Are Nontenure-Track Faculty Worse Teachers? The Short Answer Is No.

Our guest author today is Ian Robinson, Lecturer in the Department of Sociology and in the Residential College's interdisciplinary Social Theory and Practice program at the University of Michigan.

Critics of higher education’s growing reliance on nontenure-track (NTT) faculty for undergraduate teaching routinely assert that NTT faculty are inferior teachers, and, therefore, that the quality of undergraduate education is deteriorating. This is true even of critics such as Marc Bousquet, the author of How the University Works (2008), who see themselves as friends of exploited NTT faculty and supporters of efforts to organize them into unions.

I think that these critics are wrong, and that their error has two important negative consequences: first, it devalues the work that NTT faculty do; and second, it impedes our understanding of one of the major successes of the “neoliberal” model - that it has been able to introduce a two-tiered faculty system in which many newer faculty are paid half or less of what the top tier is paid per class, without dramatic decline in the quality of undergraduate education that would de-legitimize the two-track system.

To understand how this has been possible – and where the critics go wrong – we need to start by asking: What determines teaching quality? I would suggest that there are five major determinants:

1. Understanding of the substance of the material being taught;
2. Pedagogical skill;
3. Working conditions (e.g., does the university supply faculty with an office and a computer, are they commuting between two or three campuses, etc);
4. Teaching load (i.e., how many courses per term and how many hours of faculty time are devoted to each course);
5. Teacher motivation and commitment.

But the critics err in generalizing from this particular case to the entire class of nontenure-track faculty. For when we are dealing with NTT faculty who are employed full-time or part-time at one institution, then working conditions are much less bad and the teaching load can be manageable.  And if the negative impacts of (3) and (4) are reduced, the other three factors do not necessarily result in inferior undergraduate teaching. Regarding knowledge of the subject matter (number 1), NTT faculty will seldom have the detailed grasp the material that a TT faculty member has in their area of research expertise.  However, NTT teachers often grasp just as well the fundamentals that are the core of what is conveyed in an undergraduate course, and they may be better at conveying that core material because they specialize in teaching.

Max Weber put the point well in “Science as a Vocation”:

One can be a prominent scholar and at the same time be an abominably bad teacher…. [T]o present scientific problems in such a manner that the untutored but receptive mind can understand them and – what is for us decisive – can come to think about them independently is perhaps the most difficult pedagogical task of all…. [T]his very art is a personal gift and by no means coincides with the scientific qualifications of the scholar.

Are other things equal? Many point to the fact that virtually all TT faculty have Ph.D.s while only some NTT faculty do. The first thing to say about this is that a growing share of NTT faculty have Ph.D.s because the production of Ph.D.s has not diminished while the TT share of all higher education jobs has decreased. The result is more Ph.D.s going into NTT slots. On the University of Michigan’s Ann Arbor campus, for example, 47% of Lecturers have PhDs, 60 percent do scholarly research in their field, and 51 percent publish the results of this research in academic journals.

But the deeper point is that Ph.D. education typically has little to do with learning how to teach well. Ph.D. candidates may serve as teaching assistants in a few courses taught by a professor, but the process of learning how to put together and teach one’s own courses usually begins when one secures his or her first faculty job.

Others argue that the low status and unequal treatment of NTT faculty are bound to demoralize them, resulting in low levels of commitment not only to the institutions that employ them, but to the students they work with as well. This argument would be more compelling if faculty were economically rational. Few are.

Two factors generate much higher commitment to quality among NTT faculty than  we would expect if commitment levels were a function of compensation and status levels: (a) our sense that we are professionals, and that as such we must hold ourselves to professional standards of performance; and (b) our concern for the well-being and development of our students.

Treating professionals poorly does have a negative impact on loyalty to and respect for the employer, particularly those in the administration  making the decisions that result in this poor treatment. But loyalty and respect of this sort are not necessary to motivate high quality performance where factors (a) and (b) are in play.  That is why the quality of undergraduate education has not plummeted with the widespread increase in NTT faculty.

Evidence in support of these arguments comes from the Ann Arbor campus of the University of Michigan (UM-AA). This is a good place to test the arguments made above because most NTT faculty (about 60%) are full-time and long-term (the average Lecturer at UM-AA has been teaching there for 9.2 years, while one third of Lecturers have been there for 20 years or more). Working conditions are generally good in that most NTT faculty (called “Lecturers” at UM) have (shared) offices and other basics required for preparation, meetings with students and so on.

Moreover, while the typical full-time Lecturer teaches twice as many courses as the typical TT faculty member (3+3 versus 2+1), this teaching load is not so onerous that Lecturers cannot put the time needed to produce a high quality result in all of their courses. Indeed, survey data indicate that, on average, Lecturers spend about 6 more hours per course than TT faculty.  They are able to do this, despite their higher teaching load, because they do not have the same research and service responsibilities as their TT colleagues. TT faculty, on average, spent only half their time on teaching, and three times more time per week on service activities.  In other words, the working conditions and teaching load facing NTT faculty at UM-AA do not undermine the quality of teaching.

