Were People in the Past Poor and Miserable? Charles Kenny* Abstract

Transcription

1 Were People in the Past Poor and Miserable? Charles Kenny* Abstract Standard economic theory would suggest close linkages between income, broader measures of the quality of life and utility. To some extent it is hard to test the link between income and utility given the way that utility is proxied by economists, but we do have measures of objective and subjective wellbeing as a potential guide. When we look at broader measures of objective and subjective wellbeing in both rich and poor countries today, the relationship to absolute income is perhaps surprisingly weak. Turning to the past, we know that most people were in an absolute income sense very poor, and faced a considerably lower broad quality of life using objective measures. However, the link between these two factors is not as straightforward as sometimes assumed. At the same time, from the preoccupations of political thinkers and others, it does appear that relative (rather than absolute) income has long been a concern, and that concerns with absolute income at the national level appear to center around avoiding absolute deprivation rather than the advantage of ever more consumption goods. In short, there is plentiful evidence that people in the past were nearly all absolutely poor and broadly worse off according to other objective quality of life measures, less evidence that these two were intimately linked, even less that everyone was miserable, and less again that those who did feel miserable felt so because they were absolutely poor. *Charles Kenny is an economist at the World Bank (ckenny@worldbank.org). This paper was written in a personal capacity and it would be a surprise were it to reflect the views of the World Bank, its Executive Directors or the countries that they represent. Thanks to Tony and Nancy Kenny for comments.

2 Introduction Economists tend to believe that you can t measure wellbeing (at least when defined as utility) directly. But, by studying people s revealed preferences, one can make an informed estimate of what improves it. One outcome of this examination according to mainstream economics is that, ceteris paribus, more income will increase welfare. Adding strength to the theoretical claim for such a link, numerous studies within and across countries suggest that rich people not only have more goods, but also have superior access to services, better health, higher education rates and are less likely to die violently, for example (Pritchett and Summers, 1996, Gangadharan and Valenzuela, 2001, and McGilivray, 2004). If the theory of a strong link from income to utility is correct, it has significant consequences for what we must think about the wellbeing of humanity in the past. We know, with certainty, that people had less income. The average GDP per capita in Western Europe in 0 B.C. was perhaps $450. The richest 10 percent had incomes averaging perhaps $1,640 per annum. 1 Compare that to an average $32,000 income in the United States or even an income of $5,760 for the poorest ten percent of Americans (World Bank, 2001). If the link between income and utility is strong, and given that 10 percent of US citizens still remain subjectively unhappy even with all of their current wealth (Furnham and Argyle, 1998), weeping, wailing and gnashing of teeth must have been the major preoccupation of 0 B.C. Europe, while laughter and contentment must surely post-date the industrial revolution. And yet, a number of papers (including Graham, 2004) have looked at the relationship between income and subjective wellbeing (self-reported contentment) within and across countries over recent time periods, and suggest the links may be more complex than often assumed. Indeed, while it seems reasonable to assume that greater health and other benefits of technological progress should have some impact on welfare, at the same time, it might be odd if our nature was so designed that subjective wellbeing required something we didn t have for most of our history (and have had for only a fleeting instant of our evolutionary existence). This paper makes an 1 In 1870s UK, GDP/capita was $3,191, the top 10 percent lived on perhaps $10,304. the same numbers for 1820 are $1707 and $5732 respectively, for 1700 $1250 and $4282 and for 1 AD Western Europe $450 and $1,638. All calculations are based on Maddison (2001) income data and a regression based on data from the World Bank (2003) linking PPP GDP to the income share of the top ten percent. The regression suggests that the top 10 percent s share in the last five years is equal to 49.2 minus the natural log of GDP/capita multiplied by

3 attempt to push the analysis of links between income and subjective welfare further into the past and tries to evaluate if the evidence available fits the idea that income-poor people were considerably worse off in terms of measures of subjective wellbeing. Of course, it is hard to know about the past even what we know today regarding subjective wellbeing nobody surveyed a representative sample of Anglo-Saxons in 800 AD England to as how they would rate their happiness on a scale of 1-3. Nonetheless, what limited evidence we can uncover from the past may provide some additional evidence regarding a prologue for greater economic growth and wellbeing, especially in poor countries. This paper looks at what we might learn about income, quality of life and views about the good life through history to see what they suggest about how poor people were, using different definitions of the term income, other objective quality of life measures and subjective wellbeing. What is Poverty? Does the Definition Matter? Poverty measures can be narrow and input-based (for example, income in dollars) or broad and output based ( functionings ), they also can be absolute (dollars a day) or relative (percentage of average income), objective ( accounting wealth ) or subjective ( how wealthy do you feel ). The view of poverty that has become most widespread in development economics is an absolute dollar amount. The UN s Millennium Development Goals, for example would halve poverty by halving the number of people living on less than a dollar a day. This even as the goals contain a list of other targets regarding attributes that people would be poorer, in a broader sense, without. All the same, many thinkers have suggested complementing or replacing income-only measures of poverty. A great number of people have come up with lists of basic needs, or elements of the full life, or the components of well-being beyond income, that the poor would have less of. Such lists often include food and shelter, economic and physical security, health, social acceptance and freedom of choice and action (Alkire, 2002) Such broader lists are important, because there is plentiful evidence that income alone does not capture very well levels of wellbeing associated with health, education, security or freedom, for 2

