If Oklahoma’s higher education costs are going to be contained and eventually reduced, the key is improved productivity, meaning doing more with less.

Unfortunately, it appears Oklahoma has done the opposite. An examination of the teaching faculty at the University of Oklahoma (OU) and Oklahoma State University (OSU), for example, suggests that a small proportion seems to do most of the work. Large numbers of faculty carry modest teaching loads, yet also have modest research accomplishments. If the bottom 80 percent of the faculty taught as much as the top 20 percent, universities could operate with demonstrably fewer faculty members and reduce tuition costs dramatically.

In our 2014 OCPA report “Dollars and Sense: Assessing Oklahoma’s Public Universities” (from which this article is adapted), we asked: are there some teachers who teach vastly more than others? An analysis of credit hours at the two premier institutions show the answer is clearly yes. At OSU, the 20 percent of faculty teaching the greatest number of credit hours taught 54.49 percent of the total. The bottom 40 percent taught fewer than 10 percent of credit hours. Part of this might reflect some faculty performing administrative or other tasks that detract from their teaching. But the top one-fifth of the teaching staff teaches more than the other four-fifths combined. If the top 20 percent, representing well over 300 professors and other instructors, can successfully provide a majority of OSU’s instruction, the question arises: are many of the large number of professors with low teaching loads earning their salaries, particularly given the institution’s modest level of success in obtaining external research funding?

The picture at OU is even more dramatic. Student credit hours per instructor average a very low 143. Assuming students typically take three-credit-hour courses, there are about 48 students per professor per year, or 24 per semester. Again, because of the inclusion of part-time faculty, that probably understates typical teaching loads, but it suggests that there are either a lot of small classes, or teachers have low teaching loads, or a combination of both.

Moreover, at OU the disparities in teaching loads are huge—even greater than at OSU. The top 20 percent of professors by teaching loads teach 60 percent of the total—averaging well over 400 credit hours each—implying nearly 140 students taught annually, assuming three-credit-hour courses. A relatively small number (505) of the faculty do most of the teaching, while another 2,018 listed as having instructional responsibilities do far less of the total classroom teaching.

The cost of instruction depends on teaching loads, class size, and faculty salaries. Combining all three factors, we obtain a “cost per student credit hour” for each member of the instructional staff. At OU, there were 2,523 reported instructional staff in 2012. Collectively they were paid $182,318,993—an average of $72,263 for each faculty member. Some of those faculty members were probably part-time teachers.

We identified the 505, or 20 percent, of these faculty members whose cost per student credit hour was the lowest, reflecting large classes, large teaching loads (teaching many classes), and/or low salaries. We then asked the question: what if the other 2,018 faculty taught at the same cost as these “most productive” or “most efficient” 505? What would have been spent on instruction for the actual number of student credit hours generated? Implied in this exercise is the idea that the school could reduce the size of its instructional staff substantially. What, however, would have been the reduction in instructional costs? The answer: $100 million (actually, $99.577 million). That is an amount equal to well over $3,000 per student.

Perhaps more realistically, let us assume that the entire faculty at OU was twice as costly (on a student credit hour basis) as the top 20 percent. Even in that circumstance, the savings in terms of instructional costs would be $53.3 million—close to $2,000 per student. In other words, there are tremendous variations in student credit hour costs between faculty members, costs that the students ultimately cover through tuition. If the school were to lower instructional costs per student hour by a combination of increased teaching loads, larger classes, and, perhaps less realistically, lower salaries, major savings could be incurred.

Let us repeat the analysis in the previous three paragraphs for OSU. OSU is a somewhat smaller institution, with total instructional costs almost 30 percent less than at OU. Nonetheless, the disparities in student credit hour costs are, if anything, greater at OSU. We estimate that if all faculty members had student credit hour costs equal to those of the lowest 20 percent, the school could save almost $82 million on staffing costs, roughly $3,000 per student. With $82 million, the university could reduce tuition fees by over one-third. If OU and OSU simultaneously achieved the maximum savings outlined (all faculty emulating the top 20 percent in terms of costs per student credit hour), total savings in the state would exceed $181 million. Expanding this to all state universities would almost certainly increase the number.

Would not a move to achieve that objective jeopardize OU’s and OSU’s research mission? It is important to remember that of the non-adjunct faculty, about 15 percent at OU generated all external research grants in fiscal year 2012. Put differently, 85 percent of the faculty received no external research funding. The top 10 percent of faculty in terms of grants generated 96 percent of all external research grants. OU and OSU are not, despite any claims to the contrary, major institutions regarding externally funded research. Lowering instructional costs via higher teaching loads and/or larger classes could achieve material results even if those receiving external research funding were excluded from new teaching rules.

The major purpose of this exercise is to suggest that big dollars are involved in supporting faculty members who have relatively high costs per student credit hour, either because they teach little, have small classes, or in some cases, are paid huge sums of money. We identified literally dozens of employees at OU and OSU paid very large salaries, say, more than $250,000 a year. For some, the compensation levels seemed to us extremely high, given their qualifications and/or observable contributions to their institution.

While valuable gains in savings and productivity can be found by addressing faculty issues, key spending areas such as administrative spending and bloat must also be examined. Nationally, higher education staffing per student declined about 10 percent from 1999 to 2011, suggesting probable productivity improvements; however, staffing per student rose about five percent in Oklahoma, suggesting probable productivity decline. Most of the staffing increase came in non-faculty areas, likely in administration.

Have universities lost their way, losing sight of their core mission to educate undergraduate students? Are they excessively interested in maximizing the gains to the staff, rather than educating students? Are too many faculty members teaching too few students too little, often in courses on obscure, specialized subjects of little general interest? Our research suggests that many areas of higher education in Oklahoma can be improved so as to enhance productivity, efficiency, and accountability.