The investigator has developed statistical methods and data sources that permit investigation of the relative importance of individual and firm heterogeneity in observed labor market outcomes. These methods do not assume, as virtually all previous research has, that the person effects and the firm effects are uncorrelated. Using data from France in research supported collaboratively by NSF and the French National Statistical Institute, the investigator has shown that person effects are much more important in determining wage outcomes than are firm effects, but that these two effects are reasonably correlated. The investigator has assembled American data from four sources that can now be used to estimate versions of the person-effect/firm-effect models applied to French data. Because of the nature of the American data, the models must be somewhat modified but the essential character of the analysis-the possibility of correlation between separately identifiable person and firm effects-is preserved. The investigator has also been sponsored by the Bureau of Labor Statistics (BLS ) and by funding under a previous NSF grant to create an employer-link file for the National Longitudinal Survey of Youth (NLSY) 1979 Cohort. This employer-link file contains the identification numbers required to link external information about the employer to the basic NLSY 1979 data for the years 1986-1994. The investigator has access to these data as a part of a confidentiality agreement with the BLS, and other users of the NLSY will be able to obtain the data from the Bureau's Longitudinal Data Programs. The investigator will extend and modify his basic statistical models that permit him to analyze wage, employment and other labor market outcomes measured on the NLSY using the employer link file and the associated information on the employer. The other data sources available to the investigator include matched employer-employee managerial data developed by the Center for Advanced Human Resource Studies at Cornell University. These data include wage outcomes, contingent pay and a large number of performance measures on the firms. Data files developed from matched Current Population Surveys that include limited employer information and from the Worker-Establishment Characteristics Database used in the investigator's current research will also be available.
