Research
Job Market Paper
"Irrigation vs education: The long run effects of opium cultivation in British India"
The spatial allocation of public investments can have long term consequences. This paper analyses the effects of opium production in British India on colonial government spending on physical and human capital, and evaluates the relative persistence of these investments over time. My empirical strategy exploits administrative boundaries that demarcated opium-growing areas and fine-grained local variation in environmental suitability for poppy cultivation. In opium-growing districts, suitability for poppies was associated with higher colonial-era spending on irrigation and the rural police but lower investment in schools and healthcare. A century after the end of the opium trade, there are no significant effects on irrigation or police presence. However, former opium-cultivating villages have lower literacy, fewer schools and fewer health centres today. These results suggest that the physical capital investments favoured by colonial officials in opium-growing districts had less persistent consequences than the resultant underprovision of human capital. Evidence that past opium cultivation is associated with lower turnout and more votes cast for criminal politicians, suggests that differences in human capital were relatively more persistent because they affected the political equilibrium that emerged after independence.
Publications
"Building connections: Political corruption and road construction in India", joint with Jacob Shapiro and Oliver Vanden Eynde, Journal of Development Economics (2018), Vol. 131.
Politically-driven corruption is a pervasive challenge for development, but evidence of its welfare effects are scarce. Using data from a major rural road construction programme in India we document political influence in a setting where politicians have no official role in contracting decisions. Exploiting close elections to identify the causal effect of coming to power, we show that the share of contractors whose name matches that of the winning politician increases by 83% (from 4% to 7%) in the term after a close election compared to the term before. Regression discontinuity estimates at the road level show that political interference raises the cost of road construction and increases the likelihood that roads go missing.
Press coverage: VoxDev, Ideas for India, The Telegraph (India), The Indian Express, The Deccan Herald, IndiaSpend
2019 update: updated data and errata
Work in Progress
"Incumbents, minorities, and voter purges: Evidence from 120 million voters' registrations in India"
Democracies around the world routinely carry out purges of the electoral roll to remove voters who are ineligible, have died, or have moved jurisdiction. This process can create an opportunity to alter the composition of the electorate by deleting legitimate registrations. In this paper, I test for such manipulation by constructing an individual-level panel dataset of over 120 million voters in the Indian state of Uttar Pradesh. I show that voters with Muslim names face a relatively lower probability of deletion from the electoral roll in constituencies where the incumbent politician is a Muslim and a higher probability in constituencies where the incumbent is a member of a party that receives little Muslim support (the Bharatiya Janata Party). By applying two separate identification strategies - that exploit (i) close elections and (ii) the re-delimitation of constituency boundaries - I am able to analyse the effects of Muslim and BJP incumbents separately and to identify the conditions under which biased purges take place. The results are driven by constituencies where both Muslim and BJP candidates are competitive. These are the settings where the incentive to disenfranchise minority voters may be large and where their political preferences are most predictable.
“Bidding for roads”, joint with Jacob Shapiro and Oliver Vanden Eynde
In many countries, public procurement is both a large component of government expenditure and a significant source of corruption. For this project we have collected detailed, bid-level data on tenders for over 20,000 road construction projects in 6 states as part of India’s major rural-road construction scheme. We match these data on the technical and financial evaluation of bids to separate administrative records on completed roads. We aim to analyse how the institutional design of the procurement process affects the levels of competition and corruption in the allocation of contracts, and how these relate to the quality of infrastructure provided.
“Justice for all? Evaluating the impact of legal aid clinics in India”, joint with Nitin Kumar Bharti
Individuals’ experience of the justice system can depend critically on whether they are able to afford legal representation. In India, 70% of people in prison have not been convicted but are awaiting trial. Pre-trial detention imposes significant burdens on defendants and their families, and around half of detainees are ultimately acquitted. We evaluate a policy to establish legal aid clinics in all prisons which has been gradually implemented since 2011. We use detailed data on case outcomes in order to understand whether access to legal aid helps defendants navigate the judicial process and reduces the problem of undertrial imprisonment.