I estimate the effect of removing children with disabilities from Supplemental Security Income (SSI) on parental earnings and household disability receipt. Using administrative data from the Social Security Administration (SSA), I implement regression discontinuity and differencein- differences designs based on changes in SSA's budget for child medical reviews. I find that parents fully offset the SSI loss with increased earnings, and the loss of the child's SSI payment reduces disability applications by parents and siblings but does not reduce their actual disability receipt. I examine alternative hypotheses for the large parental earnings response.

We estimate the effect of losing Supplemental Security Income (SSI) benefits at age 18 on criminal justice and employment outcomes over the next two decades. To estimate this effect, we use a regression discontinuity design in the likelihood of being reviewed for SSI eligibility at age 18 created by the 1996 welfare reform law. We evaluate this natural experiment with Social Security Administration data linked to records from the Criminal Justice Administrative Records System. We find that SSI removal increases the number of criminal charges by a statistically significant 20% over the next two decades. The increase in charges is concentrated in offenses for which income generation is a primary motivation (60% increase), especially theft, burglary, fraud/forgery, and prostitution. The effect of SSI removal on criminal justice involvement persists more than two decades later, even as the effect of removal on contemporaneous SSI receipt diminishes. In response to SSI removal, youth are twice as likely to be charged with an illicit income-generating offense than they are to maintain steady employment at $15,000/year in the labor market. As a result of these charges, the annual likelihood of incarceration increases by a statistically significant 60% in the two decades following SSI removal. The costs to taxpayers of enforcement and incarceration from SSI removal are so high that they nearly eliminate the savings to taxpayers from reduced SSI benefits.

How does the expectation that a child will receive government benefits in adulthood affect parental investments in the child’s human capital? Most parents whose children receive Supplemental Security Income (SSI) benefits overestimate the likelihood that their child will receive SSI benefits in adulthood. We present randomly selected families with the predicted likelihood that their child will receive SSI benefits in adulthood. Reducing parents’ expectations that children will receive benefits in adulthood does not increase investments in children’s human capital. This zero effect is precisely estimated. Likely explanations include parents working more themselves, nonfinancial goals influencing investment, and families facing investment constraints.

We study how increases in the U.S. Social Security full retirement age (FRA) affect benefit-claiming behavior and retirement behavior separately. Using long panels of Social Security administrative data, we implement complementary research designs of a traditional cohort analysis and a regression-discontinuity design. We find that while claiming ages strongly and immediately shift in response to increases in the FRA, retirement ages exhibit persistent “stickiness” at the old FRA of 65. We use several strategies to explore the likely mechanisms behind the stickiness in retirement, and we find suggestive evidence that employers play a role in workers’ responses to the FRA.