Incorporating Longitudinal Aspects into Mortality Research Using Social Security Administrative Record Data¹

Article type: Research Article

Authors: Duleep, Harriet Orcutt^{a; *}

Affiliations: [a] U.S. Social Security Administration

Correspondence: [*] Address correspondence to Dr. Harriet Duleep, 4417 Yuma Street, N.W., Washington, D.C. 20016.

Note: [1] Although formerly employed at the Social Security Administration, the author is currently an economist at the U.S. Commission on Civil Rights.

Abstract: To develop effective policies to raise longevity and health, a clearer understanding must be achieved of what determines health and mortality. This can only be accomplished through a multivariate approach in which relevant variables are either statistically controlled for, or controlled in a natural experiments framework. Furthermore, a better understanding of the causal structure underlying variation in mortality requires longitudinal information on several explanatory variables. This paper describes the resources available in the Social Security administrative record system that can be utilized for the study of health and mortality determination. Studies that have used the administrative record data for mortality research are discussed. The most important data resource maintained by the Social Security Administration is the Continuous Work History Sample. This data set contains longitudinal information on several key variables. Matched to other administrative and survey data sources, the Continuous Work History Sample could be developed into a truly comprehensive data base for mortality research. Unfortunately, the development and even the maintenance of this file are endangered by ramifications of the 1976 Tax Reform Act.

DOI: 10.3233/JEM-1986-14203

Journal: Journal of Economic and Social Measurement, vol. 14, no. 2, pp. 121-133, 1986