NotesFAQContact Us
Search Tips
ERIC Number: ED504913
Record Type: Non-Journal
Publication Date: 2009-Feb
Pages: 12
Abstractor: As Provided
Reference Count: 3
The Madness of Weighted Mean Faculty Salaries
Micceri, Theodore
Online Submission, Paper presented at the Annual Meeting of the Florida Association for Institutional Research (Cocoa Beach, FL, Feb 25-27, 2009)
Higher education frequently uses weighted mean faculty salaries to compare either across institutions, or to evaluate an institution's salary growth over time. Unfortunately, faculty salaries are an extraordinarily complex phenomenon that cannot be legitimately reduced to a single number any more than the academic construct of skills, knowledge, intelligence and experience of a prospective student can be encompassed by a single test score. Unfortunately, USF's faculty used weighted means in 2004 (Faculty Senate, 2004) to conclude: "...Faculty compensation at USF is unacceptably low, placing us well below peers nationally and regionally, and ranking lowest among doctoral level institutions in Florida." This paper attempts to clarify some of the confusing and misleading aspects of such faculty salary analyses by providing direct comparisons across institutions and showing that the concept of a weighted mean salary for a large institution suffers from at least the following problems: (1) Weighted mean salaries reflect an ill-suited misapplication of an accounting technique designed to estimate the labor cost per widget from different hourly skill levels wage levels. Higher education's three widgets of education, research and public service result from different contributions by graduate assistants, adjunct, assistant, associate and full professors. Unfortunately, weighted mean faculty salaries reflect a simplistic weighting by rank proportions of only assistant, associate and full professors; (2) Weighted mean salaries are computed using proportions of faculty at rank across institutions. Unfortunately, different Doctoral/Research institutions have different proportions at the three ranks used, and full professors average 35-50% more than associates, who average 10-20% more than assistants. Thus, UF, with 42% full professors in FY 2007 will have a substantially greater weighted mean salary than USF (32%) even ifs have thee mean salary at rank; (3) Rather large salary differences occur among disciplines, which exacerbates the negative effects of different proportions of higher and lower salaried ranks across disciplines; (4) Mature research institutions such as UF employ greater numbers of highly paid distinguished professors and high cited scholars, which further confuses comparisons; (5) Point estimators like the mean well represent the center of a symmetrical, centrally dense and light-tailed distribution of values. Unfortunately, faculty salaries are a tri-modal and highly skewed distribution of values which invalidates a single global estimate; and (6) Finally, and perhaps most importantly, salaries are influenced by the local cost of living which can vary substantially from place to place within a state. To better explicate some of the six limitations described above, this paper conducts a few exemplary comparisons of, weighted mean faculty salaries. (Contains 2 tables, 1 figure, and 7 notes.)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: Higher Education
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Identifiers - Location: Florida