2007 Awards

No margin, No mission? Evaluating the Role of Incentives in the Distribution of Public Goods

Nava Ashraf, Assistant Professor, Negotiations, Organizations, and Markets Unit, Harvard Business School

2007-08 award of $20,000

Non-profit and public organizations increasingly rely on the services of community members to deliver and promote health goods, yet research on the motivation and performance of these agents is scarce. The key challenge for incentive design in this setting is that the task performed by community agents entails benefits for the society at large and perhaps standard financial incentives that work well for the distribution of commercial products might be ineffective or even harmful if they crowd out the agents' intrinsic motivation for the task.

This project performs a rigorous evaluation of different incentive schemes to motivate community-based agents engaged in the distribution of female condoms in Zambia. Female condoms present a valuable technology for the prevention of HIV/AIDS and other STIs and give women the ability to take protection into their own hands. While female condom uptake has been high in some parts of Southern Africa, barriers to adoption similar to those accompanying many new technologies continue to hinder usage in many communities. Society for Family Health in Lusaka, Zambia is relying on a novel distribution channel - hairdressers and barbers - to promote the female condom and to provide information to customers about its proper use and the dangers of unprotected sex.

In collaboration with SFH, we have designed a randomized control trial to evaluate the effect of different incentive schemes on the selection of hairdressers and barbers that choose to join the program and on their performance as female condom promoters. Our experimental design assigns different salons to one of four different incentive schemes:(i) a baseline "volunteer" scheme, whereby agents are given the opportunity to buy condoms from SFH to sell to their customers but receive no reward for performance; (ii) "low powered financial", whereby agents are offered a small monetary reward as a function of their sale performance; (iii) "high powered financial", whereby agents are offered a large monetary reward as a function of their sale performance; and (iv) "status", whereby agents are offered public recognition of their services as a function of their sale performance.

Improved understanding of how different incentive schemes affect both selection and effort of community agents can facilitate the delivery of health products and other goods that entail positive externalities for the community. Any agency, public or private, concerned with the design of incentives for tasks with a pro-social component can benefit from the research findings by tailoring them to its specific context.

Final progress report - submitted 090130


Ashraf, Nava, Oriana Bandiera and B. Kelsey Jack. 2014. No margin, no mission? A field experiment on incentives for public service delivery. Journal of Public Economics 120: 1-17.

The Role of Small-Scale Energy Use in Local Air Pollution Patterns in Accra, Ghana

Majid Ezzati, Associate Professor of International Health, Department of Population and International Health, School of Public Health

John (Jack) Spengler, Akira Yamaguchi Professor of Environmental Health and Human Habitation, Department of Environmental Health, School of Public Health

Matt Welsh, Assistant Professor of Computer Science, School of Engineering and Applied Sciences

2007-08 award of $29,510

The aim of the proposed research is to understand the effects of urban biomass use on local air pollution patterns, especially in relation to neighborhood socioeconomic and spatial characteristics. To achieve this, we leveraged an ongoing project on air pollution exposure in Accra, Ghana, and conducted detailed and high-quality measurement of energy supply and energy use and within-neighborhood variation of air pollution, and collected samples to analyze the composition of neighborhood particle pollution in selected low-, middle-, and high-income neighborhoods. We are using these data to map and quantify domestic and small-scale commercial energy supply and use, and their contributions to neighborhood air pollution. In addition to scientific results, the project helped with graduate and undergraduate student training at Harvard and University of Ghana. We also plan to use the results of this research as the basis for working with Ghanaian scholars and practitioners who specialize in energy policy, environmental regulation, and community development to advance policy analysis and evaluation regarding urban development, energy, and air pollution in Accra.

Final progress report - submitted 090121


Dionisio, KL, RE Arku, AF Hughes, J Vallarino, H Carmichael, JD Spengler, S Agyei-Mensah , and M Ezzati. 2010. Air pollution in Accra neighborhoods: Spatial, socioeconomic, and temporal patternsEnvironmental Science and Technology. 44(7):2270–2276

Dionisio, KL, MS Rooney, RE Arku, AB Friedman, AF Hughes, J Vallarino, S Agyei-Mensah, JD Spengler, and M Ezzati. 2010. Within-neighborhood patterns and sources of particle pollution: Mobile monitoring and GIS analysis in four Accra communitiesEnvironmental Health Perspectives 118:607-613.

Measuring Willingness to Pay for Improved Water Quality at the Source

Michael Kremer, Gates Professor of Developing Countries and Professor of Economics, Department of Economics

2007-08 award of $40,000.

Understanding how local communities value environmental services can help donors and governments gauge the benefits from implementing protection measures, and can also determine the scope for cost-recovery from these investments. In practice, however, determining willingness to pay for environmental services is difficult. Valuation approaches that rely on self-reported preferences are not based upon market transactions and may be subject to bias, and other estimation strategies may rely on structural assumptions to generate results.

We utilize a unique opportunity to compare a variety of measures of willingness to pay (WTP) for improvements in source water quality in rural Kenya using new data collected in conjunction with a randomized evaluation of improvements in source water quality. With this data and research design, we implement both a travel-cost analysis and a structural econometric method using the mixed-logit framework, exploiting the randomized nature of the intervention for econometric identification and utilizing detailed information on households' actual water collection choices. We compare experimental estimates that rely on exogenous variation in water quality to more conventional non-experimental revealed preference estimates. We also compare the revealed preference estimates to households' stated willingness to pay. These methods allow us to estimate the underlying heterogeneity in household preferences for clean water, which may be substantial.

In preliminary results, we find that stated preference valuations for improvements in source water quality yield much higher WTP estimates, by a factor of three, than do estimates based on a revealed preference travel-cost approach. These WTP figures inform water regulatory policy and resource management decisions, including decisions on user fees and public investment levels. To our knowledge these are among the first revealed preference estimates of household valuation for drinking water quality, a fundamental behavioral parameter in a rural developing country setting.


Null, Clair, Michael Kremer, Edward Miguel, Jorge Garcia Hombrados, Robyn Meeks and Alix Peterson Zwane. 2012. Willingness to pay for cleaner water in less developed countries: Systematic review of experimental evidence. London: International Initiative for Impact Evaluation.

Encouraging Take-Up of Point of Use Water Treatment Products

Sendhil Mullainathan, Professor of Economics, Department of Economics

2007-2008 award of $39,930

Behavioral economics is a natural methodology to apply to sustainability science. Sustainability challenges arise in part because people sometimes make puzzling choices. Behavioral economics seeks to understand how human behavior deviates from the perfectly rational, utility maximizing agent of neo-classical economics. By better understanding the basis for these choices, we can develop approaches to encourage people to make choices that mitigate environmental risks, reduce exposure, or adapt to changing environmental circumstances. We applied to apply techniques from behavioral economics to promote the adoption and sustained use of point of use (POU) technologies that improve drinking water quality in rural Kenya. POU treatment is medically effective at mitigating the health risks of low quality source water. Chlorine dispensers at the source promote take up, as does distribution at clinics.

Final progress report - submitted 091223


Kremer, Michael, Edward Miguel, Sendhil Mullainathan, Clair Null, and Alix P Zwane. 2009. Making water safe: price, persuasion, peers, promoters, or product design. Unpublished Manuscript.