Notes
INTRODUCTION: THE TWO CULTURES
1. Of course some competition is fun, especially in sport and games. But in most other contexts it generates tension and fear, especially within teams and within families.
2. For a powerful argument along these lines, see Brooks (2015).
4. Collier (2018).
5. Ipsos (2016). Data relate to 2016. 12 per cent had done yoga in the last month, and 28 per cent had ever done yoga. On the practice of meditation, see also Clarke et al. (2015), Table 1, which relates to 2012, and shows that 11 per cent had done deep breathing exercises and 8 per cent had meditated in the past twelve months. Comparable figures for Britain are: meditation 7 per cent in last month (youGov/University of Lancaster Survey 2013 <http://cdn.yougov.com/cumulus_uploads/document/mm7go89rhi/YouGov-University per cent20of per cent20Lancaster-Survey-Results-Faith-Matters-130130.pdf>); yoga: 6.3 per cent in the last twelve months, 3.3 per cent in the last month. Data relate to 2015/16. (Taking Part Survey, DCMS. For data access, see <https://www.gov.uk/guidance/taking-part-survey#how-to-access-survey-data>).
6. American Psychological Association (APA) poll (2004). The survey was conducted by US market research firm Penn Schoen Berland; <https://www.apa.org/monitor/julaug04/survey>. Comparable figures for Britain are: 28 per cent of adults have themselves visited a counsellor or a psychotherapist at some point in time; 54 per cent of people say that a family member, friend, work colleague or themselves have consulted a counsellor or psychotherapist. British Association for Counselling and Psychotherapy (BACP) Survey (2014). Poll conducted by Ipsos MORI; <http://www.parabl.org.uk/english/wp-content/uploads/2013/11/13381_attitudes-survey-2014-key-findings.pdf>
7. Among economists the leading group has been at MIT in the Abdul Latif Jameel Poverty Action Lab (or J-PAL for short) who have already undertaken 600 randomized control trials, mostly in poor countries.
8. On the potential speed of change, see Chapter 4.
9. See Croson and Gneezy (2009) and Brdar et al. (2009).
10. For evidence that women have more other-regarding preferences than men, see results of the Gallup World Poll in Falk et al. (2018), especially Figure 3; and for evidence from surveys in seventy countries, see Schwartz & Rubel (2005). For wider reviews, see also Croson and Gneezy (2009) and Niederle (2016), who also document that women are more averse to competition between individuals. On gender differences in emotional intelligence, see, for example, Joseph and Newman (2010); Van Rooy et al. (2005); Petrides and Furnham (2000); Craig et al. (2009). Of course, it is unclear whether these differences are driven by biological or socio-historical factors. Moreover, although such gender differences have been widely shown, the magnitude of the difference is typically smaller than might be assumed from some casual stereotypes.
11. Layard (1980).
12. Obvious areas omitted include global poverty; the preservation of peace; the solution of racial and ethnic conflicts; human rights, including the advancement of women and ethnic minorities; the social care of the elderly, the disabled and the young; and animal rights.
CHAPTER 1 WHAT’S THE PURPOSE?
1. Written on 22 June 1830, and found in a friend’s young daughter’s birthday album. Quoted in Parekh (1993).
2. For a fuller statement of the argument, see Layard (2011), Chapter 15; de Lazari-Radek and Singer (2017) and my online Annex 1.1. Of course, not everyone accepts this argument. For example, some people would say the ultimate good is to do the will of God; but the issue still remains of what He wishes to be done. Others believe that there are multiple goods each of which are ends in themselves; but this leaves unresolved the issue of choice when different ends conflict. A related issue is whether all that matters is how people feel. Nozick (1974) asked, ‘Would you plug into an experience machine that would make you feel anything you desire, and let others do the same?’ I personally would not plug in (even though I think that only experience matters) because I want to actually affect the experiences of other people. I could not do so if I was in the machine.
3. On the relative importance of reducing misery compared with increasing average happiness, see pp. 32–6, below. On minimum rights, also see p. 36.
4. ‘Everybody’ means people worldwide, including future generations, plus proper allowance for other sentient beings.
5. Unfortunately, the idea became known as utilitarianism, because actions were to be judged by their effects, i.e. their utility. No word worse describes an idea which aims to cultivate warmth of heart and generosity as supreme virtues. I shall use the term Happiness Principle throughout.
6. See, for example, Pinker (2018).
7. Helliwell and Wang (2012), Chapter 2.
8. For a discussion of different measures of happiness/wellbeing, see online Annex 1.2.
10. Mill’s essay on Liberty is not intended as a general statement about how people should behave, but as a statement about the regulatory role of the state; see Mill (1859).
11. On such measurements, see the next note. As Annex 1.2 explains, the so-called eudaemonic methods which are currently in use bear little relation to the kind of virtue that Aristotle had in mind and that we need to find ways of measuring.
12. We know a lot about children’s behaviour because we can get parents or teachers to report on them, and bad behaviour is treated as a mental illness. We know much less about the behaviour of adults, due to obvious reporting problems. We can use self-reports (e.g. measuring compassion), which are already well developed, but also records of crime, domestic violence, behaviour at work and peer-to-peer reports. To complete the picture of happiness created, we also have to place a happiness value on the kind of work people do.
13. Ricard (2015).
14. Lane (2017), Dunn et al. (2008), Aknin et al. (2012), Aknin et al. (2013), Aknin et al. (2015), Aknin et al. (2019), Otake et al. (2006). As Helliwell and Aknin et al. (2018) show on p. 10, helping behaviour gives more pleasure to the helper when it is done for the sake of the other person than when it is done for self-oriented reasons.
15. Meier and Stutzer (2008). Similarly in Japan both average altruism and average happiness rose after the 2011 earthquake – a natural experiment which presumably shows a positive effect of altruism on happiness – see Ishino et al. (2012).
16. Borgonovi (2008) and Brown et al. (2003). See also Greenfield and Marks (2004).
17. See online Annex 1.3.
18. This is broadly similar to the distribution in Germany, France and Spain, but is more equal than the distribution in the USA. See Helliwell et al. (2016), pp. 33–4.
19. For more detail, see online Annex 1.4.
20. Examples are Harsanyi (1953, 1955) and Rawls (1971). On the question in the text, they reach very different conclusions from each other and from the general case I present below. Rawls concluded that the ranking of states should be based entirely on the primary resources of the person who is least well off. This is extremely egalitarian. Harsanyi, by contrast, concludes in favour of ranking according to the average utility in the state, i.e. ∑Πi ui where Πi is the probability of being person i and ui is the utility of person i. Harsanyi measures the utility of person i’s condition as the equivalent probability of being in the best possible position and I would argue that this equivalent probability is itself a concave function of the actual happiness in condition i. See Layard (2011), p. 312.
22. On income inequality, see Chapter 11.
23. Sen (2009). On social justice, see also Sen (2017).
24. For more on this, see Layard (2011), Chapter 15. See also Anand (2016).
CHAPTER 2 WHAT MAKES PEOPLE HAPPY?
1. Moynihan and Weisman (2010).
2. Diener and Biswas-Diener (2008).
3. A. E. Clark et al. (2018).
4. Urry et al. (2004). Davidson and Begley (2012). Goleman and Davidson (2017).
5. Steptoe and Wardle (2012). The figure was privately supplied by Andrew Steptoe.
6. A. E. Clark et al. (2018), Table 16.2. See also Chapter 4 of Clark et al. (2018).
7. A. E. Clark et al. (2018), Tables 2.1 and 2.2; Stevenson and Wolfers (2008). For the best form of the function relating happiness to income, see Layard et al. (2008). In most countries the variance in income explains under 2 per cent of the variance in happiness.
8. Other factors explain up to 20 per cent of the variance. Similarly for children, family income explains under 1 per cent of the variance in emotional wellbeing at age sixteen (A. E. Clark et al. (2018), Table 10.1 and Ford et al. (2007)).
10. A. E. Clark et al. (2018).
11. Plomin (2018). This makes it clear that social scientists should always where possible include the relevant polygenic score in any behavioural equation. As regards the ‘heritability’ of happiness, there are many estimates which imply that the share of variance explained by the genes is between 30 and 60 per cent (Plomin et al. (2013), p. 322). However these estimates attribute to the genes the influence of all environmental influences (such as those in the chart) in so far as they are correlated with the genes. Nor can they handle the fact that the environment itself influences the effect of the genes.
12. This comprises some 10 per cent of the population, using the BHPS.
13. A. E. Clark et al. (2018), Table 16.1.
14. Layard (2018b), Table 2. Here misery comprises the lowest 20 per cent or so of life-satisfaction in each country.
15. Maslow (1954). In his terminology, the needs (from lowest to highest) are Physiological, Safety, Love and Belonging, Esteem, Self-actualization.
16. Tay and Diener (2011) show that, although different needs have differential impacts on happiness, the effect of satisfying each of Maslow’s needs is largely independent of whether other needs have been met.
17. This does not mean that income is unimportant, as I have laid out elsewhere.
18. A. E. Clark et al. (2018), Figure 1.2.
19. A. E. Clark et al. (2018), Figure 1.5(b).
20. Helliwell and Wang (2012), Table 2.1.
21. Income inequality as such does not show up significantly in these cross-country regressions but is discussed in Chapter 11.
22. In logarithmic form.
23. World Values Survey.
24. Rojas (2018).
25. For evidence on the effect of human rights, see Diener and Diener (1995).
27. Layard (2005b), p. 30, and World Happiness Report.
28. Layard et al. (2010), p. 149.
29. Easterlin et al. (2017).
30. Sacks et al. (2010), Figure 8.
31. See Helliwell, Layard and Sachs (2019), and for some countries online Annex 12.1. Overall world happiness has fallen, mainly due to a big drop in India.
32. Di Tella et al. (2003).
33. A. E. Clark et al. (2018), Table 2.4. However, if economic growth does not increase happiness, we still have to explain the country cross-section results in the table on p. 52. One consistent explanation here would be that people are comparing their country with other countries. But this is very difficult to test.
34. Sacks et al. (2010).
35. Online Annex 2.1 shows trends in each of the world’s regions of stress, worry, anger, sadness and the Cantril Ladder for life evaluation.
CHAPTER 3 TRAINING OUR THOUGHTS AND FEELINGS
1. Dryden (1685).
2. For more on the themes in this chapter, see Layard and Clark (2014).
3. Wolpe (1958).
4. Paul (1966).
5. It is of course important to understand the sources of your negative thoughts and feelings (and even OK for a time to feel you are a victim). But to move forward it generally requires more than this.
6. On this paragraph, see Layard and Clark (2014).
7. Goleman (1995).
8. Seligman (2002).
9. There is also of course a long-established mystical tradition in Christianity, and increasing use of retreats and silence (as in the movement centred in Taizé, France).
10. See Williams and Penman (2011) and Williams and Kabat-Zinn (2013).
11. Kabat-Zinn (1990).
12. See Davidson et al. (2003).
13. Segal et al. (2013). Other forms of meditation have also been studied scientifically. For example, a three-month retreat focused on concentrative meditation techniques and complementary practices used to cultivate benevolent states of mind increases telomerase activity and thus length of life. See Jacobs et al. (2011).
14. Their authors know this, and MBSR includes a final two sessions of compassion meditation (which is not mindfulness as normally conceived).
15. See, for example, Goleman (2003) and Singer and Ricard (2015).
16. A full list of these publications can be viewed on their website at <https://www.mindandlife.org/books/>.
17. T. Singer (2015) and Ricard and Singer (2017). Note that Ricard, Kabat-Zinn, Davidson and Goleman are routinely invited to the World Economic Forum in Davos.
18. Hanh (2008). See also his great book on anger, Hanh (2001).
19. Hanh and Weare (2017), p. xxxvi.
20. Shantideva (700 AD). The Bodhisattva Prayer for Humanity, written by Shantideva, an Indian Buddhist sage.
21. A remarkably uplifting and relieving experience is simply to imagine yourself inside the mind of some actual other person.
CHAPTER 4 CAN THE HAPPINESS MOVEMENT SUCCEED?
1. Twenge and Campbell (2010); Twenge (2017).
2. Twenge and Campbell (2010), p. 95.
3. Twenge (2017), Appendix Table E.2. 70 per cent of Americans think people are ruder than they were twenty years ago.
4. Twenge and Campbell (2010).
5. Collins (2001) and references therein.
6. See, for example, the strong downward trend in US crime statistics in data published by the FBI <https://ucr.fbi.gov/ucr-publications> as well as the Bureau of Justice Statistics <https://www.bjs.gov/index.cfm?ty=dcdetail&iid=245>. On other countries, see Tseloni et al. (2010). In some countries there has been an increase in violent crime in the last few years but this is not (yet at least) an established trend.
7. For a more long-term analysis of the causes of reduced violence, see Pinker (2011).
8. See references in the Introduction on gender differences in preferences.
9. See the British Social Attitudes Survey Reports published by the National Centre for Social Research; <http://www.bsa.natcen.ac.uk/latest-report/>.
10. Sometimes of course people are well advised not to interact. Women have always been subjected to unwanted attention, highlighted especially in recent years by the long-overdue #MeToo and Time’s Up movements.
11. OECD (2013); Durand (2018). In the USA the data have been collected less frequently (in 2010, 2012 and 2013) in the American Time Use Survey. Data have also been collected in the Panel Study of Income Dynamics (PSID) since 2009. A USA National Academy of Sciences Panel recommends annual collection. See (Mackie and Stone (2013).
12. OECD (2016). The document did not, however, define wellbeing as subjective wellbeing.
13. The meeting was held in Paris in October 2019. In the same month (24 October) the EU Council of Ministers called on member countries and the European Commission to ‘put people and their wellbeing at the centre of policy design’; <https://www.consilium.europa.eu/en/press/press-releases/2019/10/24/economy-of-wellbeing-the-council-adopts-conclusions/>.
14. In 2016 for the first time the UNDP’s Human Development Report included measures of life-satisfaction.
15. Before that Bhutan and the United Arab Emirates had done the same, with formal procedures for scrutinizing all policy proposals for their impact on happiness. See UAE (2017) and Ura et al. (2012).
16. For a fuller analysis, see Durand (2018), Table 3.1, in Global Happiness Policy Report. In Germany Angela Merkel declared in 2015 that ‘what matters to people must be the guideline for our policies’. So she launched a major national consultation, involving 100 meetings attended by ministers and intended to give new perspectives on what really matters; <http://www.gut-leben-in-deutschland.de/static/LB/>.
17. Stiglitz et al. (2009).
18. O’Donnell et al. (2014).
19. HM Treasury (2018).
20. Scotland has taken a lead in organizing an alliance of nations committed to wellbeing, including also New Zealand and Iceland. It is called Wellbeing Economy Governments; <http://wellbeingeconomygovs.org/>
21. See NatCen (2018); <http://natcen.ac.uk/our-research/research/british-social-attitudes>.
22. Wilson (2007) and Appiah (2011).
CHAPTER 5 EACH OF US
1. Einstein (1951).
2. In 2018, 22 per cent of Americans reported going to church every week, and a further 10 per cent almost every week. See <http://www.gallup.com/poll/1690/religion.aspx>. In many Muslim countries religious observance is also declining. For example, in Egypt the proportion of people praying five times a day has fallen from around 80 per cent in 2011 to around 60 per cent, see The Economist (2017).
3. US Federal Reserve Chairman, Alan Greenspan, was a long-time member of Rand’s inner circle, the so-called Ayn Rand Collective (see Michael Kinsley, ‘Greenspan Shrugged’, The New York Times, 14 October 2007; <https://www.nytimes.com/2007/10/14/books/review/Kinsley-t.html>. Trump has also professed admiration for Rand, see for example an interview with Kirsten Powers in USA Today: <https://www.usatoday.com/story/opinion/2016/04/11/donald-trump-interview-elections-2016-ayn-rand-vp-pick-politics-column/82899566/>.
