Is Presidential compensation linked to performance of Ontario’s colleges?

Abstract

This paper examines relationships between Presidential salaries and key performance indicators for Community Colleges in Ontario, Canada. This paper aims to determine the impact College performance has on executive salary levels. This study uses the annual Sunshine List to discern three years of data for Presidential salaries, up to and including the wage freeze in 2012, for a population of 21 colleges. Key performance indicators (KPIs) from surveys conducted by Colleges Ontario between 2010 and 2013, were used to assess lagged salary impacts. A number of linear regressions were conducted, with the dependent variable of most interest being Presidential compensation, and independent variables spanning the various KPIs, institutional size, full-time enrollment, region and the tenured years of the executive. Initially, compensation was found to negatively correlate with graduation rates, whereby compensation increases when graduation rates decrease. Through a Sobel-Goodman test, it was later discerned both were mediated by enrolment figures. Hence, a null hypothesis was found; specifically, that Presidential compensation does not significantly influence the performance of a college. With Ontario’s Colleges actively consulting the public on College CEO compensation, this paper provides an important foundation for those considering pay-per-performance mechanisms. To increase school performance, College boards may find their efforts limited should they only focus on the President. Above all, this paper highlights the need that CEOs of higher educational institutions may, through engagement and leading faculty, academic programming and research, indirectly enhance college performance, including student satisfaction. To the best of the author’s knowledge, this is the first study of its kind focussed on Colleges, and builds of only one other study to consider the nature of performance-based compensation in Ontario’s broader public sector empirically.