“Political Violence Cycles: Electoral Incentives and the Provision of Counterterrorism“ (2017), Comparative Political Studies
How do electoral incentives affect the counterterrorism policies chosen by reelection-seeking incumbents? This article tests the argument that governments alter their choice of security strategies as elections approach to signal competence to potential voters. Which policy they select should depend on the intended audience of the signal. Governments seeking support from their partisan base should select different policies than those courting the support of moderates. Using data on Israel-West Bank checkpoint closures and casualties in the Palestinian territories between 2000 and 2013, I find evidence that Israeli governments manipulate security strategies in the run-up to elections in a manner consistent with an attempt to attract support from core voters. As elections approach, left governments become more dovish on security, while right governments become more hawkish. The relationship between partisanship and policy choice raises concerns that electoral incentives may induce democratic governments to select inefficient or suboptimal security strategies around election time.
Race and Representative Bureaucracy in American Policing (2017), Palgrave Macmillan. [With Brandy Kennedy, Adam Butz, and Nazita Lajevardi]
What are the causes and consequences of racial representation in American policing? We create a new dataset of police officer demographics in thousands of departments across the US between 1993 and 2013 to ask two questions. First, why does the degree of racial representation vary so much from one department to the next? Second, what are the systematic effects of variation in racial representation in the police? We show that variation in the extent of representation from one jurisdiction to the next is associated with institutional factors like residency requirements and union presence, as well as political factors like the race of the city’s mayor. In turn, jurisdictions in which the police are more representative of the population exhibit fewer citizen complaints of excessive force and more citizen-friendly procedures for dealing with complaints that do occur. We argue that while many of the processes which we identify are themselves the result of historical pathways, identifying causes and consequences of representation more narrowly provides a clear path forward for policymakers seeking to increase bureaucratic representation and improve the quality of policing services.
Working Papers and Chapters:
How does sectarian integration of the police affect intergroup conflict? The police shape the citizen-state relationship by enforcing laws and distributing government services. I argue that in divided societies, the inclusiveness of rank-and-file officers affects individual-level incentives to use violence against the government. I test this argument using a survey with an embedded experiment of 800 Baghdad residents. I find that providing members of the Sunni minority with an informational prime that the police are integrated reduces support for anti-government violence. I then test the relationship between police integration and several mechanisms which existing research identifies as motivators of conflict. Sunnis who perceive the police as more integrated are less fearful of future repression, hold fewer grievances over government service provision, and feel less excluded from desirable employment opportunities.
Countries characterized by deep, politically-salient ethnic or religious divisions face a unique challenge: How can a single set of institutions effectively govern a fragmented population? While most research on institutions in divided societies focuses on those that select leaders and make policies, I argue that the design of institutions responsible for policy-implementation is also critical for the quality of governance. I test whether integrating one such institution, the police, affects the provision of public safety. Using panel data on police officer demographics in Israel, I show that increases in local police inclusiveness are associated with decreases in crime victimization. I also find evidence of a threshold effect, with inclusion decreasing the likelihood of crime victimization only at higher levels. I then provide evidence of one potential mechanism linking integration with crime prevention: citizens who perceive the police as integrated are more willing to provide them with information about problems in the community. These findings highlight the importance of the design of policy-implementing institutions in determining the quality of governance in divided societies.
“The Extremes of Preference Alignment: Israel’s Management of Proxy Agents in Lebanon and Gaza” [book chapter – volume under review] Draft available upon request
Draws on the theory of proxy-agent management presented in Padro i Miquel and Yared (2012). Provides evidence that Israel has engaged proxy agents to prevent attacks from neighboring territories on at least two occasions: southern Lebanon between 1975 and 2000, and Gaza beginning in 2007 and continuing through today.
“Family Matters: Social Embeddedness and Police-Community Relations in the Philippines” [with Dotan Haim and Michael Davidson]
[Project Overview] How does officer embeddedness affect the ability or willingness of the police to carry out their duties? On one hand, greater ties with the community they serve may make police officers more effective by providing access to information. On the other hand, these same ties may make officers more likely to engage in selective enforcement or corruption. We test the effects of officer embeddedness in the Philippines using a novel dataset of family networks and an original survey on citizen experiences with the police.
“Police-Community Engagement, Information Communication Technology, and Citizen Trust in State Institutions” [with Dotan Haim and Nico Ravanilla]
Part of the EGAP Metaketa IV: Community Policing initiative.
We test the effects of a large scale police engagement program in the Philippines using a randomized controlled trial. Due to an ongoing insurgency, nation-wide scandals of police abuse, and perceptions of neglect, trust in state law enforcement agencies in the Philippines is low. We evaluate the effectiveness of a community engagement program designed to improve trust and legitimacy by randomizing the rollout of the program. We implement two additional components aimed at improving the efficacy of the engagement program, an SMS hotline that provides citizens with a low-cost method of providing information to the police, and an officer incentive program aimed at improving program implementation.