clear policy interest.</p><p>Deeply happy people, on average, enjoy sufficiency in 81.5% of the domains. However it can still be interest-</p><p>ing to look at the domains in which even they lack sufficiency. Interestingly, there are some insufficiencies in</p><p>each domain, although these are very low in health. Overall, deeply happy people have the lowest deprivations</p><p>across the four groups of happiness in health, living standards, time use, and psychological well-being. They</p><p>have the highest relative (not absolute) contributions from deprivations in governance and culture.</p><p>The many faces of GNH</p><p>The GNH Index, like the philosophy of GNH that motivates it, is very much a living experiment, seeking to convey</p><p>more fully the color and texture of people’s lives than does the standard welfare measure of GNI per capita. It reflects</p><p>the fact that happiness is a deeply personal matter and people will rarely agree on a set definition. Indeed, happiness</p><p>has many faces, as the GNH survey shows. Here are the stories of just some happy people whose experiences of GNH</p><p>were captured in the 2010 survey and who were identified as happy by the GNH Index.</p><p>These profiles help to enrich our understanding of happiness according to GNH and show that different</p><p>groups – literate or illiterate, urban or rural, young or old, monk, farmer, or corporate worker, can all be happy</p><p>according to these models.</p><p>One such happy person in the GNH survey was a married corporate employee aged 35 living in urban Chukha. He</p><p>has completed 10th class, and has achieved sufficiency in nearly all indicators. He was a bit sleep deprived, and did</p><p>not feel a deep sense of belonging to his community, but was overall very satisfied with his life. When</p><p>asked what</p><p>contributed most to happiness he said: to be healthy, to meet basic needs, to have peace in the family, to be religious.</p><p>Another happy person whose experiences were captured in the GNH survey was a married woman farmer</p><p>aged 44 living in rural Tongsa. She was illiterate, and was deprived due to wildlife damage to her crops, and</p><p>thought she never felt forgiveness among the positive emotions – yet was happy. She mused that she felt</p><p>happy when she was able to do her household work, when she was harvesting potatoes, and as she wove.</p><p>Another happy person in the GNH survey was a widowed gomchen aged 70 living in rural Thimphu. He had</p><p>no formal education, and was deprived in education, housing, sleep and did not participate politically. He</p><p>observed that getting good agricultural products from the land contributes to happiness.</p><p>65409_Earth_Chapter5v2.indd 140 4/30/12 3:57 PM</p><p>141</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Another happy person as defi ned by the GNH Index is an unmarried young woman aged 26 living in urban</p><p>Tashigang. She completed a bachelor’s degree and is a civil servant living alone. She scores highly</p><p>across domains, although she misses a sense of belonging. When asked what contributes to her happiness</p><p>she replied: love, family, friends, education, and enough money.</p><p>Increasing Happiness: Policy implications</p><p>Aside from deepening our understanding of happiness, the GNH Index is formulated to provide an incentive</p><p>to increase happiness. Civil servants, business leaders, and citizens of Bhutan may ask, “how can I help to</p><p>increase GNH?” The GNH Index can help them answer this question in practical ways. It also enables the</p><p>Government and others to track changes over time. In general, there are two mechanisms by which public</p><p>policy action can be directed so as to increase GNH; it can either increase the percent of people who are</p><p>happy, or increase the percent of domains in which not-yet-happy people enjoy suffi ciency.</p><p>Insuffi ciencies by domain</p><p>To improve GNH we can look at people who are not-yet-happy and look at the areas where they lack</p><p>suffi ciency – 59% of Bhutanese are not-yet-happy, and they are deprived in roughly four domains each. The</p><p>not-yet-happy people are more deprived in all 33 indicators than the happy people (Figure 17). The biggest</p><p>deprivations are in education, living standards and time use. Among the not-yet-happy, women are unhappier</p><p>than men.</p><p>Rural people are less happy than urban people although their intensities are similar. But the composition of</p><p>insuffi ciencies varies. The urban groups have bigger insuffi ciencies in governance, time and culture and in</p><p>rural areas the biggest problems are education and living standards. The difference here is thus in terms of</p><p>the more material domains versus those that are about community, culture and spirituality. In Thimphu, the</p><p>capital, for example, the biggest deprivations are in community vitality.</p><p>Figure 17: Comparing percent of people who are insuffi cient among the happy and not-yet-happy</p><p>65409_Earth_Chapter5v2.indd 141 4/30/12 3:57 PM</p><p>142</p><p>Across all indicators we see that there is no indicator in which “happy” people have less suffi ciency than</p><p>not-yet-happy people. Looking at psychological well-being, health, and time, we see that the “not-yet-happy”</p><p>always have higher insuffi ciency. The groups are closest in sleep. In education, culture, and governance, the</p><p>groups are least different than in value, language, Driglam Namzha, and political participation. Both have</p><p>highest deprivations in education. In community, ecology, and living standard, the strong differences are in</p><p>wildlife damage and in living standard. Happy people’s insuffi ciencies in community and ecology are otherwise</p><p>rather close, and in urbanization, almost equal.</p><p>Health is the lowest contributor to unhappiness followed by community vitality. Education is the highest</p><p>contributor to unhappiness. In turn, we can break apart each domain and see how its individual indicators</p><p>look, to see where the biggest sources of unhappiness are coming from.</p><p>Figure 19 illustrates this for the education domain. The highest insuffi ciency is in the knowledge indicator. Bhuta-</p><p>nese experience low levels of knowledge in cultural and historical aspects of the country and in health and politics.</p><p>Figure 18: Understanding what constitutes unhappiness</p><p>Figure 19: Contribution of Education indicators to unhappiness</p><p>65409_Earth_Chapter5v2.indd 142 4/30/12 3:57 PM</p><p>143</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Who can increase GNH?</p><p>Increasing happiness is not only the business of government. The GNH requires civil servants, people in</p><p>their personal lives, business leaders and others to ask how they can increase the GNH. It tries to offer the</p><p>Index as a public good. His Majesty the King Jigme Khesar Namgyel Wangchuk clearly mentions that:</p><p>“Our nation’s vision can only be fulfi lled if the scope of our dreams and aspirations are matched by</p><p>the reality of our commitment to nurturing our future citizens.”</p><p>The people who are not-yet-happy are an important policy priority and thus it is important to look at the areas</p><p>in which they enjoy suffi ciency and the percent of domains in which they still lack suffi ciency. Government,</p><p>monasteries, communities, and individuals and households efforts can contribute to increasing GNH.</p><p>While responsibility for some indicators is shared across government, community and households, there is</p><p>a lot of overlap between the areas of actions.</p><p>Insuffi ciencies by Happiness Group</p><p>Figure 21 shows the percent contribution of each domain to the insuffi ciency of the four population groups</p><p>that we identifi ed. As can be seen, clearly the average insuffi ciency is lowest, as we would expect, among</p><p>the deeply happy group. We can also see that the absolute contribution of each indicator is the lowest in the</p><p>deeply happy group. The biggest contributions to insuffi ciency among the unhappy are living standards, edu-</p><p>cation, and psychological well-being – a combination of traditional and innovative measures of well-being.</p><p>Time pressures and a lack of governance including access to services is also very high. Deprivations in com-</p><p>munity and ecology contribute relatively less to insuffi ciencies of the not-yet-happy.</p><p>Figure 20: Overlapping responsibilities for increasing happiness</p><p>65409_Earth_Chapter5v2.indd 143 4/30/12 3:57 PM</p><p>144</p><p>The Unhappy</p><p>Those who achieve suffi ciency in less than half of domains are consider unhappy. In 2010, 10.4% of Bhuta-</p><p>nese were unhappy. Who are these people? 69% of the unhappy people are women and thirty one percent are</p><p>men. 84% of unhappy people live in rural areas. Although the unhappy come from every age cohort, 57% of</p><p>the unhappy are over 40 years old. Samtse, Tashigang, and Chukha are home to the most unhappy people,</p><p>followed by Thimphu and Samdrup Jonkhar, but there are some in each district nationally. And 76% of un-</p><p>happy people are married. While 90% of unhappy people have no formal education, others pertain to every</p><p>other educational category except that there are zero unhappy people who have completed a diploma or post-</p><p>graduate studies. 79% of unhappy people are farmers, but unhappy people are drawn from all occupations,</p><p>except that there are zero unhappy people among the monks, anim, GYT and DYT.