ERIC Number: ED414492
Record Type: Non-Journal
Publication Date: 1997-Nov
Reference Count: N/A
Employer Learning and Statistical Discrimination. National Longitudinal Surveys Discussion Paper.
Altonji, Joseph G.; Pierret, Charles R.
The relationship between employer learning and statistical discrimination was explored through a statistical analysis that included a test for statistical discrimination or "rational" stereotyping in environments where agents learn over time. The test is used to study the working hypothesis that, because firms have only limited information about the quality of workers in the early stages of their careers, they distinguish among workers on the basis of easily observable variables that are correlated with productivity, such as years of education or degree, the quality of the school the person attended, race, and gender. The hypothesis was investigated by using panel data on education, the Armed Forces Qualification Test (AFQT), father's education, and wages for young men and their siblings from the National Longitudinal Survey of Youth. The empirical implications of statistical discrimination on the basis of race were also examined. The analysis results supported the hypothesis of statistical discrimination; however, they were inconsistent with the hypothesis that firms fully utilize the information on race. (Thirty-eight references and 10 tables are included. Appended are the standard deviations, minimum and maximums of the variables used in the analysis and data on the relationship among wages, schooling, AFQT, and parental education.) (MN)
Publication Type: Reports - Research
Education Level: N/A
Authoring Institution: Bureau of Labor Statistics, Washington, DC.; National Science Foundation, Arlington, VA. Directorate for Computer and Information Science and Engineering.
Identifiers - Assessments and Surveys: National Longitudinal Survey of Youth; Armed Forces Qualification Test