Longitudinal Analysis of Salary Discrimination in Higher Education

Abstract

Legal and statistical issues associated with the use of multiple regression models in faculty discrimination cases in higher education are presented in this paper. Faculty salary models as a function of gender, rank, tenure status, race, academic discipline, and age variables are analyzed in a longitudinal study covering three years (1982-84) at the University of Northern Colorado (UNC). Declining student enrollment during the period saw the size of the faculty drop from a high of 492 in 1982 to a low of 380 in 1984. Results of the exploratory data analysis indicate declining roles for gender, race and age variables in explaining salary differences. While the contribution of academic discipline variables in the regression models was statistically significant, results seem consistent with institutional salary policies which were in effect at each point in time.