4 example. 2 Vietnam in 2000 had the same income as the UK in the early Nineteenth Century but compared to the UK in the 1800s, modern Vietnamese live an average of 28 years longer and the infant mortality rate is only a quarter as high for example (Kenny, 2005a). Income differences also have a limited impact on objective quality of life measures within developing countries at a particular time. In India, 60 percent of adults who are illiterate are above the national income poverty line and 63 percent of those with chronic health problems are above the income poverty line while 38 percent of those below the national poverty line are literate and 91 percent are not chronically unhealthy, for example (Laderchi, Saith and Stewart, 2002). Furthermore, a number of recent papers have suggested that the link between income growth and improvements in a range of other quality of life variables at the national level is also far weaker than frequently assumed. Kenny (2005b) shows both that most quality of life variables, unlike income, are rapidly converging and that significant improvements have occurred even in countries that have seen no economic growth. Easterly (1999) shows that there is no correlation between the speed of improvement in most quality of life variables and the speed of GDP per capita growth across countries. 3 There is also evidence that relative income measures sometimes correlate better with objective wellbeing measures than do absolute measures. Relative income measures are frequently found to have a more significant impact on health outcomes than absolute income effects, for example (Thorbecke and Charumilind, 2002,who also note a strong relationship between income inequality and crime rates although see Deaton, 2004). The importance of relative income to people s subjective view of their own income poverty is also clear from numerous studies. Today, more than half of Americans say that they cannot afford everything they really need (Graham et. al. 2004). Indeed, between 1958 and 1999 in the US, while personal disposable income per capita rose 131 percent in real terms, people s need for goods rose even faster personal consumption expenditures per capita rose 140 percent 2 It should be noted that even when we decide that income will dictate our measures of poverty, altering survey assumptions can dramatically change the number in poverty (Laderchi, Saith and Stewart, 2002). 3 While GDP per capita growth isn t related to a lot of good things, it is related to a load of bad things: man slaughter, rates of carbon dioxide, sulfur dioxide and nitrogen oxide pollution, output of waste paper and percentage of the population smoking (Easterly, 1999). See also Kenny (2005a) for a further examination of the link between income and quality of life spanning the Twentieth Century. 3

5 (Redmond, 2001). 4 Dickens in David Copperfield suggested in 1850 what this imbalance might do to happiness: Annual income, twenty pounds, annual expenditure nineteen nineteen six, result happiness. Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery. But the population of the United States are all the rich of Beverly Hills compared to a Tanzanian living on a dollar a day, or a Western European at the time of Christ living on only a little more. 5 It appears that relative income concerns must dominate over the benefits of any additional absolute income increase in the United States, then. 6 Even in poor communities there can be considerable divergence between subjective and objective measures of poverty. While actual measures of consumption are a statistically significant predictor of perceived consumption adequacy in Jamaica, a model including actual measures was only able to explain 12 percent of the variation in perceived consumption adequacy (Pradhan and Ravallion, 2000), suggesting that even in developing countries, it is difficult to predict who will perceive themselves as poor. A survey reported in Kenny (2005c) regarding Tanzania found that relative income effects played a role in that country as well. In areas around Lindi, where household incomes averaged $330 a year, the proportion of people who felt poor or very poor was lower than the proportion who felt poor or very poor around the city of Arusha, where the average household income was $2,120 (see also Laderchi, Saith and Stewart 2002, on Peru). 7 These results would be less of a surprise to anthropologists than to economists. Anthropologists see poverty even in poor countries not as the possession of few 4 Ask those with a net worth of between $1 and $4 million what rich is, and more than half say it is more than $5 million, while only 9 percent will admit to being rich themselves (Coniff, 2002). 5 Meanwhile, in 1990, the average American felt that 21 percent of his fellow citizens were rich. But less than 0.5 percent labeled themselves as rich this in the country that is the richest on World history (Myers, 1992). 6 So much are people concerned about relative income, that in experiments where people can pay some of their own money to reduce the amount of money that other (richer) subjects have been given, two thirds of subjects will pay to burn richer subjects money (Zizzo and Oswald, 2000). Evidence of the benefits of egalitarian distributions include a study of two communities in Israel, one with a very egalitarian distribution of income, another where there were significant differences in incomes between members, which found that those in the egalitarian community were considerably happier, that they were far more concerned with protecting comparatively egalitarian income distributions, and considerably less likely to think that having more money would make them happier (Morawez, 1977). A second study, by Graham and Felton (2004), finds that in Latin America, controlling for absolute income, the poor are made considerably more miserable by their relative poverty than the wealthy are made happy by their relative wealth. 7 It is clear that the reference group matters, then in Tanzania one significant reference group is the region, Graham and Felton (2004) report findings which suggest this may not be the case in Russia, where regional gini coefficients have no impact on subject wellbeing but inequality between Moscow and the rest of the country does. There is also plentiful evidence of relative income effects within firms (see Brown et. al., 2004). 4

6 goods, or some relationship between means and ends, but about relationships between people poverty is a social status (Geremek, 1997, Sahlins 1972). Moving even further from narrow, objective, income-based measures of poverty, we can look at subjective wellbeing self-measured utility as it were as captured in answers to happiness surveys of the type taking your life as a whole, do you consider yourself unhappy, somewhat happy, or very happy. Happiness survey responses, it has been shown elsewhere, do appear to capture elements of most people s conception of the term. People who score highly on such measures of subjective wellbeing appear happier to friends, family and psychologists, they smile more often, they remember more positive life events in their past, they have more friends and interact socially more, they are more likely to respond to requests for help and less likely to be involved in disputes, they respond physiologically less to stress, and they have different electroencephalogram measures of prefrontal brain activity. Hospitalized alcoholics, new prison inmates, new psychotherapy clients, South African blacks during apartheid, the unemployed and students living under politically oppressive regimes are all less happy than average (see Kenny, 1999, and Konow, 2002, for brief reviews). Perhaps one reason for the limited crossover of happiness studies into the academic mainstream is that the results suggest little that a cliché d sonnet, or any number of self-help books hasn t said before. Happiness isn t about money, its about friends, helping people, loving relationships and liking your job (Ahuvia and Friedman, 1998). Within the United States, it appears that differences in income account for perhaps 2 to five percent in the variation in subjective wellbeing across people at a single time (Ahuvia and Freidman, 1998), with a dramatic decline in the marginal utility of income (Helliwell, 2002). 8 Blanchflower and Oswald (2000) note that one of the interesting conclusions, from the economist s point of view, is how influential nonfinancial variables appear to be in human welfare. A lasting marriage is worth the same as $100,000 a year in terms of happiness, they estimate, while loosing a job, even after allowing for direct income effects of not getting a paycheck, costs $60, Conniff, 2002, asked a trust fund beneficiary to explain why the rich weren t happier and was rewarded with a list of woes: social isolation, resentment from peers, rich-bashing from society, betrayal or exploitation by friends, unrealistic expectations from family and society, unequal financial status in marriage, and an absence of all the usual factors (like worrying about the rent) that cause the rest of us to drag ourselves out of bed most mornings in search of bread and a modicum of self-worth. 5

7 At the same time, what causes happiness is, perhaps unsurprisingly, itself a social construct. Given that the number of likes and dislikes we re born with is limited (sweet tastes, certain types of touch on the positive side, pain and bitter tastes on the other), it is clear that most likes and dislikes are acquired. And it is also clear that, as we change our positions in life, our likes and opinions can change fairly rapidly. 9 We know that preferences can be re-anchored up and down in response to environmental stimuli such as display. 10 We also know that people are adaptable downwards when it comes to the impact on subjective wellbeing of declines in broader measures of the quality of life (Cummins et. al., 2002). 11 It is not surprising, then, that the relative importance of the factors that make up a person s view of the happy life can change. One example of recent change in people s utility calculus is opinions about the importance of money itself. In the United States in 1970, only 39 percent of college freshmen said that being very well off financially was an essential or very important life goal. By 1995, this had risen to 74 percent, making it the most important life objective and beating out others such as raising a family (Myers and Diener, 1997). 12 It also appears that the happiness impact of unemployment, for example, is in large part the result of its social status the impact is greater where there are stronger norms of living off one s own income. Furthermore, right-wingers are upset more about inflation than unemployment, while left-wingers feel the reverse (Frey and Stutzer, 2002). 9 There is evidence that people can be trained to act as the market suggests they should, for example. While novices do not trade as much as they should between goods of equal market value, those used to trading goods do so (List, 2003). It is also easy to reduce the intrinsic joy that people feel in performing a particular activity, for example just pay them to do it (Bowles, 1998). 10 In a 1981 survey, only 17 percent of those who were content with the amount of money they had also said they had less money than their friends compared to 59 percent who labeled themselves money troubled (Furnham and Argyle, 1998), although apparently, over time, what the Joneses have in their house has become a less important anchor while television and advertising is playing a larger role (Redmond, 2001). Myers (1992) reports that there was a strong correlation between the introduction of television in US cities and increases in theft where TV became widely available in 1951, thefts jumped in that year; where it became widely available in 1955, that year instead saw significant jumps in the rate of thefts. As younger, poorer people saw images of wealthier people on TV, resentment led to crime. 11 Even with health, self-reported ratings correlate far better with happiness than objective health ratings by physicians, and people can quickly recover in terms of subjective wellbeing from maiming accidents although adaptation is rarely complete (Cummins et. al. 2002, Zautra and Hempel, 1984 and Frey and Stutzer, 2002). And we also know that people themselves are not the best judge of the long-term impact of change on their contentment. If they have truly internalized the values of the new system, they have repudiated the values of the old. We have plenty of examples of cases where many of us are willing to be skeptical of the newfound contentment --religious converts, those who have lived through revolutions. 12 This might help to explain some of the decline in average US happiness over that period, as materialism is negatively correlated with subjective wellbeing even in the US (Ahuvia and Friedman, 1998). 6

8 Across wealthy countries, we see similar results to the US regarding subjective wellbeing and its determinants both that income is not terribly important and that social differences may lead to differences in what makes for a happy person. While some papers have suggested that there some relationship between income growth and happiness in rich countries (Tella, McCulloch and Oswald (1997), Hagerty and Veenhoven (1999), Veenhoven, 1991) it remains fair to say that the relationship is weak (Kenny 1999 and Easterlin, 1974, 1998, 2004). 13 Helliwell (2002) argues that [t]hose who have the highest levels of subjective well-being are not those who live in the richest countries, but those who live where social and political institutions are effect[ive], where mutual trust is high, and corruption is low (see also ILO, 2004). Regarding differences in what makes people happy between countries, it appears that Europeans are made considerably more unhappy by inequality than are most Americans, for example (the US exception is rich leftwingers see Alesina, Di Tella and MacCulloch, 2004). Turning to poorer countries, that Nigeria reported average happiness on a ten-point scale of 6.82 in 1995 while Japan s average was 6.61 and South Korea s was 6.69 (Frey and Stutzer, 2002) suggests that there may be a lot more to happiness than income. Graham (2004) concludes that her work in Russia and Peru suggests that our results strongly support the important role played by relative income differences, reference norms, and other non [absolute]-income factors... Indeed, for the most part, the determinants of happiness were very similar in the developing economies [to determinants] in the advanced economies. Age, income, education, marriage, employment and health effects were all broadly the same (see also Graham and Felton, 2004). The other thing that appears similar about the results is the small percentage of the variation explained by all of these variables put together R-Squareds of 0.08 (or 8 percent of variation explained) in the US suggest that income actually explained more of the variation in happiness in this rich country than it did in the available middle-income samples, where only seven percent in Latin America and four percent of the variation in Russia was explained by income, health, marriage and other factors all together (See also Cummins et. al., Rojas, 2002, Vitterso et. al. 2002, Graham et. al., 2004). 13 Methodological differences will be one reason for differing results. Estimates that include fixed effects suggest a far lower impact of income (see Di Tella, MacCulloch and Oswald (2001) for somewhat mixed results depending on method). 7

