Our proposal is directed toward a set of questions bearing on the way in which the labor market substitutes skills of one type for another. Where workers are not perfectly substitutable, changes in their relative supplies will change their relative prices. To the extent that demographic shifts such as birth rate variations cause changes in relative quantities of skills, e.g. experienced versus inexperienced workers, they form natural experiments for estimates of substitution relationships. The rising participation of women workers and the changing educational attainment of the population also trace wage shifts that may yield measures of economy-wide substitutability between skill types. Our interest is not in substitution per se rather it is in finding ways to summarize the labor market impact of changes like the baby boom and the growth in female participation. Our approach is to view individual labor types (defined by age, experience, sex, education, race, etc.) as composites of a small number of basic skills. Rather than there being as many factors of production as there are classifications of workers, we model the separate classes as if they are distinguished only by different quantities of the same basic building blocks. Using 1968 to current CPS data, as well as the micro-files of the 1940 to 1980 Censuses, our empirical work will examine substitution by age, gender, and education. The pooling of time series of cross-sections will allow us to investigate important hypotheses about the extent of geographic substitution of labor--its mobility and sensitivity to local labor markets. Finally, our methodological studies attempt to understand the information content of the statistical experiment being performed when demographic shifts affect changes in relative wages. Our ultimate goal is to introduce concise descriptions of demand side considerations into the modelling of changes in relative wage.
