My research covers three related substantive areas:

  1. Policing in Divided Societies: How does the demographic makeup and organization of the police affect prospects for peace in divided societies? I explore the role of group-based representation on minority and dominant group attitudes towards one another, the quality of police service provision, and prospects for peace. I focus primarily on Iraq and Israel, with additional coauthored research in the United States.
  2. Community Policing: A rich line of research explores the effectiveness of “community policing” in the United States and other developed countries. My coauthors and I ask whether these same strategies might improve service delivery in developing states. This research is centered around my role in EGAP’s Metaketa Initiative and draws data primarily from a series of field experiments with the Philippine National Police.
  3. Counter-Terrorism: How do political institutions, incentives, and interactions shape the perpetration of terrorism and the implementation of counterterrorism? I draw on micro-level data from Israel.

I also work on methodology surrounding the use of surveys for measuring sensitive topics, both for academic audiences and for practitioners of PVE/CVE.

Policing in Divided Societies



Working Papers:

Societies plagued by group-based divisions face a key challenge: how should the police include individuals from each societal group? On the one hand, including individuals from previously marginalized groups in key state institutions might increase attachment to the state and reduce feelings of insecurity, making them less likely to turn to violence. On the other hand, these individuals may use their new-found power to turn their weapons on the state and renew hostilities.

Community Policing


Working Papers:

  • "Fire Alarms for Police Patrols: Experimental Evidence on Co-Production of Public Safety." With Nico Ravanilla and Dotan Haim. (Draft available)
  • “Evaluating Community Policing in the Philippines: Communication, Trust, and Service Provision.” With Dotan Haim and Nico Ravanilla. (Draft available)

Community policing proposes that policing will prevent crime most effectively when officers connect with the local community, identify local problems using citizen-provided information, and develop and implement solutions to target those problems.

In conjunction with the EGAP Metaketa Initiative, my coauthors and I ask whether this model can be applied to developing contexts. We study community policing in Sorsogon Province, Philippines, where a long-running anti-government insurgency competes with government law enforcement institutions for authority. We propose that one of the key challenges for the police in the Philippines is a lack of trust which prevents citizens from providing them with useful information about problems in the community.

The heart of our study involves a randomized controlled trial (RCT) with two primary research arms. The first, Community Oriented Policing (COP), addresses the problem of trust and information flows between citizens and police. As part of our research, the Sorsogon PNP implemented randomized the roll-out of a program which increased informal contacts between officers and citizens across 298 villages. The second treatment arm created a task forces for a random subset of villages in the province. The PNP assigned pairs of officers to a specific village and tasked them with developing and implementing a solution to a pressing local problem in conjunction with the community. We measured the effects of these experiments through survey and administrative data.

We implemented a number of other projects alongside the main experiment. For example, during One Sorsogon, we studied barriers to citizen information reporting to the police by distributed more than 100,000 stickers publicizing an SMS Police Hotline in a randomly-selected subset of villages. In related work, we focus in on the role of family ties in citizens’ willingness to engage with public safety institutions. We find that while citizens who are personally related to officers are more likely to report information about criminal activity to those officers, officers who are too embedded can elicit backlash from unconnected citizens. The resulting backlash and perceived bias hinders officers’ abilities to mediate disputes between citizens.



Working Papers:

How do political institutions, incentives, and interactions affect counterterrorism policies? In one paper, I show that reelection-seeking politicians’ electoral incentives cause them to provide counterterrorism policies that are suboptimal for peace and security. As elections approach politicians move towards ideological extremes, with right-wing politicians implementing more hawkish policies and left-wing politicians implementing more dovish ones. These politicians are not the only leaders to pursue inefficient security strategies, however. In another chapter, I apply Padro i Miquel and Yared’s (2012) theory of principal-agent proxy management to Israel’s relationship with two proxies in its fight against terrorism: The South Lebanon Army from 1975 until 2000, and Hamas from 2007 until present day. Finally, in a working paper with Trevor Bachus, we explore the way that “hardening” potential terror targets by making them more difficult for would-be attackers to reach affects the spatial and temporal distribution of attacks in the absence of solutions to the underlying political conflict.