9 At the same time, it appears that the non-income correlates with happiness may be different in developing than developed countries. Self esteem is much more closely correlated with happiness in Western cultures whereas in collectivist cultures happiness is more closely related to feelings of social harmony (Frey and Stutzer, 2002). 14 In developing countries there is apparently little correlation between trust in government or more objective measures of political and economic freedom and happiness (whereas there is in Europe) perhaps because expectations are lower in LDCs (Frey and Stutzer, 2002). As with the United States, there is also evidence that values are changing during growth in LDCs, in a way that might minimize the benefits of that growth for happiness. Abramovitz argues, for example, that economic development is associated with an increased tendency to base value on price (Abramovitz, 1958). We can watch consumer cultures developing in countries as they get richer (see for example, Hart, 2001 on Korea). Perhaps growth itself leads to a central concern with income as a quality of life measure. 15 Overall, then, the evidence for the modern world suggests that links between income and both broader objective quality of life measures and subjective measures may be weaker than often assumed. Relative income concerns may play a considerable role in what weak relationship there is between income and happiness worldwide. And given a less developed consumer culture in poorer countries, it may even be that income as a source of subjective wellbeing is more important in rich countries than in poor ones. It is important to note at this point that the poorest of the poor in the developing world live lives of deprivation by anyone s standards. In these cases, it would be surprising indeed if more incomes and dramatically improved physical quality of life did not lead to higher subjective wellbeing (indeed, it would suggest great weaknesses in the subjective wellbeing measures themselves). And evidence from the slums of Calcutta suggests that, at this level of deprivation, 14 As Christopher (1999) points out, the subjective wellbeing measure is grounded in an egalitarian, utilitarian atomistic world view. This is not the viewpoint that was common in the past, or indeed universal at the dawn of the Twenty-First Century. 15 Although even post-modern ennui may in fact predate the modern. Schama (1997) quotes de Tocqueville suggesting about Nineteenth-century America that the strange melancholy which often haunts the inhabitants of democratic countries in the midst of their abundance, and the disgust at life which sometimes seizes upon them in the midst of calm and easy circumstances. 8

10 increased income does notably improve subjective wellbeing (Diener and Oishi 1997). Turing to the history of the West, perhaps one question is what percentage of people in European history lived lives similar to those faced in Calcutta s slums and a second question is what evidence is there of widespread misery based on absolute income deprivation? Income and Objective Quality of Life in History In the last few hundred years for parts of the World, and in the last 100 globally, objective measures of the quality of life have begun to improve, surely not unrelated to technological advance, globalization and (even) economic growth. As Richard Easterlin (2000) notes [b]y most measures, a revolution in the human condition is sweeping the world. Most people today are better fed, clothed, and housed than their predecessors two centuries ago. They are healthier, live longer, and are better educated. Women s lives are less centered on reproduction and political democracy has gained a foothold. Real income worldwide has risen dramatically. Global GDP per capita averaged perhaps $444 in 1 AD, and a little less 1000 years later. But by 1820, it had risen to $667, with Western European GDPs per capita averaging about $1,232. In 1998, global incomes averaged $5,709, with OECD countries averaging well above $20,000 (Maddison, 2001). In 1820, perhaps 84 percent of the World s population of 1 billion lived on below a dollar a day. By 1910, this had dropped to 66 percent of a population of 1.7 billion. In 1992, just 24 percent of a population of 5.5 billion lived on below a dollar a day (Bourguignon and Morrisson, 2002). At broadly the same time, life expectancy, which hovered above 20 years for the whole world for most of history, began to make impressive strides in the West in the second half of the Nineteenth Century and worldwide in the Twentieth (Simon, 1995). World life expectancy has increased from an average of 24 years in 1000AD to 31 in 1900 and 66 in 1999 (Maddison, 2001). A range of other measures of the quality of life are rapidly improving worldwide and faster in developing countries than in the developed world (Kenny, 2005b) There has also been a dramatic extension of the rights of man over time. Engerman (1995) reports estimates that about one third of the population of ancient Greek states were slaves, and a similar proportion in Roman Italy in 28 9

