Blog
Global Patterns of Income and Health

by Angus Deaton

Inequalities in income and inequalities in health

Global inequality takes many dimensions. Not only is there great inequality across the peoples of the world in material standards of living, but there are also dramatic inequalities in health. The inhabitants of poor countries not only have lower real incomes, but they are also more often sick, and they live shorter lives. These international correlations between income and health should affect the way that we think about the level and distribution of global wellbeing. They also need to be understood if we are to be effective in reducing global poverty, in incomes, or in health. Wellbeing should never be thought of only in terms of income, or only in terms of health. Health scientists and economists need to come together if we are to fully understand global poverty and inequality and if we are to design policies that will be effective in making the world a better place, particularly for its most deprived inhabitants.

Income, life expectancy, and global wellbeing

In a justly celebrated paper, Samuel Preston drew a graph of life expectancy against national income, showing that, in the poorest countries, small changes in average income were associated with large improvements in life-expectancy, while among the rich countries, the protective effects of income, although still present, were less pronounced. Figure 1 shows the relationship for the year 2000. Life expectancy is on the vertical axis, and per capita national income in purchasing power parity dollars is on the horizontal axis; the diameter of each circle is proportional to the population of each country. The point at which the slope flattens out, at an income per capita of around $5,000 in 2000, is the ‘epidemiological transition’, where deaths from the ‘diseases of the bowels and lungs of children’ are replaced by deaths from ‘diseases of the arteries of the elderly’. Preston also showed that income is not the only factor at work. Even countries with stagnant incomes typically have increases in life expectancy, which he attributed to the adoption of new methods of public health, or more precisely in most cases, to the new adoption of old methods of public health.​

If we were to make an adjustment to income to take into account life chances, scaling down for those with poor health, and scaling up for those with good health, the world distribution of this compound of income and health would be much more unequal than the world distribution of either health or income taken separately. The case for humanitarian action to improve the lot of the world poorest is stronger once we recognize that the poor are doubly deprived, not only in material living standards, but also in lower chances of living a long and healthy life.

The joint distribution of global health and income has undergone remarkable changes over time. Before the industrial revolution, there was relatively little inequality between countries in either health or income, although there was a great deal of inequality within countries. As the material living standards of the countries of northwest Europe, particularly Britain, began to pull away from the rest of the world, so did their life expectancy rates. This growth in cross-country inequality of incomes that began in the eighteenth century shows no sign of reversal to this day, but the same is not true of health. The sanitary and preventive measures associated with an understanding of the germ theory of disease spread to the countries of southern and eastern Europe in the early years of the twentieth century and, after the Second World War, were rapidly brought to the rest of the world. As a result of this diffusion of health knowledge, cross national inequality in life expectancies fell quite sharply in the years after 1950. If, for example, we take a simple compound measure of income and health by multiplying income by life-expectancy, then inequality over countries declined for most of the post-war period, driven not by any reduction in income inequality between countries, but by the convergence of life-expectancy.

Things fell apart after 1990. HIV/ AIDS in sub-Saharan Africa undid almost all of the increases in life-expectancy of the previous 40 years and there was a significant, although smaller, reduction in life expectancy in the countries of the former Soviet Union. Huge gaps in life expectancy opened up again between many African countries and the rich, healthy countries of the world, and health (and total) inequality increased once again. If, as we might hope, the AIDS pandemic is a historical anomaly, international health convergence will resume. 

The cross-country convergence in health is arguably overstated by focusing on life expectancy. The lives that are being saved in poor countries are mostly the lives of babies, and these reductions in infant mortality have a dramatic effect on increasing life expectancy. Indeed, in the years immediately after the Second World War, in the heyday of malaria eradication and child vaccination, some countries increased their life expectancies by several years each year. By contrast, the current increase in life expectancy in rich countries is largely driven by reductions in the mortality rates of the middleaged and elderly, partly through reductions in smoking, partly through improvements in medical prevention and treatment, particularly for cardiovascular disease, and partly from improvements in the health and nutritional standards of children 50 years ago. It is far from clear how we should compare the value of mortality reductions at different ages. Focusing on the effects on life expectancy is one way of comparing the two, but this essentially arbitrary ‘solution’ gives much higher weight to reductions in infant and child mortality, and requires a better justification than habit and the convenient availability of information on life expectancy.

