Research and politics in policymaking for Social Security.

Abstract

BRIAN Gratton's article (1997) in this issue gives us the unsurprising news that there is a politics of statistics — that research is not a neutral form of fact-gathering but rather a political weapon. His article vividly indicates that the Social Security Board used tainted data support its efforts establish the constitutionality of the program in 1937. Indeed, Mr. Justice Cardozo's opinion provided an eloquent defense of the program, arguing that the purpose of Social Security was to save men and women from the rigors of the poorhouse as well as from the haunting fear that such a lot awaits them when journey's end is near (Cardozo, 1937/1985, p. 33). To my knowledge, the defenders of Social Security never regretted their tactics in gathering data for the Supreme Court brief (Berkowitz, 1995). On the contrary, they often celebrated the decision, just as I am sure that Thurgood Marshall had reason celebrate the Court's 1954 decision in the school desegregation cases. That case, too, rested on social science research that was subject challenge. The strategy of using social science in a legal brief went back at least as far as the Brandeis brief that was used in wage-andhour cases at the turn of the century (Goldmark, 1953). In all of these cases, legal advocates never doubted the justice of their causes. They simply employed social science theory as a forensic weapon. Does that mean that research done by the Social Security Administration comes from a tainted tradition and is inherently suspect? I think not. The Annual Statistical Supplement of the Social Security Bulletin contains page after page of reliable program statistics, reflecting a practice that began in part with Ida Merriam who, for many years, ran the Office of Research and Statistics. She helped train such outstanding practitioners in the field of health and social welfare statistics as Dorothy Rice, who remains one of the world's experts on disability statistics. The actuaries at Social Security also come from a long line of dedicated people who have tried make reliable estimates of future program expenditures in the face of many political inducements, in some cases from inside the agency but more often from outside political advocates in Congress, shade the numbers one way or the other (Derthick, 1979). Recently, for example, the actuaries accepted a challenge help prepare cost estimates for a plan advocated by members of the Social Security Advisory Council who favored the privatization of the program. In this case, the actuaries were simply doing their jobs, even though their jobs conflicted in some sense with the political sentiments of their immediate superiors. Nonetheless, the history of Social Security illustrates that advocacy and objectivity are often antithetical one another. The times of greatest conflict come during eras of political uncertainty in which, as in Dr. Gratton's example, the program is struggling for its survival or in which program advocates are attempting move the program into new realms. Disability insurance, first proposed in 1938 but not adopted as part of Social Security until 1956, serves as a good case in point. In 1938 an open rift developed within the Social Security Board (as it was then called) over the cost of adding disability insurance the program. The head of the research and statistics department wanted the agency announce a definite cost estimate for disability protection, but the chief actuary refused give a single figure. The actuary said he could not possibly give one estimate and maintain his professional integrity (Cohen, 1938). A decade later another actuary warned against the use that the agency was making of disability cost estimates. He said the issue was not the initial cost of the program, which would be quite low, but rather the fact that all restrictions would have be removed, and costs will hence become very large (Myers, 1948). The actuary's words went unheeded. When the program passed in 1956, it initially cost very little and then experienced huge cost increases in the 1970s and again in the 1990s (Mashaw, 1997). Here was a case of the agency ignoring its own research in favor of making what it perceived as an important social advance. More recently, the effort preserve Social Security in the face of calls for its retirement has led renewed controversy over the use of statistics. Here we might consider the events leading the 1983 amendments. Created by President Ronald Reagan in 1981, the National Commission on Social Security Reform eventually produced the deal that led these program-saving amendments. In the