11 Along with recent and dramatic improvements in the objective quality of life, there is also a good deal more (human) life around. The population of the planet has increased from four million 10,000 years ago to one billion in the Nineteenth Century to six billion today (Simon, 1995). 17 At the same time, the link between economic advance and broader quality of life advance has not always been straightforward. First, poor communities in the past have managed to provide basic elements of the quality of life to all --and proposals have long been around to go considerably further than that --John Comenius, in 1650, was arguing for universal education, pre-natal clinics, marriage counseling and geriatrics (Barzun, 2000) when UK income was below $1, A second sign of a weak relationship is that economic advance is not always (usually?) associated with advances in the broader quality of life. Even at the dawn of history, civilization was hardly an unmitigated good for the masses. Vernon Smith (1987) notes that early huntergatherer societies (much like those still extant today) had reliable frequently abundant food supplies, and life was far from brutish. Indeed, the very fearsome abilities of early hunter BC. In the US Southern States in 1750, 38 percent of the population were slaves, while in the Caribbean, the proportion reached as high as 90 percent. Legal slavery died out in 1970, with final emancipations on the Arabian peninsula. At the same time, the percentage of the world s population that lives under what might broadly be defined as a democratic state has risen from effectively zero in 1 AD to 52 percent in countries labeled fully democratic by UNDP s 2002 Human Development Report. In the UK, the percentage of the population above age 20 that was enfranchised climbed from below 10 percent in the 1820s to nearly universal levels 100 years later. 17 Mill opposed ever increasing populations on similar grounds to Malthus that more would not be better. In On Liberty (1869) he argued that "To bestow a life which may be either a curse or a blessing, unless the being on whom it is to be bestowed will have at least the ordinary chances of a desirable existence, is a crime against that being." This is a tradeoff that no longer applies in most of the world, potentially suggesting a gargantuan utilitarian benefit to further greater population growth (although Collard (2003) argues that as far as Bentham, at least, was concerned, the present population of the polity was all that was important for utilitarian calculations). 18 Looking at measures of nutrition, even in the 1500s in the UK, starvation was extremely rare. Fogel (1995) points out that in the period the UK saw famines accounting for less than 0.6 percent of total mortality (all crisis events together only accounted for 6 percent of mortality). For much of modern British history, there appears to have been a food surplus which has allowed for conspicuous consumption even as the majority of peasants ate reasonably well. The rich have had to find other ways to display their differences from the common crowd than basic nutrition and eating too much (conspicuous over-consumption) was one method favored by Henry VIII (Conniff, 2002). There was a rise in malnutrition during the Industrial Revolution in part a result of economic change and harsher policies towards a growing legion of the absolutely poor based on a Smithian concern with laissez-faire. Smith (in the Wealth of Nations) argued that scarcity of corn could be the result of the waste of war or by the weather, but that a famine has never arisen from any other cause but the violence of the government attempting, by improper means, to remedy the inconvenience of a dearth by putting in price controls. Perhaps, but the Irish potato famine suggested at the least that some interventionist policy was needed to prevent starvation. 10

12 gatherers put at risk seemingly limitless supplies of game. 19 The adoption of arable farming worldwide allowed for population expansion and the birth of cities, but was far from an unmitigated good. On the Mississippi floodplain, for example, the move to maize cultivation required collective effort that gave rise to storage facilities, large fields, earth moving, elites to oversee it all and (so) growing inequality. Indeed, it may be that crop domestication was actually started as a way to generate intoxicants and delicacies for the privileged the first domestic crops were chili peppers (still a status symbol in some Maya cultures), wheat (first used for beer) and the bottle gourd, used as a serving vessel at feasts (Conniff, 2002). Overall, early farmers had to work harder and had a higher rate of starvation and of chronic disease than their hunter-gatherer forebears (Fernandez-Armesto, 2001, see also Harlan, 1992) in sum, there is something to the noble savage model, at least compared to early civilization. 20 Looking at the impact of globalization, while it is probably responsible for much of the rapid improvements in the worldwide objective quality of life that we have seen in the past Century, its record was not always so positive. As with the dawn of civilization, the dawn of globalization was accompanied by significant suffering. Perhaps 90 percent of the Mexican Indian population was wiped out by "the white man and his fellow-traveling pathogens" as Landes (1998) puts it. 21 Meanwhile, only ten million African slaves reached the New World out of probably over 20 million who left Africa. When they got there, many died from disease, violence and suicide --an indication of the harshness of the life was that they died far faster than they reproduced. Life expectancy at landing was estimated at below seven years in 1682 (Fernandez-Armesto, 2001). Back in Northern Europe, average heights fell from 173 cm in the AD period to 166 cm 19 Although, regarding the distant past, we should avoid romantic images of the noble savage in a contented stasis with nature. The Kalahari bushmen, a frequent subject of noble savage paeans, see very high infant mortality rates, for example. 40 percent of children die before the age of 15 because of malnutrition and limited to non-existent access to healthcare (DeGregori, 1998). 20 Indeed, it is clear that life was pretty grim for the Russian serf or English peasant throughout most of history. It is likely that infant mortality as the cap on population growth was the world norm prior to 1790, and life expectancies over 30 only came about in the last years. And disaster could strike on a catastrophic scale one third of the European population was killed by the Black Death in the few decades after it reached Messina in 1347 (Bernstein, 2004). 21 Having said that, Aztec blood orgies at the time of Columbus involved thousands of victims at a time, suggesting that many native Americans were damned either way. 11

