One of the fastest ways to get into an argument at a university is to suggest that there is some necessary trade-off between research and teaching.  Really.  The words will hardly be out of our mouth before someone comes charging at you, claiming the opposite.  It’s not an empirical argument or anything, but an article of faith.  Frankly, you wouldn’t want it to be an empirically testable position because if it was, someone might start asking some pretty difficult questions.

There have been attempts to quantify this relationship but these attempts are not super-high quality mainly because there are multiple possible different measurements of “research” and “teaching quality” and in the case of the latter, measures are neither plentiful not uncontested.  Not surprisingly then, studies come up with wildly different correlations depending on (among other things) the methodology used.   The last reasonably broad synthesis of data that I know of was done about fifteen years ago in the UK by the Department of Education and Skills (available here), which reviews a very large number of studies. It notes that most studies suffered from deep methodological challenges and inconsistencies which, even if you accepted each of them as valid, makes a nonsense of most meta-analyses.  That said, the review found that the preponderance of studies showed either a null or positive relationship, and that the average relationship was very modestly positive, at about 0.1. 

I mean, I suppose you could call that a victory for the “no trade-off” position.  But you’d have to be straining pretty hard to do so.

The thing is, even if there are some complementarities between research and teaching, there are still hard trade-offs.  Time, for instance.  An hour in the laboratory or the archives is an hour not spent on pedagogy.  So yes, an hour of research might improve teaching, but does it improve it as much as a more direct application of talent to the task at hand?  Or, take money.  Being a serious research university requires you to spend tens if not hundreds of millions of dollars every year in construction and maintenance of laboratories.  That’s money an institution is not spending on shrinking class sizes.  Is all that money worth your (methodologically suspect) 0.1 correlation?   

And then there is the fact that the actual relationship between teaching and research is probably non-linear.  Below is a graph produced by William Massy, one of the giants in the field of economics of education.  It’s a sensible way of visually suggesting the relationship between teaching and research at the institutional level – up to a certain point, they are complentary, but beyond it they are substitutes.  This is, I think, a pretty accurate way to describe things, and it mostly works at the level of the individual faculty member, too.  

In fact, this is such a reasonable depiction of the true relationship between teaching and research I suspect the no-trade-off crew would probably agree with it as well.  And in fact, therefore, what I suspect the no-trade-off crew is saying when it rages against trade-offs is “we here at institution X are at the apex of this curve”.   And maybe that’s true.  

But how would they know, exactly?

I mean, look, if being at the apex was really a priority, then deans and department heads would be managing staff to make sure that on aggregate, the balance between research and teaching was at the apex and not too far to one side or the other.  And provosts would be managing faculties to reach an institution-wide apex.  And Ministries or an intermediary body would be managing across institutions to achieve a system-wide apex.  But of course, no one does this.  No one.  What we have instead is a system where nearly all the status and financial rewards are on the research side, and so there is a perpetual race to the right side of the curve, all the while claiming that there is no trade-off even though it is almost certainly the case that at some places the apex point is a long way in the rear-view mirror.