The claim of a widespread incidence of “deadwood” in academia has always been popular, not only with the general public (largely comprising individuals who are not really familiar with academia), but also within academia itself, especially among graduate students, postdoctoral associates and even (less frequently) faculty and administrators. I myself, when I was a postdoc, was convinced of the reality and seriousness of the problem.
What is deadwood, anyway ? According to Wiktionary, one of the definitions of deadwood is people judged to be superfluous to an organization or project. In other words, deadwood is any individual from whose presence the workplace does not benefit in any concrete, factual way, whose contribution to the overall enterprise is intangible.
Is there a lot of that in academia ? Well, if one listens to some, one would think that being a piece of deadwood is almost a pre-requisite to be a scholar. Such a claim is typically made with little or no substantiation whatsoever, as if identifying deadwood were straightforward, i.e., deadwood were self-evident.
Funnily enough, however, if one searches through the literature, one finds that, while its existence is always acknowledged and lamented, the extent of the deadwood problem is always assessed in the few percent  — hardly enough to make it sound as an endemic that is eating academia alive, or in any case affecting it to a far worse degree than other professional environments .
Of course, many a proponent of the pervasiveness of deadwood dismiss such figures as hugely underestimating the problem, often countering them with much higher figures (of unclear origin other than personal observation). In one of her comments to my post, Mad Hatter makes the suggestion that perhaps this apparent disagreement has to do with different definitions of “deadwood”. I agree in principle, but it seems to me that nobody performing effectively one or more of the functions required by an “organization or project” should ever deserve to be called “superfluous”, and have a label of “deadwood” attached. This is especially the case when those functions, while perhaps not the primary ones, are still mission critical, and require that a faculty perform or oversee them.
It is true, as Mad Hatter observes, that at a research university faculty are expected to establish and maintain a vigorous, internationally recognized research program, capable of producing novel and significant results (thereby giving prominence to the institution), with an important educational component for graduate and undergraduate students, as well as of attracting extramural funding.
Any tenure-track faculty at a research university had better have this base covered, if (s)he is aiming at putting forward the strongest tenure bid. Excellence in teaching and effectiveness in service will not compensate for a deficient research portfolio. Even after tenure, excellence in research will be the decisive factor in promotion to full professor, as well as merit-based salary increments. But does that mean that any tenured faculty near retirement, whose research effort is dwindling, or no longer research active, should automatically be regarded as “deadwood” ?
No, it does not. Not by itself, anyway.
Even though research is the main focus, teaching and service are also part of the profession. They are both crucially important pieces of the mission of the institution. To this aim, it might be useful to remind ourselves of the following:
Neither can or should be eliminated.
To the extent that it is possible, neither should be assigned to personnel outside the faculty.
Both can be very time-consuming, and a huge energy drain.
Neither is particularly highly rewarded, especially at a research university.
Neither has a way of magically taking care of itself.
The notion that each faculty, probationary or tenured, should be engaged in equal measures in research, teaching and service is either disingenuous or surprisingly naive. Anyone who has taught in a single term one or more sections of large, introductory courses (each one with, say, two hundred of students, or maybe with no teaching support in the form of graders, for example), or served as undergraduate advisor, or has been in charge of departmental teaching assignments, or served as department chair, knows perfectly well what a significant setback each one of these tasks can deal to one’s research program.
Probationary faculty asked to perform any of the above functions would object, rightfully requesting to be relieved from them (to the extent that that is possible), in order to spend the bulk of their time on building their research program. In fact, I think I can comfortably state that at any research institution, a decision on the part of a department chair of inflicting upon a probationary faculty a major piece of committee work, for example, or a heavier-than-average teaching load, would be regarded as odd, and raise eyebrows among faculty and administrators .
By the same token, however, calling “deadwood” a senior faculty who, for example, consistently teaches the most onerous courses and/or picks up the most time-consuming service assignments, hardly seems fair.
The fact is, in most professions the focus of one’s activity naturally shifts at different stages of one’s career.
 Just kidding — we all know that academia is by far the worst place to be. It is so obvious that no arguments should ever be brought forward to substantiate this claim. In fact anyone who insists with requiring some quantitative evidence must be … well, deadwood.
 It is quite common for a probationary faculty undergoing annual evaluation, to be reminded by the college’s reappointment, promotion and tenure committee, of the importance of scholarly research, and warned about spending too much time with committee work.