January 08, 2005

Do Institutions Cause Growth?

Readers of this site—all three of you!—know that I'm a big fan over the liberalization/democratization debate. The discussion boils down to one question, really: is it better for a developing country to become a democracy first and then work on economic development, or should the economic stuff come first so that people can become rich enough and educated enough to have a democracy?

One way this plays out is when the "democracy first!" folks say that you can't have good economic growth without things like transparency in your financial institutions, the rule of law, low corruption, and that these things all require democracy. (More on this here.) But that argument always struck me as odd. After all, Singapore's not really a democracy, but its government doesn't go around seizing people's stuff and monkeying with the law. Same with the East Asian "tigers" during their boom years in the 1950s. Not democracies. At a glance, democracy is probably more likely to lead to good institutions and the rule of law, but so can a benevolent dictatorship.

Anyways, I just found this NBER paper (PDF) by Glaeser, Shleifer, La Porta, and Lopez-de-Silanes that goes into this critique in more depth. After doing the analyses, they find that the good democratic stuff everyone loves—constitutional limits on sovereign power, judicial independence, etc.—have no predictive power for the growth of per capita income. "[I]nstitutions," the authors claim, "have only a second order effect on economic performance." So what does have an effect? Human capital—stuff like primary school enrollment, which shapes both institutional and productive capacities of a society." Most importantly, they're all very skeptical of pushing countries with low levels of human capital into democracy.

Let me just add that the work of Shleifer, La Porta, and Lopez-de-Silanes on these sorts of subjects has come under heavy fire. Previously, they've argued that a country's economic success depends on good deal on what kind of legal tradition it has—British-style common law or French-style civil law. Nicholas Thompson discussed their work in Legal Affairs, along with some of the criticisms. Still, it's interesting.
-- Brad Plumer 3:37 PM || ||