Daniels, Kennedy, and Kawachi have written an important and deeply humane essay. They take as their starting point the well-known observation that wealthier populations are on average healthier and that the more evenly wealth is distributed within a population, the better the average health. Thus, life expectancy in rich countries is higher than in poor countries, and it is higher in countries like Sweden, where wealth is spread relatively evenly, than it is in similarly rich countries, such as the United States, where there are vast disparities. The authors conclude, reasonably enough, that there are social determinants of health, and that these determinants vary across and within populations. Less reasonably, they conclude that there is something about inequality itself, quite apart from poverty, that is a risk to health. On this basis, they propose that lessening income disparities within populations would improve the overall health of a population. Their philosophical underpinning is Rawls’s theory of justice, but that is hardly necessary for the case that the reforms they suggest would make the United States a healthier, wiser, and more decent nation.

Poverty is crippling not only physically but intellectually and spiritually.

There is no question that socioeconomic status and health are tightly linked, and the effect of one on the other can be huge. For example, in a study of the use of aspirin to prevent heart attacks in male physicians, the modest benefit of aspirin was swamped by the benefits apparently conferred by the high socioeconomic status of these men. The overall rate of fatal heart attacks in the physicians was only 12 percent of what would be expected in men from the general population. (To be sure, we don’t know whether doctors might do better than other privileged groups, such as lawyers, because of their medical education, but even if that were true, medical education would still have to be counted as a social determinant.) Very few medicines or interventions can offer such a benefit.

Indeed, the fact that social advantage correlates so closely and powerfully with health can make it extremely difficult to interpret the results of clinical research. Studies of the effect of passive smoking on childhood asthma, for example, are impossible to interpret unless they attempt to control for socioeconomic status. Without such control, it is impossible to know whether the increased prevalence of asthma in the children of smokers is really because of passive smoking or because smokers are more likely to be poor and poverty itself is associated with asthma for other reasons. Similarly, studies of the effect of lead exposure on intelligence are confounded by socioeconomic status. The children of well-educated parents are more likely to do well on IQ tests and are also less likely to be exposed to lead. It is hard to know, then, what causes low IQ scores: lead or lack of parental education.

Yet, despite the undoubted importance of socioeconomic status to health, no one knows which aspect of social standing matters—wealth or education or occupation or some other condition—much less how it operates. We are dealing here with a black box—the most mysterious and powerful of all determinants of health. Differences in medical care seem to account for only a small part of the effect, as pointed out by Daniels, Kennedy, and Kawachi. The lion’s share of the effect is caused by other factors, mostly unknown. Since it is inconceivable that money in the bank or a sheepskin on the wall could directly affect health, they must be markers for the real factors that matter.

What might those factors be? Most good studies of the subject—and there are lamentably few—try to control for the usual suspects, such as cigarette smoking and heavy drinking, both of which are more frequent among people of lower socioeconomic status. Even after controlling for them, the health disparities across social strata persist, although they are lessened. The increased frequencies of trauma, substance abuse, and HIV infection among the disadvantaged cannot explain the differences, either, since death rates from other causes, such as heart disease and cancer, are also higher in poor people. One can imagine a host of influences that might affect health—such as diet, stress, exposure to infectious agents or toxins—that are related to socioeconomic status, but there is very little evidence to point to any of them as a major cause of the health disparities across income groups.

Some people suggest that in analyzing the association between health and socioeconomic status, we tend to confuse cause and effect. They believe that privilege does not lead to better health, but rather the reverse—healthier people tend to become richer and better educated, because they are more energetic and competitive. A variation of this view holds that both health and wealth stem from good genes. Whatever the answer to the question of why socioeconomic status is correlated with health, the question deserves serious study. Good research in this area will undoubtedly yield enormous dividends in understanding human biology and health, and Daniels, Kennedy, and Kawachi are on solid ground in pointing us in that direction.

They are on less solid ground in their contention that inequality somehow contributes to poor health directly, above and beyond the effects of poverty itself. Although there is some evidence of that from international comparisons, it is by no means consistent. Denmark, for example, has about the same per capita wealth as the United States with less inequality, but its life expectancy is lower. Kerala and Costa Rica, which provide the strongest support for a direct benefit of equality, are such outliers that it is risky to generalize from them.

Unequal societies, by definition, have pockets of poverty and pockets of great wealth. If the pockets of poverty contribute disproportionately to population measures of health—such as average life expectancy—that would explain the apparent correlation between inequality and poor health. I believe that is the likely explanation. Inequality just seems to be a direct contributor to poor health, whereas the real cause is poverty. Daniels, Kennedy, and Kawachi base their argument for a direct effect of inequality on the notion of a linear health gradient that operates equally across all socioeconomic strata, so that the wealthy benefit as much as the poor lose. But the evidence for that is weak. There have not been sufficient studies of a broad enough range of income levels to know what the shape of the curve is. The best information may come from international comparisons showing that with increasing wealth, the health benefits become smaller and smaller until a plateau is reached.

One need not invoke some mysterious effect of inequality on health to make a very strong argument for lessening inequalities that lead to deprivation at the low end of the scale. Poverty is crippling not only physically but intellectually and spiritually. It cripples any wealthy society that tolerates it on a large scale, as does the United States. In addition to the loss of human potential and the social pathology that grows out of poverty, the costs include the callousness that inures the rest of society to its presence, even as many people enjoy extraordinary riches.

The fact that there are also health consequences of poverty, whether they are exacerbated by inequality or not, is doubly punishing and adds greatly to the injustice. Daniels, Kennedy, and Kawachi are right about that. F. Scott Fitzgerald famously pointed out that the rich are different from the rest of us. But what is less well known is that he observed that no difference that divides people is so important as that between the well and the sick. I agree.