This research proposes to model and test empirically the ways in which changing patterns of industrial demand affect interindustrial wage differentials and result in differing job exit patterns of older workers. Based primarily on the March CPS files from 1968 through 1986, both time series and cross sectional analyses will be used to track wage shocks and trends. This work will attempt to discover the extent to which industrial wage differentials are work-experience or age-related, the extent to which they persist through calendar time, and whether there are common patterns of evolution. CPS data on employment by industry, possibly combined with industrial production data, will be used to examine the question of whether the wage movements are in fact demand-generated. A second major focus is to estimate how accumulated demand shocks affect retirement behavior of older workers. In contrast, most recent studies of retirement have almost exclusively examined supply-side aspects, especially Social Security and pension considerations. The research will also examine sources of the surprisingly frequent practice of approaching retirement via exit from the primary work force and into self employment. The March CPS matched files (nine sets from 1977-1978 through 1985- 1986) are to be the major data source for this portion of the project, with the Retirement History Survey used to verify findings. A final goal is to assess the usefulness of the CPS matched files as a quasi-panel. These data may enjoy substantial advantages over competing sources because of the way in which the sample is continually replenished. Parallel analyses conducted with CPS match files and with the Retirement History Survey will aid in this assessment.