4. Dalai Lama (2012).
5. Dalai Lama (2012).
6. For some others, see p. 271 below.
7. V. King (2016). Also, for children there is V. King et al., 50 Ways to Feel Happy (2018).
8. The Ten Keys to Happier Living align well with a number of other frameworks. These include the twelve elements identified by Sonja Lyubomirsky (2008); Martin Seligman’s PERMA – positive emotion, engagement, relationships, meaning and accomplishments (Seligman (2011), and the thirteen elements identified by Seldon (2015).
9. Bullock (2000).
10. Foresight Mental Capital and Wellbeing Project (2008). The work was done by the New Economics Foundation; see <https://neweconomics.org/2008/10/five-ways-to-wellbeing-the-evidence>. See Aked et al. (2008).
11. Gottman (1994).
12. V. King (2016). See also Baumeister et al. (2001).
13. Quoted in Ricard (2015).
14. M. E. P. Seligman (2011).
15. Dolan (2014).
16. P. Gilbert (2010). There is much evidence that happy people spread happiness – happiness is contagious; see Fowler and Christakis (2008).
17. Kok et al. (2013). See also Fredrickson (2013).
18. V. King (2016) and references therein.
19. Ehrenreich (2010).
20. The calendars have already been translated by volunteers into sixteen languages.
21. Krekel et al. (2020).
22. In social science a more common measure of the effect of an intervention is its ‘effect size’. This measures the change in a variable relative to its standard deviation. For changes of the size we are typically discussing in this book, the change in percentage points equals approximately forty times the ‘effect size’. For example, an effect size of .25 corresponds to a change of 10 percentage points. However, this is an approximation and the numbers in the charts and tables in this book are calculated exactly.
23. The largest teaching movements are the International Positive Psychology Association (IPPA) and its offshoot the International Positive Education Network (IPEN), each with thousands of members. Teaching movements based in Britain include the School of Life, a modern humanist enterprise; and the School of Economic Science, which focuses on Eastern philosophy, and its Western counterparts, mostly pantheist. Other individual teachers with large followings include: Dan Goleman, Tal Ben-Shahar, Mo Gawdat, Gretchen Rubin, Rick Hansen and Ed Diener (with the programme called Enhance).
24. Gautier (2008).
25. On humanism, see Pinker (2018). Like many good people who have qualms about happiness, Steven Pinker uses the word ‘flourishing’, which he says has many dimensions. But, as we have seen, this leaves unresolved the issue of how we are to combine them. The same is true of the five elements in Seligman’s PERMA; see Seligman (2011) and of the capabilities identified by Amartya Sen (2009).
26. Suppose I give $1 to a person in a poor country. And in each country happiness depends positively on own income and negatively on average income:
Then, when I give $1 to a person i in a poor country, the change in aggregate happiness in that country is:
This is positive if:
which would normally hold (with the left-hand side greater than 1 and the right-hand side smaller than 1).
And what about the change of happiness in my own country? The change of happiness when I lose $1 is positive if:
which might also often be true if I was reasonably well off.
On top of this the recipient in the poor country is likely to be initially less happy than the donor in the rich country (see World Happiness Report). So even if the changes summed to zero across the two countries, social justice would still call for the income transfer to be made.
27. This is inspired by the long-standing ideas of Peter Singer of Princeton University (P. Singer (2015)), and it has a strong presence at the Centre for Effective Altruism in Oxford.
28. Found online at <https://www.givewell.org/>.
29. Found online at <https://80000hours.org/>.
CHAPTER 6 TEACHERS
1. A. E. Clark et al. (2018), Chapter 1.
3. A. E. Clark et al. (2018), Table 14.5.
4. Flèche (2017).
5. For the past fifty years, Phi Delta Kappa have run an influential annual poll of the American public’s attitude towards public schools (see <http://pdkpoll.org/results>). In the 2017 poll, 82 per cent said it was highly important for schools to develop the interpersonal skills of pupils and 39 per cent of the respondents said it was ‘extremely important’ to develop skills like teamwork and persistence. This is compared to just 13 per cent who consider standardized test scores an ‘extremely important’ measure of school quality. A further 82 per cent also supported more job/career skills classes, even if it had to come at the expense of time on traditional academic education.
6. Durlak et al. (2011), Hanh and Weare (2017), Adler (2016), see also Fredrickson and Branigan (2005).
8. See, for example, Blanden et al. (2018).
9. Palmer (2007) and Palmer (2016). Barber and Mourshed (2007).
10. Einon et al. (1978).
11. OECD (2017). Annex 6.1 shows the country rankings for children’s life-satisfaction.
12. Collishaw et al. (2004), West and Sweeting (2003), Sadler et al. (2018).
Percentage of children suffering from diagnosable mental illness in England, 2017:
| Age | Boys | Girls |
|---|---|---|
| 5–10 | 12.2 | 6.6 |
| 11–16 | 14.3 | 14.4 |
| 17–19 | 10.3 | 23.9 |
Source: Sadler, et al. (2018), Table 1. | ||
13. Sadler et al. (2018).
14. McManus et al. (2016), p. 302. The figures are higher for young women than for young men.
15. Twenge et al. (2010) and Twenge (2017), Chapter 4. In 2015 over half of all US college students attended counselling for mental health concerns; see Haidt and Lukianoff (2018), p. 156.
16. On GRIT, see Duckworth (2016).
17. See <https://www.onderwijsinspectie.nl/onderwerpen/sociale-veiligheid/toezicht-op-naleving-zorgplicht-sociale-veiligheid-op-school>.
18. ISI (2017), section C.
19. The former British Prime Minister Theresa May promised that in England the government would produce a government-approved questionnaire which schools could use voluntarily. The government of South Australia already does this.
20. The Government of South Australia runs the online administration of the questionnaire and tabulation of the results. No individual is identified, but schools and classrooms are provided with benchmark data for comparison. Further information at <https://www.education.sa.gov.au/wellbeing-and-engagement-census/about-census>.
21. For some suggestions, see online Annex 6.2.
22. Below age nine it is generally found that children’s replies are unreliable.
23. For wellbeing (as for academic achievement) the school should be looking at its ‘value-added’ – compared with a national reference norm. Some organization from outside a school should organize the measurement (typically online). In addition, the scope for gaming will be reduced if secondary schools judge themselves by how they augment pupils’ wellbeing beyond the level already measured somewhere else, i.e. at primary school.
24. Hawkes (2013) and Hawkes and Hawkes (2018).
25. Bullock (2000).
26. In England, schools have a duty to promote the spiritual, moral, social and cultural development of their pupils, but without any clear guidance on what this might involve. In practice, most formal moral education happens in assemblies, or in the required weekly lesson called Religious Education.
27. Hanh and Weare (2017); Kuyken et al. (2013).
28. Durlak et al. (2011); Hale et al. (2011).
29. Hale et al. (2011).
30. Layard et al. (2018); Lordan and McGuire (2018). In most large preventative interventions, the effect-sizes are what might be considered small (here 0.18 on life-satisfaction). But the costs are small and the population affected large (see Greenberg and Abenavoli (2017)). So in this case the estimated cost per QALY is only £1,000 compared with the critical value of around £30,000.
31. For example, the English programme of Social and Emotional Aspects of Learning (SEAL) introduced in the early 2000s was found in controlled trials in secondary schools to have no impact – largely because the teachers were not trained to use the materials. Humphrey et al. (2010).
32. Greenberg et al. (1995); Domitrovich et al. (2007); and Kam et al. (2003). In the UK the results have been somewhat less impressive, e.g. Humphrey et al. (2016). This may be because PATHS is normally being compared with the alternative of treatment-as-usual. For results of the trial in Northern Ireland, see PATHS (2013).
33. Seligman and Adler (2018).
34. Algan et al. (2013). They use data from the Civic Education Study, from the TIMMS study and from the Progress in International Reading Literacy (PIRLS) which between them allow comparisons of countries, schools and classrooms. See especially their Figure 3.
