A nice article in the Journal of Politics by John Gerring, Strom Thacker and Rodrigo Alfaro reminds us of the value of looking at phenomena over time to gauge their effects. In the article “Democracy and Human Development” they test the controversial proposition that democratic states produce beneficial social outcomes for its citizens. Using infant mortality as a key measure and a 21 point scale that looks at features like checks and balances, the selection of an executive, etc., they find that year to year democratization matters little, but over time democratization is significantly associated with low infant mortality. Here’s a key passage:

Contrary to much recent work, this article argues that there is no strong or robust relationship between a country’s current regime type and its subsequent human development, as measured by infant mortality rates. In this respect, we agree with recent critiques of the received view (Gauri and Khaleghian 2002; McGuire 2004; Ross 2006; Shandra et al. 2004). However, we argue that a robust causal relationship does appear if democracy is considered as a long-run, historical phenomenon.

I like this research because it points to the limits of “snapshot” cross sectional data analysis. If we look at democratization at a point in time, its effects are bound to be constrained by what preceded it. However, if looked at more completely across time, it’s cumulative effect on a society is more likely to come into focus.

CITATION: