Corruption, Distortions and Development

Abstract: Essay I: Does increased government transparency make bureaucracies more meritocratic? I study the impact of politician disclosure requirements on the assignment of bureaucrats to public posts. I collect detailed information on bureaucrats' qualifications and construct a novel measure of bureaucratic mismatch—an index measuring the extent to which a bureaucrat is under- or overqualified to perform a specific task. Using the staggered implementation of Indian state elections for identification, I find that information disclosure is associated with a mismatch reduction of 2.5% of a standard deviation. This effect is substantially larger (15–45% of a standard deviation) in posts that are more important for policy implementation and tasks that bureaucrats perceive to be more prestigious. In addition, I find that information disclosure increases returns to bureaucrat skill. Bureaucrats with greater skill are more likely to reach prestigious positions in more transparent states. Taken together, the results show that government transparency promotes meritocracy within public organizations.Essay II (with Roza Khoban): Political distortions are prevalent in many developing countries and can imply substantial productivity losses. Theory is ambiguous as to whether greater openness to trade amplifies or reduces the effects of such distortions. This paper shows that trade liberalization in India decreased the value of firms' political connections, suggesting a reduction in political distortions. First, using variation in firm connections stemming from political turnover, we identify that political connections increase firm performance by 10–20%. Second, we evaluate how the value of political connections changed after India's externally imposed tariff reductions, using a triple-difference and difference-in-discontinuities design. We find that political connections became substantially less valuable when tariffs on input goods were reduced. Our findings imply that access to international markets reduces firms' dependence on political connections to source input goods, thus reducing the distortionary effect of such connections. Our results suggest a new margin for gains from trade in the presence of political distortions through a direct effect of trade liberalization on the prevalence of such distortions.Essay III: Social unrest is a pervasive problem in the developing world. Yet, the causes of social unrest are not fully understood. This paper studies the impact of electing dominant party representatives on social unrest in South Africa. Since the end of apartheid, the African National Congress (ANC) party has held a hegemonic position in South African politics. At the same time, the party has been criticized for poor implementation of basic service delivery and accused of corruption and government malpractice. Combining geo-referenced data on riots and protests with a regression discontinuity design, I find that the prevalence of violent social unrest is substantially lower in ANC-controlled areas. The findings provide a nuanced perspective on the political determinants of social unrest in a dominant party setting.