35. Elliott and Dweck (1988).
36. Webster-Stratton et al. (2011); Reinke et al. (2012); Davenport and Tansey (2009); Webster-Stratton et al. (2001); Baker-Henningham et al. (2012); Hutchings et al. (2007). For an intervention to prevent bullying, see Bonell et al. (2018).
37. Flook et al. (2013).
38. A. E. Clark et al. (2018), Table 14.4. See also Leuven and Løkken (2018) and Angrist et al. (2017).
39. Universities offering courses on wellbeing at some level or other include Northwestern, Vermont, Dartmouth, Michigan, Miami and MIT in the US, and Warwick and the London School of Economics in Britain.
40. Seldon and Martin (2017).
41. See, for example, Kessler et al. (2005).
42. Ideally mental health professionals work under supervision in a team. So the mental health service of an educational institution (if the service is small) should be part of some wider service.
43. See note 6 of this chapter.
CHAPTER 7 MANAGERS
1. E. J. Hughes (1963).
2. Krueger (2009).
3. Bryson and MacKerron (2017).
4. Jahoda (1982).
5. See Kay (1998). See also the US Business Roundtable’s ‘Statement on the Purpose of a Corporation’ (August 2019); <https://opportunity.businessroundtable.org/wp-content/uploads/2019/09/BRT-Statement-on-the-Purpose-of-a-Corporation-with-Signatures-1.pdf>.
6. Companies Act 2006.
7. E.g. limited liability, contract enforcement, etc.
8. Edmans (2011) and Edmans (2012). For later work, see Edmans et al. (2017).
9. A. E. Clark (2001).
10. Oswald et al. (2015).
11. De Neve and Oswald (2012). Happiness is measured here by life-satisfaction. Similar results are found using positive affect measures.
12. For surveys of the evidence, see Tenney et al. (2016) and Walsh et al. (2018).
13. Deci and Ryan (2012).
14. On the value of non-financial recognition, see Ashraf et al. (2014).
15. De Neve (2018).
16. In an interesting experiment workers labelling medical images were divided into three groups. One group were told they were ‘labelling tumour cells in order to assist medical researchers’; the second group were given no context for their work; and the third group were told their labels would be discarded on submission. The first group did much more work than the other two, with no loss of precision (Chandler and Kapelner (2013)). On the importance of meaning in work motivation, see also Grant (2008); Chadi et al. (2017) and Ariely et al. (2008).
17. See Chapter 2, Figure 2.2. That figure was based on data from the European Social Survey, but data from the ISSP shows a similar impact of the overall quality of work.
18. Managers also completed a self-paced computer-based course.
19. Moen et al. (2016). Voluntary quits were 7.6 per cent in the treatment group and 11.3 per cent in the control group.
20. Bloom et al. (2015), p. 167. This compares with 40 per cent in 1970.
21. Even in call-centre work there can be problems of loss of control over malpractice (where customer payouts are involved) and loss of teamwork.
22. The shift length was unchanged. Productivity rose by 9 per cent due to more minutes actively worked and it rose by 4 per cent due to more calls per minute worked.
23. The same principles apply elsewhere. For example C. Knight et al. (2010) found that care-home residents empowered to decide the decor of their floor had improved wellbeing compared with other residents who were not empowered to do so on their floor.
24. For evidence on their effectiveness, see Blasi et al. (2008), Kruse et al. (2010) and Bryson, Clark et al. (2016).
25. On productivity, see Lazear (2000), Bloom and Van Reenen (2010) and Bandiera et al. (2017). On life-satisfaction, see Böckerman et al. (2016). Even in such cases pay based on individual performance can demotivate those who are less productive if they can compare their wages with those of others who are paid more; see Breza et al. (2017). However, this negative effect ceases when workers also know the productivity of their colleagues.
26. Bandiera et al. (2013).
27. Especially in combination with good work organization, including especially autonomy and job security. See Blasi et al. (2008). Also see Kruse et al. (2010).
28. Böckerman et al. (2016). This is with wages held constant.
29. Dahl and Pierce (2019). PRP did not include piece-work pay.
30. Card et al. (2012).
31. For example, Cohn et al. (2014) and Breza et al. (2017).
32. Barankay et al. (2012). Similarly for truck drivers, the public ranking of individual performance had a negative effect on performance once workers had received instruction in teamwork (Blader et al. (2016)). Even the ranking of teams can be counterproductive (Bandiera et al. (2013))
33. Card et al. (2012).
34. Similarly, after all Norwegian income tax records became publicly available online in 2001, the gap in happiness between rich and poor increased by 21 per cent. Bandiera et al. (2013); Perez-Truglia (2019).
35. Kellam et al. (2011).
36. Two other arguments are often brought against pay based on individually assessed performance. One is that it can distort the direction of effort. The other is that it can lead to over-arousal which can undermine performance, see online Annex 7.1.
37. World Economic Forum (WEF) (2012).
38. These can be used in-house, or businesses can hire consultants to use them and then suggest how worker wellbeing can be improved. See, for example, Robertson and Cooper (2011) and Lundberg and Cooper (2011).
39. Layard and Clark (2014); OECD (2012).
40. For example, Action for Happiness offers a two-day course called ‘Doing Well from the Inside Out’. This has good before-and-after evaluation results, but without comparison to a control group.
41. Hulsheger et al. (2013).
42. See Hsieh (2012).
CHAPTER 8 HEALTH PROFESSIONALS
1. National Institute for Health and Care Excellence.
2. The World Health Organization also measures the burden of disease in terms of a similar concept: Disability-Adjusted Life Years (or DALYs).
3. A. E. Clark et al. (2018). The health measures are as described in the text except that for the UK the mental health question is ‘Have you been to the doctor for an emotional problem?’ and for Australia the physical health measure is the physical component of the SF36.
4. Layard (2018b), Table 2.
5. Helliwell, Layard and Sachs (2017) and Layard and Clark (2014), p. 41.
6. For UK evidence, see McManus et al. (2016). Gambling addiction is also a serious problem affecting some 0.8 per cent of adult Britons; see Conolly et al. (2017).
7. J. M. G. Williams (2001).
8. Suicide and murder figures from the Global Burden of Disease Study (2015). Battle death figures from Pinker (2011), p. 50.
9. Case and Deaton (2017). Drug overdose now costs the USA 2.8 per cent of GDP; see Council of Economic Advisers (2017) at <https://www.whitehouse.gov/sites/whitehouse.gov/files/images/The%20Underestimated%20Cost%20of%20the%20Opioid%20Crisis.pdf>.
10. For the UK we have the following percentages with any mental illness:
| 1993 | 2014 | ||
|---|---|---|---|
| All | 16–24 16–64 |
13.7 14.1 |
17.3 17.5 |
| Men | 16–24 16–64 |
8.4 10.5 |
9.1 13.6 |
| Women | 16–24 16–64 |
19.2 17.7 |
26.0 21.4 |
Source: McManus et al. (2016), Table 2.2. CIS-R score of 12 or more; see <https://files.digital.nhs.uk/excel/9/s/apms-2014-ch-02-tabs.xls>. | |||
For the US we have the following percentages with any mental illness.
| 2008 | 2017 | |
|---|---|---|
| 18–25 | 18.5 | 25.8 |
| 26–49 | 20.7 | 22.2 |
Source: SAMHSA (2018), Figure 48; see <https://www.samhsa.gov/data/sites/default/files/cbhsq-reports/NSDUHFFR2017/NSDUHFFR2017.pdf>. | ||
On younger adolescents in the US, see Chapter 13.
11. For more on this whole section, see Layard and Clark (2014) and Hollon et al. (2006).
12. Layard and Clark (2014), pp. 51–3.
13. This is despite the fact that the majority prefer psychological therapy (McHugh et al. (2013)).
In the USA the proportion in treatment for depression who receive psychotherapy has fallen (see these figures from Olfson et al. (2002) and Marcus and Olfson (2010)).
| 1987 | 1997 | 1998 | 2007 | |
| Percentage of population receiving any treatment for depression. | 0.73 | 2.33 | 2.37 | 2.88 |
| Percentage of those in treatment who receive: | ||||
| Psychotherapy | 71.1 | 60.2 | 53.6 | 43.1 |
| Medication | 44.6 | 79.4 | 80.1 | 81.9 |
In the UK the proportions of people with common mental disorders receiving each type of treatment are:
| Percentage of mentally ill adults who receive | 2000 | 2007 | 2014 |
| No treatment | 76.9 | 75.6 | 60.6 |
| Medication only | 14.4 | 14.1 | 26.8 |
| Counselling or therapy only | 3.8 | 4.9 | 4.9 |
| Both medication and counseling | 4.8 | 5.5 | 7.7 |
| Any counselling or therapy | 8.6 | 10.4 | 12.6 |
| Any medication | 19.3 | 19.6 | 34.5 |
| Any treatment | 23.1 | 24.4 | 39.4 |
Source: McManus et al. (2016), Table 3.11. | |||
14. Layard (2018b).
15. For example, to calculate the effects of an illness upon the quality of life (in QALYs), you have to weight the importance of five main factors (mobility, self-care, usual activities, physical pain and mental pain). The weights currently used in the British system give much less weight to mental pain than the weights obtained by regressing life-satisfaction on the five factors. See Layard and Clark (2014), pp. 65–6.
16. OECD (2012).
17. McHugh et al. (2013). See also Chilvers et al. (2001), Deacon and Abramowitz (2005) and van Schaik et al. (2004).
18. D. M. Clark (2018).
19. D. M. Clark (2018); D. M. Clark et al. (2018).
20. Parts of Norway, Sweden, Lithuania, Australia and Israel already have services, and Ontario and Quebec (both in Canada) will soon do so.
21. Andersson (2016).
22. Singla et al. (2017).
23. Kim-Cohen et al. (2003) and Kessler et al. (2005).
24. Layard and Clark (2014).
25. Layard and Clark (2014).
26. Department of Health and Social Care and Department of Education (2017).
27. Jefferson (1809): ‘The care of human life and happiness, and not their destruction, is the first and only legitimate object of good government.’
28. In technical terms, it depends on the distribution of QALYs across persons, with each person’s QALYs equal to the length of their lifetimes its average quality.
29. See Human Mortality Database, Period Life Table for Total Population of England and Wales. Figures refer to the standard deviation of the age of death in 1910 and 2016. If one measures the standard deviation in age of death over time, it has narrowed progressively over the time period, and continues to narrow.
30. For a different and more widely accepted definition of health inequality, see Marmot et al. (2010).
31. In Britain, the standard deviation of length of life is around 17 per cent of the average length of life; the comparable figure for life-satisfaction is around 28 per cent (in the Gallup World Poll).
32. The most striking feature of the data is that the proportional reduction in age-specific death rates has been greater at young ages. If the main factors at work were general improvements in living standards, one might expect the effect on death rates to be similar at each age. Since this is not what we see, the changes in age-specific death rates suggest a big role for modern medicine.
33. Gawande (2014).
34. French et al. (2017), Exhibit 3.
| Spending on hospital care in last 12 months of life (per cent of overall spending) | |
| Denmark | 10.0 |
| England | 14.6 |
| France | 15.0 |
| Germany | 21.2 |
| Japan | 8.2 |
| Netherlands | 8.9 |
| Quebec | 22.7 |
| Taiwan | 15.5 |
| USA | 9.9 |
35. Too many people die without having discussed death with their loved ones, or having exchanged feelings of gratitude, love and forgiveness.
36. As an additional feature, the new law proposed in the UK by Dignity in Dying requires that a doctor should be present when the fatal dose is taken, to ensure it works well; see the Assisted Dying Bill of 2016 which was defeated in the House of Commons.
37. Battin et al. (2007). Oregon introduced the Death with Dignity Act in 1997; California passed the End of Life Option Act in 2015. On the case for assisted dying, see Campaign for Dignity in Dying (2015).
38. For example, in Britain extra money spent on health would generate around sixty times more benefit than the corresponding cost in terms of income foregone by taxpayers (see pp. 240–41).
39. These are UK figures, from Drummond et al. (2016), p. 239, and M. Williams (2014), p. 48.
40. On the relative dangers of different substances, see Nutt (2012), Global Commission on Drug Policy (2014).
41. Nearly 1 per cent of the UK population are addicted to them; see Roberts et al. (2016), p. 266.
42. For example, the United Nations Office on Drugs and Crime (2005), Chapter 2, refers to the valuation of the illegal drug market in 2003 at $320 billion. See also Reuter and Trautmann (2009).
43. Estimates of violent deaths in Iraq since 2003 are typically under 200,000. By contrast, the following figures for intentional homicide between 2000 and 2016 relate to just five of the countries affected by violence in which drugs played a major part: Mexico, 85,000 deaths; Colombia, 281,000; the Philippines, 313,000; Guatemala, 130,000; Brazil, 821,000 (see United Nations Office on Drugs and Crime (2018) at <https://dataunodc.un.org/crime/intentional-homicide-victims>).
44. Even in the US, the prohibition of alcohol (1920–33) only applied to production and sale.
45. For heroin, the first step is to stabilize the person’s life on a reducing supply of uncontaminated heroin, or a heroin substitute like methadone or buprenorphine. Where the aim is total abstinence, this can be helped by taking naltrexone to reduce craving. Once the drug situation is under control, relevant psychological therapy is crucial. For alcohol the first step is towards abstinence, which can be facilitated by taking benzodiazepines to reduce withdrawal symptoms, and other drugs like naltrexone to inhibit relapse. With alcohol addiction, psychological treatment is vital and should begin quite soon. For more on this, see Layard and Clark (2014), pp. 172–4.
46. For a description of this and policy innovations in other countries, see the Global Commission on Drug Policy (2017). When Western Australia decriminalized cannabis, there was no increase in consumption relative to other Australian states.
47. Domoslawski (2011); C. E. Hughes and Stevens (2010).
48. On Europe, see EMCDDA (2015).
49. Csete (2010).
50. Ribeaud (2004).
51. United Nations Office on Drugs and Crime (2016). The document also favours the use of evidence in policy design. See also Meacher and Warburton (2015), All Party Parliamentary Group for Drug Policy Reform, prepared for the UN General Assembly Special Session on Drugs in 2016, at which the new UN stance was adopted. More recently, the UN Chief Executive Board for Coordination have committed to ‘promote alternatives to conviction and punishment in appropriate cases, including the decriminalization of drug possession for personal use, and to promote the principle of proportionality, to address prison overcrowding and over-incarceration by people accused of drug crimes …’ See <https://www.unsceb.org/CEBPublicFiles/CEB-2018-2-SoD.pdf>.
52. National Academies of Sciences, Engineering and Medicine (2017).
CHAPTER 9 FAMILIES
1. Rutter et al. (2010).
2. National Family and Parenting Institute (NFPI) (2000).
3. For data in the USA going back to the 1960s, see the US Census Bureau’s ‘Historical Living Arrangements of Adults’; at <https://www.census.gov/data/tables/timeseries/demo/families/adults.html>.
4. Ibid.
5. See US Census Bureau, Living arrangements of Children 2017; at <https://www.census.gov/data/tables/2017/demo/families/cps-2017.html>. Data is drawn from the Annual Social and Economic Supplement (ASEC) of the Current Population Survey (CPS). In 2017, 68.9 per cent of 0-to-17-year-olds were living with both parents.
6. Rosenfeld and Thomas (2012).
7. Rosenfeld and Thomas (2012), Figure 1.
8. Cacioppo et al. (2013). See also Rosenfeld (2017).
9. One concern is that an increase in ‘swiping culture’ may increase stress and anxiety. (An analogous point is often made in consumer psychology: there can be a ‘paradox of choice’ wherein a greater number of options increases stress and ultimately lowers satisfaction.)