</p><p>Across domains, the unhappy people show markedly higher percent contributions to their deprivations from</p><p>living standards, health deprivations, and psychological ill-being. This profi le of unhappiness, when con-</p><p>trasted with the profi le of the deeply happy people, is quite striking, in showing that no single category fi nds</p><p>happiness unattainable, but in the same way very few categories leave one “immune” from unhappiness, with</p><p>the possible exception of post-graduate education and the monastic or spiritually committed life. 31</p><p>Building GNH</p><p>The GNH has been presented to provincial district-level leaders to</p><p>allow them to review their policies against</p><p>the district-level results and see how they could alter policies according to the results. The wider goal is</p><p>to promote a public dialogue around the index so people share their own understandings and appreciate how</p><p>they could increase their own GNH. Policy and program screening tools have already been in use since the</p><p>2008 index, and all agencies, whether public or private, are encouraged to think holistically.</p><p>0%</p><p>10%</p><p>20%</p><p>30%</p><p>40%</p><p>50%</p><p>60%</p><p>70%</p><p>80%</p><p>90%</p><p>100%</p><p>Deeply Happy Extensively Happy Narrowly Happy Unhappy</p><p>Living</p><p>Standards</p><p>Ecology</p><p>Community</p><p>Governance</p><p>Culture</p><p>Education</p><p>Time Use</p><p>Health</p><p>Psych WB</p><p>Insufficiency</p><p>across</p><p>domains</p><p>Figure 21: Insuffi ciencies across the Happiness Gradient</p><p>65409_Earth_Chapter5v2.indd 144 4/30/12 3:57 PM</p><p>145</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>As His Majesty the King said, “GNH has come to mean so many things to so many people but to me it signifies</p><p>simply - Development with Values.”</p><p>“We strive for the benefits of economic growth and modernization while ensuring that in our drive to acquire</p><p>greater status and wealth we do not forget to nurture that which makes us happy to be Bhutanese. Is it our</p><p>strong family structure? Our culture and traditions? Our pristine environment? Our respect for community</p><p>and country? Our desire for a peaceful coexistence with other nations? If so, then the duty of our government</p><p>must be to ensure that these invaluable elements contributing to the happiness and well-being of our people</p><p>are nurtured and protected. Our government must be human.” (The Madhavrao Scindia Memorial Lecture</p><p>delivered by His Majesty the King, 23 December 2009 in New Delhi.)</p><p>65409_Earth_Chapter5v2.indd 145 4/30/12 3:57 PM</p><p>146</p><p>Appendix: Methodology: GNH Index</p><p>Let dn</p><p>M</p><p>, denote the set of all dnu matrices. The typical element dn</p><p>My</p><p>,� is the matrix of</p><p>achievements of n people in d different dimensions. For every</p><p>��</p><p>i 1,2,...,n and dj ,...,2,1 , the typical entry</p><p>ij</p><p>y of y is individual i´s achievement in dimension j. The row vector ),....,,( 21 idiii</p><p>yyyy contains individual</p><p>i�´s achievements in the different dimensions; the column vector ),....,,(. 21 njjjj</p><p>yyyy ' gives the distribution of</p><p>achievements in dimension j across individuals. Let 0!</p><p>j</p><p>z be the sufficiency cutoff value in dimension j. The sum</p><p>of entries in any given vector or matrix v is denoted by |v|, while P(v) is used to represent the mean of v (or |v|</p><p>divided by the number of entries in v).</p><p>For any matrix y, it is possible to define a matrix of deprivations from sufficiency ][ 00</p><p>ij</p><p>gg , whose</p><p>typical element 0</p><p>ij</p><p>g is defined by 10</p><p>ij</p><p>g when</p><p>jij</p><p>zy � , and 00</p><p>ij</p><p>g when</p><p>jij</p><p>zy t . That is, the th</p><p>ij entry of</p><p>the matrix is 1 when person i has not achieved sufficiency in dimension j, and 0 when he/she is sufficient.</p><p>For each of the d dimensions we apply a weighting vector Ȧd such that</p><p>1</p><p>1</p><p>j</p><p>j</p><p>Z ¦ . The insufficiency</p><p>profile of person i is then generated by summing the weights of the dimensions in which person i has not achieved</p><p>sufficiency.</p><p>Following the methodology to identify the multidimensionally poor proposed by Alkire and Foster (2007),</p><p>let</p><p>k</p><p>U be the identification method such that 1),( zy</p><p>ik</p><p>U when kc</p><p>i</p><p>t , and 0),( zy</p><p>ik</p><p>U when kc</p><p>i</p><p>� .</p><p>That means that a person is identified as not having achieved happiness if he or she does not have sufficiency in at</p><p>least k dimensions. Once identification is applied, a censored matrix )(0</p><p>kg is obtained from 0</p><p>g by replacing the</p><p>i</p><p>th row with a vector of zeros whenever 0),( zy</p><p>ik</p><p>U . This matrix is used to generate the GNH Index and to</p><p>analyze how happiness might be increased.</p><p>To construct the GNH Index, we first construct an Adjusted Headcount, given by 0</p><p>0 ( ( ))M g kP , which</p><p>is the sum of the weighted indicators of those people who do not enjoy sufficiency in any indicator ( |)(| 0</p><p>kg )</p><p>divided by total the number of people ( n ). It can also be expressed as HA where H is the Headcount Ratio</p><p>);( zyHH defined by nqH / , where q is the number of people in set</p><p>k</p><p>Z . A is the average percentage of</p><p>dimensions in which people who are not yet happy experience insufficiency, and is given by | ( ) | /( )A c k q . M0</p><p>summarizes information on the incidence of unhappiness and the average proportion of dimensions in which a not</p><p>yet happy person lacks sufficiency. It satisfies dimension monotonicity and is also decomposable by population</p><p>groups.</p><p>The GNH is constructed by subtracting M0, from unity; that is, it is GNH = 1- M0.</p><p>The measure M0, like all members of the );( zyMD family, is decomposable by population subgroups.</p><p>Given two distributions x and y, corresponding to two population subgroups of size )(xn and )(yn</p><p>correspondingly, the weighted average of sum of the subgroup poverty levels (weights being the population shares)</p><p>equals the overall poverty level obtained when the two subgroups are merged:</p><p>0 0 0</p><p>( ) ( )( , ; ) ( ; ) ( ; )</p><p>( , ) ( , )</p><p>n x n y</p><p>M x y z M x z M y z</p><p>n x y n x y</p><p>�</p><p>Clearly, this can be extended to any number of subgroups such as Dzongkhags, women and men, rural and</p><p>urban, and so on.</p><p>Additionally, once the identification step has been completed, the 0M index can be broken down into</p><p>indicator. To see this, note that M0 can be expressed in the following way: 0</p><p>0 *1</p><p>( ; ) ( ( ))n</p><p>ji</p><p>M y z g kP</p><p>¦ , where</p><p>0</p><p>* j</p><p>g is the jth column of the censored matrix 0 ( )g k . Thus 0</p><p>* 0( ( ( ))) / ( ; )</p><p>j</p><p>g k M y zP is the contribution of indicator</p><p>j to the overall shortfalls in Gross National Happiness. Itemizing these shortfalls clearly provides information that</p><p>can be useful for government policy.</p><p>32</p><p>65409_Earth_Chapter5v2.indd 146 4/30/12 3:57 PM</p><p>147</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>1 The guide was carried out in 2010 with funding from UN organizations based in Thimphu particularly UNDP. The analysis of GNH data and</p><p>reporting was funded by IDRC of Canada. A complete and detailed report on 2010 GNH Index will be printed in May 2012, also funded by IDRC</p><p>of Canada. The Royal Government of Bhutan and the Centre for Bhutan Studies express their gratitude for the generous support given by IDRC.</p><p>2 Extracted from RjeMkhan-po 10, Bstan ’dzinChosrgyal, Lho’ichos ‘byungbstan pa rinpoche’i ‘phromthud ‘jam mgonsmonmtha’i</p><p>‘phrengbazhesbyaba. Written during the years 1755-59. The Legal Code dated 1729 (earth bird year) is attributed to the 10th Desi-</p><p>MiphamWangpo while he was serving on the Golden Throne of Bhutan, as representative of the Shabdrung Rinpoche, and based on</p><p>the Shabdrung’s earlier work. KMT, Thimphu has reprinted this book 2004. See p. 253.</p><p>3 http://www.educatingforgnh.com/</p><p>4 Opening Address of “Educating for Gross National Happiness” Conference: Lyonchhen Jigmi Y. Thinley, Thimphu, Bhutan 7th December, 2009</p><p>5 The 10th plan of Bhutan specified GNH by focusing on four pillars: “In order to translate the multi-dimensional concept of GNH into</p><p>core objectives… four strategic areas were initially defined” (p.16). These areas, called the “four pillars of GNH,” are: 1. Sustainable and</p><p>equitable socio-economic development; 2. Environmental conservation; 3. The preservation and promotion of culture; and 4. Good</p><p>governance.</p><p>6 Royal Government of Bhutan (2008).</p><p>7 Royal Government of Bhutan (2009).</p><p>8 The report narrates an extensive review of the composition of subjective well-being into two major components; first, the evaluation of</p><p>a person’s life as a whole or of various domains and second, the measurement of the actual feelings. Both the components are reflected</p><p>in the psychological well-being domain of GNH and were computed separately. The report states, “that these measures provide informa-</p><p>tion about the determinants of quality of life at the level of each person. These determinants include both features of the environment</p><p>where people live and their individual conditions,</p><p>and they vary depending on the aspect considered.” Further, it highlights that these</p><p>subjective measures provide information beyond what is being given by income.</p><p>9 A five item Likert scale was used rather than the single item question on life satisfaction because dissatisfaction in life is usually due to</p><p>dissatisfaction in any of multiple areas of life. One of these areas can pull down the satisfaction level (Diener, 2006).