Friday, October 26, 2007

Wealth or Health?

Check out this 20 minute presentation by a gentleman named Hans Rosling on "TED Talks." Dr. Rosling is a doctor does research in the medical profession on public health in the "Developing Wolrd" and maintains an analysis tool known as "Gapminder." One interesting contribution he makes is his ability to express trends in data in a very intuitive, visual way.

The econometrician inside of me has about a thousand "buts" for the presentation, but it represents a great start- it does the essential surface "strip mining" that helps us see what things need better testing, and which things can be used as controls. It reminds me of something I learned from Roger Koenker: "If you torture the data enough, it will confess." In Dr. Rosling's case, he sheds quite a lot of light on the nature of health and wealth with very little of the usual waterboarding that economists usually do.

The most intersting part of it is that growth does not seem to "trickle down" to aspects of welfare such as life expectancy, infant and child mortality, etc. on average. Conversely, it seems to be the case that things like public health, evolution in institutions and society, etc. act as a precursor to growth, something about which Daron Acemoglu has been on the economics profession's case for some time now. And, it shouldn't be surprising-- the industrial revolution in Britain followed a similar pattern, as I've mentioned here before.

Yet, Dr. Roslings conclusions are similar to those that economists, even the most neoclassically trained among us, would readily agree with. First, that things like trade and markets are helping a number of countries pull themselves out of despair and poverty; second, that political institutions and good government play a critical role in ensuring that the benefits of growth are distributed in a way that respects the median citizen; third, that these factors HAVE contributed to a considerable normalization and flattening in the distribution of world income over the last half-century.

No comments:

Post a Comment