The dual divergence: growth successes and collapses in the developing world since 1980

Cross-country econometrics has been the preferred tool of empirical growth analysis over the past 20 years. This literature has tended to focus on long time periods, ignoring the instability and volatility of growth patterns in developing countries. Because shocks play a central role in explaining variations in growth patterns in the developing world (Easterly et al. 1993), the meaning of the statistical coefficient estimated for long periods of time is unclear. In this sense, the use of panel data econometrics for shorter periods may be more appropriate, although, for many reasons, not ideal (Pritchett, 2000). Another route is to examine factors that determine growth spurts (accelerations) and collapses. Country-specific historical analysis is a third route (see, for example, Rodrik, 2003, among others), but the comparability of such analyses is a significant limitation in this regard.(This abstract was borrowed from another version of this item.)