Stubborn Attachments

by Tyler Cowen

  Thus even a constant level of reported happiness implies growth in real happiness over time, because the word “happy” takes on ever more ambitious meanings as society accumulates more wealth and richer experiences. Life improvements do generally make us happier, while both our expectations for happiness and our reporting standards for “being happy”—our use of language—adjust upwards.

  The belief that greater wealth correlates with greater happiness is supported by direct observation. Many individuals strive to earn higher incomes, even after they have experienced the strength of aspiration and treadmill effects. It’s not that they are all being tricked, but rather that they know at a gut level that money will help them achieve valuable ends or that it will help their families. Individuals value happiness for their families as well as for themselves, even if this particular channel of happiness is quite indirect and is not always reflected in their daily moods or their moment-to-moment self-reports.

  It is also the case that within a country, wealthier people report unambiguously higher levels of happiness, on average, than do poorer people.21 For all the talk about how some happiness studies present a revisionist view of material wealth, this result has not been challenged, and it pretty decisively demonstrates that, at least on average, wealth brings more happiness. To some extent, the greater self-reported happiness of the wealthy may reflect a zero-sum relative status effect, namely that the wealthier people feel better, but their possessions make the poor feel worse off. Nonetheless, it is unlikely that all of the gains from wealth, or even most of the gains, dissipate in zero-sum games. Wealthier lives are easier and happier in absolute terms in numerous ways, as discussed above, and as evidenced by a quick look at where most immigrants wish to migrate, namely to the wealthier countries.

  If a wealthy man buys a Mercedes, his neighbor may express greater dissatisfaction with his Volkswagen. That same neighbor, if he had a Lada in Leningrad circa 1976, might express a high or at least decent level of satisfaction on a happiness questionnaire. In absolute terms, however, he is still much better off having the Volkswagen in contemporary America than the Lada in 1970s Leningrad. The Volkswagen is a pretty good car, and the Lada broke down a lot and was hard to start. So while the neighbor might envy the wealthy man’s Mercedes, the happiness gains from wealth (and from better cars) do not dissipate through envy. Better cars really do make us better off. To put it another way: it is better to envy your neighbor’s Mercedes than to envy his horse and buggy. Envying his supersonic transport would be better still.

  On top of all of these considerations, happiness isn’t a single, simple variable which can be measured unambiguously. Happiness means a lot of different things to different people. Some people may seek temporary stimulations while others may want to feel fulfilled at the end of their lives, and others may seek to maximize the quality of their typical day. Some will seek happiness through the process of out-competing their peers for status, while others will look inward for contemplative delights. Most likely we seek some mix of these ends, but with varying emphases and weights. Wealthier societies offer greater opportunities and freedoms to pursue preferred concepts of happiness, even if this privilege does not always show up in the measurement of a single, aggregate number.

  The happiness literature takes a relatively limited view of human well-being. Usually the contemporary empirical literature on happiness starts with the operational definition of whether an honest, self-aware person would report himself or herself as being happy, if asked. Even if this accurately captures one notion of happiness, and perhaps it does, it is only one component of human well-being. A wealthier economy will offer greater options for structuring choices of work vs. friends, thrills vs. long-term satisfaction, enjoying children vs. a life of theater and travel, and so on.

  A wealthier economy also gives us more “fleeting” experiences of happiness. An individual will admit to being happier if he has recently found a dime, or if his soccer team won a big game. These sources of happiness will likely be more frequent and more consistent in the wealthier society. A diverse commercial economy offers more sources of temporary stimulation and more short-term turns of good fortune. This means more new gadgets, more entertaining videos, and more serendipitous encounters with interesting new people. That sounds a bit superficial, and indeed it is, but it is yet another reason why economic growth will boost happiness in its more complex and plural forms.22

  At most, the happiness literature shows that many changes in individual conditions are irrelevant for our well-being, due to habituation and expectation effects. This conclusion would not, however, eliminate the major benefits of economic growth. Even if many small changes in income are irrelevant or nearly irrelevant to happiness, sufficiently large changes in life circumstances may still boost or harm our welfare.