</p><p>10 A number of different time frames have been used in various studies (Green, Goldman and Salovey 1993; Watson, Clark and Tellegen</p><p>1988; Watson and Tellegen 1999). The use of a “few weeks” reference period is not ideal; ideally we would have information on aver-</p><p>age emotional experiences throughout the past year. But this may be too difficult to recall accurately. The GNH emotional indices will</p><p>be partly inaccurate as a reflection of annual emotional states at the individual level because “the past few weeks” will not have been</p><p>representative for all respondents. However they were the best that could be constructed from the available data.</p><p>11 Jeffrey Hopkins defines karma as “A general term used loosely for behavioral cause and effect. Also called: karmic impulse.” On another</p><p>occasion, Hopkins has stated that “karma has the dual meaning of past actions that shape the present, and present intentions and actions that</p><p>will shape the future. Intention is the heart of karma, the very heart. What does intention mean? …In the teachings, there are descriptions of</p><p>a mind basis of all, the alaya-vijnana, which serves as a medium for karma. There are also descriptions of a subtle mental consciousness that</p><p>serves as the medium for the infusion of karma. And then interestingly, there is the description of the person as the medium of karma, which</p><p>is rather fascinating.” Available at Accessed on [2.14.2012]</p><p>12 An examination of the underlying factor structure resulted in a single factor with loadings above 0.5. Internal consistency was sufficient</p><p>(Cronbach’s alpha of .65) to allow computation of an indicator.</p><p>13 It may be that in future surveys the response categories might be altered.</p><p>14 Ura, K., 2012. Dialogue on Time and Time Use, forthcoming.</p><p>15 Work encompasses the following activities: Agriculture related activities; Guarding crops from wild animals; Livestock related activities;</p><p>Forestry related activities and related travels; Horticulture related activities; Processing of foods and drinks; Construction or repair of</p><p>private infrastructures in GNH 2010 data; Construction or repair of public infrastructure; Weaving and related works; Carpentry and</p><p>masonry; Others crafts; Business, trade and related travels; Services and related travels; Ferrying, carrying, transporting and related travels;</p><p>Cooking; Serving or entertaining; Dishwashing; Cleaning or upkeep of dwellings; Building fire; Fetching water; Laundry; Shopping; Ar-</p><p>ranging, mending household objects; Consultations with, engaged during the visits of official or office visits to professionals; Mining and</p><p>quarrying related activities; Care of children, old, sick and disabled; Woola (labor contribution to community works); Voluntary works and</p><p>informal helps. Since time spent on this activities is calculated separately, the classification of work and non-work can be changed easily, if</p><p>necessary, e.g., care of children, old, sick and disabled can be taken as an activity under social and cultural activities.</p><p>16 There are numerous studies that have used different stages of performance indicators such as input, output, outcome etc. (Boyne and</p><p>Law 1991; Sorber 1993; Duckett and Swerissen 1996; Hedley 1998; Stone and Cutcher-Hershenfeld 2001). A strong association be-</p><p>tween subjective and objective indicators for outcome performance indicators has been confirmed by Torenvlied and Akkerman (2009)</p><p>in their multi-stage performance indicator research paper. For Bhutan, the performance index is based on outcome indicators.</p><p>17 The response category also has the option of “don’t know” that has been re-categorized into mid-value “average,” which is considered</p><p>a deprived category. This has no major impact on the results since individuals are expected to have some knowledge of the functioning</p><p>of the institutions and so “don’t know” is inherently deprived.</p><p>65409_Earth_Chapter5v2.indd 147 4/30/12 3:57 PM</p><p>148</p><p>18 Similar concepts can be found in the following reports: Doolittle and McDonald 1978; Ahlbrandtand Cunningham 1979; Wanders-</p><p>man and Giamartino 1980; Riger and Lavrakas 1981; Bachrach and Zautra 1985; Davidson and Cotter 1986.</p><p>19 The poverty line given here is a measure for absolute poverty developed by the National Statistical Bureau of Bhutan in 2007 and is</p><p>based on food and non-food needs.</p><p>20 See for example, Gordon (2006) and Hillyard et al (2003).</p><p>21 The GNH data poverty line has been adjusted for the difference in the medians between BLSSR data and GNH data. Poverty line for</p><p>GNH data = Poverty line (PAR 2007)*Median (BLSSR data)/Median (GNH data).</p><p>22 The questionnaire for income and expenditure in the GNH Survey differed from the BLSS, and the GNH data had different median</p><p>and mean values from the BLSS as well as different district rankings by poverty and average per capita income. As a result, in the</p><p>income indicator, we implemented the sufficiency threshold of 1.5 times the poverty line in the original BLSS 2007 dataset, to obtain</p><p>the percentage of people who enjoyed sufficiency in income. We then mapped the same percentage onto the GNH income per capita</p><p>data. In using the percentage from BLSS data we are assuming that the distribution in both surveys is equivalent and that the percent-</p><p>age of people who enjoy 1.5 times the poverty line in 2010 is the same as in 2007, both of which are strong assumptions.</p><p>23 The asset index developed by Filmer and Pritchett (1999) has been used in Demographic and Health Surveys (DHS) to estimate</p><p>reasonable wealth effects.</p><p>24 Enumerators of the GNH surveys pointed out that the asset index was more accurate since it is easier for respondents to reflect on</p><p>their ownership than on income. Additionally, enumerators could confirm the ownership by actually seeing goods in the household.</p><p>So, the asset index is less likely to contain reporting bias.</p><p>25 The analysis is based on focus group discussions conducted by Dr. Alkire, Tshoki Zangmo and Tshering Phuntsho in Wangdipho-</p><p>drang and Punakha in 2011.</p><p>26 Many studies have confirmed that good housing is at the top of the hierarchy of human needs (Burns and Grebler 1986; Kiel and</p><p>Mieszkowski 1990).</p><p>27 These are just some of the studies that show the impact of housing quality on welfare. For example, Housing, Health and Climate</p><p>Change: Developing Guidance for Health Protection in the Built Environment: Mitigation and Adaptation Responses, World Health Organ-</p><p>isation (2010).</p><p>28 Note that this is a simplification: the actual calculation uses 33 indicators and calculates an individual deprivation profile based on</p><p>these rather than only nine domains, but the same principles apply.</p><p>29 This is a very simple re-arrangement as follows: GNH =1- HnAn = 1- HnAn –Hn + Hn = (1-Hn) + (Hn- HnAn) = (1-Hn)+ (Hn)(1- An) = Hh +</p><p>(Hn x As), since (1-Hn)= Hh and (1- An)= As.</p><p>30 The GNH is subgroup consistent and decomposable and satisfies dimensional monotonicity. It is related to Alkire and Foster’s</p><p>M0 measures which satisfy key additional properties such as Symmetry, Scale invariance, Normalization, Replication invariance,</p><p>Poverty Focus, Weak Monotonicity, Deprivation Focus , Weak Re-arrangement, as well as Dimensional Monotonicity, and Decom-</p><p>posability. See Alkire and Foster 2011.</p><p>31 Recall that sample sizes are such that the decompositions by occupational group and higher education cannot be taken to be repre-</p><p>sentative but are shared for illustrative purposes only.</p><p>32 Note that in some cases the sufficiency cut-offs are</p><p>used as a current emotional report- “How happy are you now?,” sometimes as a remembered emotion, as in</p><p>“How happy were you yesterday?,” and very often as a form of life evaluation, as in “How happy are you with your</p><p>life as a whole these days?” People answer these three types of happiness question differently, so it is important</p><p>to keep track of what is being asked. The good news is that the answers differ in ways that suggest that people</p><p>understand what they are being asked, and answer appropriately. Thus when people are asked about their</p><p>happiness now or yesterday, the answers are closely correlated with current activities and events in their lives today</p><p>or yesterday. By contrast, when people are asked how happy they are with their lives a whole these days, their</p><p>answers match very closely the answers to other similar evaluations of life as a whole.6</p><p>We shall return later to more detailed discussions of the meaning and validity of different measures. The</p><p>introduction above is intended to provide a springboard for our initial description of world happiness. The specific</p><p>data we use are drawn from the Gallup World Poll (GWP), the World Values Survey (WVS), the European Values</p><p>Survey (EVS), and the European Social Survey (ESS). We shall start by presenting data from the Gallup World</p><p>Poll, since it provides far greater country coverage than is currently available from any other source. The Gallup</p><p>World Poll contains measures of positive and negative affect (yesterday) as well as a life evaluation. We shall</p><p>start with life evaluations, since they will be shown to depend much more on life circumstances, to have larger</p><p>and more stable international differences, and to be more readily and systematically explained. We shall then</p><p>consider the levels and uses of affect measures, and compare affect measures and life evaluations from other</p><p>surveys covering fewer countries.