Malaise, slump, deadwood -- there are lots of words for what supposedly happens to professors’ research outputs after tenure. A forthcoming study in the Journal of Economic Perspectives doesn’t use any of those terms and explicitly says it must not be read as an “indictment” of tenure. But it suggests that research quality and quantity decline in the decade after tenure, at least in economics.
The authors of the paper -- Jonathan Brogaard, an assistant professor of finance at the University of Washington at Seattle; Joseph Engelberg, professor of finance and accounting the University of California, San Diego; and Edward Van Wesep, associate professor of finance at the University of Colorado at Boulder -- started with a question: “Do academics respond to receiving tenure by being more likely to attempt ground-breaking ‘homerun’ research and in this way ‘swinging for the fences?’”
After all, they wrote, “the incentives provided by the threat of termination are perhaps the starkest incentives faced by most employees, and tenure removes those incentives.” (The question is sure to annoy academic freedom watchdogs. In the authors’ defense, they do cite the benefits of tenure, including job stability’s potential to encourage risk taking.)
Looking for answers, Brogaard, Engelberg and Van Wesep collected a list of academics who worked and were tenured in economics or finance departments at 50 top-ranked institutions at any time between 1996 and 2014. The final sample included 980 professors, all of whom were tenured by 2004.
Next, the authors considered two variables in the years before and after each listed professor received tenure: their overall number of publications in 50 prestigious economics and finance journals and their number of “homerun” publications therein. The paper defines the latter as being among the 10 percent most cited of all publications in a given year; about one-seventh of the publications considered in the study qualified as home runs. Those variables are stand-ins for a professor’s academic effort and degree of risk taking, according to the study, since widely cited, “highly influential output” is “presumably more likely to result from risky ventures.”
Both variables had values that peaked at tenure and declined thereafter, according to the study. On average, the number of annual publications fell by approximately 30 percent over the two years after tenure was granted and by an additional 15 percent over the next eight years.
Home-run publications also fell by 30 percent within two years of professors earning tenure and by an additional 35 percent over the next eight years.
Combining these facts, the study says, “we find that not only do both the overall publication rate and the homerun rate fall, but the likelihood of a given publication being a homerun falls by approximately 25 percent during the 10 years following tenure.”
Conversely, papers in the bottom 10 percent of citations were published more frequently in the years following tenure than in the tenure year.
As for why risk taking appears to decrease after tenure, the authors say that professors might start to add co-authors to their publications, work less at promoting their work at conferences or work on easier topics that may be published by good journals but have less impact.
The paper also considers whether professors increase their nonresearch service work after tenure or whether they branch out into new subject areas. But neither fully explains the initial findings, the authors say, in part because high-impact work output would presumably increase as professors doing more service concentrated their research efforts.
“This paper does not evaluate the broad and multidimensional case for and against tenure,” the authors say. “But it does suggest that at least for economists, tenure is not providing incentives to undertake research in the same quantity and quality that led up to the tenure decision.”
Numerous studies support the economists' general conclusions about tenure. Yet a number of recent studies also challenge the "dead weight" stereotype, such as a recent paper suggesting great variability in peak research activity among individual scientists, despite what their aggregate productivity curve shows.