Cameron, Trudy AnnFitch-Fleischmann, Benjamin2015-08-182015-08-182015-08-18https://hdl.handle.net/1794/19232The first essay considers the relative effectiveness of government and non-governmental organizations (NGOs) as channels to allocate resources. I use a catastrophic climate-related shock--Hurricane Mitch--to examine the political economy of these channels of aid distribution at the micro level. I combine extensive data on aid received by Nicaraguan households with data on municipal election outcomes and an exogenous, precipitation-based measure of hurricane impact. I find that the hurricane had long-lasting effects on the aid received by households from both NGOs and the government. In the short term, however, the government did not provide aid according to the objective measure of hurricane damage but instead provided aid along political lines. The second essay presents estimates of a relationship between extreme hot temperatures during gestation and a child's subsequent physical well-being in a sample of children in Peru, thus extending existing evidence constructed from U.S. data. Estimates are constructed using high-resolution gridded climate data and geo-coded household surveys. The results suggest that a period of extreme heat (a month whose average temperature is more than 2 standard deviations above the local average) in the period 1 to 3 months before birth is associated with lower weight at birth and a reduction in height (measured 1 to 59 months after birth) that cannot be fully explained by birth weight. There is no evidence of differential maternal investment, as measured by duration of breastfeeding, according to a child's exposure to extreme heat during gestation. The third essay asks whether improved treatment of HIV/AIDS in Africa can be achieved simply by paying health workers to do more. I present estimates of the impact of financial incentives paid to individual workers at public health facilities in Mozambique. The results suggest that piece-rate incentives increased the delivery of five out of fourteen health services for which treatment effects can be identified, with estimated increases ranging from 34 to 157 percent, depending on the particular service. I find no evidence of a corresponding decrease in the delivery of services that are not financially incentivized, suggesting that there is no "crowding out" of intrinsic motivation.en-USAll Rights Reserved.Birth weightDevelopment aidHurricaneNicaraguaNon-governmental organizationsElectronic Thesis or Dissertation