"Using annual, repeated cross-sections from national household survey data, we estimate how the January 1997 termination of federal disability benefits for those with Drug Addiction and Alcoholism affected labor market outcomes, health insurance, health care utilization, and arrests among individuals targeted by the legislation. We employ propensity score methods and a difference-in-difference-in-difference approach to mitigate potential omitted variables bias. Declines in SSI receipt accompanied increases in labor force participation and current employment, but had little measurable effect on insurance and utilization. In the long-run, (1999-2002), rates of SSI receipt rebounded somewhat, and short-run gains in labor market outcomes waned"--National Bureau of Economic Research web site.