The National Bureau of Economic Research this month issued a working paper  containing a preliminary report on a study of the learning outcomes of students in courses taken during their first term at Northwestern University. The study considers an impressively large sample, "15,662 students taking 56,599 first-quarter classes" and, its authors claim, offers clear statistical evidence that the students learned more in courses taught by non-tenure-track faculty members than in courses taught by tenured and tenure-track professors.
Not surprisingly, the study received extensive coverage in higher education publications like Inside Higher Ed  and The Chronicle of Higher Education , as well as attention from the mainstream media . Unfortunately, that coverage tended to misrepresent the study’s findings by claiming it shows either that "adjuncts are better teachers" or that "tenured faculty are worse." A close look at the working paper yields no evidence that the study demonstrates these things in a broadly generalizable way.
What thereby shifts into the background — though it does not go unmentioned — may, in fact, be the most important finding reported in the paper, that this successful cohort of "non-tenure-track faculty" were not short-term temps, but rather long-term employees. Admittedly, it was also downplayed by the study's authors. They remind their readers that Northwestern is an elite institution, and that "its ability to attract first-class non-tenure-track faculty may be different from that of most institutions." But the only details they give about these faculty appear in a footnote, which tells us only that "[a]lmost all classes taught by non-tenure-track faculty at Northwestern are taught by those with a longer-term relationship with the university."
Why is this so significant? In the first place, because the paper strongly hints that the study may call into question the applicability of a number of previous studies showing that the widespread use of adjunct faculty negatively impacts students at many types of institutions. The authors are careful to say that their work bears only upon research institutions, where tenured and tenure-track faculty members have significant research duties, but rhetorically this qualification merely serves to reinforce the suggestion that tenured and tenure-track research faculty are inferior teachers.
On the other hand, if we recognize that the contrast being drawn actually involves two groups composed principally of long-term employees, it becomes considerably more difficult to account for the differences between them on the basis of "tenure."
But there is another feature of the "non-tenure-track" cohort that the paper only alludes to in passing: they are full-time employees. When queried by my colleague at NewAPPS, Eric Schliesser, the paper’s lead author, David Figlio, made this explicit:  “we are comparing long-term full-time lecturers versus tenure-track professors.” Moreover, "these lecturers have long-term contracts and the same benefits as tenure-track faculty."
In his e-mail, Professor Figlio congratulates Eric and me for "picking up on" this. He is being far too generous. From the working paper alone, I found it impossible  to determine with any confidence the conditions under which these faculty members were working. Having clarified this, it becomes evident how wrong much of the initial reporting on the study has been. Far from showing that they are better teachers, it provides no evidence whatsoever about the effectiveness of “adjunct” faculty in the more common sense of part-time employees hired to teach on a course-by-course basis. Nor can this be easily blamed on lazy reporters.
By describing the "non-tenure track" cohort in vague terms, and by making multiple references to studies of “adjuncts” in its introduction, the rhetoric of the working paper does a great deal to sow confusion on this point, especially among those unfamiliar with the nuances of academic hiring terminology.
So what does the study establish? Broadly, it shows that full-time, relatively stable and (presumably) well-compensated non-tenure-track faculty do well in the classroom.
Specifically, the paper outlines evidence for two conclusions: 1) that students "learn" better during their first term at Northwestern in classes led by non-tenure-track faculty, and 2) that students taking classes with non-tenure-track faculty are more likely to take another course in the same discipline.
"Learning" is operationalized here in terms of grades earned by students in a subsequent course taken in the same discipline. This, it is suggested, offers a good way to measure whether "instructional quality has a lasting impact" or "deep learning."
The most precise comparison listed shows the difference between the two groups of faculty to be less than a 0.1 grade point (gp) improvement, with slight variations for classes within and outside a student's intended major. Relaxing the statistical controls makes the effect appear to grow to around 0.2 gp.
So there is a consistent, though modest, improvement relative to faculty on tenure lines. This is a genuinely interesting result, and it would obviously be valuable to understand how it was achieved. But we cannot do so without knowing more about the working conditions of those being compared.
How many courses per semester are those in each group expected to teach? How many times have they taught the courses being studied? How recently and how often do they do so? How large are their classes? Do they differ materially in format? And, of course, what other responsibilities are they expected to balance?
All of these factors, were they found to consistently vary between the two groups, could plausibly be seen to contribute to the improvements in student learning the study identifies. Indeed, it is far more likely that they should be doing so than the bare fact of not being on a tenure line — especially with regard to full time faculty who are long-term employees.
Sadly, while it goes into considerable detail regarding differences among various groups of students, the working paper provides no information about these matters bearing upon faculty. Worse, it does not appear that the authors have even conducted much analysis of these factors. When asked about the effect of variations in class size, in particular, Professor Figlio acknowledged to Professor Schliesser that he and his co-authors had not yet looked into this factor.
All of which leads one to wonder why the authors have chosen to circulate a vaguely framed paper based on such incomplete results. It is difficult to avoid the conclusion that this is a rhetorical exercise. But if so, what do these authors intend to suggest?
As usual in these instances, it is helpful to look at their conclusion, the final two sentences of which read as follows: "Our results provide evidence that the rise of full-time designated teachers at U.S. colleges and universities may be less of a cause for alarm than some people think, and indeed, may actually be educationally beneficial. Perhaps the growing practice of hiring a combination of research-intensive tenure track faculty members and teaching-intensive lecturers may be an efficient and educationally positive solution to a research university’s multitasking problem.”
Especially read in the context of what preceded them, these statements involve a remarkable amount of equivocation and deliberate obscurity.
In the first place, while people are very much alarmed at the growth in the ranks of adjuncts, there is less concern about full time, long-term teaching faculty. And where questions do arise, they tend to have more to do with the welfare of these faculty or the institutional devaluation of research than with their educational effectiveness. In which case, a study showing that such faculty are effective teachers answers a question that virtually no one is asking.
Why, then, make such a big deal of answering it? Because, as the media and everyone else recognized, the answer suggests questions about the pedagogical effectiveness of tenured faculty.
Indeed, it does so despite the paper’s careful qualifications on this precise point: "the evidence that non-tenure-track faculty produce better outcomes may not apply to more advanced courses.”
It may not, but no matter. The seed of doubt has been planted. This leads us to the last sentence, which is an equivocal tour de force. At the end of a report on a study contrasting non-tenure-track to tenured and tenure-track faculty, we are told that hiring "teaching-intensive lecturers" in addition to "research intensive tenure-track faculty" is an "efficient" solution to the problems facing administrators at research institutions.
Notice how the operative distinction has shifted to one between teaching and research-intensive faculty — but without quite disconnecting it from the issue of tenure.
Given this shift, the claim may be true. But there is nothing about teaching-intensive faculty that is incompatible with their being eligible for tenure — especially if one fully intends to build long-term relationships with them and keep them around.
Why, one is led to ask, can we not have "efficiency" and tenure? The answer, if there is one, must have to do with other ways in which non-tenured faculty differ from those with tenure. The authors mention academic freedom — an important consideration.
But they otherwise ignore the degree to which non-tenured faculty lack a secure position from which to question, criticize, or oppose the actions of university administrators.
And here, indeed, is another sense of "efficiency" that administrators at many institutions might well wish to cultivate, allowing them to enjoy a pedagogically effective, but largely vulnerable, and therefore easily controlled faculty.