This implies that at UM-AA, it is the other factors identified above – substantive understanding of course subject matter, pedagogical skills, and levels of commitment to high quality teaching – that will be the main determinants of any differences in the teaching quality by faculty type. For the reasons noted above, these factors do not point to lower NTT faculty teaching performance.

How can we assess this argument?  We need  ways of comparing the quality of TT and Lecturer teaching at the UM-Ann Arbor.  In our recent report, Teaching Quality: What the Principle of Equal Pay for Equal Work Means for Lecturer Pay at the University of Michigan (2012), available on-line here, we report the results of three such measures.

The first measure is the quantitative student evaluations that every student in the School of Literature, Science and the Arts (LS&A) is required to complete. In their study of ten U.S. universities, Off-Track Profs, Goldenberg and Cross (2009, 124) found that none had ever done a systemic comparison of the performance of TT and NTT faculty as measured by student evaluations. However, the authors were able to draw upon their LS&A connections to generate University of Michigan data for that school. They found that in every department for which comparable data were available, Lecturers had significantly higher average student evaluation scores than TT faculty.

Our second measure of quality also relies on students’ subjective assessment of their learning experience, but does so in a very different way. Our survey of UM students conducted in the Fall of 2011 asked UM-Ann Arbor Sophomores, Juniors and Seniors to identify the best course they had taken at UM to date.  Students were then asked to provide the name of the faculty member who taught these courses. We used the UM web Directory to determine whether the faculty so identified were tenure-track or Lecturers. We were able to identify the faculty for 1,608 (or 83%) of the 1,944 courses for which a faculty name was provided by the student. Lecturers accounted for 37.3% of these courses, while they accounted for about 32% of all Student Credit Hours (SCHs) in LS&A undergraduate courses. Thus, Lecturers were disproportionately likely to teach what the students considered the best course they have taken at the University of Michigan.

Our third and final measure of quality looked at 11 types of teaching and/or evaluation methods (“non-banking methods”) that could be used to move courses away from simple memorization and regurgitation methods: exams with one or more essays as an important component; shorter essays over the course of the term; a term research paper; multiple drafts of written work; oral presentations; group projects; student evaluations of others’ work; performance, reflection, and critiques; entry and exit surveys; student journals; lab and/or field reports; and reading responses.  

No course will employ all of these methods, but most courses employ a number of them. The more of these non-banking methods that faculty employ, the broader, deeper and more enduring the learning experience is likely to be, for two reasons: first, the effect of each innovative teaching technique is cumulative; and second, the different approaches call upon, and help to develop, different student capacities.  

Our survey data showed that the larger classes become, the fewer of these more labor-intensive methods are employed, regardless of faculty type: classes with 25 students or less employ almost twice as many of these methods as classes with more than 150 students. However, it is also true that, taken as a group, Lecturers make more use of non-banking teaching and evaluation methods than do their TT faculty counterparts in two out of three of the class size categories. We found that in classes of 25 students or less, Lecturers employed an average of 4.6 of these methods, compared with 3.8 for TT faculty.  For classes with 26-150 students, the corresponding figures were 3.7 and 3.3, respectively.  Together, two types of course accounted for 96.5% of all classes taught by the faculty in our survey sample.

These findings make sense when we remember that Lecturers spend about a third more time per course than TT faculty. The main reason they take more time is that they are using more labor intensive (and generally superior) teaching and evaluation methods (e.g., write an essay on this topic as opposed to answer a multiple choice questionnaire).

The above measures do not come close to capturing all of the types of student learning outcomes we would like to compare.  But the evidence considered is clear on at least this much: it offers no basis for claiming that Lecturers’ teaching is inferior to that of the TT faculty. On the contrary, the preponderance of the evidence points to somewhat higher quality teaching by Lecturers, though the differences are not large and may not always be statistically significant.  The burden now lies on those who believe that increased reliance on NTT faculty necessarily lowers the quality of undergraduate education to provide evidence – as distinct from a priori arguments – to support their conviction.

Suppose that the arguments and evidence advanced here are correct. So what? One important implication is that we must stop talking about nontenure-track faculty as though it were a uniform category about which meaningful generalizations can be made on subjects like teaching quality. On any given campus, there is likely to be much more heterogeneity within the category of NTT faculty than there is within the category of TT faculty.   It seems unlikely that NTT faculty who fit the description of “freeway fliers” can provide the same quality teaching that other faculty can.   But that is not because they are NTT faculty; it is because of the appalling conditions under which they work.   And the share of all NTT faculty that work under these conditions varies greatly among schools, campuses and universities.

A second important implication of the argument advanced here is that inequalities in Lecturer pay per course cannot be justified by differences in the quality of the teaching done by the two types of faculty. TT faculty at UM-AA are, on average, paid twice as much per course as NTT faculty, and the gap is even larger in many other places, so this is no small matter.  What about the radical differences in rights and status that exist between the two types of faculty?  If teaching were the most important thing done in higher education, equal quality teaching would also make it hard to justify these other inequalities. But some universities, like UM-AA, tend to see research as more important than undergraduate teaching. In my next blog, I’ll interrogate the defensibility of that position.  

- Ian Robinson