NGO reputation and effectiveness

In a given year, OECD countries and private individuals channel tens of billions of dollars in overseas development assistance through nongovernmental organizations (NGOs). A recent study evaluates the effectiveness of NGOs in the interventions they implement. The study’s central thesis is that NGOs reduce transactions costs and influence household choice to drive policy-relevant outcomes, especially in those cases when the NGO has previously been active locally.

The authors test this thesis in a particular context, namely an intervention designed to promote improved cookstoves (ICS) in rural India. Nearly three billion people globally rely on traditional stoves and solid fuels for their primary energy needs. ICS are designed to increase the efficiency of solid-fuel combustion. Widespread uptake has proven difficult.

The study starts with the creation of two different treatment samples of villages, one with prior exposure to a local development NGO and one without. Altogether, the experimental intervention covers nearly 1,000 households across approximately 100 geographically distinct hamlets in rural Uttarakhand, India.

The results suggest that ICS promotion increases adoption: nearly half of all households targeted by the promotion campaign purchased an ICS. Equally important, by comparing the treatment samples, the authors estimate a large, positive, and statistically significant “NGO reputation effect”: purchase rates are approximately 14 percentage points (31%) higher in treated communities with prior NGO engagement. The “NGO reputation effect” is even higher when one measures continued use of the more efficient stoves over time.

The figure below summarizes the results. (A village is considered an adopter if someone in that village purchased at least one intervention ICS in response to the ICS-promotion intervention.) The vertical axis measures the mean intervention ICS purchase rate. The light gray bar corresponds to non-NGO villages, that is, villages where the NGO did not have prior experience. The dark gray bar corresponds to villages with prior NGO experience. Error bars represent 95% confidence intervals for the means.

The authors also find that households in villages with prior exposure to the NGO reduce daily fuelwood use and time spent collecting fuels. By contrast, no variation is found in non-NGO villages. Finally, these effects are strengthened when the authors adopt measures of NGO experience (that is, when the authors go beyond the binary classification NGO vs non-NGO villages).

All in all, the results suggest that NGOs foster reputational capital and lower transaction costs, which in turn enhances the effectiveness of interventions in communities in which they are active.

Faraz Usmani, Marc Jeuland, Subhrendu K. Pattanayak

NGOs and the Effectiveness of Interventions Unavailable

The Review of Economics and Statistics (2024) 106 (6): 1690–1708.