One important consideration is that fertility falls with infant mortality, albeit with a lag. If parents adjust their fertility so that the number of surviving children is the same after the reduction in infant mortality as it was before, the new age structure of the population will eventually settle down to the previous one, except for the absence of those young children who were previously destined to die. In the new demographic equilibrium, the babies who used to cease to exist very soon after they were born, because of lack of vaccinations or clean water, are now not born at all or, to use parallel language, now cease to exist immediately before they are born. Before the health improvement, these now unborn children would have been born, and most would have led lives beyond infancy. These possibly good lives are lost. While not everyone accepts the legitimacy of including the potential lives of unborn children in welfare calculations, the example should at least make clear the illegitimacy of assuming that mortality reductions at different ages are of the same value if they have the same effects on life expectancy. In any case, the calculus of life expectancy takes no account of the benefits to mothers’ lives of not having to bear so many children and of having to watch so many of them die.

Growth as a means to better health?

The correlation between health and income opens up the possibility that economic growth is one route to improving global health. If this is true, then economic growth would be twice blessed, because it would simultaneously reduce both income and health poverty.

That low incomes are a primary cause of ill-health is plausible enough. The vast majority of deaths in poor countries are from diseases from which almost no one dies in rich countries. In the countries that the World Bank classifies as ‘low income’, nearly one-third of all deaths are among children who have not reached their fifth birthday. In the ‘high income’ countries, less than 1 per cent of deaths are deaths of children. Worldwide, there are around four million deaths a year from acute respiratory infections, nearly two million from diarrheal disease, and more than a million from diseases that are preventable by childhood immunization. How to prevent or cure these diseases is well-known, and often long known, so that these deaths come not from lack of knowledge, but because of some other factor, among which poverty is clearly a leading suspect. In further support of the poverty hypothesis, it has long been known that countries whose economies have grown faster have had the largest (proportionate) reductions in infant and child mortality. (What happens to adult mortality is not something that we know with any certainty, given the dearth of vital registration systems in the poorest countries of the world.)

But the correlation between growth and child mortality is not all that it seems. Perhaps surprisingly, there is no correlation at all between the absolute rate of mortality reduction and economic growth, even over the 40 year period from 1960 to 2000. Countries that grow faster have lower levels of infant mortality so that, even though their reductions in infant mortality are on average no higher, the proportional declines are larger. This correlation between economic growth and the level of infant mortality cannot come from a causal link from poverty to mortality, which would lead to a relationship of changes with changes, and almost certainly reflects omitted third factors that influence both. Indeed, the literature on economic growth presents a number of immediately appealing candidates, such as the quality of governance, which not only creates the environment for economic growth, but also helps states provide an effective system of public health and healthcare delivery. Even more important is the level of education, particularly women’s Angus Deaton is Dwight D. Eisenhower Professor of Economics and International Affairs at Princeton University. education which is more important for child mortality and health than is either economic growth or poverty reduction. Better educated women are better caregivers to their children, as well as more effective users of and lobbyists for better health services.

Economic growth is much to be desired because it relieves the grinding material poverty of much of the world’s population. But economic growth, by itself, will not be enough to improve population health, at least in any acceptable time. Instead, governments must tackle the often difficult task of increasing education, and of providing better public health and health services, particularly in the poorest and least served areas. As far as health is concerned, the market, by itself, is not a substitute for collective action.

Angus Deaton is Dwight D. Eisenhower Professor of Economics and International Affairs at Princeton University.