</p><p>Happiness Across the World</p><p>In the Gallup World Poll respondents are asked (using fresh annual samples of 1,000 respondents aged 15 or</p><p>over in each of more than 150 countries) to evaluate the quality of their lives on an 11-point ladder scale running</p><p>from 0 to 10, with the bottom rung of the ladder (0) being the worst possible life for them and 10 being the</p><p>best possible. We begin with this ladder measure, which we sometimes refer to as the Cantril ladder,7 because</p><p>it currently covers the widest span of countries. Figure 2.1 gathers together the responses from all available</p><p>Gallup World Polls, from 2005 through mid-2011, and weights them by each country’s population aged 15 and</p><p>up to show the state of world happiness. There are 11 columns in the figure, one for each possible answer to</p><p>12</p><p>the question. The total height of each bar represents the number of people in the world, aged 15 and over (the</p><p>population being surveyed), who give that score for their evaluation of life today. Because of the large number</p><p>of countries covered, Figure 2.1 provides the broadest measure of the level and distribution of world happiness</p><p>in the second half of the first decade of the 21st century.</p><p>What do the data show? Over one-fourth of the world’s population give answers of 5, which is exactly the mid-point of</p><p>the range of possibilities. In every country there are life evaluations covering the whole range of possible</p><p>answers, from 0 to 10. The differences within each country reflect differing life circumstances and personalities,</p><p>and perhaps whatever else was in the respondents’ minds when the question was asked. It has even been argued</p><p>by some that individuals have their own personal set points for their happiness, as determined by their personali-</p><p>ties. In this view, while good or bad experiences might push people away from their set points, they eventually</p><p>adapt to the new circumstances, and revert to their set point. If this were generally true, the world distribution of</p><p>happiness answers in Figure 2.1 could tell us little about the economic and social circumstances of people’s lives.</p><p>It would reflect instead the distribution of more and less happy personality types.</p><p>On the contrary, Figure 2.2 presents the distributions of answers in each of nine country groupings to show</p><p>that life evaluations reflect much more than individual personality differences. Average life evaluations differ</p><p>a great deal from continent to continent, as shown in Figure 2.2, and even more from the top to the bottom</p><p>of the country rankings, as shown in Figure 2.3. There are very large differences in average life evaluations</p><p>across world regions, with a difference exceeding 3 points on the 11-point scale between a group of industrial</p><p>countries and sub-Saharan Africa. Even more striking is the ability of just a few differences in average life</p><p>circumstances, including per capita incomes, healthy life expectancy, having friends to count on in times of</p><p>need, having a sense of freedom to make life choices, and absence of corruption to explain almost all (more</p><p>than 95%) of these inter-regional differences.</p><p>Happiness, like income, is unequally distributed within and among nations. As is shown in Table 2.1, however,</p><p>the variation of happiness across the world’s population is largely within countries, while this is much less so</p><p>for incomes. Thus 42% of the worldwide variation in log of household incomes is between countries, much</p><p>higher than the corresponding percentages for subjective well-being, which are 22% for the Gallup World Poll</p><p>ladder and 7% for happiness yesterday. The primary reason for the difference is that income is but one of the</p><p>supports for happiness, and most of the other supports are much more evenly spread across countries. However,</p><p>some of the economically poorest regions and countries also have lower trust and weaker social relations, both</p><p>of which have strong links to happiness.</p><p>Also apparent from Figures 2.1 and 2.2 is that in every region there is a broad range of life evaluations. To some</p><p>extent these simply reflect different personality types.8 But to a much greater extent they reflect different circumstances</p><p>of life, and predict different future life courses, within the same neighborhoods and nations.</p><p>Happiness Averages by Country</p><p>The several panels of Figure 2.3 show ranked ladder averages by country, with horizontal lines showing the</p><p>95% confidence bands. Data from several years are combined, so that the sample size is several thousand for</p><p>most countries. This large sample size, coupled with the fact that the year-to-year changes in happiness averages</p><p>are small relative to the inter-country differences, means that it is possible to establish many significant</p><p>inter-country differences.</p><p>Because of the variety of national experiences, the range of national happiness averages is even greater than</p><p>for groups of countries. The top four countries (all in Northern Europe) in Figure 2.3 have life evaluations averaging</p><p>7.6, compared to 3.4 in the bottom four (all in sub-Saharan Africa). As described in more detail in Chapter 3,</p><p>13</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>about 80% of these inter-country differences can be attributed to the same few variables measuring the</p><p>material, social and institutional supports for a good life. All of these supports are stronger in the high-ranking</p><p>countries. Comparing the top four to the bottom four countries, average incomes are 40 times higher, healthy</p><p>life expectancy is 28 years greater, people are much more likely to have someone to call on in times of trouble</p><p>(95% vs. 48%), to have a sense of freedom (94% vs. 63%), and are less likely to perceive widespread corruption</p><p>in business and government (33% vs. 85%).</p><p>Returning to the question of the distribution of subjective well-being within countries, the panels of Figure 2.4</p><p>show the standard deviations of each country’s distribution of life evaluations. The countries are listed in the</p><p>same order as in Figure 2.3, so as to illustrate the great extent to which inequality of well-being differs</p><p>identified as weak rather than strong; this is explained in the domains and indicators sec-</p><p>tion.</p><p>65409_Earth_Chapter5v2.indd 148 4/30/12 3:57 PM</p><p>149</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Cited References</p><p>Afsa, Cédric; Blanchet, Didier; Marcus, Vincent; Pionnier, Pierre-</p><p>Alain; Rioux, Laurence; Mira d’Ercole, Marco; Ranuzzi, Giulia</p><p>Schreyer, Paul (2008). ‘Survey of Existing Approaches to Measuring</p><p>Socio-Economic Progress’, Paris: Commission on the Measurement</p><p>of Economic Performance and social Progress (April) [available at</p><p>http://www.stiglitz-sen-fitoussi.fr/documents/Survey_of_Existing_</p><p>Approaches_to_Measuring_Socio-Economic_Progress.pdf accessed</p><p>on 14 September 2011].</p><p>Ahlbrandt, Roger S. and Cunningham, James V. (1979). A new public</p><p>policy for neighborhood preservation, New York: Praeger.</p><p>Albritton, Robert B. and Bureekul, Thawilwadee (2009). ‘Are Democracy</p><p>and ‘Good Governance’ Always Compatible? Competing Values in</p><p>the Thai Political Arena’, Santiago: International Political Science</p><p>Association - Asian Barometer Project Office, (Working Paper</p><p>Series: No. 47) [available at http://www.asianbarometer.org/</p><p>newenglish/publications/workingpapers/no.47.pdf accessed on 10</p><p>October 2011].</p><p>Alkire, Sabina. (2002). ‘Dimensions of human development’, World</p><p>Development, 30, 2, 181-205.</p><p>Alkire, Sabina. (2008). ‘Choosing dimensions: The capability</p><p>approach and multidimensional poverty.’ In N. Kakwani & J. Silber</p><p>(Eds.), The Many Dimensions of Poverty (pp. 89-119). New York:</p><p>Palgrave Macmillan.</p><p>Alkire, Sabina and Foster, James (2011a). ‘Counting and Multidi-</p><p>mensional Poverty Measurement’, Journal of Public Economics, 95,</p><p>p. 476-487.</p><p>Alkire, Sabina and Foster, James. (2011b). ‘Understandings and</p><p>Misunderstandings of Multidimensional Poverty Measurement’,</p><p>Journal of Economic Inequality, 9, 2, 289-314.</p><p>Andrews, Rhys; Boyne, George A. and Walker, Richard M. (2006).</p><p>‘Workforce Diversity in the Public Sector: An Evaluation of the</p><p>Performance of English Local Authorities’, Policy and Politics, 34, 2,</p><p>pp. 287-306.</p><p>Aris, Michael and Hutt, Michael (eds.) (1994). Bhutan: Aspects</p><p>of Culture and Development, Gartmor: Kiscadale (Kiscadale Asia</p><p>Research Series No.5).</p><p>Arthaud-Day, Marne L.; Rode, Joseph C.; Mooney, Christine H. and</p><p>Near, Janet P. (2005). ‘The Subjective Wellbeing Construct: A Test</p><p>of its Convergent, Discriminant and Factorial Validity’, Social Indicators</p><p>Research, 74, pp. 445–476.</p><p>Bachrach, Kenneth M. and Zautra, Alex J. (1985). ‘Coping with</p><p>a Community Stressor: The threat of a hazardous waste facility’,</p><p>Journal of Health and Social Behavior, 26, pp. 127-141.</p><p>Baker, Michael; Stabile, Mark and Deri, Catherine (2003). ‘What do</p><p>Self-reported, Objective Measures of Health Measure?’, Journal of</p><p>Human Resources, 39, 4, pp. 1067-1093.</p><p>Bardasi, Elena and Wodon, Quentin (2009). ‘Working Long Hours</p><p>and Having No Choice. Time Poverty in Guinea’, Washington, D. C.:</p><p>The World Bank (Policy Research Working Paper, 4961).</p><p>Beach, Betty A. (1987). ‘Time Use in Rural Home-Working Families’,</p><p>Family Relations, 36, 4, October, pp.412-416.