| I study the effect of historic land inequality and presence of landed elites on electoral competition and public goods provision in Pakistan. First I use household data to document that landowners can make transfers to sharecropping tenants at a relatively low cost, thereby gaining tenants' electoral support and swaying public policy in their favor. Specifically, I show that when an election is introduced after a military regime, politician landlords offer concessions on input costs to their tenants. Then I postulate that in the presence of moral hazard, technical change in agriculture increases the cost of sharecropping tenancy, attenuating landlords' electoral advantage. To test this, I exploit the introduction of high yielding variety seeds for plausibly exogenous variation in productivity and proxy for initial landlord dominance with the colonial prevalence of large land grants. Consistent with my hypothesis, I find that increased productivity lowers the likelihood that landlords are elected to office, which in turn, improves electoral competition and shifts the composition of public goods away from those preferred by landlords. |