10. For the latest official numbers in the UK, see ONS Statistical Bulletin: Sexual identity, UK, 2016; at <https://www.ons.gov.uk/peoplepopulationandcommunity/culturalidentity/sexuality/bulletins/sexualidentityuk/2016>.
11. Rojas (2018).
12. On the relation between child outcomes and family conflict/break-up, see A. E. Clark et al. (2015); Jekielek (1998); Hanson (1999); Amato et al. (1995); Epstein et al. (2015); A. E. Clark et al. (2018).
13. Harold and Sellers (2018). On the negative effects of easier divorce on children, see Gruber (2004). On the positive effects of easier divorce on the couple, see Stevenson and Wolfers (2006).
14. Stuart (1969); Gottman (1994), p. 57. Gottman also characterized marriages where there were five positive interactions for every one negative interaction as ‘happy marriages’.
15. Baucom et al. (2015). Marriage counselling is not new, but only recently do we have interventions whose outcomes have been successfully evaluated. Note that IAPT also provides Systemic Couples Therapy based on Hewison et al. (2014).
16. Baucom et al. (1990) and Atkins et al. (2010).
17. Fischer and Baucom (2018). See also Baucom et al. (2018).
18. Epstein et al. (2015).
19. US data from Epstein et al. (2015).
20. See World Health Organization (2009) for a summary of some successful community- and school-based interventions that seek to tackle domestic violence. See also Chapter 6 above.
21. Epstein et al. (2015).
22. The difference was statistically significant – see Schulz et al. (2006), Table 3 and Figure 2. The project is called the Becoming a Family project. The Cowans also developed a successful intervention for couples whose children were beginning primary school. It lasts eighteen weeks. One set of parents are placed in groups which focus mainly on issues between the parents, while the other set are placed in groups which focus mainly on issues between the parents and the child. In a trial they were followed for ten years. For both the parents and the children, the most effective intervention was the one focused on the parents’ relationship with each other. See Cowan and Cowan (2008).
23. Cowan et al. (2007). There were sixteen meetings. In the fathers’ groups, as in the control group, the usual pattern emerged – people became (on average) less satisfied. The programme was called Supporting Father Involvement. The couples version has been repeated in Britain under the name the Parents as Partners programme; see Little (2016). Follow-up was after eighteen months in the USA and after six months in Britain.
24. See Acquah et al. (2017) and Harold et al. (2016).
25. Feinberg et al. (2010). There was no significant effect on the girls.
26. Bowlby (1969); Rutter et al. (2008).
27. Rutter et al. (2010).
28. Baumrind (1971).
29. Twenge and Campbell (2010).
30. Plomin (2018).
31. Layard and Mincer (1985).
32. J. Scott and Clery (2013). Women spend on average thirteen hours on housework and twenty-three hours on caring for family members each week. For men this is eight and ten hours respectively.
33. See, for example, A. E. Clark et al. (2018), Tables 11.1 and 11.3.
34. For a natural experiment, see Carneiro et al. (2015) – in Norway an increase in parental leave boosted the children’s educational attainment and their wages.
35. See Eisenberg et al. (2004); Jacob et al. (2008).
36. Webster-Stratton et al. (2001). The programme is for children with mild to moderate problems. If children have severe behavioural problems, they need to be treated individually, often together with a parent. See, for example, Kazdin (2009).
37. Webster-Stratton et al. (2001).
38. See Menting et al. (2013) for a meta-analysis.
39. That is, ‘oppositional defiant disorder’. See S. Scott et al. (2014). Anti-social character traits were also reduced, and emotions expressed by the parents were warmer.
40. Layard and Clark (2014).
41. Other well-evidenced courses of parent training include the so-called ‘Triple P’ programme – the Positive Parenting Program.
CHAPTER 10 COMMUNITIES
1. From Martin Luther King Jr.’s lecture, 11 December 1964, to mark his acceptance of the Nobel Peace Prize; at <https://www.nobelprize.org/prizes/peace/1964/king/lecture/>.
2. TNS Loneliness Omnibus Survey for Age UK (April 2014). For people over sixty-five the figure is 1 million.
3. Steptoe and Lassale (2018), Figures 9.2 and 9.3.
4. McDaid et al. (2017), p. 32.
5. Holt-Lunstad et al. (2010); Holt-Lunstad et al. (2015).
6. See Figures 7.1 and 7.2.
7. Helliwell and Wang (2011), Table 3, shows clearly how strongly in Canada life-satisfaction depends on the sense of belonging to your community, your province and your country, and on your trust in your neighbours and co-workers.
8. For time devoted to these activities, see Krueger et al. (2009) and Gershuny and Halpin (1996).
9. Putnam (2007) describes social capital as ‘features of social life – networks, norms and trust – that enable participants to act together more effectively to pursue shared objectives’. These objectives must, in my view, include the sheer pleasure of being together. For evidence that social capital increases happiness, see Putnam (2000), Section IV. On social capital, see also Halpern (2004, 2015).
10. People underestimate the trustworthiness of others; see Helliwell and Wang (2011).
11. In Britain about one third of the income of NGOs comes from public funds; see Hall (1999).
12. Brown et al. (2003). Needless to say we want volunteers to volunteer because they want to help – not because it looks good on a CV. A sad example of the latter is the flyer for the LSE Careers Centre with the rhyme ‘Boost Your Career, Volunteer’.
13. Tan et al. (2006).
14. Carlson et al. (2015).
15. McDaid et al. (2017).
16. Pitkala et al. (2009). The costs of health care per person per year were €1,522 in the experimental treatment group compared with €2,465 in the control group. This statistically significant difference was larger than the total cost of the intervention (€881 including the group rehab and the programme costs, transport, food and the training and tutoring of the group leaders).
17. Through time-use surveys, the OECD estimates the economic value of volunteering for Germany in 2013 to be around USD 118 billion, or 3.3 per cent of real GDP. See Table 5.4 in OECD (2015). The figure is roughly comparable in other countries like the UK (2.5 per cent) and the USA (3.7 per cent).
18. Lawlor et al. (2014). Note that just visiting lonely people is often not particularly effective.
19. Halpern (2010), Figure 2.1.
20. Dustmann and Fasani (2016).
21. Halpern (2010).
22. Leong (2010).
23. Singleton et al. (1998), Table 12.1.
24. Barnoski and Aos (2004).
25. Oesterle et al. (2018).
26. Bentham (1789).
27. Putnam (2007); Alesina and La Ferrara (2000); Alesina and La Ferrara (2002); Alesina and Glaeser (2004); Glaeser et al. (2000).
28. Langella and Manning (2016). See also Longhi (2014), who finds that this result applies only to the white British population. However, if we look only at the effects of immigrants coming from Eastern Europe, Ivlevs and Veliziotis (2018) find that life-satisfaction is actually increased for residents who are younger or employed or on higher incomes, and reduced for people who are older or unemployed or on lower incomes.
29. Akay et al. (2014).
30. Betz and Simpson (2013).
31. Helliwell, Layard and Sachs (2018), p. 39.
32. Many studies have failed to find significant average effects on wages or employment, but there are clear individual instances of such effects. These can create great resentment both among those affected and also often in areas of very low immigration.
33. Helliwell, Huang, Wang and Shiplett (2018), p. 35.
34. Rao (2019).
35. There is a huge literature on whether people are happier living in villages or towns/cities of different sizes. But it is so far inconclusive.
36. Around London, land in the Green Belt with no special amenity value and within 800 metres of a tube or train station would be able to accommodate 1 million homes; see Stringer (2014).
37. See the minutes of the All Parliamentary Policy Group on Wellbeing Economics meeting, held on 12 June 2014: at <https://wellbeingeconomics.wordpress.com/2014/06/10/meeting-4-planning-transport-and-well-being/>. This is not atypical: most government departments have somewhat arbitrary and unrelated targets.
38. For a pioneering controlled study of the wellbeing effects of an urban improvement scheme, see Anderson et al. (2017).
