UCLA

In India, politicians facing criminal charges are routinely elected at higher rates. In this dissertation, I investigate three primary questions to better understand criminal politicians’ electoral success and performance in office: 1) Do criminal politicians deliver superior access to social welfare programs relative to clean politicians? 2) Do criminal politicians target benefits to co-partisans at higher rates than clean politicians? 3) Do voters reward criminal politicians for delivering more constituency service than clean politicians? On the one hand, powerful dons may be less responsive to voters’ needs, banking on clout to keep voters in line. On the other hand, previous literature and my fieldwork suggest a more Machiavellian strategy, where criminal politicians use both violence and deep pockets to distribute resources to voters.

I present two key arguments to explain criminal politicians’ distributive advantages. First, I contend that criminal politicians core assets of money, muscle and networks make them particularly suited to both deliver more state benefits and target co-partisans. Second, I identify a trade-off that candidates face between accruing enough capital to fund campaigns and remaining rooted in the constituency to provide personalized service to voters. I argue that criminals’ muscle-power allows them to sidestep this trade-off and optimize on both dimensions. Muscle enables criminals to establish lucrative protection rackets in their home constituencies. In effect, protection rackets turn muscle into money. To protect this money, criminals invest in networks for delivering resources to voters. Constituent service networks help criminal politicians maintain political power, which proves useful for protecting their illegal enterprises.

To measure criminal politicians’ in-office performance, I focus on how India’s state legislators influence the delivery of the world’s largest public works program, India’s National Rural Employment Guarantee Scheme (NREGS). Specifically, to determine if criminal politicians translate their assets of money, muscle and networks into superior social welfare delivery, I construct and combine three original datasets. First, to measure criminality, I scraped self-disclosed affidavits listing 87,000 candidates’ criminal charges. The dataset details the criminal histories, wealth, and electoral results of all state legislative candidates in India between 2003 and 2017 (N = 87,000). To measure criminal politicians’ benefit distribution, I combine the candidate dataset with original data on the geo-locations of over 20 million NREGS local public works projects. Finally, to determine if criminal politicians are more likely to target resources to co-partisans, I map the geotagged NREGS projects to over 400,000 polling stations. Methodologically, I use causal inference and machine learning techniques to analyze this data and strengthen the validity of my estimates.

Overall, I find that criminal politicians deliver more NREGS benefits in safe seats, though not necessarily in competitive constituencies. Second, I find suggestive evidence that criminal politicians target welfare benefits to co-partisans at higher rates relative to clean politicians. By remaining embedded in the constituency, I argue that criminals are better positioned to identify, and then meet, supporters needs. Finally, and perhaps unsurprisingly, I find criminals’ core advantage derives from their capacity for violence. Both qualitative and quantitative evidence speak to criminal muscle as a necessary input for improved constituency service and benefit delivery. Empirically, I find that criminals with violent charges are associated with increased NREGS delivery. Whereas, non-violent criminals are not.