An Analysis of a Small, Closely Observed Labor Market: Starting Salaries for University of Chicago M.B.A.'s

Abstract

This paper is an initial report on a continuing study of the behavior of a small but very interesting labor market-the market for new (M.B.A.) graduates of The University of Chicago Graduate School of Business (UCGSB). Obviously, salaries of business executives are of substantive interest to economists. But beyond this, the data in the records of UCGSB's placement office afford a rare opportunity to study the interrelations among salary, ability, prior work experience, job location, and various employer characteristics as these appear in a very well-structured labor market. In a later paper it is hoped to compare the determinants of starting salaries for This paper analyzes differences in starting salaries of UCGSB M.B.A.'s, 1971-74. Starting salaries were regressed upon a variety of characteristics, both of M.B.A.'s and employers. The principal findings are: (i) starting salaries are positively associated with previous work experience and grade-point average; (ii) black males had higher average starting salaries, lower gradepoint averages, and more prior work experience than whites: (iii) female starting salaries were somewhat lower, owing mainly to smaller rewards for female work experience; (iv) jobs located in New York City paid higher than elsewhere, and jobs located outside the U.S. paid less; (v) accounting firms paid less, and consulting firms paid more, than others in the same location. * An empirical study of this kind always involves the work of many individuals in addition to its director-reporter. This is no exception. My research assistant, Michael Bradley, literally created the data base upon which this paper rests; he copied the numbers, coded, and keypunched them. In the early stages of this work he was assisted by John Bloom. In addition to creating the data set, Bradley also did all of the computations, in the course of which he made many important suggestions about concepts and statistical procedures. This study could not even have commenced without the cooperation and support of Richard J. Thain, the director of the placement office of The University of Chicago's Graduate School of Business, and of his administrative assistant, Mary Neal. Their continuing advice and encouragement have been indispensable. Many other members of the Graduate School of Business have given counsel and encouragement; Barry Plotkin and Lubos Hale have been especially helpful. Financial support was provided by a Faculty Research Grant from The University of Chicago Graduate School of Business and by the Charles R. Walgreen Foundation. This support is gratefully acknowledged.