The distribution impact of the social security program, 1962-1972

This dissertation investigated the effect of old age insurance (OAI) benefits on the distribution of lifetime income for persons retiring between 1961 and 1973. An actuarial standard of fairness was employed to assess the direction of redistribution and the extent to which the OAI program redistributed income across subgroups comprising the retirement cohort. It was hypothesized that socioeconomic status influenced the absolute size of the redistributional component because of differential survivorship rates and the differential treatment of certain identifiable groups, such as women, nonworking persons age 65-71, traditional family structures, and low-income household units. In addition, the distributional consequences of a "maturing" pay-as-you-go retirement system were examined;The distributional impact of the OAI program under legislation in effect in 1972 was isolated by decoupling the insurance portion of the OAI benefits from the redistribution portion. The decoupling was accomplished by estimating a series of annuity-type counterfactuals--person-specific estimates of actuarially fair benefit payments. The size of the annuity benefit received was dependent on the accumulated value of OAI contribution, the comprehensiveness of the income insurance purchased, and the degree to which the insurer could discriminate among beneficiaries. The annuity-type estimates and 1972 OAI benefit levels used to measure redistribution were based on data from the 1973 Current Population Survey--Administrative Record Exact Match File and the Longitudinal Social Security Earnings Exact Match File;All 1972 beneficiaries received more than their "money's worth" from the OAI program. In addition, the OAI program was found to be "mildly" and "generally" progressive across income groups, but it also exhibited strong regressive features, resulting in lower relative returns to middle-income beneficiaries. The regressive features were more pronounced with the introduction of the earning test and socioeconomic-adjusted mortality rates. Overall, the lower-income groups received the largest relative gains from the program, whereas the middle-income groups received the largest share of the intergenerational transfer. The size of the intergenerational transfer was found, in many cases, significantly associated with the following worker characteristics: lifetime earnings, service length, sex, age at retirement, year of retirement, and marital status.