Colonial Institutions, Geography, or Post-Colonial Policy? Evidence on Long-Term Comparative Development from a Natural Experiment in South Asia

Abstract

The ‘divergence debates’ revolve around the relative importance of two exogenously predetermined factors— colonial institutions and geography— in shaping the long-term development prospects of different countries. As a part of the same research agenda, albeit a focus on economic divergence between peripheral economies, this paper enquires about the extent to which comparative development is shaped not only by the two inherited factors, but a third, endogenous channel: economic policies pursued after independence. I propose that a ‘twin study,’ modelled within a Difference-in-Difference experimental design, with evidence from natural experiments of partitioned economies provides an ideal setting, allowing a decomposition of the impact of history/geography (‘nature’) from policy (‘nurture’) via the aid of a counterfactual. I present evidence from one South Asian case, the agrarian Punjab region partitioned between India and Pakistan in 1947. The two sub-regions have been exposed to very different policy environments, despite their historical similarities. To examine the impact of these policy differences on agricultural performance, I constructed a district-wise dataset on colonial land-revenue institutions, geography, and yields of food crops from 1900-2016, and setup a pre- and post-treatment juncture at the time of partition in 1947, with one state acting as ‘control’ for the other. The results indicate a strong policy effect beyond history/geography: given identical initial conditions, if a Pakistani district were assigned to India in 1947, it would enjoy a 53% and 70% higher yield of wheat and rice, respectively. The comparative analysis reveals that the Indian land reform, combined with public investments in the provision of cheap credit/inputs, created optimal conditions for private investments to flourish during the Green Revolution, leading over time to significantly better development indicators and lower wealth inequality, in comparison with Pakistani Punjab which chose a fairly laissez-faire approach to agrarian policy.