  As an example of how large changes matter, most life catastrophes create significant misery. Very sick individuals have less autonomy, experience more pain, and face the stress of dealing with their condition. The death of a child or close family member profoundly damages happiness for most individuals, and these effects can persist for many years. Torture, extreme stress, rape, and severe physical pain also produce depression, trauma, and persistent unhappiness. Individuals who have been through wars, revolutions, and collapses of civil order very often experience recurring flashbacks, nightmares, irritability, depression, alcoholism, troubled relationships, and an inability to concentrate. However psychologically troubled our modern, wealthy societies may seem, poverty is not the solution to these problems, and in fact it makes them worse.23

  Some commentators have doubted whether even extreme catastrophes make people less well-off. For instance, the onset of a severe disability or physical handicap may cause individuals to alter their expectations for what their lives will be like. Some or even all of their initial levels of happiness may be reattained by lowering their aspirations, and for that reason the loss in happiness may be blunted. Nonetheless, victims of catastrophes still report lower levels of happiness than do comparable healthy individuals, and many of these victims endure years of significant suffering. Coping mechanisms appear to work the least well when an individual’s condition is worsening steadily over time.24

  So, to the extent that a poorer society results in an ongoing worsening of conditions for many individuals, the associated human suffering will be greater, and this does indeed represent a true loss of human well-being. Once again, there are significant benefits to ongoing and sustainable economic growth.

  Extra wealth also serves as a cushion against very bad events, or at least against later declines in wealth. Ten or fifteen years ago, it was common to hear the claim that once a nation reaches the level of material wealth found in Greece, happiness more or less flatlines. Indeed, this was more or less where the flatlining point seemed to be. Yet since the Greek economic crisis, dating from 2009, no one uses the Greek example to make a point about the flatlining of the happiness-income relationship. The country lost almost a quarter of its economic output, unemployment has risen to over twenty percent, there have been riots in the streets, a neo-Nazi party was elected to the legislature, and at times basic medicines have been unavailable. Having some additional reserves of wealth prior to the crisis would have helped the country a good deal, and might have prevented the troubles altogether by easing debt repayment.

  Finally, even if we accept the “flatline” empirical result on happiness and wealth, these self-reported happiness questionnaires are given to individuals in normal life circumstances. The answers will not pick up the ability of wealthier economies to postpone or mitigate extreme tragedies. For instance, happiness measures cannot pick up the benefits of greater life expectancy. The dead and incapacitated cannot complain about their situation from the grave, at least not on questionnaires. Life expectancy rises with wealth, but self-reports of happiness miss this benefit because researchers do not poll the dead. If an immigrant, or a child of immi
grants, fills out a happiness questionnaire, no comparison is made between their current and pre-immigration state of affairs, either as it was or as it might have been. By design, happiness research draws upon a fixed pool of people living under relatively normal circumstances. This will limit its ability to measure some of the largest benefits brought about by economic growth. If we want to be around to even have the option of answering happiness questionnaires, wealth is extremely important.

  1. In terms of providing operational guidance in calculating Wealth Plus, the two best efforts I know of are Jones and Klenow (2016) and Becker, Philipson, and Soares (2005). For a good piece that puts production and productivity at the center of moral theory, see Stanczyk (2012), and also works by the philosopher David Schmidtz.