39. For more on the Copenhagen project, see Gehl and Gemzøe (2004); Gehl and Rogers (2010).
40. Richard et al. (2008).
41. Gehl (2011). See also Whyte (1980).
42. Cohen and Spacapan (1984).
43. Halpern (1995).
44. Gifford (2007).
45. Office for National Statistics (2016). Living in high-rise compared with terraced housing reduced life-satisfaction by 0.3 points. In a study in the USA, more space was good for wellbeing but only when measured relative to the space which others had; see Bellet (2017).
46. See for example Davies-Cooper et al. (2014) and Vartanian et al. (2015).
47. Krekel et al. (2016).
48. 0.007 x 6,000. See Krekel et al. (2016). For a similar estimate, see White et al. (2013).
49. Life-satisfaction (LS) = 0.3 log income (Y) + etc. Thus 
50. Holding a constant distance from the city centre.
51. White et al. (2013). See also Alcock et al. (forthcoming).
52. Ulrich (1984).
53. Montgomery (2013).
54. Kuo and Sullivan (2001).
55. Weinstein et al. (2009).
56. MacKerron and Mourato (2013). See also Testoni and Dolan (2018).
57. Montgomery (2013), p. 110.
58. See Rajan (2019) and Collier (2018).
CHAPTER 11 ECONOMISTS
1. Ben Bernanke at the 32nd General Conference of the International Association for Research in Income and Wealth, Cambridge, Massachusetts; at <https://www.federalreserve.gov/newsevents/speech/bernanke20120806a.htm>.
2. This was the way economists framed their objective up till the 1930s. After that they redefined ‘utility’ in terms of preference-satisfaction, but, as I shall show, it needs to revert to its earlier meaning.
3. A slightly better measure of income is Net National Income, but this is less commonly used.
4. I was also shocked by the way economics textbooks like The Theory of Price by George Stigler discussed policy conclusions without having discussed welfare economics. Layard and Walters (1978) tried to rectify this.
5. Layard et al. (2008).
6. See, for example, Layard and Glaister (1994).
7. See Layard and Glaister (1994).
8. Quality is measured on a scale of 0–1. For some problems with QALYs, see Layard and Clark (2014), pp. 188–90.
9. For a detailed analysis, see Layard and O’Donnell (2015).
10. This unethical procedure is usually justified by the so-called Hicks/Kaldor criterion, which says that if the gainers could compensate the losers there is a net improvement, even if no compensation is paid. No ethical justification for the rule has ever been provided. A dreadful example of the use of this criterion is the analysis done to justify a new high-speed rail link from London to the North of England (HS2).
11. Durand (2018). The UAE approach is the most all-embracing, see UAE (2017). For the UK’s new Green Book, see HM Treasury (2018), paragraphs 2.3, 4.15, 4.16, 6.21 and 6.22.
12. Lucas (2003).
13. A. E. Clark et al. (2018), Tables 4.2 and 5.2. Data for the UK.
14. See Kahneman (2011), De Neve et al. (2018) and Boyce et al. (2013) – the last two for confirmation from real-world time series using happiness data at the country- and individual-level respectively.
15. See Layard et al. (2012), which summarizes the controversy between Easterlin (2010), Chapter 5, and Sacks et al. (2010).
16. In addition, people also adapt to higher levels of income. But this is generally a less important factor; see A. E. Clark et al. (2008) and A. E. Clark et al. (2018), Chapter 2. See also Vendrik (2013).
17. Baumol (1967).
18. This is also one of the advantages of immigration.
19. A. E. Clark et al. (2018), Table 16.2.
20. Layard and Nickell (2011).
21. To our great satisfaction, the explanation we gave for unemployment up to 1990 worked equally well in explaining unemployment in different European countries between 1990 and 2002; see ‘Introduction to New Edition’ in Layard et al. (2005).
22. A. E. Clark et al. (2018), pp. 74 and 63, show that one standard deviation of job quality raises life-satisfaction by 0.4 points, while unemployment lowers it by 0.7 points. So if it were better for a person to remain unemployed than to accept a job, the job would need to be very bad relative to the one they could get by waiting.
23. See, for example., Knabe et al. (2017), which finds ‘workfare’ participants have a level of life-satisfaction above that of those who are left unemployed (though lower than that of employed workers).
24. Layard and Philpott (1991).
25. Layard (2000), Blundell et al. (2004).
26. Marlow et al. (2012), p. 70.
27. Layard and Nickell (2011).
28. This should be automatic, but with a right for the client to opt out.
29. Autor et al. (2013).
30. Acemoglu and Autor (2011); Acemoglu and Autor (2012). Foreign trade has a major impact on the pattern of jobs (Autor et al. (2013)) but is not the main influence on increased wage inequality.
31. For Britain, see McIntosh and Morris (2016) and Walker and Zhu (2013).
32. For early analyses of the race between the demand for skill and the supply of skill, see Jackman et al. (1999) and Manacorda and Petrongolo (1999).
33. OECD data.
34. This both reduces the number of unskilled people and makes them scarcer, thus raising their wages.
35. Thomas and Dimsdale (2017), version 3.1.
36. This argument relates to an ongoing situation. If the degree of redistribution is increased at some time, we have also to take into account the extra cost due to loss aversion.
37. Atkinson and Stiglitz (1980).
38. Layard (1980); Layard (2005a); Layard (2006).
39. See Kahneman (2011), Thaler and Sunstein (2008), and Thaler (2015).
40. For example, Haushofer and Shapiro (2016) report on an unconditional cash-transfer trial in Kenya. For the ‘large transfer group’ their monthly income increased by a factor of three and their life-satisfaction increased by 0.36 standard deviations. The cost was high relative to the benefit in terms of wellbeing.
41. At present this means the respondent is in the bottom 10 per cent of the adult population.
42. This calculation of cost ignores the savings which flow back from reduced mental illness.
43. Wilkinson and Pickett (2009), Wilkinson and Pickett (2018).
44. Wilkinson and Pickett (2008), Wilkinson and Pickett (2018). On mental illness, the Kessler data which they use are very weakly correlated with the WHO estimate in World Health Organization (2017).
45. Both indirectly and directly. The direct effects would be attributable to diminishing marginal utility of income. Doubtless this effect exists, but it is too small to be detected. For example, if Hi = 0.3 log yi, then the difference in average happiness between Sweden and the USA on account of the differences in income equality would be roughly 0.075 points (out of 10). To derive this, we use a Taylor’s series expansion. So if Hi = 0.3 log yi and pi is the frequency of income yi, then
We assume SD (y) / y– = 0.4 in Sweden and 0.8 in the US.
46. See, for example, Goff et al. (2018), who find small or insignificant effects. For an earlier exploration of this issue, see Stevenson and Wolfers (2010).
47. Goff et al. (2018). A one-point increase in the standard deviation of happiness is associated with an increase in average happiness of 1 point (out of 10).
48. Goff et al. (2018). Table 6 shows that the relation between social trust and happiness inequality is closer than between social trust and income inequality.
49. Layard (2018b).
50. Easterlin (2010), Figures 5.3 and 5.4.
51. Easterlin et al. (2017).
52. Easterlin et al. (2017).
53. Knight and Gunatilaka (2018).
54. Tay and Diener (2011).
55. Only 10 per cent of migrants are refugees and, due to small sample numbers, the report includes migrants of all sorts in a single analysis.
56. If there is a flow of migrants between any two countries, the migrants gain in happiness nearly 90 per cent of the difference in happiness between the two countries.
57. Helliwell, Huang, Wang and Shiplett (2018), p. 35.
58. Helliwell, Layard and Sachs (2018), p. 9.
59. Collier (2013).
60. The Happiness Principle requires that we adopt many lower-level rules of thumb, which we generally respect (Hare (1981)). These include the rights of a country’s residents to control entry to their country, except in the case of refugees.
