You can find my full CV here.
Works in Progress
Electoral Accountability and State Violence: The Political legacy of the Marikana massacre (with Daniel de Kadt and Melissa Sands) pdf
Abstract: Democratic states often wield coercive force against ordinary people, yet little is known about the electoral consequences of such violence in the communities directly effected. Using geo-coded polling station data and a difference-in-differences design that leverages temporal and spatial variation, we analyze how incumbent electoral support was affected by South Africa's Marikana massacre, one of the most high-profile examples of state violence in contemporary democratic Africa. We find evidence that in the communities directly affected by the massacre the incumbent party was dramatically punished at the polls. Using geo-referenced survey data to investigate the sources of this change, we find that our results are almost exclusively driven by voters switching to an opposition party that formed in the wake of the massacre, rather than (de-)mobilization. We also find, contrary to some academic and popular narratives, no evidence of attitudinal shifts around institutional trust or views of democracy and participation.
Bureaucratic embeddedness and Inequality in Local Public Goods Provision (Dissertation Paper)
Abstract: Scholars of political economy argue that politicians focus the provision of public goods to their supporters. Others argue that citizens drive distribution of resources, and that those who are more engaged solve collective action problems, and can provide public goods for themselves. Yet neither politicians nor citizens/groups actually provide public goods themselves; instead it is bureaucrats that implement actual service provision. I argue that characteristics of bureaucrats predict who they provide goods and services to. Specifically I focus on `bureaucratic embeddedness', how much of a bureaucracy comes from a group. Like politicians and citizens, bureaucrats are interested in distributing goods to groups they belong to. I posit that bureaucracies with a high proportion of bureaucrats from one group are more likely to provide goods to their group. I test my expectations in the context of the Nigerian Federal civil Service. I create a dataset of federal civil servants from 1905 to 2015 and collect data on public goods projects for a more contemporary period. I find that historical bureaucratic embeddedness does not matter for the allocation of projects but for places with representation in the bureaucracy, completion rates of projects are higher.
Historical State Capacity and Development: Evidence from the Green Revolution (with Aditya Dasgupta)
Abstract: One of the major economic transformations of the twentieth century was the green revolution, a technological revolution in agricultural productivity owing to the spread of newly developed cultivars known as high-yielding variety (HYV) crops. But why did some countries, notably those in Asia, experience the green revolution but not others, like most countries in Africa? In this paper, we argue that historically transmitted state capacity played a critical role. Successful implementation of the green revolution depended upon the pre-existence of powerful agricultural bureaucracies that developed under pre-colonial and colonial regimes that relied upon the taxation of agriculture. The argument is tested with a comparative analysis of HYV crop adoption across countries as well as a spatial analysis of the impact of pre-colonial state boundaries in Africa on remote-sensing measures of agricultural modernization based on satellite imagery. The findings highlight the central role that historically transmitted state capacity plays in long-term technological change and economic development.
Works in Progress
Electoral Accountability and State Violence: The Political legacy of the Marikana massacre (with Daniel de Kadt and Melissa Sands) pdf
Abstract: Democratic states often wield coercive force against ordinary people, yet little is known about the electoral consequences of such violence in the communities directly effected. Using geo-coded polling station data and a difference-in-differences design that leverages temporal and spatial variation, we analyze how incumbent electoral support was affected by South Africa's Marikana massacre, one of the most high-profile examples of state violence in contemporary democratic Africa. We find evidence that in the communities directly affected by the massacre the incumbent party was dramatically punished at the polls. Using geo-referenced survey data to investigate the sources of this change, we find that our results are almost exclusively driven by voters switching to an opposition party that formed in the wake of the massacre, rather than (de-)mobilization. We also find, contrary to some academic and popular narratives, no evidence of attitudinal shifts around institutional trust or views of democracy and participation.
Bureaucratic embeddedness and Inequality in Local Public Goods Provision (Dissertation Paper)
Abstract: Scholars of political economy argue that politicians focus the provision of public goods to their supporters. Others argue that citizens drive distribution of resources, and that those who are more engaged solve collective action problems, and can provide public goods for themselves. Yet neither politicians nor citizens/groups actually provide public goods themselves; instead it is bureaucrats that implement actual service provision. I argue that characteristics of bureaucrats predict who they provide goods and services to. Specifically I focus on `bureaucratic embeddedness', how much of a bureaucracy comes from a group. Like politicians and citizens, bureaucrats are interested in distributing goods to groups they belong to. I posit that bureaucracies with a high proportion of bureaucrats from one group are more likely to provide goods to their group. I test my expectations in the context of the Nigerian Federal civil Service. I create a dataset of federal civil servants from 1905 to 2015 and collect data on public goods projects for a more contemporary period. I find that historical bureaucratic embeddedness does not matter for the allocation of projects but for places with representation in the bureaucracy, completion rates of projects are higher.
Historical State Capacity and Development: Evidence from the Green Revolution (with Aditya Dasgupta)
Abstract: One of the major economic transformations of the twentieth century was the green revolution, a technological revolution in agricultural productivity owing to the spread of newly developed cultivars known as high-yielding variety (HYV) crops. But why did some countries, notably those in Asia, experience the green revolution but not others, like most countries in Africa? In this paper, we argue that historically transmitted state capacity played a critical role. Successful implementation of the green revolution depended upon the pre-existence of powerful agricultural bureaucracies that developed under pre-colonial and colonial regimes that relied upon the taxation of agriculture. The argument is tested with a comparative analysis of HYV crop adoption across countries as well as a spatial analysis of the impact of pre-colonial state boundaries in Africa on remote-sensing measures of agricultural modernization based on satellite imagery. The findings highlight the central role that historically transmitted state capacity plays in long-term technological change and economic development.