Budget cuts, data and evaluation

Abstract

c oncern about the Reagan administration's recent cuts in federal budgets for administrative and statistical data systems has focused almost entirely on one expected impact of those cuts: that is, a decrease in the amount of information directly available to the public. (The term administrative data as I use it refers to those data that are designed and collected by administrative agencies in the normal course of their work to facilitate the administration of a program---e.g., to guide agency operations and policy decisions. The term statistical data refers to those data that are designed and collected to improve scientific knowledge.) Researchers, businessmen, members of Congress, farmers, government planners at federal, state, or local levels, labor unions, and many other groups have expressed fears that they would not continue to find the data they need for performing their various functions. Given the large number of radical changes brought to government programs by the present administration, many data users have manifested surprise at these cutbacks in the quantitative tools that will eventually be needed by the administration both to describe its innovations and demonstrate their success. Administration spokesmen have responded that, as particular federal programs are reduced, eliminated, or transferred to the states under its New Federalism initiatives, the policy is to cut back or abolish federal data collection efforts in those program areas, as well as in any other areas in which the data are not collected specifically and directly for executive branch policy purposes. They have pointed out that, as programs get smaller or disappear from the federal locus of responsibility, the policymakers' need for data in those areas is also simultaneously reduced; that much of the federal government's current costly data collection effort is both unnecessary and burdensome: states, local governments, and business firms can easily fill the federal government's data collection role; that the appetites of data users arc insatiable; and that special evaluation studies can substitute for existing data systems by conducting original data collection and developing their own independent information. To the contrary, I argue that: