Kudamatsu, Masayuki
(2007)
Political economy of development: health as a development outcome, micro evidence, and heterogeneity of democracies and autocracies.
PhD thesis, London School of Economics and Political Science.
Abstract
The thesis explores whether and how democratic and autocratic political institutions affect the welfare of people in developing countries. First, we empirically investigate whether democracy improves people's health, by using time-series country-level aggregate statistics. We find that there is a robust cross-sectional correlation between democracy and life expectancy at birth. Country fixed effects estimation, on the other hand, does not yield a statistically significant correlation between the two. This empirical approach, however, does not disentangle the effect of democracy from country-level confounding factors. To overcome this, I empirically examine whether democratization has reduced infant mortality in sub-Saharan Africa in the 1990s, by using micro data on child survival. Mother fixed effects estimation shows that mothers see their infants more likely to survive after democratization than before. This result may suggest that African dictatorships are particularly bad compared to those in other regions. To shed some light on this possibility, we theoretically investigate under what condition autocracy yields good policy outcomes. We show that such a condition is that those enfranchised in autocracy can retain the right of leadership selection after overthrowing a dictator for his bad performance. We also show that such a successful autocracy outperforms a democracy if distributional issues are so important that voters in democracy cannot discipline policy-makers in the general interest policy outcomes. What affects the salience of distributional issues, therefore, needs to be understood. One such factor may be ethnic favoritism by the government, which has rarely been empirically investigated in a systematic way. By using micro data on infant mortality and by exploiting one-time unexpected change in the president's ethnicity in Guinea, I provide evidence on whether the ethnicity of those in power affects infant mortality for each ethnic group under an autocratic rule.
Actions (login required)
 |
Record administration - authorised staff only |
Download statistics
Download statistics