Since the 1960s British researchers have been conducting longitudinal studies of British civil servants, the Whitehall studies.8 A twenty-five-year follow-up from the first Whitehall study found that the higher the position in the occupational hierarchy, the lower the mortality rate from all causes, from coronary heart disease, and from a range of diseases not related to smoking (Exhibit 1⇓).9 It should be borne in mind that none of these men was poor in any usual sense of that word. They were all in stable, office-based employment in and around Whitehall, London.
The civil service excludes the richest and poorest of society. Yet among these men there is more than a twofold difference in mortality rates, over the twenty-five years of follow-up, between top and bottom. Also, among these men there is a social gradient in health that runs from top to bottom of the social hierarchy. There is no clear point with good health above and poor health below.
One might assume that something peculiar to the class-based nature of the British civil service accounts for this remarkable social gradient. Not so. For example, a recent publication on British national data classified areas into twentieths according to degree of deprivation.10 There is no evidence of a threshold, but a clear gradient in mortality for the general population runs from the least to the most deprived.
I have argued, on the basis of the Whitehall findings, that the problem for the rich countries today is inequality in health rather than poverty and health.11 For me the distinction is important. Door-keepers and messengers in the British civil service are not poor compared with the working-class poor of late nineteenth-century York. If we are dealing with poverty in Britain and the United States, it is in general a different type of poverty that that of nineteenth-century York or of the poorest countries today. Because health follows a social gradient, if we wish to talk of deprivation, we have to appeal to the concept of relative deprivation.
The question of whether money matters can be approached in two ways: comparison of countries; and studies of the relationship between income and mortality within countries.
Comparisons of countries.
A 1993 World Bank report examined the relationship between life expectancy and gross national product (GNP) per capita in more than 100 countries from about 1900 to 1990.12 It extended the work of Preston, who showed these relationships for an earlier period.13 The report makes two key points. First, at low levels of GNP, a small increase in GNP corresponds to a large increase in life expectancy. As GNP increases, the relation levels off. Above about $5,000 per capita in 1991, there is a shallow relationship between a country’s average income and life expectancy.
Second, for a given GNP, life expectancy increased during the twentieth century. This suggests that the finding in York was part of a general pattern. In 1900 rich people in York had high infant mortality rates compared with those rates 100 years later. Something was responsible for the improvement that was not related to income. I speculate that even the servant-keepers would have been subject to some of the same environmental insults that we now associate with deprivation. With the improvement of water and sanitation, for example, this was removed.
That improvements in life expectancy in rich countries can happen for reasons unrelated to income was further emphasized by Amartya Sen.14 He looked at improvements in life expectancy in Britain by decade, from 1901 to 1960. The decades 1911–1921 and 1940–1951 had the fastest increases in life expectancy—decades that embraced the World Wars. These decades of fast expansion in life expectancy corresponded to slow growth of per capita GDP. Sen doubts that it is simply a time lag between economic growth and reduction of mortality rates. He attributes the rapid improvement in life expectancy in the two decades to policies of support: sharing of means of survival, including sharing of health care and the limited food supply (through rationing and subsidized nutrition). The psychology of sharing in beleaguered Britain made radical public arrangements for the distribution of food and health care acceptable and effective. Even the National Health Service (NHS) was born during the war years of World War II.
In the introduction I suggested that the commonalities between rich and poor countries might be greater than they appear. Sen attributes improvements in life expectancy in poor countries despite sluggish economic growth to “support-led” strategies; these include spending on public goods such as education, public health, and basic health care.