</p><p>Belcher, Anne R.; Dettmore, Diane and Holzemer, Stephen Paul</p><p>(1989). ‘Spirituality and Sense of Well-Being in Persons With</p><p>AIDS’, Holistic Nursing Practice, 3:4, pp. 16-25.</p><p>Bergner, M.; Bobbit, R. A.; Carter, W. B. and Gilson, B. S. (1981).</p><p>‘The Sickness Impact Profile: development and final revision of a</p><p>health status measure’, Medical Care, 19, pp. 787-805.</p><p>Blanchflower, David G. and Oswald, Andrew J. (2011). ‘Internation-</p><p>al Happiness’, Cambridge: National Bureau of Economic Research</p><p>(NBER Working Paper Series, 16668) [available at http://www.nber.</p><p>org/papers/w16668 accessed on 9 September 2011].</p><p>Bohen, Halcyone and Viveros-Long, Anamaria (1981). Balancing</p><p>jobs and family life: Do flexible work schedules help?, Philadelphia,</p><p>PA.: Temple University Press.</p><p>Boyne, George A. and Law, Jennifer (1991). ‘Accountability and local</p><p>authority annual reports: The case of Welsh district councils’, Financial</p><p>Accountability and Management, 7, pp. 179–194.</p><p>Bradburn, Norman M. (1969). The structure of Psychological Wellbeing,</p><p>Chicago: Aldine Publishing.</p><p>Brazier, J. E.; Jones, N. and Kind, P. (1993). ‘Testing the validity</p><p>of the Euroqol and comparing it with the SF-36 Health Question-</p><p>naire’, Quality of Life Research, 2, pp. 169-180.</p><p>Bruni, Luigino and Porta, Pier Luigi (eds.) (2007). Handbook on the</p><p>Economics of Happiness, Cheltenham: Edward Elgar.</p><p>Burchardt, Tania (2008). ‘Time and income poverty’, London: London</p><p>School of Economics, November (Centre for the Analysis of Social</p><p>Exclusion Report, 57).</p><p>Canadian International Development Agency (CIDA) (1996).</p><p>‘Policy for CIDA on Human Rights, Democratization and Good</p><p>Governance’, Quebec: CIDA, December. [available at http://acdi-</p><p>cida.gc.ca/INET/IMAGES.NSF/vLUImages/HRDG2/$file/HRDG-</p><p>Policy-nophoto-e.pdf accessed on 15 November 2011].</p><p>Canadian International Development Agency (CIDA) (1999).</p><p>‘CIDA’s Policy on Gender Equality’, Quebec: CIDA [available at</p><p>http://www.sice.oas.org/Genderandtrade/CIDA_GENDER-E_Policies.</p><p>pdf accessed on 9 November 2011].</p><p>Case, Karl E.; Quigley, John M. and Shiller, Robert J. (2005).</p><p>‘Comparing Wealth Effects: The Stock Market versus the Housing</p><p>Market’, Advances in Macroeconomics, 5:1, Article 1.</p><p>Centers for Disease Control and Prevention (CDC) (2000). ‘Measuring</p><p>Healthy Days. Population Assessment of Health-Related Quality of</p><p>Life’, Atlanta: CDC, November.</p><p>65409_Earth_Chapter5v2.indd 149 4/30/12 3:57 PM</p><p>150</p><p>Chavis, David M. (1983). ‘Sense of community in the urban envi-</p><p>ronment: Benefits for human and neighborhood development’,</p><p>doctoral dissertation, Nashville, TN.: Vanderbilt University.</p><p>Choden, D. and Namgay, K. (1996). ‘Report on the findings and</p><p>recommendations of the wild boar survey: project for assessment</p><p>of crop damage by the wild boar’, Thimphu: National Plant</p><p>Protection Centre–Ministry of Agriculture-Royal Government of</p><p>Bhutan.</p><p>Chophel, Sangay (2010). ‘Cultural Diversity and Resilience’,</p><p>Thimphu: The Centre for Bhutan Studies [available at</p><p>http://ns6.asphostserver.com/surveyReports/culture/culture.</p><p>pdf accessed on 9 November 2011].</p><p>Chouguley, Ulrike; Naylor, Richard and Rosemberg-Montes,</p><p>Cristina (2011). ‘Edinburgh Festivals Impact Study. Final Report’,</p><p>London: BOP Consulting, May.</p><p>Chowdhury, N. and Skarstedt, C.E. (2005). ‘The Principle of</p><p>Good Governance’, Oxford: Centre for International Sustainable</p><p>Development Law (CISDL), March.</p><p>Clark, Andrew E. and Senik, Claudia (2010). ‘Who Compares</p><p>to Whom? The Anatomy of Income Comparisons in Europe’,</p><p>The Economic Journal, 120 (May), pp. 573–594.</p><p>Commission of the European Communities (1992). ‘Euroba-</p><p>rometer 38’, Brussels: Commission of the European Communi-</p><p>ties [available at http://ec.europa.eu/public_opinion/archives/</p><p>eb/eb38/eb38_en.pdf accessed on 15 November 2011].</p><p>Constitution of Bhutan (2008). The Constitution of the Kingdom</p><p>of Bhutan, Thimphu: Constitution of Bhutan [available at</p><p>http://www.constitution.bt/TsaThrim%20Eng%20%28A5%29.</p><p>pdf accessed on 15 November 2011].</p><p>Cosmides, Leda and Tooby, John (2000). ‘The cognitive neuro-</p><p>science of social reasoning’, in Gazzaniga, M. S. (ed.), The New</p><p>Cognitive Neurosciences, Second Edition. Cambridge, MA: MIT</p><p>Press, pp. 1259-1270.</p><p>Cox, Dianne; Frere, Marion; West, Sue and Wiseman, John.</p><p>(2010). ‘The development and use of community wellbeing</p><p>Indicators: Learning from Community Indicators Victoria’,</p><p>Australian Journal of Social Issues, 45:1, pp. 71-88.</p><p>Cummins, Robert A.; Collard, James; Woerner, Jacqui; Weinberg,</p><p>Melissa; Lorbergs, Markus and Perera, Charini. (2008). ‘The</p><p>Wellbeing of Australians – Who Makes the Decisions, Health/</p><p>Wealth Control, Financial Advice, and Handedness. Part A: The</p><p>Report’, Victoria: Australian Unity Wellbeing Index (Report 22.0)</p><p>Davidson, Richard J. (2004). ‘Well-being and affective style:</p><p>neural substrates and biobehavioural correlates’, Philosophical</p><p>Transactions</p><p>among</p><p>countries. If the degree of within-country inequality in the distribution of happiness were the same in each</p><p>country (as measured by the coefficient of variation) then the standard deviations in Figure 2.4 would follow</p><p>the same gradual downward slope displayed by the Figure 2.3 country rankings of average life evaluations.</p><p>Figure 2.4 shows that this is far from reality. In general, there is no significant cross-country correlation</p><p>between country means and standard deviations for the global sample of countries. However, among the</p><p>OECD countries the correlation between county means and standard deviations is significantly negative, while</p><p>for the rest of the world the correlation is on average positive. Among those countries with high average scores,</p><p>some have quite high degrees of equality in the distribution of happiness (e.g. Denmark and the Netherlands),</p><p>while in some other fairly high-ranking countries (e.g. Costa Rica and the United States) there is much more</p><p>dispersion, and a higher proportion of the population has low life satisfaction. The OECD has recently reported</p><p>a growing inequality of income in almost all OECD countries over the past 20-odd years.9 There are not yet</p><p>sufficiently large and long samples of happiness data, and of social support data, to see whether this growing</p><p>inequality of income has been matched by growing inequality of happiness and its non-income drivers.</p><p>Comparing Different Measures</p><p>After starting with the ladder, since it is the data set with the best international coverage, we now turn</p><p>to illustrate the differences between the ladder, life satisfaction, overall happiness, and measures of positive</p><p>and negative affect. If all of these measures were collected from the same respondents, in the same surveys,</p><p>in comparable ways, then it would be relatively easy to see how they differ, both in average values and in the</p><p>stories they tell about why some lives are happier than others.</p><p>But most surveys only ask one or two subjective well-being questions, so it is instead necessary to proceed in</p><p>stages, using a number of pair-wise comparisons. But first some further distinctions need to be made between</p><p>different ways of getting individuals to report on their well-being. A first distinction is between experienced</p><p>and remembered well-being. Experienced well-being depends on moment-by-moment reports, usually of pleasure</p><p>or pain, and remembered well-being is reported subsequently, and is hence based on memory. A second</p><p>distinction relates to the time span of the emotion or event being experienced or remembered. For experienced</p><p>well-being, the time span is momentary, but for remembered well-being, the report can relate to a past</p><p>moment (how did you feel when something happened, or at noon), to the average for any particular event or</p><p>time period, whether yesterday, last week, your last holiday, your just-finished colonoscopy, or to your life as a</p><p>whole these days. A third distinction is between evaluations and emotional reports. An evaluation is inherently</p><p>a judgment about something, while an emotional report is more simply the description of an emotional state.</p><p>All three of these distinctions are potentially important. The reference to colonoscopies was deliberate. In a</p><p>well-known study, colonoscopy patients were asked to report their moment-by-moment pain levels, and were</p><p>later asked for a retrospective evaluation using the same scale.10 There was a systematic difference between the</p><p>average of momentary assessments and the retrospective evaluations, with the latter closely tracking the aver-</p><p>age of the peak and final momentary pain levels. If these two reporting forms give different answers, then</p><p>which should be taken to represent the true pain level? Some have argued that the true total pain is the sum of</p><p>the momentary assessments, and that therefore the retrospective reports are mistaken.11 Others have argued</p><p>14</p><p>that the ability to frame experiences into summary memories that help to inform future judgments is an</p><p>essential rather than a mistaken part of human nature.12 And when future decisions are made about having</p><p>colonoscopies, choosing where to go on spring break, or deciding to have another child, it is retrospective</p><p>evaluations that govern decisions.13 But most agree that “the remembering self and the experiencing self must</p><p>both be considered, because their interests do not always coincide.”14 This distinction does not impinge</p><p>directly on this current analysis of world happiness measures, since all of the affect reports currently available</p><p>are of remembered rather than experienced affect. This includes all of the Gallup World Poll affect measures</p><p>presented here, which refer to memories of emotions on the previous day.