Accepted Manuscript

The focus of policy reform in developing countries has moved from getting prices right to getting institutions right, and accordingly countries are increasingly being advised to move towards "best-practice" institutions. This paper argues that appropriate institutions for developing countries are instead "second-best" institutions --those that take into account context-specific market and government failures that cannot be removed in short order. Such institutions will often diverge greatly from best practice. The argument is illustrated using examples from four areas: contract enforcement, entrepreneurship, trade openness, and macroeconomic stability.

Few would doubt the proposition that political institutions matter for economic development. Yet we lack robust generalizations and systematic evidence on how exactly they do so. In this short paper, I draw attention to a regularity in the cross-national data that has received little attention to date: participatory political regimes are associated with significantly lower levels of aggregate economic instability. After presenting some of the evidence in the next section, I speculate that the reason has to do with the propensity of democracy to moderate social conflict and induce compromise. I discuss three distinct arguments as to why this may be the case. I. Some evidence The relationship between democracy and economic growth has been studied extensively. The data tend to show that democracy has no systematic effect on long-run growth rates. The top panel of Figure 1 shows a typical result: the partial correlation between an index of democracy during the 1970s and subsequent economic growth is virtually zero. The relationship between democracy and volatility in economic performance, on the other hand, is negative, statistically significant, and quantitatively large. This is shown in the bottom panel of Figure 1:

There exists a positive correlation between an economy's exposure to international trade and the size of its government. The correlation holds for most measures of government spending, in low- as well as high-income samples, and is robust to the inclusion of a wide range of controls. One explanation is that government spending plays a risk-reducing role in economies exposed to a significant amount of external risk. The paper provides a range of evidence consistent with this hypothesis. In particular, the relationship between openness and government size is strongest when terms-of-trade risk is highest.