  2. For a formal look at the concept of sustainability, see Heal (1998).

  3. See Shermer (2002). Rees (2003) and especially Posner (2004) both focus on issues of civilizational collapse. S.E. Finer (1999, 30–34) provides an alternative look at civilizational survival, under the heading “Total Life-Spans.” He defines a civilization in broader terms than does Shermer, so for him ancient Egypt is one civilization, not eight. Civilizations have correspondingly longer life spans. Some of the longer-lived civilizations include Egypt (2,820 years), China (2,133 years), and the Byzantine Empire (1,962 years). The Venetian Republic lasted 1,112 years. The shorter examples include the Achaemenid Persian Empire (220 years) and the Sassanian Persian Empire (427 years). Finer also develops the more finely grained category of civilizational breakdowns, which occur more frequently. A breakdown is the “disintegration of a previously united state” (32), in contrast to the more severe examples of total civilizational collapse. Egyptian breakdowns occur after periods of 675, 184, 206, 215, and 1,238 years. Chinese breakdowns occur after periods of 400, 500, 442, 360, 326, 69, and 936 years. Assyrian breakdowns occur after periods of 157, 82, 38, 143, and 312 years. Arnold Toynbee, in his classic A Study of History, classifies world history into twenty-six civilizations. By his count, sixteen of these civilizations no longer exist. Samuel Huntington’s The Clash of Civilizations (1996, 44–45), citing Matthew Melko (1969), refers to twelve major civilizations, seven of which have perished.

  4. Samuel Huntington (1996) cites a variety of definitions of “civilization,” including the concepts of “settled, urban, and literate” (40) and “the broadest cultural entity” (43). Fernández-Armesto (2001, 16–20), referring to the work of Kenneth Clark, designates a civilization as a society with the confidence to build for the future. Clark (1969) noted that while he did not know exactly what civilization was, he could recognize it when he saw it. Matthew Melko (1969, 113) remarks that “when a civilization is operating effectively, it is likely to grow.”

  5. Diamond (2005) focuses on how ecological catastrophes have destroyed or damaged civilizations of the more distant past. On the brutality of primitive warfare, see Keeley (1996) and LeBlanc (2003).

  6. These figures are drawn from the U.S. Census Bureau, via an unpublished paper by Bradford DeLong, “Lecture 2: Slow Growth and Poverty in the North Atlantic, 1800–1870.” On work hours, see Huberman and Minns (2007).

  7. See Fogel (2004, 8, 9, 34).

  8. See Fogel (2004, 70).

  9. I am indebted to Don Boudreaux for this framing of the point.

  10. See for instance Dollar and Kraay (2000).

  11. See Helpman (2004, 84).

  12. Statistical work on trade, investment, and growth cannot always sort out which variables are endogenous. Nonetheless, the available evidence shows definite correlations between openness to trade and growth; see Helpman (2004, 70–71) for a survey.

  13. On the decline in global inequality, see for instance Lakner and Milanovic (2014).

  14. See for instance Venkatraman (2009) and Tsai (2006).

  15. Cowen (2004).

  16. See Lucas (1988, 5).

  17. See Friedman (2006). On the connection between the arts and economic growth, see Cowen (1998). For a recent defense of GDP maximization, see Oulton (2012). For a recent philosophic defense of the importance of economic growth based on pluralistic considerations, see Moller (2011). Stanczyk (2012) considers the importance of “productive justice,” namely making sure the output gets produced in the first place.

  18. See for instance Stevenson and Wolfers (2008, 2013), Sacks, Stevenson, and Wolfers (2010), and Deaton (2007).

  19. On the flattening of the curve, see for instance Helliwell (2002, 28). Other relevant sources are Argyle (1999), Oswald (1997), and Myers (2000). Wealthy countries, when they become wealthier over time, do not in general become happier in the aggregate. In some cases (e.g., the United States 1946–1991), greater wealth is correlated with lower levels of self-reported happiness; see Dieter (1984), Blanchflower and Oswald (2000), Diener and Oishi (2000), Myers (2000), Kenny (1999), Lane (1998), Frey and Stutzer (2000), and Easterlin (1995). On the United States, see Frey and Stutzer (2002a, 76–77). For a criticism of the income-happiness link from a time use perspective, see Kahneman et al. (2006).

  20. See Deaton (2007).

  21. See for instance Dieter (1984), among many other sources.

  22. See Schwarz and Strack (1999) on some of these dilemmas. Levinson (2013) offers some interesting observations on the role that projection (assuming excess permanence of current events) may play in these evaluations.

  23. On the link between catastrophes and unhappiness, see Dyregrov (1990), Lehman et al. (1987), Weiss (1987), Frederick and Loewenstein (1999), Lehman, Wortman, and Williams (1987), Archer (2001), and Wortman et al. (1992). For evidence on the difficulty of recovering from rape, see Meyer and Taylor (1986) and Wirtz and Harrell (1987).

  24. On coping, see Brickman, Coates, and Janoff-Bulman (1978), Bulman and Wortman (1977), Kessler, Price, and Wortman (1985), and Wortman and Silver (1987). On the “core of distress” idea, see Frey and Stutzer (2002a, 56), Wirtz and Harrell (1987), and Stroebe et al. (2001). On once-and-for-all changes vs. ongoing deteriorations, see Frederick and Loewenstein (1999).

  3—Overcoming disagreement

  The overwhelming benefits of economic growth help us resolve clashing preferences, and thus we are able to overcome what I have called aggregation problems. A higher rate of economic growth, of course, will not make everyone better off; some individuals living today might be better off with more consumption and less work, and thus in a regime with lower rates of economic growth. And even if economic growth makes many people better off in the longer run, it is unlikely to make each and every person better off along the way.

  But still, consider how the problem of clashing preferences has been transformed. When rates of economic growth are sustainably higher, there is an overwhelming preponderance of benefits on one side of the scale.

  Consider a one percentage point boost to the growth rate, starting at parity. We would need a time horizon of 110.4 years to establish a 3:1 ratio of one GDP (throughout, adjusted for leisure time and sustainability) to another. If we consider a two percentage point boost in the growth rate, we would need a time horizon of 55.5 years. For a more ambitious boost of five percentage points to the growth rate, the time horizon must stretch for only 22.5 years to obtain that 3:1 ratio.

  Keep in mind that the indicated number of years expresses when a 3:1 ratio will be reached, comparing a higher to a lower growth rate. Over time, if the higher growth continues, that 3:1 ratio will be exceeded on an ongoing basis and that gap will increase continually. For growth boosts of one, two, and five percentage points net, we would need 161, 81, and 33 years, respectively, to reach a quintupling of GDP. So the 3:1 requirement is only a temporary milestone on the path toward even greater discrepancies of wealth and (most likely) well-being.

  Economists sometimes argue that a construct known as Arrow’s impossibility theorem makes it impossible to resolve clashes of preference. But once we consider happiness and wel
l-being as relevant features for comparing one choice to another, Arrow’s impossibility theorem does not hold. Arrow’s original theorem assumes, and has to assume, that “happiness talk” (e.g., cardinal, interpersonal utility information) is absent from the comparisons of social states. Think of Arrow’s theorem as looking only at the kind of information which can be reflected in observed votes or ordinal rankings. Once a richer information set is introduced, the standard impossibility theorems fade away, and that is why I do not pay them much heed. They were only a big problem in some rather limited theoretical frameworks, and they were never a big problem in the real world in the first place.1

  The main point is simply that if the gains to the future are significant and ongoing, those gains should eventually outweigh one-time costs by a significant degree, and they will likely carry along other plural values as well.

  Think about South Korea today vs. sub-Saharan Africa. As argued above, a sustainable increase in economic growth, properly understood, will boost many plural values in the medium and long runs. To be sure, some people will be worse off, and some values, in the short to medium run, will not be favored. In these respects, aggregation problems do not disappear. Nonetheless, the competing options do not generally offer a deadlock of roughly equivalent values and interests on each side of the scale, with one side looking better by some moderate amount. South Korea is much better off than, say, the Democratic Republic of Congo by a considerable margin. The higher growth alternative will eventually offer a clear and ongoing preponderance of plural values in its favor, whether it be living standards, women’s rights, freedom of choice, the fight against poverty, or other important values. So why not choose that option and recognize that we have rational grounds for preferring it?2