61. See, for example, my address at the American Economic Association meetings in 2018; Layard (2018a).
CHAPTER 12 POLITICIANS AND PUBLIC SERVANTS
1. Jefferson (1809).
2. For other similar arguments, see Bok (2011) and Diener et al. (2009).
3. Ward (2015; forthcoming). The explanatory variables also included average income, the unemployment rate, the inflation rate (all three insignificant), as well as the share of the governing parties at the previous election, a fixed effect per country, and a fixed effect per year. All variables were standardized using their standard deviations over the whole sample of countries and elections. If a different regression is run where variables are standardized using their standard deviations in each country, the partial correlations for life-satisfaction and economic growth are equal in size. In a separate analysis, greater individual life-satisfaction increases the probability of an individual supporting the government. On this, see also Liberini, Redoano and Proto (2017).
4. Ward et al. (2019). See also Ward (2019). The explanatory variables include the average share of the vote for the Republican Party in the previous four presidential elections.
5. ‘Eventually, the poor woman’s mother-in-law solved the problem by promising to pawn a set of gold earrings and so Nasam Velankanni got to hold her newborn baby.’ See Holmberg and Rothstein (2011).
6. Norris (2003), Table 2.
7. Helliwell, Huang, Grover and Wang (2018). One hundred and fifty-seven countries were surveyed. Fixed effects are included in the analysis. See also Ott (2011).
8. See Helliwell, Huang, Grover and Wang (2018). But this controls for many variables which may themselves be affected by democracy.
9. There was a reform in 2007, but it was evolutionary not revolutionary, see Zhang and Oyama (2016).
10. Franks and Mayer (1996); Bruner (2002). However, the shareholders of the company which is bought do on average gain, even though their workers usually lose.
11. Ipsos MORI (2017), p. 55.
12. For example in the British Rhine Army huge prestige went to the troop which was the very first to range on to an Army target.
13. This statement is less true in some developing countries. For an assessment of the effects of privatization in Britain, see Bishop and Kay (1989).
14. See O’Neill (2002).
15. Syed (2015).
16. Hagen (2015); EUROCONTROL (2006).
17. Aviation Safety Network.
18. Blair (2010), p. 216.
19. The Economist (2018).
20. On economic growth and public spending, there is a large and inconclusive literature, see Nijkamp and Poot (2004).
21. Flavin et al. (2011), Table 1, Column 2. Data from World Values Survey.
22. This is provided the technical quality of government is good enough. Using data for 130 nations from the Gallup World Poll, Ott (2011) confirms these results. Similar results were also found in Alvarez-Diaz et al. (2010) for a cross-section of US states, using the DDB Lifestyle Survey.
23. Using Gallup World Poll data on 106 countries, we regressed average happiness on average log GDP and (separately) (a) the share of tax in GDP, (b) the share of government services in GDP, and (c) the share of welfare spending in GDP. In none of the three cases did tax/spending have any significant effect in the equation. For all of our measures, we averaged the data over the three-year period between 2014 and 2016.
24. On the role of social comparisons, see Layard et al. (2010).
25. Suppose LS = 0.3 log Y and income is £30,000. Then 
So if £15,000 is raised in small amounts from many people, 
. 15,000 = 0.15. This ignores the role of comparator income.
26. Claxton et al. (2015) estimate that the spending of £15,000 in today’s money generates one extra Quality Adjusted Life Year where quality is measured on a scale of 0–1. But life-satisfaction is measured on a scale of 0–10, so we can roughly equate 1 QALY to 10 life-satisfaction years; see Peasgood et al. (2018).
27. Some people argue that higher taxes will reduce private charitable giving. The evidence here is ambiguous; see Bredtmann (2016).
28. Populism can be defined in different ways, but is typically characterized by a strong anti-establishment view, often combined with nationalist sentiment, see Mudde (2007).
29. There is tentative evidence that those supporting Brexit are less contented than other citizens (e.g. Liberini, Oswald, Proto and Redoano (2017)).
30. Data on individual European countries are in online Annex 12.1. Since 2006, when these data were first collected in the Gallup World Poll, there has been a rise in negative emotion (average of anger, worry and sadness) experienced the previous day and a decline in positive emotion (average of happiness, laughter and enjoyment) in Western Europe; see Helliwell, Huang and Wang (2019). See also online Annex 2.1.
31. For a summary of the evidence that news of bad events generally attracts more interest than good events, see Baumeister et al. (2001). See also Marshall and Kidd (1981).
32. Leetaru (2011). Data is taken from Summary of World Broadcasts, which tracks newspaper articles, conference proceedings, television and radio broadcasts, periodicals and non–classified technical reports in their native languages in over 130 countries.
33. Pinker (2018).
34. Lewis (2010); M. A. King (2016).
35. The Federal Communications Commission (FCC) fairness doctrine was introduced in 1949.
36. Even among professional academic economists this is also the case. A recent study showed a shocking amount of abusive language and behaviour on the Economics Job Market Rumors forum for economists, on which people post anonymously; see Wu (2018).
37. Murthy and Lakshminarayana (2006).
38. Pinker (2011).
39. Haidt and Lukianoff (2018), p. 58.
40. An egregious error occurred when Tony Blair sacked Home Secretary Charles Clark for a long-standing failure of the Home Office to deport immigrants convicted of crimes, after their release from prison. This happened to come to light when he was Home Secretary.
CHAPTER 13 SCIENTISTS AND TECHNOLOGISTS
1. Bacon (1597).
2. We may also need new ways to extract carbon dioxide from the atmosphere.
3. D. King, Browne et al. (2015).
4. Mazzucato (2015).
5. By 2050 Britain has committed to reduce emissions to no higher than 20 per cent of their level in 1990.
6. D. King, Browne et al. (2015).
7. It is run by the annual meeting of Energy Ministers of the member countries and a seven-person international secretariat. It has seven working parties on the different technological challenges.
8. This situation could only be avoided if cost-effective ways are discovered for extracting carbon dioxide from the atmosphere.
9. D. King, Schrag et al. (2015).
10. The British government uses a discount rate of 3.5 per cent. Of this, 2 per cent is due to higher income, on the basis that the marginal utility of income is inversely proportional to income, so that a forecast 2 per cent growth rate per annum makes next year’s marginal pound sterling worth 2 per cent less in happiness than this year’s pound. The remaining 1.5 per cent is due to ‘catastrophic risk and pure time preference’. A 3.5 per cent discount rate means that £1 in fifty years’ time is worth only 17 pence today.
11. Unless one assumes pure time preference – for which there is no obvious ethical justification.
12. See ranking of countries in Helliwell, Layard and Sachs (2018).
13. We are now at roughly 1 degree Celsius above pre-industrial temperatures, and the current rate of warming is 0.2 degrees per decade (IPCC (2018)).
14. United Nations Environment Programme (2018).
15. For example, these gadgets may well include deep brain stimulation for conditions like depression, Parkinson’s disease and dystonia; see Perlmutter and Mink (2006).
16. Also, perhaps, deep brain stimulation and transcranial magnetic stimulation.
17. Kramer (1994).
18. Currently in Britain, if a public body considers that parents’ choices risk harming a child, it can challenge them in court. Doctors and social workers can ask a judge for an order to override the legal state of parental responsibility.
19. Plomin (2018); see also Okbay et al. (2016).
20. For parents with fertility problems, we already allow IVF to fertilize many eggs and choose the one most free of disease.
21. See Layard et al. (1994). This explains why unemployment fluctuates with aggregate demand but is not generally trended.
22. See Twenge (2017). Monthly active users of Facebook went from 250 million in 2009 to 500 million in 2010 and then in a straight line to 2 billion in 2017.
23. Tromholt (2016). Lonely: 16 per cent versus 25 per cent; depressed: 22 per cent versus 35 per cent. See also Deters and Mehl (2013).
24. Shakya and Christakis (2017); Kross et al. (2013).
25. Haidt and Lukianoff (2018).
26. An example would be to ban all anonymous writing.
27. Schools which ban the carrying of mobile phones in school achieve better results; see Beland and Murphy (2016).
CONCLUSION: ACTIONS FOR HAPPINESS
A NOTE OF THANKS
1. O’Donnell et al. (2014).