Growth outside the stable path: Lessons from the European reconstruction

This paper exploits a natural experiment, the large destruction of capital in continental Europe during World War II, to characterize the transitional dynamics of an economy that begins with a capital stock below its steady state level. We use these regularities as a benchmark to discriminate among competing growth specifications. A model that combines non-separabilities in preferences with a technology that restricts the degree of substitutability between inputs outperforms the widely used AK and Cobb–Douglas specifications with time-separable preferences. Our results suggest that policy evaluations based in growth models that overlook non-separabilities in preferences or impose strong restrictions on the technological structure might be grossly misleading.