It is my purpose here to join the current debate about management compensation, giving special emphasis to the tournament theory explanation of high CEO compensation. Except for work on tournament compensation theory, most current studies of management compensation set aside questions about the level of CEO compensation. They consider instead questions about the sensitivity of CEO compensation to firm performance [Lewellen (1971); Jensen and Murphy (1990); Murdoch (1991)]. Their technique, consequently, has been to investigate the statistical association between changes in firm equity value and changes in CEO compensation, using the former as a measure of CEO productivity. The present study examines the level of CEO compensation and the structure of compensation within the management hierarchy, but a few words may be said about the sensitivity of CEO compensation to firm performance.

Some issues related to the sensitivity of management compensation to firm performance

The discussion to this point should have revealed just how difficult it is to measure productivity. Certainly, firm performance is an important aspect of this measurement insofar as top management productivity is concerned, but gross oversimplifying is involved in studies that take stock price measures of performance so seriously as the sole arbiter of CEO productivity. The working assumption used by the more important of these studies is that a dollar change in performance should translate to a dollar change in CEO compensation if proper incentives are to prevail. The assumption strains credulity. Interdependence between the productivities of team members, including leading shareholders and members of the board, argues against a dollar-for-dollar translation. But beyond these considerations is another.