Public goods often are poorly provided in states characterized by violence and weak institutions. In several papers, I use field or lab-in-the-field experiments to examine how ethnic identity and corruption affect the provision of public goods in Afghanistan.

How does individuals’ willingness to aid non-coethnics change when confronted by their physical presence?

Kabul hotel lobby (2014)

We investigate the role of other-regarding preferences in ethnic bias in an experiment in Kabul, Afghanistan. Drawing on interdisciplinary behavioral research, we argue that in-person exposure to non-coethnics affects ethnic identification differently than descriptive cues of non-coethnics. Non-Pashtun minorities give equally to a hospital serving the Pashtun majority as a hospital benefitting non-Pashtuns when Pashtuns are not physically present, but decrease giving to the Pashtun hospital by 25% when they waited alongside Pashtuns for the experiment. This decrease is driven by subjects who were randomly assigned to wait longer with Pashtuns, but only if they cannot speak Pashto, suggesting that interethnic contact highlights differences in descent-based attributes and increases in-group identification. Our findings provide evidence on the role of the psychological mechanism leading to in-group bias in casual, everyday interactions in a post-conflict society, with implications for studying the nature of ethnic bias in political and economic behavior. (Condra et al. 2017a)

How should we understand the power of Muslim clerics to motivate collective action?

Mosque in Kabul (2012)

We theoretically disentangle the psychological influence of a Muslim cleric’s power to mobilize for collective action through contributions to a public good (a hospital). The same cleric requests contributions when dressed as a civilian and as a cleric. In Civilian condition, 50% contribute and 17% make large contributions; in Cleric condition, 83% contribute but average giving did not increase as most gave the smallest possible amount. Inclusion of a recitation of Qur’anic verses in the Cleric condition maintains the 82% contribution rate while increasing large contributions to 30%, doubling average contributions. Formal education and subjective feeling of poverty appear to drive the opposing effects of cleric and scripture. These results suggest that the power to activate spiritual channels lies in the scripture, not with the human wielding religious authority, who instead appears to induce minimal compliance with Islamic norms of charitable giving. (Condra et al. 2017b)

How can governments and NGOs enable more effective development spending in poorly governed and dangerous places?
Corruption and rent capture are pervasive in many developing countries, necessitating audits and monitoring of project spending. Yet international aid increasing flows to violent, unstable spaces where foreign donors cannot access projects. We experimentally evaluated a program that trained Afghans to monitor the quality of rural road construction adjoining their villages, funded by the World Bank. Our experiment spans five Afghan provinces and four years of road quality measurement using our own trained technical teams. We find that trained monitors cause dramatic improvements in road quality, producing roads that are better able to endure difficult Afghan winters. Our two-level randomization design shows these effects are not concentrated near trained villages, but spill over to the entire road. Four years after the training, treatment effects have attenuated, potentially because the monitoring program was removed. Finally, interviews with trained monitors suggest the program was effective due to the combination of the technical training and the newly created channels of accountability, helping to reconcile mixed findings in past literature where programs often provided only one of these. All told, the experiment shows that scholars can measure and experimentally evaluate community-based monitoring programs in places like rural Afghanistan, a relatively inexpensive and safe program of community monitors improved road quality through advocacy and increased information flow. (Berman et al. 2017)