Estimating Faculty Salary Distributions: An Application of Order Statistics

Abstract

A knowledge of the distribution of income in different professions has a variety of policy implications and uses. Important applications include being able to make competitive salary offers for potential employees as well as helping to retain values employees through appropriate financial incentives. In some professions, salary data are readily available, and in others, data are limited. In academics the amount of salary data available is very diverse. In one study, Oklahoma State University publishes the highest, lowest and mean salaries by rank and discipline for different Carnegie Research Classifications. This paper outlines and approach, using order statistics and a Burr distribution, to utilize the information available from the Oklahoma State Survey to estimate underlying probability density functions for salary distributions. Given the estimated distribution, deciles can be estimated which facilitate an analysis of salary structure. The methodology is applied to estimate the salary distribution of statistics professors.