The globalization of food markets is accompanied by nutritional and epidemiological transition in developing countries — often manifest as increased dietary diversity but also increased consumption of sugars and fats, and an accompanying increase in obesity and diabetes. In this paper, I examine how improved rural accessibility to major markets brings unhealthy junk food to rural households in Nepal, increasing consumption of those junk foods even among young children. Nepal's dramatic topography has historically served to isolate many rural communities from major markets, but in recent years the government has engaged in large-scale road and bridge construction, reducing travel costs dramatically. This road construction, mostly concentrated in urban areas, has improved rural accessibility to Kathmandu, where almost all junk food is manufactured or imported. Working with OpenStreetMap data and identifying on the plausibly exogenous reductions in travel time to Kathmandu due to non-local road construction, I ask whether rural accessibility to Kathmandu increases expenditure on and consumption of junk food. I find that a 10% decrease in travel time to Kathmandu from rural communities increases in household junk food expenditure by 2.4% and increases junk food consumption among children by 2.7%. No impact is found on other types of food consumption or on household income, as expected if the identifying variation is exogenous to local economic conditions.

The United States experienced a massive and unexpected influx of Cuban refugees in 1980; called Mariel Boatlift. According to the “Public Law 1980”, the US government provided permanent residence and financial assistance to Marielito for their stable resettlement. However, none of the previous Mariel Boatlift economic studies considers their refugee status in their analysis. This paper re-examines the housing rental market impact of Mariel Boatlift, considering Marielito as refugees. First, I estimate difference-in-difference (DD) specification by constructing a control group based on the housing market supply elasticity, and I find that rental housing price increases in the short run. Next, Event History estimation confirms that the rise in rental housing price happens right after the Mariel Boatlift. Last, I utilize the Synthetic Control Method (SCM), to deal with the concern on non-parallel pre-trend and a potential sample bias of DD with fixed effect. These empirical specifications consistently confirm that the influx of Cuban refugees increased housing rents in Miami in the short run. Notably, the natives who had lived in the second-lowest quartile rental units were affected most harshly by the housing demand shock. A revised theoretical model and suggestive evidence of the evolution of Hispanic/white ratio and the receipt of welfare income or social security income propose a possible mechanism to explain why the second lowest quartile rental units are most affected by the refugee shock.

We study how an early life nutrition program affects health and economic outcomes in adulthood. The Supplemental Nutrition Program for Women, Infants, and Children (WIC) is intended to improve the nutritional well-being of low-income pregnant and postpartum women, infants, and children, by giving food vouchers for specific items with key nutrients. We utilize variation across counties in the date of implementation of the WIC program to identify the impact of the program on later life health and economic outcomes. Using geocoded data from the Panel Study of Income Dynamics, we find that early life exposure to the WIC program caused a lower incidence of high blood pressure and asthma in adulthood, while no clear improvement is shown in economic outcomes.

Since 1997, many states in the US have introduced a Quality Rating and Improvement System (QRIS) to improve the quality of child care. One important goal of a QRIS is to improve child development, but no studies have examined the effect of QRIS's on child development. In this paper, we study the effect of QRIS's on reading and math achievements at kindergarten, utilizing the variation in implementation dates of QRIS's across states as a source of identification. We estimate a difference-in-difference model, using data from three cohorts of the Early Childhood Longitudinal Study. We find that reading and math achievements are reduced by 38% and 23% of a standard deviation, respectively, for five or six years of exposure to a QRIS. The negative effect is possibly because QRIS could increase the price of child care; and because QRIS could be similar to imposing stringent regulations, and so decrease the number of child care centers or family day care homes. As a result, parents may shift into low quality child care due to these two channels.