This dissertation studies the impacts of local labor market changes on the US family structures and disability benefit take-up. The dissertation uses unexpected time series changes in energy prices, together with the pre-existing county exposure to the unexpected time series changes, to identify causal links between the changes in county's economic conditions and changes in marriage outcomes, fertility outcomes, and disability benefit payments.,The first chapter is motivated by the decline in marriage and the rise in out-of-wedlock births, which are among the most important changes in American lives over the last half century. A notable hypothesis to explain both trends, sometimes dubbed as the “unmarriageable men hypothesis,” attributes these changes to the disappearance of well-paying industrial jobs for the less-educated men. Exploiting county-level variation in oil-producing and coal-producing areas from the shocks to world oil and coal prices in the 1970s and 1980s, I find that a decrease in men's earnings leads to lower marriage prevalence, lower marital fertility rates, and higher non-marital fertility rates, confirming the hypothesis. When men's earnings increase, however, marriage outcomes and non-marital births respond differently across regions: while higher earnings lead to higher marriage prevalence and lower non-marital births in coal-producing counties, I find little marriage response and more non-marital births in oil-producing counties during the oil boom, which is inconsistent with the implications of the unmarriageable men hypothesis.,The second chapter, coauthored with Kerwin Charles and Melvin Stephens Jr., examines the extent to which US Social Security Disability Benefits (SSDI) and Supplemental Security Income (SSI) payments, the two largest Federal programs that provide assistance to people with disabilities, are affected by local labor market earnings from 1970 to 2011. To circumvent the potential reverse causality problem in the literature, we exploit county-level variation in oil-producing areas from shocks to world oil and gas prices to identify the causal effects. We extend well-known previous work using a similar research design by analyzing a different price shock, a larger, more representative set of labor markets, and a more recent period marked by skyrocketing disability payments. Our estimated elasticity for SSDI payments with respect to earnings of −0.29 is surprisingly similar to earlier findings. Our preferred SSI elasticity estimate of −0.16 is smaller than previous findings, but we show that SSI programmatic changes explain most of the difference.