This dissertation studies the effects that government spending has on the well-being of individuals and on community-level economic outcomes. The first chapter examines federally funded disaster relief with the aim of explicitly quantifying the role that the government has in propping up labor markets after large storms that damage and destroy communities. The next two chapters are about welfare. The second chapter uses administrative data from the state of Michigan to study one of the largest, and most sudden, changes to a cash welfare program in the country's history. The aim of this piece is to quantify the holistic impact of losing welfare on the financial well-being of the affected mothers. The final chapter revisits one of the most consequential welfare-to-work experiments of the late 20th century with modern empirical tools to determine whether work-first retraining programs or remedial coursework benefit the marginal welfare participant more in the long-run. Together these essays highlight the role that the federal government plays in the lives of its citizens when they are at their most vulnerable. It is the hope of the author that economists and policymakers can use the conclusions herein when considering and drafting future programs that aim to assist those at the margin of society or those who will suffer the consequences of catastrophic climate disasters.

This dissertation consists of three chapters exploring challenges that many developing countries face in augmenting state capacity and accumulating human capital. In particular, I focus on difficulties in developing state capacity and human capital induced by political violence, natural disasters, and over-reliance on income from foreign countries. The first chapter explores the effects of losing local politicians on the fiscal and personnel capacity of local governments using the outcome of the assassination attempts on mayors in Mexico. The second chapter investigates the effects of exposure to natural disasters on birth outcomes in Indonesia, using the Indian Ocean Tsunami as a natural experiment. In the final chapter, I use a cross-country analysis to study the link between reliance on remittances and the capacity of a country to collect taxes efficiently. The first chapter investigates the effects of losing mayors to successful assassinations on the capacity of local governments. By leveraging the randomness in the outcomes of assassination attempts against mayors in Mexico in 2002-21, I find that the loss of mayors negatively affects the fiscal and personnel capacities of the local governments. Municipal tax collection decreases by 29\%. The share of expenditure on primary services falls by 3 percentage points and is crowded out toward investment in construction. Municipal workers at productive stages in their careers leave the position. The back-of-the-envelope calculation shows that wages should increase by 13\% to retain them after assassinations. Organized criminal groups take advantage of the loss of mayors by increasing their presence in municipalities with successful assassinations. The results are not explained by non-political violence, levels of economic activities, or population changes. The results speak to the significance of leaders in maintaining fiscal capacity and retaining capable personnel in the workforce even in a violent environment. In the second chapter, co-authored with Elizabeth Kayoon Hur (Michigan State University), I evaluate the effect of in-utero exposure to the 2004 Indian Ocean Tsunami on short-term childbirth outcomes in Indonesia. Exploiting variation in the timing of exposure, I find that the probability of successful pregnancies drops by 5.9 percentage points (pp), while miscarriages increase by 5.5 pp for those exposed in the earliest stage of pregnancy. I find suggestive evidence that post-disaster health investments by households may have shielded later cohorts from harmful effects. The results suggest the importance of considering fetal loss in developing countries and highlight that facilitating household investment in health through various policies may mitigate negative birth effects in the aftermath of natural disasters. The third chapter investigates the relationship between a country's reliance on remittances from abroad and its ability to collect taxes from various domestic sources. Despite the increasing flow of remittances in volume and proportion, particularly among developing countries, their role in determining the state's capacity to collect taxes has received little attention. This chapter explores the link between remittances and various tax revenue categories using country-level data. Two-way panel regressions suggest that a 1 percentage point (pp) increase in the inflow of remittances explains a 0.12 pp rise in consumption tax revenues. The same estimate derived from IV methods proxying for migrant network strength and openness of borders increases to 0.9 pp. Decomposing this result reveals that the increase in household consumption expenditure explains all of the statistical association, not the efficient tax-collecting mechanisms such as VAT. Subsample regressions by income category suggest that the association between remittances and consumption tax revenue is stronger in countries with lower income.

This study examines the recovery process of businesses in Homestead, Florida after Hurricane Andrew in 1992. The goal of this study was to determine which organizational characteristics were useful in predicting the level of physical damage and the length of time to reopen for affected businesses. The organizational characteristics examined were age, size, pre-disaster gross sales, ownership of the business location, membership in the Chamber of Commerce, and property insurance. Three-hundred and fifty businesses in the area were surveyed. Because of the complexity of the recovery process, the disaster experiences of businesses were examined using three paradigms, organizational ecology, contingency theory, and configuration theory. Models were developed and tested for each paradigm. The models used the contextual variables to explain the outcome variables; level of physical damage and length of time to reopen. The SIC was modified so that it could form the framework for a taxonomic examination of the businesses. The organizations were examined at the level of division, class, subclass, and order. While the taxa and consistent levels of physical damage, the length of time needed to reopen varied greatly. The homogeneous level of damage within the groups is linked to similarity in assets and transformation processes. When examined using the contingency perspective, there were no significant relationships between the level of physical damage and the contextual variables. Only predisaster gross sales and level of physical damage had moderate strength associations with the length of time to reopen. The configuration perspective was applied by identifying clusters of organizations using the contextual variables. Clusters were identified and examined to determine if they had significantly different disaster experiences. The clusters varied significantly only by the length of time to reopen. The disaster experience of businesses is conceptualized as a process of accumulation-deaccumulation-reaccumulation. The level of physical damage is driven by selection while the lenght of time to reopen is determined by both adaptation and selection.

In my dissertation, I investigate the relationship between household consumption expenditures and transitory income shocks. In the first two essays, I pay particular attention to household expenditures in the aftermath of natural disasters, which are becoming more frequent and costly in the U.S. since 1980. Additionally, I study specialty farm producers' risk attitudes after an income shock due to natural disasters. Although the permanent income hypothesis predicts that households smooth consumption over their lifetimes, credit-constrained households may find consumption smoothing impractical. This dissertation brings forth evidence regarding heterogeneity in the effect of income shocks on household expenditures. First, I find that floods and hurricanes affect food-at-home (FAH) spending in different ways. The average 15-day decrease in FAH spending is about $2 in the 90 days after a flood and about $7 in the 30 days after a hurricane. In other words, floods have a prolonged effect and hurricanes have an immediate effect. I find that floods and hurricanes remain a threat to the FAH expenditures of vulnerable households, for instance, low-income households and households in coastal states. Second, Indiana specialty farm households reduce their monthly expenses of food and miscellaneous categories by about $119 and $280, respectively, after an income loss of 20%-32%. I also find that Indiana specialty producers are less willing to take financial risk after an income loss experience, i.e., they have a decreasing absolute risk aversion. Finally, in the third essay, I show that Australian households exhibit loss aversion in consumption expenditures which also means that they behave asymmetrically in their consumption response to income shocks. However, it is only working-age younger households that show asymmetric consumption behavior as opposed to the symmetric behavior of retirement-age households. The main message of these various findings is clear: after an income shock, the magnitude of change in consumption expenditures and the saliency of certain expenditure categories for adjustment are context- and population-dependent. Hence, income support policies and post-disaster relief programs may benefit from a better understanding of the consumption behavior of beneficiary population, to achieve maximum impact through better targeting.