13 in the Seventeenth and Eighteenth Centuries, not recovering to previous levels until as late as the mid-20 th Century. Changing climate, growing inequality, urbanization, increased trade, colonization and wars all take part of the blame (Steckel, 2001). 22 Overall, what did less than nothing for the average European in the age of discovery also did immense harm to those who were conquered and enslaved. At the same time, more income has also allowed more effective violence, with the percentage of people dying in wars rising from a little under one percent of the global population in the Nineteenth Century to over four percent in the Twentieth. 23 Under these circumstances, it is not surprising that the poor peasantry that tended to make up revolutionary cannon-fodder were usually reactionary thinking that their life was life likely to be better as it was than in changing. Skocpol (1976) argues that: When peasants rose during historical social revolutionary crises, they did so in highly traditional rebellious patterns: bread riots, defense of communal lands or customary rights, riots against hoarding merchants or landlords, social banditry. Peasants initially drew upon traditional cultural themes to justify 22 Komlos (1998) notes the evidence of declining stature from the 1830s to the 1860s in the US, and from as early as 1730 through 1850 in the United Kingdom, despite considerably rising incomes in general suggests the importance of growing inequality, income variability, child labor, urbanization absent refrigeration and sanitation to health outcomes. Even at the low level of incomes extant in Eighteenth Century Britain, then, any impact of increases in income on health were swamped by (associated) changes such as increased urbanization. 23 Technological and economic advance has also allowed more people to be involved in more bloody wars. In the 1470s, during the Wars of the Roses, the size of UK-based armies was approximately 0.6 percent of the population. By 1810, during the Napoleonic Wars, this had risen to 5 percent (Ferguson, 2001). By World War One, the UK had 9 percent of its citizens under arms (calculated from Maddison, 2001 and Ferguson, 2002). Connected with this capacity was a generally more powerful state The government s revenues rose from perhaps 2 percent of national income under Elizabeth in the late Sixteenth Century to four percent 100 years later. By 1788, tax revenues had reached 12.4 percent of GDP. By the late Twentieth Century they were close to 45 percent (Ferguson, 2001). Under the wrong leadership, more powerful states were capable of ever greater atrocities. Matthew White has collected data on the history of wars, massacres and atrocities of history that attempts to calculate death tolls through time. His list of the 22 worst things that people have done to each other, measured in terms of numbers of deaths (the list includes wars, the slave trade, famine in British India amongst others) suggests that ten of the twenty-two, along with about half of the total death toll (in the region of 190 million people) occurred in the Nineteenth and Twentieth Centuries. The rest span the Seventh through Eighteenth Centuries. Looking in more detail at the Nineteenth and Twentieth Centuries, White estimates that between four to five percent of all deaths in the Twentieth Century worldwide were 'overtly caused by other people' in genocides, wars, and other large acts of state-sponsored violence. In the Nineteenth Century, this same figure he estimates at 0.9 percent ( Mary, in Sixteenth Century Britain, killed only about 300 people to earn the title bloody --compare this to the many millions killed by revolutionaries such as Mao or Stalin, with powerful state machines at their back, acting in the name of the people. 12

14 rebellion. Far from being revolutionaries through adoption of a radical vision of a desired new society, revolutionary peasants have typically been backwardlooking rebels incorporated by circumstances beyond their control into political processes occurring independently of them... The record of revolutions suggests that, however unhappy people were in the past, as a rule they were probably made more miserable by attempts to make them happier. Burke (who stood up for continuity even at the cost of minor disruption such as American independence) may have been the more helpful defender of the little man than Abbe Sieyes (champion of the French Revolution), then. 24 Turning to the last five hundred years in the United Kingdom, we can take a closer look at the link between income growth and quality of life growth in recent history. Table One suggests that the quality of life using a number of indicators has been improving in the UK since as early as the 1500s. Income indicators really began to take off in the early Nineteenth Century, with health and other indicators following towards the end of the same century. Nonetheless, a cursory examination of the evidence in the table suggests that the link between income change and improvements in health is not straightforward. 24 Regarding violent death outside of war, the picture is also mixed while murder appears to have declined, suicide has risen. Murder rates in the UK have seen a dramatic decline over the period from the 13 th Century until today, with rates dropping from around 20 homicides per 100,000 people to under one per 100,000 today (Chesnais, 1995). At the same time, Chesnais notes that [i]n the Middle decades of the last century, no country outside the German cultural area of Central Europe had a [suicide] rate of above 10 per 100,000 inhabitants. Currently [i]n the Western World, most rates are between 10 and

15 Table One Historical Indicators of UK Quality of Life Date Income ($) Stature (cm) Life Expectancy Infant Survival (per 1,000 live births) Calorie Intake Literacy Source: Kenny, 2005b except stature 1300 from Steckel, 2001, life expectancy estimated from Floud and Harris, 1996 and literacy from Floud and Harris, 1996 (dates are within one year). The industrial revolution was a particularly mixed bag. Adam Smith thought that the great majority of Britain s population was able to afford the absolute necessities of life prior to the Industrial Revolution. The rich, by employing the poor for their own vain and insatiable desires... make nearly the same distribution of the necessities of life... had the earth been divided into equal portions among all its inhabitants he argued in the Theory of Moral Sentiments (Smith, 1982). A little further into the Industrial Revolution Malthus Essay on Population (1798) was arguing that things had got worse. The increasing wealth of the nation has little or 14

