Conventional wisdom in the development community includes the following two ideas:
1) There is good evidence that health improvements in poor countries lead to significant increases in GDP per capita.
2) Idea (1) is an important consideration for policy making.
I would like to propose a heretical take on these questions:
1) Available data and theory do not support the conclusion that health improvements in poor countries lead to significant increases in GDP per capita.
2) Idea (1) is not relevant for policy making.
Idea (1) in its conventional form is most closely associated with Jeffrey Sachs. It has found support in papers by Bloom, Canning, and Sevilla, among others. As discussed in the paper that I posted earlier, there are significant problems of endogeneity and omitted variables in the inferences that Sachs and company draw from cross country data. Two alternative research approaches have found different answers. Acemoglu and Johnson (2007, Journal of Political Economy) use instrumental variables to control for the endogeneity of health improvements, finding that such improvements actually lower GDP per capita at a horizon of 40 years. Ashraf, Lester, and Weil (2008, NBER Macroeconomics Annual) put microeconomically based estimates of the effects of health on labor supply and human capital accumulation into a demographic-economic simulation model and show that, while health improvements do increase income per capita, these increases are small and very slow in coming.
Now, just as there are problems with the Sachs et al. approach, there are plenty of problems with the alternatives. For example, Ashraf, Lester, and Weil point out problems with the results that Acemoglu and Johnson get regarding the effect of health improvements on population growth (which they see as the main reason why health improvements lower GDP per capita). And Ashraf, Lester, and Weil are upfront about the fact that they do not account for all possible effects of health on growth. So I would definitely not want to assert that health improvements don’t have a significant impact on economic growth. My claim is simply that there is not strong evidence for it at present.
Let me now turn to idea (2): that the effect of health on economic growth is important for policy-making. The question, presumably, is whether scarce resources (political, financial, human) should be devoted to improving health in developing countries or to some other pressing need. The conventional take on idea (2) is that the fact that they provide a two-for-one benefit (less death and suffering, on the one hand, and more economic growth, on the other) will make policies to improve health all the more attractive. Indeed, under some of the more dramatic estimates of the growth effect of improved health, one can look at health improvement solely as a pro-growth policy which has better bang-for-the-buck than available alternatives (building roads, etc.).
My heretical take on idea (2), which is summed up in the paper I posted earlier, takes a different tack to end up in the same place. If we as economists are going to get involved in advising policy makers on cost-benefit calculations, we should keep in mind what we are trying to achieve. Somewhere early in his or her education, every economist learns that what people care about, and what we should help policy makers try to maximize, is not income but utility. In the current context, utility would be a function not only of consumption (which is closely linked to income), but of not suffering with or dying from disease as well as not watching one’s loved ones suffer and die. Putting a price on life and good health strikes many non-economists as distasteful. But to an economist, it is not putting a price on these things that is distasteful. Economists have developed a body of technique for measuring how much people value life and health. Preliminary calculations suggest that the utility benefits of health improvements are enormous. In particular, the direct utility benefits of health improvements far outweigh any indirect utility benefits that may flow through the channel of better health raising income. Thus any positive effects of health improvements on economic growth are going to be of very marginal import in cost-benefit calculations regarding health improvements.
