This dissertation uses econometric models to reevaluate key questions in public finance economics and estimate causal effects of funding changes on safety net programs. In particular, this dissertation explores various funding to federal programs and their impact on the social welfare of vulnerable Americans. The main thread connecting these chapters is their relevance to policy and programs that affect disadvantaged communities.

The first chapter studies the effects of government funding on private funding to charitable organizations. I exploit a recent release of tax microdata from electronic filers spanning an eight-year period and analyze the financial responses of food banks and similar charitable organizations. Using a U.S. federal program that allocates funding to states on a formula-basis, I implement an instrumental variables design to analyze the aggregate state-by-year change in private funding to emergency food providers as a result of an increase in exogenous government funding. Contrary to predictions from the standard crowd-out theory, I find that private funding increases by 1% to 1.6% as a result of a 1% increase in government funding. This increase is driven by growth on the intensive margin, or growth among existing organizations rather than an increase in number of organizations, and is robust to various specifications. I also find crowd-in of fundraising expenditures, which suggests the increase in private funding is driven by increased fundraising efforts. My results highlight the importance of accounting for heterogeneity in financial responses across types of charitable organizations and how government funding may help food providers increase their scale to help solve social needs.

The second chapter evaluates the effects of low-income housing on crime rates in affluent neighborhoods, which are frequently excluded in similar analyses. I exploit a change in policy on how the Low Income Housing Tax Credit (LIHTC) is awarded in Texas, which incentivizes the construction of subsidized housing by private developers. The policy change created a rule to award more generous tax credits to developments located in affluent communities. Using the quasiexperimental variation on project location generated by this rule, I find that eligible areas see an increase of 8% in their flow of units over comparable areas that do not qualify for the tax credit boost. Additional LIHTC units do not appear to affect property or violent crime rates in their low-poverty neighborhood. I also do not find evidence of an effect on drug related offenses.

The third chapter builds on the second chapter's empirical setting and the first chapter's theoretical framework to estimate the effects of the change in LIHTC construction in affluent neighborhoods and consequent effects on the housing market. I use a fuzzy difference-in-difference design to evaluate the crowd out effects of subsidized housing on privately owned housing. I also analyze the effects on other market outcomes. I find that subsidized housing does not crowd out market rate housing in affluent areas. I also find that the rental vacancy rate increases as a result of new subsidized housing, which may suggest that the change in housing is driven by increased rental housing rather than homeowner-occupied housing.