Imposing institutions: Evidence from cash transfer reform in Pakistan

Institutions are recognized as critical for development, but there is limited evidence on whether policies can improve them. In this paper, we quantify the impact of an outside agency imposing on a developing country a new policy designed to improve a key government function: the allocation of public spending. Donors imposed a new system for selecting recipients of government transfers in Pakistan as a condition of financial aid. Before this reform, households in winning politicians' villages were 200-400% more likely to receive cash transfers than those in rivals' villages. The reform reduced favoritism for the best connected households: politicians' clan members in their villages. Some connected households continued to receive more transfers, likely because politicians assisted them in overcoming administrative hurdles. However, the reform improved targeting substantially in politicians' own villages and province-wide. As a result of this imposed policy, the public legitimacy of government transfer programs increased by about 40%.