This dissertation consists of three essays, each covering very distinct topics under the broad umbrella of Development Economics, each set in a different region of the developing world (Latin America, Sub-Saharan Africa, and South Asia). The one element that loosely ties them together is that they each seek to add, in a small way, to our understanding of factors that contribute to, and in some cases may entrap people in, poverty: factors such as (lack of) geographic mobility, hunger, and disease.
In the first chapter, I use the natural experiment of military conscription in Argentina, which randomly assigned not only military service, but also the location of service, to study the effect of this temporary displacement on long-run migration rates. I then use a rich source of administrative earnings and employment data to investigate the labor-market implications of conscription and, in particular, displacement. I find that conscription on the whole caused a small increase in the likelihood of appearing in the formal labor force, and a small increase in earnings particularly for those who were assigned to serve in the Navy. Assignment to military service outside of one's province of origin increased the likelihood of living outside the province of origin by 2.5 percent, and while the net effects of this displacement on earnings and employment are imprecisely estimated, the evidence suggests that there are modest long-term benefits of conscription in Argentina that are not fully attributable to displacement.
In the second chapter, I investigate the effects of Ramadan on calorie consumption and labor supply among Muslim households in rural Malawi. Across four rounds of household survey data, I find no evidence of a decrease in calorie consumption during Ramadan on average. I do, however, find evidence that working-age people reduce their weekly work by about three hours, or nearly 20 percent, on average. This finding on calories shows substantial variation across the different rounds of data. The evidence presented calls into question the hypothesis that consumption during Ramadan should fall more dramatically when the holiday overlaps with the harvest (when baseline consumption levels are relatively high compared to the rest of the year), compared to when Ramadan falls near the annual hunger season (when baseline consumption levels tend to be much lower). I discuss potential implications of this variation for our understanding of seasonal consumption patterns.
The third and final chapter, which is authored jointly with Kaivan Munshi and Nancy Luke, discusses a randomized intervention conducted in rural South India aimed at improving rates of treatment completion for tuberculosis. Tuberculosis (TB), despite being a highly treatable disease, kills well over 1 million people every year, with 95 percent of cases and deaths appearing in developing countries. India bears the largest TB burden of any country, with more than 25 percent of the world's total yearly cases. A key factor for successful management of TB is ensuring that patients complete the full six-month (or more) treatment regimen: missing even a few doses of the prescribed medications increases the likelihood of relapse and development of a drug-resistant strain of TB, which is much more difficult and costly to treat effectively. We conduct an intervention allowing patients to select a community member to serve as a Directly Observed Treatment (DOT) provider to help ensure compliance with the full treatment regimen. Although patients assigned a Community DOT provider report significantly more frequent visits and higher rates of satisfaction compared to our control group, we do not find any significant improvement in treatment outcomes among those assigned this intervention. We explore several potential explanations for this finding and suggest potential avenues for future research.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/d8-tbe4-hn21 |
| Date | January 2021 |
| Creators | Weiner, Scott |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0132 seconds