Postrecovery experience of disabled-worker beneficiaries.

This study followed a 1972 cohort of newly entitled beneficiaries from January 1, 1981, when they left the Disability Insurance program because of a recovery, to June 1986. This recovery group represents about 11 percent of the cohort of disabled-worker beneficiaries entitled in 1972. Three outcomes or next events were of specific interest: Return to the DI program, death, and attaining age 62 (considered retirement for the purposes of this study). The tendency toward reentitlement and the tendency toward death were modeled and then combined with retirement age to project the percentages of recovered beneficiaries who end the postrecovery period by reentitlement, death, or retirement. About 43 percent of these recovered beneficiaries are expected to become reentitled some time after leaving the program, and 52 percent are expected to reach retirement age before dying or becoming reentitled. Of the 43 percent projected too return to the program, 23 percent are expected to become reentitled within the first 5 years of recovery. The tendency to return to the DI program drops sharply at the fifth year of the postrecovery period. There may be program-based reasons for this pattern. The Social Security Amendments of 1980 may provide incentives to return to the DI program within 5 years. When covariates were examined, the covariate PIA had a strong effect on the reentitlement tendency. For those in the high PIA group ($500 or more), it is projected that 65 percent will return to the DI program. This projection is considerably higher than the 34 percent projection for those in the low PIA group (less than $500). Projected median time to reentitlement is quite different for the two PIA groups. The median time to reentitlement is 10 years for the low PIA group and only 3 years for the high PIA group. The PIA was also important in modeling the death tendency as the next event in the postrecovery period. For those in the low PIA group, death is the next event projected for 3 percent of the individuals; in the high PIA group, the projected proportion is 11 percent. It is suspected that the PIA may be acting as a proxy for the severity of the disabling condition, but this hypothesis cannot be tested with the available data. These descriptions of the reentitlement and death tendencies and the project percentages provide a global picture of the DI program reentitlement process. Further research will continue by comparing this pre-1980's cohort with a later cohort and studying the causal mechanisms underlying the recovery and reentitlement processes.