Delivery Agents, Social Ties and Group Identity: Theory and Evidence from a Two-Layered Field Experiment


Imran Rasul (University College London)
Wednesday 27 September 2017 at 3:30 - 5:00 pm (Hong Kong time, GMT +8)
IAS 4042, 4/F, Lo Ka Chung Building, Lee Shau Kee Campus, HKUST

Event Summary

In his Academic Seminar, Imran Rasul, Professor of Economics at University College London, presented research on the role of delivery agents in the delivery of poverty alleviation programs. The study looks at determinants of delivery agents’ behavior as being their social ties to potential recipients and the social context in which they operate. Many development organizations use delivery agents to allocate scarce resources across recipients. Delivery agents often have social ties to some recipients and not others, which can lead to biased allocations and can motivate agents to exert effort. Prof Rasul studies these issues in the context of an anti-poverty program in Western Uganda. 4,741 farmers of over 168 villages were surveyed with regards to their knowledge and adoption of improved seeds and modern techniques. BRAC (an NGO), offers a program delivered by locally recruited agents tasked to sell improved seeds to farmers and train poor farmers in modern techniques. The study employs a novel two-layered field experiment that exploits: (1) experimental variation in the social ties of delivery agents to farmers within villages that they engineer; and (2) non-experimental variation in group identity and conflict between groups across villages. The study finds that delivery agents are significantly more likely to target households in which they are socially tied rather than those in rival groups, a distinction exacerbated in villages where inter-group conflict is most salient. The study then measures the allocation and total coverage of the anti-poverty program relative to three benchmarks: (1) if delivery agents were self-interested; (2) if delivery agents have social ties but display no group biases; and (3) the social planner’s objective of targeting the poorest households. The study then documents when these social contexts lead to allocative biases, and highlights that contexts where group rivalry is most salient lead to greater coverage. The study finds that greater conflict between groups can enhance targeting within groups by delivery agents (parochial altruism). Such behaviors will potentially skew the allocation of the program and therefore, need to be taken into account when designing programs, recruiting delivery agents and incentivizing agents. Development programs that leverage to use of locally hired agents need to account for the network structure of the agents they employ. This network structure not only includes their social ties and connections to their “in-group”, but also the groups they are not connected to.

About the Speaker

Imran Rasul is Professor of Economics at University College London, co-director of the Centre for the Microeconomic Analysis of Public Policy at the Institute for Fiscal Studies, and research co-director of the Entrepreneurship Research Group of the International Growth Centre. His research interests include labor, development and public economics and his work has been published in leading journals such as theJournal of Political EconomyQuarterly Journal of EconomicsEconometrica and theReview of Economic Studies. He is currently a co-editor of the Journal of the European Economic Association, and he been a co-editor and director of the Review of Economic Studies (2009-17). He was awarded the 2007 IZA Young Economist Prize, the 2008 CESIfo Distinguished Affiliate Award, an ERC-starter grant in 2012, and a British Academy Mid-career Fellowship in 2018.

Get updates from HKUST IEMS