For written analysis of completed projects, see my Writing page.


+ Building Trust and Improving Police Effectiveness: A Field Experiment in Urban Liberia

Monrovia, Liberia is plagued by rampant crime. According to the 2015 Afrobarometer survey, 65 percent of the city’s residents reported that they or someone they knew was a victim of theft in the past year, while 35 percent reported that they or someone they knew was physically assaulted. These figures are considerably higher than those for urban, sub-Saharan Africa overall, and place Monrovia among sub-Saharan Africa's most crime-ridden places to live. This project aims to address a key factor believed to contribute urban Liberia's high rate of crime: the reluctance of citizens to cooperate with the police through activities such as crime reporting, information sharing, and evidence provision. In collaboration with the Liberian National Police, the project will test whether community policing can build confidence in the police, increase crime reporting, improve information sharing, and ultimately reduce the incidence of crime.

+ How Confirmation Bias Undermines Compliance and Impedes Statebuilding: Evidence from the Ebola Crisis in Liberia

This paper proposes that confirmation bias – the tendency for individuals to let their prior beliefs and affections inform their judgments – can undermine efforts to build trust and engender compliance in fragile, illegitimate states. In these settings, citizens who distrust their government and exhibit confirmation bias will be reluctant to credit their government for good performance and more likely to punish it for bad performance; they will also be resistant to the government’s efforts to engender voluntary compliance through persuasion. Together, these tendencies reduce the likelihood of virtuous cycles of trust in government, citizen compliance, and government performance capacity, and at the same time increase the risk of vicious cycles of distrust, non-compliance, and state decline.

I provide empirical support for this argument using data collected during the height of the 2014-2015 Ebola crisis in Monrovia, Liberia. Using voting for the opposition in the 2011 Presidential eleciton as a proxy for pre-crisis distrust of government, I show that distrustful individuals were more likely to believe in misinformation about Ebola at the start of the crisis and less likely to offer voluntary compliance with preventative measures. They were also more likely to have had friends or family become infected with Ebola at the height of the crisis. Conditional on knowing someone with Ebola, they were more likely to also know someone with Ebola in a subsequent month, suggesting that the reproductive rate of Ebola was higher within distrustful social networks. None of these patterns, however, are observed after the height of the crisis, when Ebola’s severity was beyond dispute.

I argue that these patterns are unlikely to reflect differences in access to information or healthcare. Instead, I suggest that citizens' mistrust of government conditioned how they responded to information during the crisis. In particular, mistrust made citizens more likely to believe in misinformation and more resistant to government efforts at persuasion, resulting in widespread non-compliance and the spread of Ebola.

I next consider how mistrust conditioned citizens’ response to a) hardships precipitated by the epidemic, and b) government-led community outreach to combat the epidemic, showing that distrustful citizens were more likely to downgrade their evaluations of government in response to hardships and less likely to upward revise their evaluations in response to government-led outreach.


+ Establishing the Rule of Law in Weak and War-Torn States: Evidence from a Field Experiment with the Liberian National Police

(with Rob Blair and Sabrina Karim)

Can exposure to a newly-reformed security sector increase citizens’ trust in the police? Can it mitigate local violence and reduce reliance on extrajudicial mechanisms for dispute resolution? We address these questions through an experimental study of the “Confidence Patrols” community policing program in Liberia. Designed to build trust in the police and raise awareness about institutional reforms in the justice and security sectors, the program involved repeated visits from teams of 10-15 officers of the Liberian National Police (LNP) over a period of 14 months. We find that the program increased knowledge of the police and Liberian law, reduced the incidence of some types of crime (notably assault and domestic violence), and increased reporting of crimes to the LNP. We also find, however, that the program failed to enhance trust in the LNP or the government more generally, and may have reduced victims’ satisfaction with the LNP’s handling of reported crimes.

+ Patterns of Trust and Cooperation in the Fight Against Ebola

(with Lily Tsai and Rob Blair)

[Academic paper in progress]

In collaboration with the Liberian government and the MIT Governance Lab, we conducted a survey of over 1,500 Monrovia residents in December 2014 to help target Ebola recovery and response efforts, and to understand the social and political factors that contributed to the epidemic.

We first focus on understanding the correlates of compliance with control policies that have been essential to halting Ebola’s spread. Our results reveal a robust link between trust in government and compliance: Individuals who are less trusting of government are less likely to support control policies and less likely to adopt preventative measures. These differences persist even after accounting for key demographic and socio-political characteristics and knowledge about EVD.

Our analysis also shows that experiences of hardship and trauma during the epidemic are negatively associated with trust in government and compliance with control measures. These traumatic experiences include losing a job, forgoing healthcare for serious illness, knowing Ebola victims, and witnessing dead bodies in the streets. A follow-up survey in March 2015 shows that these effects do not persist. This may be the result of emotive mechanisms operative during conflict that reduce trust and cooperation.

We also identify an important and underreported success story in the effort to win citizens’ confidence and cooperation in the fight against Ebola. Throughout the crisis, government and NGO workers conducted community-level outreach to build trust and increase compliance with control measures. We find that government and NGO outreach efforts are associated with greater trust in government, support for control policies, and uptake of preventative measures. Follow-up qualitative research suggests that outreach was especially effective when it deliberately incorporated pre-existing community networks and institutions. Outreach involving government was considerably more effective than NGO outreach, especially in building trust in government.

+ Local integration as an alternative to camp-based management: Evidence from the 2010-2012 Ivoirian refugee crisis

In light of the growing incidence of protracted refugee crises, policymakers have increasingly called for local integration of refugees as an alternative to camp-based management. Yet there exists a large gap between the growth of local integration policies and the availability of research to inform these policies. Several questions stand out as particularly important. First, what is the effect of local integration on refugee welfare relative to camp-based management? Second, what are the social, economic, and political effects of local intergration on host communities? And finally, what types of NGO programming are most effective in supporting host communities?

To address this gap, we collaborated with the Norwegian Refugee Council in Liberia to conduct a panel survey of 64 communities in the Liberia-Cote d'Ivoire border region. Using matching methods and difference-in-differences analysis, we evaluate a) the effect of a government policy to concentrate refugees in particular villages on refugee and host community outcomes, and b) the effect of NGO programming on the recovery of host communities after the crisis. This project resulted in a 2014 policy report produced for the the NRC that addressed the aforementioned questions. The data has also been used to explore the link between past experiences of violence and willingness to host refugees (link). Further write-up and analysis for academic publication is ongoing.