16 no tendency to better the conditions of the labouring poor. They have not, I believe, a greater command of the necessaries and conveniences of life, and much greater proportion of them than at the period of the revolution is employed in manufactures and crowded together in closse and unwholesome rooms. Mill, in the Principles of Political Economy, written some fifty years later (1848, this edition 1985) recognized that more and more was being produced in England, but still argued that it is questionable if all the mechanical inventions yet made have lightened the day s toil of any human being. They have enabled a greater proportion to live the same life of drudgery and imprisonment. 25 Average rates of consumption did not begin to rise until the late 1840s (Mokyr, 1988), and this during a period of growing inequality. Lindert (1986) estimates that the income share of the top ten percent of the population in England and Wales rose from 44.1 percent in 1688 to 48.8 percent by 1803 and 53.4 percent by 1867 (it is 27.3 percent today). 26 Furthermore, the risk of being thrown into absolute destitution in the growing urban areas appears to have been much higher than in rural communities (Fernandez-Armesto, 2001). 27 This all as people were forced to work longer hours. At the start of the Industrial Revolution, working hours rose from under 3,000 in 1750, to over 3,200 hours per year by 1800, and did not fall below the 1750 level again until over 100 years later (Voth, 2003) In 1845, Freidrich Engels claimed that English society daily and hourly commits what the workingmen s organs, with perfect correctness, characterise as social murder (quoted in Floud and Harris, 1996). The Communist Manifesto, in part based on Engels experience of mid-century Manchester, highlighted conditions in the English city perhaps most brutalized by the Industrial Revolution at the nadir of its socioeconomic development. Dickens visited in 1838 and said that what I have seen has disgusted and astonished me beyond all measure. (Boyer, 1998). 26 Hoffman et. al. (2000) argue that real inequality rose during the 16 th Century and remained high until the 20 th Century. O Rourke and Williamson (2002) present evidence that wage-rental ratios in England were falling (suggesting growing inequality) from at least the 1500s until the early to mid Nineteenth Century, when they began to reverse. According to data in Clark, 2001, wages and salaries rose from 61 percent of GDP in 1265 to 67 percent in From then until 1805, the percentage declined to 55 percent. In 1825, the share was 58 percent and by 1865 it was 65 percent. 27 It should be noted that the idea of the Industrial Revolution has come under some attack. While, between , the population of England increased from 5 to 17 million and the number employed in agriculture fell from 61 to 29 percent (Floud and Harris, 1996), Clark (2001) argues that there was no revolution in economic output until after 1860, which would to some extent re-establish the correspondence of the revolution in social and economic progress. 28 An illustration of the disruption caused by the start of the industrial revolution, and the weak response of the government to that disruption, is provided by poor relief records from the village of Compton and the town of Shefford in Bedfordshire. Expenditures in both communities approximately trebled between 1794 and 1830, 15

17 Health impacts were also negative. During the Industrial Revolution, debilitating hunger was seen by medical authorities as a major cause of workers susceptibility to disease (Fernandez- Armesto, 2001). 29 Easterlin (1995, 2000) argues that life expectancy in the late Nineteenth Century was very similar to levels reached in Elizabethan England. The life expectancy figures reported in Table One also suggest very slow progress in health indicators in the first half of the Nineteenth Century, with rapidly increasing per capita GDP in the middle of the century accompanied by stagnant life expectancies and declining statures. And as with income inequality, health inequality also increased. Fogel (2004) argues that from the beginning of the industrial revolution to the end of the Nineteenth Century, the gap in life expectancy between rich and poor expanded by about ten years (by 1875, the total gap was 17 years, it has since dropped to four years). 30 It was technology and institutional advance (rather than more income) that played the major role in reversing the health declines associated with industrialization. In the early Nineteenth Century, rapidly expanding English and Welsh provincial cities saw declining life expectancies reflecting a considerable rise in the number of paupers (as much as a quadrupling in Shefford, approximately a doubling in Compton) (Williams, 2005). The government s response to a national situation that reflected this local situation was the Poor Law Amendment Act of 1934 that stipulated authorities should give no help to able-bodied individuals outside a workhouse, that the workhouse should operate on the principle of 'less eligibility' (workhouse conditions should be made less preferable than those of the lowest paid laborer), and that groups of paupers (including husbands and wives) should be separated. 29 Contrast the Seventeenth Century Dutch Republic where incomes were approximately the same as 1840s Britain Dutch were well fed, and food shortages were rare --to the extent that even poorhouse food included bread, milk, cheese, vegetable and meat stews as well as some fowl (Schama, 1997). Workers in towns were able to afford fresh and cured meat and fish, was well as fruit, vegetables, butter, eggs and cheese. The real social divide, according to Schama, was between those who could only afford to spread butter on their bread and those that could substitute it for lard in cooking. Having said that, life expectancies were still in the low 30s, suggesting the overwhelming importance of improved scientific knowledge to improved health. 30 Rich people have always lived longer. In Nineteenth Century Glasgow, the height of the obelisk under which richer people tended to get buried was correlated with the lifespan of the cadaver underneath each extra meter of obelisk was worth two additional years of life. Fathers able to invest in the largest dowries for their daughters in fifteenth-century Florence had less than half the mortality rates of poorer fathers (Conniff, 2002). Looking at the indicator of infant mortality, for the British peerage, child mortality rates started declining as early as the mid- Eighteenth Century. From the 1550s until 1749, rates of male child mortality amongst the privileged fluctuated between 256 and 338 per 1,000 live births. In the period, that rate declined to 192, and continued falling after that to a rate of 109 in the period (Hill, 1995). (The Hobbesian level of might be that achieved in refugee camps, where rates as high as 600 per 1,000 live births have been recorded (White, 2004). Some developing countries at the start of this century saw consistent rates as high as 400 per 1,000 births). These rates were somewhat (although not dramatically) below the 156 rate reported for the country as a whole in 1840 (see table 3). At the same time, and in common with modern cross-country evidence, it is clear that the effect of income was not as large as might be supposed that the rich in the 1850s saw considerably higher infant mortality rates than those who are absolutely poorer today (compare Vietnam s rate of 37 per 1,000 to the UK aristocracy s rate of 109). 16