</p><p>The second and third distinctions, relating to the time span covered by the question, and whether the question</p><p>invites an evaluation of life or an emotional report, both remain important for our discussions here. As already</p><p>noted, the Gallup affect measures all relate to yesterday. Most life assessments are explicitly evaluative, asking</p><p>respondents to think of their lives as a whole, often nowadays, or alternatively at some past or future time.</p><p>Thus the three separate Cantril ladder questions in the Gallup World Poll ask respondents to evaluate their</p><p>lives “at the present time,” five years ago and five years in the future. The data in this report are all based on</p><p>answers to the “at the present time” question.</p><p>The life satisfaction question in the European Social Survey asks “All things considered, how satisfied are you with</p><p>your life as a whole nowadays?” (on a 0 to 10 scale). The World Values Survey asks almost the same life satisfaction</p><p>question, except that it uses “these days” instead of “nowadays,” and the response scale runs from 1 to 10.</p><p>Happiness measures sometimes relate to a specific moment or day, in which case it is appropriate to regard</p><p>them as reports of affect. This is the case with the positive affect question now being used widely by the U.K.</p><p>Office for National Statistics (ONS), which asks “Overall, how happy were you yesterday?.”15 By contrast, the</p><p>happiness question in the European Social Survey, and that asked in the World Values Survey, are</p><p>both evaluative in nature, and broader in their time coverage. For example, the European Social Survey asks</p><p>“Taking all things together, how happy would you say you are?” (on a scale of 0 to 10), while the World Values</p><p>Survey asks “Taking all things together, would you say you are: Very happy, Quite happy, Not very happy, or</p><p>Not at all happy?.”How do the various evaluation measures compare? Figure 2.5 shows the country-by-country</p><p>rankings for life satisfaction based on the Gallup World Poll and Figure 2.6 shows them for the combined</p><p>World Values Survey/European Values Survey. These rankings are very similar to those shown for the ladder</p><p>in Figure 2.3. If the Figure 2.5 country rankings for life satisfaction are compared with those for the Gallup</p><p>ladder responses asked of the same respondents, and in the same survey, the correlation is very high (r=0.94).</p><p>Analysis of the resulting data show that while there were significant differences in average scores, with the</p><p>mean of life satisfaction being higher by about 0.5 on the 11-point scale, the two variables are explained by the</p><p>same factors, including the same effects of income.16</p><p>What about comparisons between the answers from life evaluations based on happiness and those based on</p><p>satisfaction with life? This is best answered using data from the European Social Survey, which asks happi-</p><p>ness and life satisfaction questions in similarly evaluative ways, and on the same scales, for large samples of</p><p>respondents in 29 countries. Figures 2.7 and 2.8 show the country averages for life satisfaction and happiness,</p><p>respectively. Although the means of the two series differ significantly, with life satisfaction generally being</p><p>rated higher by 0.4 points on the 11-point scale,</p><p>tests of explanatory equations show that the same variables</p><p>explain happiness and life satisfaction, with generally similar coefficients, including the effects of income. In</p><p>addition, the ESS country rankings for happiness and life satisfaction are almost identical (r=0.987). Thus</p><p>when happiness is asked about in a life-evaluative mode, the answers have the same structures across</p><p>individuals and countries as do the answers to life satisfaction questions. Indeed, these structures are so similar</p><p>that taking an average of the life satisfaction and happiness answers for each respondent gives a combined</p><p>evaluation of life that is explained significantly more accurately than is either on its own. The same is true for</p><p>life satisfaction and ladder responses in the Gallup World Poll.17</p><p>15</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>Bhutan has not yet been included in the Gallup World Poll, but has used the European Social Survey happiness</p><p>question in its recent large (n=7,000) national survey, so that Bhutanese average happiness, equal to 6.05 on</p><p>the 0 to 10 scale, can be compared with that in the ESS countries. This is lower than the 7.01 average in the</p><p>latest ESS survey, but higher than in Russia, Ukraine and Bulgaria, significantly so in the latter two cases.</p><p>Compared to Bhutan’s near neighbors, less precise calculations rank Bhutanese happiness levels slightly above</p><p>those in India, and significantly above those in Nepal, China and Bangladesh.18</p><p>The bottom line of our comparisons among life evaluations is that when life satisfaction, happiness and ladder</p><p>questions are asked about life as a whole, they tell very similar stories about the likely sources of a good life.</p><p>The information base for these comparisons is still growing, however, so there may be some systematic differences</p><p>that appear in larger samples.</p><p>But when happiness is seen as an emotional report, and measured at a point in time, then it looks very like</p><p>other measures of positive affect. Thus “happiness yesterday” measured on a 0 to 10 scale as a positive affect</p><p>measure (as currently being asked by the ONS in the U.K.) has very different properties from life satisfaction,</p><p>asked on the same scale of the same respondents. The affect measure of happiness is much less correlated</p><p>with major life circumstances than are the life satisfaction answers, and the effects of income are much smaller,</p><p>and often statistically insignificant. This is similar to what was found when the Gallup ladder answers were</p><p>compared to the Gallup “yesterday” affect answers.19 The ONS and Gallup data both show that life evaluations</p><p>are much more closely related to life circumstances than are affect measures, and that positive affect is more</p><p>easily explained by life circumstances than is negative affect.</p><p>Figure 2.9 shows country rankings based on the 4-point evaluative happiness answers in the combined World</p><p>Values Survey/European Values Survey, while Figure 2.10 shows rankings, for a much larger number of countries,</p><p>of the average Gallup World Poll answers to a question asking about the respondent’s happiness yesterday</p><p>(using a yes/no 2-point response scale). We would not expect these country rankings to be very similar, and</p><p>they are indeed much less correlated than were the ESS happiness and life satisfaction rankings. The reasons</p><p>for difference relate to answer scales (binary for the Gallup happiness vs. a 4-point scale for the WVS/EVS</p><p>happiness question), to the difference between yesterday and some longer period, and to the related and more</p><p>fundamental distinction between evaluations and emotional reports.</p><p>Although short-term emotional reports carry much less information about life circumstances than do life</p><p>evaluations, they are very useful at revealing the nature and possible causes of changes in moods on an hour-</p><p>by-hour or day-by-day basis. They are hence of most use when asked in the context of time-use surveys that</p><p>provide scope for explaining these short-term changes.20</p><p>How do international differences in measures of affect compare with those for the more cognitive life evaluations?</p><p>Figure 2.11 shows Gallup World Poll country rankings for positive affect (the average of yes/no answers on the</p><p>frequency yesterday of enjoyment, happiness and laughter). Figure 2.12 does the same for negative affect (averages</p><p>for worry, sadness, anger and depression), and Figure 2.13 shows net affect (positive affect minus negative</p><p>affect).21 We should expect that proportionate differences between average country scores would be larger for</p><p>life evaluations than for affect, because the effects of income are relatively larger for the evaluative measures,</p><p>and the international differences in income are much greater than those for the non-economic supports for a</p><p>better life. This is consistent with the results in Figures 2.11 to 2.13, which show some interesting differences</p><p>in country rankings both between evaluations and affect, and between positive and negative affect. In all parts</p><p>of the world, the frequency of positive affect is two to three times greater than for negative affect. The countries</p><p>of Latin America and the Caribbean have higher than average rankings for the ladder and for positive affect, to</p><p>a slightly greater extent for the latter.