In this dissertation, I study the effects of state and federal programs and policies aimed to help vulnerable populations.

In the second chapter, I estimate the effect of immigration enforcement on prenatal safety net programs and birth outcomes. I compare participation in prenatal WIC and Medicaid between likely undocumented mothers and US born non-Latina mothers. I find an increase in immigration enforcement lowers participation in Medicaid but has a null effect on WIC participation. Because undocumented people are ineligible for Medicaid except in special circumstances, using Medicaid to pay for the delivery of a newborn may signal a person's immigration status. WIC eligibility does not have any restrictions regarding a person's citizenship or legal status. This may explain the chilling effect observed in Medicaid but not in WIC. I find that undocumented mothers reduce their prenatal care. There are also improvements in infant birth weight and a decline in undocumented women's birth rate. This suggests positive selection into birth when immigration enforcement intensifies.

Chapter three examines the effect of Deferred Action for Childhood Arrivals (DACA) on living arrangements. DACA provides temporary relief from deportation for undocumented immigrants who arrived to the United States as children. DACA recipients receive a social security number, work permit, and may obtain a driver's license in their home state. Previous studies have found that DACA improves beneficiaries' economic well-being. Since housing decisions are closely linked to economic security, I compare DACA eligible and ineligible immigrants to estimate changes in living arrangements. I find DACA increases the incidence of living with a parent in a rented home by 1.9 percentage points and lowers the incidence of living with other family members by 2.4 percentage points. The economic benefits and mobility that DACA provides along with the lowered fear of deportation may change beneficiaries' living arrangements but does not increase the likelihood of moving into a home without a family member.

In chapter four, I re-examine the effect of Naloxone Access Laws on opioid mortality. In response to the opioid crisis, states adopted laws increasing the availability of Naloxone, an overdose reversal drug. The theoretical effect of these policies is ambiguous due to the potential for moral hazard. The current literature contains mixed results when using a difference-in-differences model to estimate the effect of Naloxone Access Laws on mortality. I revisit these studies and establish that the discrepancies in the findings stem from both different time periods studied and the policy definitions used. I then make a methodological correction by adjusting for the staggered policy adoption and find that Naloxone Access Laws increase opioid mortality by 39%. Finally, I discuss the validity of the results. A sharp rise in opioid mortality preceded the adoption of Naloxone Access Laws. Therefore the estimated results from a difference-in-differences model will not be causal. I propose using a contiguous county-border model to establish the causal effect of Naloxone Access Laws. The findings from this chapter emphasize the challenges of establishing causal estimates when evaluating public policies that inherently are not exogenous.