18 between the 1820s and 1840s these dropped from 35 to 30 years, compared to a national average that stayed level at 41 years. 31 Easterlin (2004) provides some graphic illustrations as to why, quoting contemporaries who describe town dairies dens and cellars in which cows were kept for the greater part of the year, standing knee-deep in filth or a girl making cigars while [smallpox] scabs were falling from her skin tailors making soldiers clothing, having their children, from whom scabs were falling, wrapped in the garments and a dunghill that contains a hundred cubic yards of inpure filth on which flies would settle and then pass into neighboring houses. 32 Only with the sanitation and health reforms of the later 1800s did the rural-urban gap decline, reaching only four years in the UK by the 1890s (see also Komlos, 1994 and Steckel, 2001). Deaths from infectious disease in particular in the UK fell from accounting for 31 percent to 6 percent of the total over the period as a result of such advances -- in the period , causal agents for typhoid, leprosy, malaria, tuberculosis, glanders, cholera, streptococcus, diptheria, tetanus, pneumococcus, malta fever, gas gangrene, plague and dysentery were all discovered and in the 40 years , vaccines for diphtheria, cholera, pertussis, tuberculosis, tetanus, yellow fever and typhoid fever were developed (Easterlin, 2004). 33 Today, as a result of these advances, urbanization worldwide is often associated with higher life expectancy (Moore et. al., 1999 although the relationship does not appear to be robust). In short, long-term global and more recent British history suggests that huge economic advance has taken place as well as huge overall improvements in the quality (and quantity) of life. However, economic advance has not always been correlated with broader social improvement 31 In the first half of the Nineteenth Century, life expectancy in the major cities of France was seven or eight years lower than France as a whole (Easterlin 2004). At the same time, Haines, Craig and Weiss (2000) calculate that in the United States in the 1850s wealthier counties actually had higher crude death rates any health advantage to wealth was outweighed by the fact that these counties tended to have higher rates of urbanization and transport links. Kunitz (1983) also suggests that higher infant mortality rates were associated with higher urbanization at least until the 1880s (see also Boyer, 1998). 32 Every article of food and drink must be covered, otherwise, if left exposed for a minute, the flies immediately attack it, and it is rendered unfit for use, from the strong taste of the dunghill left by the flies. Finally, Easterlin passes on a description of the Thames in the summer of 1858 when for the first time in the history of man, the sewage of nearly three millions of people had been brought to seethe and ferment under a burning sun in one vast open cloea lying in their midst. 33 Ferrie and Troesken (2005) suggest that as much as half of the 60 percent decline in crude death rates in Chicago is due to the introduction of pure water alone. 17

19 civilization, globalization and industrialization all expanded output, trade, disease and warfare at the same time. For the great majority of people in the World, it would be hard to make the case that civilization, globalization or industrialization had any significant net positive impact on objective quality of life indicators at least until the mid Nineteenth Century, and in many cases later than that. What might we conclude regarding the history of income and objective measures of quality of life? Certainly, in terms of health and education, the last 150 years has seen revolutionary improvements worldwide, in some part fostered by developmental states and global technological progress, even while the ability of the state to starve or murder millions has also dramatically increased. Furthermore, the link between economic advance and advance in these broader areas of objective wellbeing has sometimes been tenuous and sometimes negative especially prior to the era of sustained economic growth. The invasive state has created the power to do some significant good in terms of improving the quality of life: it is probably behind urbanization now being a force for improvements in the quality of life rather than reverses, for example, but it is also a power for considerable harm. Overall, it is clear that technologies and institutions have been required to improve the broader quality of life not only, nor primarily, income. A recent sign of the weak historical link between income and broader measures of the quality of life over the last 150 years is that it takes one tenth the income to achieve the same life expectancy in 1999 as it took in 1870 (Kenny, 2005b). Happiness and Income in the History of Thought And what of the link between income changes, objective quality of life changes and subjective feelings about that quality through history? Better objective health measures, education and civil liberties are correlates with happiness in rich countries today and these variables have all improved. At the same time, some correlates with subjective wellbeing have declined (including church attendance, see Frey and Stutzer, 2002), suicide rates have risen dramatically (Chesnais, 1995) and we have seen that correlates of wellbeing themselves can change over time and across 18

20 cultures. Given this, it might be worth examining what historical writers actually said about subjective wellbeing and in particular the link with income. First, it is important to note that some people in the past did manage to have a good time or at least write about good times. To take but one example, can anyone really imagine Chaucer s Wife of Bath as an unhappy soul? Far from sacred or gloomy, the mood portrayed in much of the popular literature of the Middle Ages is jollity argues Jacques Barzan (Barzun, 2000). It is also clear that the entire world did not share a culture of poverty as Lewis (1959) would have it lack of integration with major institutions, minimal organization, weak family structures, feelings of helplessness. This despite all being incredibly poor in any modern Western monetary sense of the term. At the same time, there is a limit on relativism in respect to the good life. A healthy life, a peaceful death, plentiful food and a limit to toil are all elements of the quality of life that have both long been present and we can still understand. Hesiod (8th Century BC) provides an early example of the perfect life (he places it in the past): Nothing for toil or pitiful age they cared, but in strength of hand and foot still unimpaired they feasted gaily, undarkened by sufferings They died as if falling asleep; and all good things Were theirs, for the fruitful earth unstintingly bore Unforced her plenty... (translation from Claeys and Sargent, 1999). The universality of social peace and of the perfect society satisfying simple needs are also currents in utopian visions from very early on. 34 Furthermore, some historical literature does also suggest income or broader progress was an important goal. The fountain of eternal youth and the idea of turning lead into gold both suggest preoccupations with modern concerns of longer life and greater wealth. 34 There is, of course, considerable evidence that they remain universal today a recent survey of poor areas in South Africa found that over 98 percent of respondents endorsed the importance of work, good health, avoiding hunger, knowledge and friendship as parts of the good life, for example (Clark, 2002). But it is worth noting in this survey that income as an essential aspect of a good life ranked below jobs, access to water and sanitation, housing and shelter, family and friends, personal safety, education, happiness, good health, sleep and rest, self-respect, living in a clean natural environment, electricity and political rights. 19