</p><p>16</p><p>To summarize, life evaluations, whether they are general questions about life satisfaction, the ladder question</p><p>in the Gallup World Poll, or overall happiness questions of the sort used in the European Social Survey, all</p><p>give similar answers about the relative importance of the economic and social supports for a good life. When</p><p>overall happiness and life satisfaction questions are asked on the same scales, and of the same respondents,</p><p>the answers have very similar distribution, as shown in the two panels of Figure 2.14. Although the mean of</p><p>the ESS happiness answers is 0.4 larger than for SWL, the two measures are very highly correlated at both the</p><p>individual (r=0.67) and national (r=0.96) levels, are explained in the same way by the same variables, and are</p><p>usefully averaged to produce even more robust life evaluations.</p><p>Measures of positive and negative affect contain much less that differs from one community or country to</p><p>another (as shown in Table 2.1), but if collected in suitable ways can unravel important aspects of life as it is</p><p>actually experienced.</p><p>Table 2.1 Inter-Country Shares of Total Variance</p><p>Data Source Well-Being Measures Inter-Country Share of Total Variance</p><p>GWP 05-11 Cantril Ladder (life evaluation) 0.222</p><p>GWP 07-10 Life Satisfaction (life evaluation) 0.327</p><p>GWP 05-11 Happiness (yesterday) 0.068</p><p>GWP 05-11 Positive Affect (yesterday) 0.072</p><p>GWP 05-11 Negative Affect (yesterday) 0.042</p><p>GWP 05-11 Net Affect (yesterday) 0.061</p><p>GWP 05-11 Log of Income 0.422</p><p>ESS round 4 Life Satisfaction (life evaluation) 0.172</p><p>ESS round 4 Happiness (life evaluation) 0.146</p><p>ESS round 4 Log of Income 0.384</p><p>WVS 3-5 Life Satisfaction (life evaluation) 0.143</p><p>WVS 3-5 Happiness (four point scale) 0.115</p><p>Notes: (1) To construct numerical income from the categorical income class in ESS round 4, we use midpoints for non-top income categories</p><p>and 1.5*(bottom boundary) for the top income category. Household income in local currency units in ESS round 4 is converted to</p><p>international dollars by multiplying by the PPP conversion factor from WDI (2011). Note that the PPP conversion ratio for Slovakia is for</p><p>euro and international dollar, however, the household income is measured by Slovak crown in the survey. Household income for Slovak</p><p>respondents is then divided by 30.126, the official exchange rate between Slovak crown and euro, before applying the PPP conversion</p><p>factor. (2) WVS 3-5 refers to the WVS round 3-5 and EVS round 4-5.</p><p>Happiness has been shown to play a double role, sometimes appearing as an emotional report and at other</p><p>times in an evaluative role. Life satisfaction and other life evaluations, by contrast, always relate to life as a</p><p>whole, and show much less short-term</p><p>variation but much more linkage to life circumstances. Having</p><p>presented a range of measures of subjective well-being, we turn now to consider the extent to which they can</p><p>provide valid and policy-relevant guides to the quality of life.</p><p>Making the Case for Measuring Subjective Well-Being</p><p>Although almost 40 years have passed since Richard Easterlin advocated using measures of happiness to</p><p>assess the quality of people’s lives, systematic collection and use of subjective well-being data at the population</p><p>level have been slow to follow.22 In the meantime, several decades of research, mainly in psychology, have dug</p><p>17</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>deeper into the meaning, reliability, and validity of various measures of subjective well-being. The results of</p><p>this research strongly support wider collection and use of subjective well-being data.23</p><p>Why has it taken so long for subjective well-being to become more widely and routinely measured as part of the</p><p>statistical base for public information and decision-making? One reason is that in the absence of some crisis in</p><p>existing ways of collecting and using information, people tend to simply and often unconsciously24 apply and</p><p>use information and decision rules that have served them well in the past.25 It took many decades to establish</p><p>national systems of accounts for income and expenditure, and even then the developments were often driven by</p><p>the imperatives of wars or depressions, and the meaning and uses of the data were frequently contested. Hence</p><p>it should be no surprise that it has taken many years to raise baseline awareness to the point where widespread</p><p>official and private collection of subjective well-being data is starting to happen. Nor should it be surprising that</p><p>there are many skeptical questions posed about what the data mean and whether they are useful.</p><p>Here are some of the questions that have been asked, and how they have been answered:</p><p>Are subjective well-being measures reliable?</p><p>Within psychology, reliability is gauged by the extent to which the same questions yield identical answers when</p><p>administered in the same conditions. The replicability of subjective well-being measures has been tested in</p><p>a variety of ways, all of which combine to produce a reassuring picture.26 For example, life evaluations asked</p><p>of the same person in a sequence of surveys start high, and become less correlated as the intervening time</p><p>grows.27 This is just what should be expected, since underlying circumstances are more likely to have changed</p><p>over the longer period. Furthermore, multi-item measures average over random errors, and hence produce</p><p>higher reliability measures at the individual level.</p><p>At the group or national level, reliability is very high, since individual-level random variations and personality</p><p>differences are averaged away, while the underlying year-to-year changes in average life circumstances are relatively</p><p>modest. Hence the year-to-year correlations of country rankings of the ladder in the Gallup World Poll are</p><p>very high, averaging between 0.88 and 0.95. Similarly the wave-to-wave country-ranking correlations of both</p><p>happiness and life satisfaction in the European Social Survey are between 0.92 and 0.98. These correlations</p><p>gradually drop, as they ought to do, when the comparison dates become further apart.</p><p>Are subjective well-being answers valid?</p><p>There are three quite different ways of judging the validity of happiness measures. The first is to see to what</p><p>extent they are plausibly explained in terms of life circumstances and other candidate variables. The second is</p><p>to assess the extent to which they are correlated with other subjective and objective measures of well-being. The</p><p>third is to see how and whether the measures predict subsequent outcomes and behavior.</p><p>As will be shown in the next chapter, more than three-quarters of the cross-country differences in national</p><p>average measures of happiness can be explained by variables already known through experimental and other</p><p>evidence to be important. The fact that different measures of subjective well-being are explained by different</p><p>patterns of other variables represents a strength rather than a weakness, because in general the differential patterns</p><p>take exactly the form they should if the measures are valid. For example, Maslow’s theory of the hierarchy of</p><p>needs would suggest that the relative importance of income and social factors might differ between richer and</p><p>poorer countries.28 Research using data from the Gallup World Poll shows country differences of just this sort.</p><p>Although both social and economic conditions are important supports for life evaluations in all countries, the</p><p>relative importance of the social factors is higher in OECD countries.29</p><p>Second, subjective measures of well-being have been correlated with a variety of objective measures including</p><p>facial expressions, brain-wave patterns and cortisol measures at the individual level, and community and national</p><p>suicide patterns. Some have regarded these correlations as a necessary pre-requisite to taking subjective</p><p>measures more seriously. But why should this be necessary? In the case of happiness the subjective measure</p><p>18</p><p>itself is primary, with the coincidental movements of physical measures being reassuring but of less consequence.</p><p>Indeed, certain patterns of electrical activity in the brain became established as measures of happiness because</p><p>they tended to be present when people reported themselves to be happy. 30 Clearly what matters are the subjective</p><p>experiences, and not any associated electrical patterns. Nonetheless, the correlations are reassuring to those</p><p>who are concerned about interpersonal and intercultural differences in how people use words and scales when</p><p>making their reports.</p><p>The ability of measures of subjective well-being to predict subsequent events and behavior is relevant for two</p><p>main reasons. First, predictive power is a straight forward test of validity. The ability of life evaluations in</p><p>large populations to predict subsequent suicide frequencies provides strong evidence that life evaluations are</p><p>important to behavior.31 The same point can be made for the ability of measures of positive affect to predict a</p><p>variety of good outcomes32 especially including health33 and mortality.34</p><p>Second, the fact that happiness measures are predictive of sickness and death feeds back to strengthen the case</p><p>for collecting measures of happiness as a regular part of health maintenance and the delivery of health care.</p><p>How sensitive are results to question wording and placement?