Impacts of solar lanterns in geographically challenged locations : experimental evidence from Bangladesh

Despite continuous efforts to improve the coverage, the access to electricity remains insufficient in many developing countries, particularly in geographically challenged locations, due mostly to the high cost of grid extension. To rigorously investigate the effectiveness of solar products as an alternative in remote areas, we conducted a randomized controlled trial in river islands of northern Bangladesh where no grid-based electricity is available. We found that solar lanterns significantly increased home study hours among schooled children, especially in the night and before exams. School attendance rate also initially increases due to the provision of solar lamps, although such effects fade away over time. The increased study time and initial school attendance rate, however, did not improve children's exam results. We also found marginal improvements on health-related indicators, such as eye redness and irritation, but negligible impacts on respiratory indicators. Households that received solar lanterns substituted the traditional lighting sources with modern technology, leading to a significant decrease in annual biomass fuel consumptions, particularly kerosene. Finally, treated households showed a greater self-reported willingness to purchase solar products compared with the control group.

Impacts of an HIV counseling and testing initiative -- results from an experimental intervention in a large firm in South Africa

We have run experimental interventions to promote HIV tests in a large firm in South Africa. We combined HIV tests with existing medical check programs to increase the uptake. In the foregoing survey we undertook previously, it was suggested that fears and stigma of HIV/AIDS were the primary reasons given by the employees for not taking the test. To counter these, we implemented randomized interventions. We find substantial heterogeneity in responses by ethnicity. Africans and Colored rejected the tests most often. Supportive information increased the uptake by 6 to 16% points. A tradeoff in targeting resulting in stigmatizing the targeted and a reduction of exclusion error is discussed.