Essays on political economy and governance

This article-based thesis aims to study political economy and governance in developing countries through the lens of crime, policing, and politics. Chapter 2 studies the impact of election timing on crime and the nature of law enforcement in a developing country known for electoral violence. Using administrative crime data from India in the 2008-2019 period, I undertake fixed effects and difference-in-differences analyses, and provide detailed and rigorous causal evidence for change in the nature of crime and the composition of policing effort as elections approach. I show that upcoming elections increase crimes against public order while reducing crimes against individuals, such as those related to bodily offences, property, or women. I suggest that these cycles are driven by changes in police behaviour – as elections approach, the police sharply increase preventive activity and the enforcement of crimes related to public order, and reduce the enforcement of other crimes. These changes are consistent both with the multitasking of bureaucrats as well as political incentives for re-election, and contribute to our understanding of the nature of crime, politics, and policing during elections. One in three elected representatives in the lower house of national and state legislatures in India is criminally accused. Using detailed and granular crime data from India, Chapter 3 estimates the impact of politicians with criminal backgrounds on local crime. On one hand, criminal politicians may be elected not despite of, but because of, their ability to control crime and get things done in an environment with weak rule of law. On the other hand, given their criminal backgrounds and political control over local police, criminal politicians may seek to consolidate their rent-seeking interests during their time in office. I use a close election regression dis- continuity design and find that having a criminal politician in office lowers weekly counts of serious crimes by 23.6%, even though the total number of complaints re- mains unchanged. This reduction in crime seems to be partly driven by reduced recording of crime. Further, this chapter suggests that the specific nature of a politician’s criminal background matters for outcomes. Chapter 4 focuses on low participation of women in politics, and explores the impact of gender-specific polling stations on turnout. Using data from Pakistan, which has one of the lowest rates of female voter turnout in the world, we analyse the impact of creating separate-gender polling stations for males and females on voter turnout. The results show that separate gender polling stations reduce turnout for females by 2 percentage points on average, and for males by 1 percentage point on average. The results remain consistent after controlling for the role of resources, norms and mobility-related costs. This paper indicates that the policy of separate- gender polling stations does not alleviate gender-based voting constraints, and may in fact, worsen women’s access to voting. Together, these three papers improve our understanding of crime during elections, criminal governance, and the participation of women in the electoral process, and contribute towards the literature in the fields of economics and political science.