</p><p>A well-known study35 hypothesized, following influential philosophical work on the logic of conversation,</p><p>that if a general question follows a related specific one the answer to the specific question will help set the</p><p>context for the general question, and will hence be likely to influence the answer to it.36 Thus the researchers</p><p>presumed, and found, that when students in Illinois were asked about how happy they were with their recent</p><p>dating experiences and how happy they were with their lives as a whole, the answers were more closely</p><p>correlated when the dating question was asked before the general question. But when the two questions were</p><p>presented as relating to one another, the ordering effect shrank to insignificance. The first part of the result</p><p>has been used by some to question the reliability of subjective assessments, but the two parts seen together</p><p>might equally well be seen to show that respondents are adept at seeing the conversational context and giving</p><p>answers that are most useful when seen in that context.37</p><p>Two other results help to show that respondents are generally able to understand the questions asked, and to</p><p>give the answers requested. The first relates to subjective health evaluations. Many surveys ask respondents,</p><p>on a 5-point scale, to report the state of their physical health, with 0 being very poor and 5 being very good. The</p><p>answers</p><p>to this question always show a significant decline as age increases. The designers of one large Cana-</p><p>dian survey, trying to be more precise, used the same response scale but asked respondents to compare the</p><p>state of their health with that of others of the same age. The answers showed no age trend at all. This suggests</p><p>strongly that respondents are able to assess the states of their own health, and to make, if asked, appropriate</p><p>comparisons with the age-adjusted states of health of others living in the same community.</p><p>The second example comes from the Gallup Daily Poll which reveals strong day-of-week effects for affect questions</p><p>that apply specifically to “yesterday,” but no daily patterns in life evaluations.38</p><p>There has also been a substantial literature testing and assessing order effects, with one meta-analysis of 16</p><p>studies showing small effects.39 But they can be dramatically large, as recently found in the Gallup-Healthways</p><p>U.S. Daily Poll.40 Split samples showed that respondents asked about their attitudes to government (which</p><p>were very negative at the time) immediately before the ladder question gave significantly lower answers</p><p>(by almost 5%), than when the political questions were absent, or were separated from the ladder question</p><p>by some less upsetting buffer questions. This effect is very large relative to the modest changes in national average</p><p>happiness that would normally happen from day to day or year to year, even during a major recession. These</p><p>results are very useful in underlining three points already implicit in the data we have presented. First, the</p><p>day-to-day and year-to-year changes in national average subjective well-being are likely to be very small</p><p>relative to the differences across individuals, communities and nations. Second, as will be shown in more</p><p>19</p><p>W O R L D H A P P I N E S S R E P O R T</p><p>detail in Chapter 3, although incomes are important supports for life evaluations, their effects are relatively</p><p>small compared to other factors, especially in terms of national average changes from one year to the next.</p><p>Third, shared changes in sentiment, whether triggered by question order or changes in the stock market, can</p><p>have large effects on average scores. The daily frequency of the Gallup-Healthways poll, and Gallup’s use of split</p><p>samples, made it easy to spot and correct the issue, and to convince others to test for question order and other</p><p>framing effects. For all of these reasons, subjective well-being data are not suitable for use as guides to short-</p><p>term macroeconomic policy, where in any case there are many other more relevant data.</p><p>Framing effects are important, but they exist for behavior as much as for survey answers. For example, experiments</p><p>showed that student subjects showed some modest tendency (less than 7%, but nonetheless greater than the 5%</p><p>noted in the previous paragraph) to mark in their own favor, but had no tendency to cheat if they had previously</p><p>been asked to write down as many as they could remember of the Ten Commandments.41 All human behavior,</p><p>whether evidenced by thought, opinions or action, is influenced by the social norms and contexts in which</p><p>people live. This does not diminish the validity of subjective answers, but does show the need for careful and</p><p>experimental data collection, and demonstrates the advantages of large and repeated samples.</p><p>How can happiness be compared across individuals, nations, and cultures?</p><p>Since the social and institutional contexts are such important supports for well-being, then we would expect</p><p>to find that there will be corresponding differences in reported well-being across communities, nations and</p><p>cultures. But what if there are cultural differences in response styles, so that people in different cultures might</p><p>report different answers to the same question, even if in other respects their life quality is the same? If these</p><p>differences in the interpretations of questions, or in the use and meaning of response scales, were very large,</p><p>they might affect subsequent judgments about where and why subjective well-being is higher.42</p><p>More generally, it has been argued that for a broad range of psychological findings, conclusions are based on</p><p>experiments undertaken using WEIRD subjects (those from Western, Educated, Industrialized, Rich Democracies),</p><p>and do not represent well what happens in the much larger populations in other countries and cultures.43</p><p>For both of these reasons, it is important to assemble data from different cultures and nations in ways that</p><p>permit researchers to make judgments about the likely extent of difficulties in making comparisons in happiness.</p><p>One basic check, once comparable data are assembled, is to see to what extent the answers drawn from different</p><p>nations and cultures appear to be influenced by the same factors, and to the same extent. As it turns out, the</p><p>cross-national commonality of the correlates of life evaluations is substantial.44</p><p>How much do aspirations and standards change?</p><p>Endowment effects, changing aspirations, adaptation, and relativities pose complications rather than road-</p><p>blocks to the use of happiness data as measures of the quality of life. Life on earth has, at least on average,</p><p>become much less brutish, nasty and short over the past 500 years. The evidence for this ranges from falling</p><p>murder rates to rising life expectancies. There are no long-standing happiness measures available to track</p><p>these life improvements, but it would be no surprise if individual and community-level aspirations and</p><p>standards have risen over the same centuries, even if at a lower rate. The empirical basis for adaptation and</p><p>relativities will be discussed in the next chapter. Our summary view of the available research is that adaptation</p><p>and relativities can truncate the average happiness increases that accompany human progress, that some</p><p>comparison effects are helpful and others harmful to average happiness, and that happiness tells a valid story</p><p>both across communities and over time.</p><p>Is there a happiness set point?</p><p>It is sometimes argued that human capacities for adaptation are so strong that even major changes in life</p><p>circumstances will have no lasting impact on subjective well-being. The most cited reference to this effect is</p><p>20</p><p>a study of subjective well-being among accident victims and lottery winners.45 However, even based on the</p><p>small number of cases analyzed in that paper, accident victims were significantly less happy than the control</p><p>group. Subsequent research has consistently confirmed that individuals with long-term disabilities have lower</p><p>subjective well-being, to an extent that varies with the severity of the disability.46 As might be expected from</p><p>other research reported in Chapter 3, the extent to which a disability affects subsequent well-being depends</p><p>not just on the severity of the disability, but also on the extent to which patients are enabled to maintain their</p><p>social connections.47</p><p>If each individual had his or her own set point based on stable personality traits, and eventually returned to</p><p>that point after any change in circumstances, there could not be such large and long-lasting international</p><p>differences in subjective well-being as are shown in this chapter. For example, average life evaluations in the</p><p>top 10 countries of Figure 2.3 are twice as high as in the bottom 10 countries, and these differences are largely</p><p>explained, as shown in Chapter 3, by measured differences in life circumstances. Nor would there be such a</p><p>systematic U-shape in happiness over the life course for each individual, as shown in Chapter 3.</p><p>Studies of identical and fraternal twins have also been used to estimate the extent to which happiness depends</p><p>on genetically based personality differences rather than differing circumstances. For example, studies of U.S.</p><p>twins have estimated that one-third to one-half of within-country variance of happiness can be explained by</p><p>genetic differences between individuals.48</p>