Being a college professor, especially at a research university, is a privilege. Faculty position openings are few and far between, and the number of qualified candidates high. The situation can be quantitatively assessed, for example, by the number of applications that are routinely received for tenure-track faculty positions in the United States, essentially at any major research universities (and even at four-year colleges with lesser emphasis on research).
Since it is a privilege, the relatively few of us who are lucky enough to have that kind of a job, have a moral duty to honor it, by keeping high standards and by engaging primarily in those activities that pertain directly to the basic mission of our institutions, namely research and teaching.
Any undertaking on the part of a faculty, not clearly classifiable as part of either research or teaching and/or not directly impacting either, should be therefore regarded as in principle extraneous to the scope of the profession — not part of the overall motivation put forward (e.g., to taxpayers) when seeking support for higher education.
The above is, of course, a statement of principle. It does not mean that the work day of the average university professor should be filled exclusively with either teaching or research. In fact, there are activities that cannot be directly identified as either, but are nonetheless necessary to the functioning of a department, a college, a university, and which require faculty involvement. These activities can be broadly classified under the label of service.
The most obvious example of a service contribution is the coordination, supervision and running of a program of study offered by the department of affiliation, either at the undergraduate or graduate level. While much of the day-to-day work can be performed by a non-academic, faculty must be involved in all decisions of strictly academic nature — e.g., what courses should be required of an undergraduate major ? How many different options and/or specializations should be offered ? What should be the course requirements for graduate students ? All of these decisions are arrived at through a never-ending process involving lengthy and substantive debates among faculty, experimentation, and a constant revisitation.
Other examples of tasks that clearly belong to faculty are the administration of entrance and qualifying exams, organization and running of departmental seminars, management of common research and teaching infrastructure, mentoring of junior colleagues, as well as occasional fund-raising and outreach initiatives.
There are also service tasks whose scope goes beyond the academic unit of affiliation, extending all the way to the College or University level. The most obvious example is the periodic (e.g., annual) evaluation of the performance of fellow faculty, but there are others .
All of the above simply must be done, which means that, at any given time, some of the faculty at a research university will be asked (by the university administration or by faculty colleagues) to set some time aside to engage in university service, e.g., serve as department chair, or as a faculty senator, or in some departmental or university committee, and so on and so forth. Naturally, in order to devote to these tasks the time that they require (often quite a lot), those faculty will have to subtract that time from their research and teaching activity. Therefore, even though all of the service tasks do ultimately, in the long run impact the research and teaching that is done by the university, because one is talking about using precious faculty time to do things that do not intrinsically belong with the stated mission of the organization, service work should always be regarded as a mean, never as an end.
To use the words of a colleague of mine, service is a necessary evil. It should be kept down to a minimum, and faculty who are asked to engage in it should be properly rewarded, as one must assume that they would rather spend their time doing that for which they sought to become professors in the first place, namely teaching and research.
By the same token, service tasks should not be assigned randomly, by rotation and/or with no consideration given to an individual’s ongoing, broad scholarly effort, but rather precisely with the aim of making the smallest possible negative impact on the overall research and teaching program of a department, or academic unit.
This means, among other things, that tenured faculty whose research program is slowing down, or being wrapped up (as assessed for example by the number of students and postdocs under their supervision), should be considered the most obvious candidates for service tasks; on the other hand, while some special circumstances may require it, a tenure-track assistant professor should generally be kept away from service [1,2].
Likewise, the number of committees on which faculty serve, should be constantly kept in check. No committee, no matter how cool or reasonably-sounding its name, should ever be created for which no clear mission can be defined right away; conversely, if the original mandate of a committee, i.e., the premise upon which it was established, no longer exists, or if for whatever reason the committee never seems to find the need to hold a meeting, it should be recognized as useless, and disbanded.
And, if at any give time, only a small fraction of faculty within a given academic unit are engaged in service, and the academic unit still manages to function smoothly, administrators should congratulate its head, not inquire as to why so many of the faculty have no service work assigned to them.
It has been my observation over the past few years, however, that universities may be re-defining (de facto, if not on paper) what the proper scope of service is, and may be trying to elevate service, especially in the evaluation of faculty, to a role comparable to that of teaching and research. In other words, a substantive “service component” is expected of all faculty, regardless of their different levels of teaching and research effectiveness.
That is simply absurd. If it is not opposed firmly by the faculty, it may in time have nefarious consequences. If you have not paid attention to this phenomenon, you might want to find out what is happening at your own institution. How many active committees does your department have ? How many associate vice chairs ? Go down the list, and if you find yourself thinking “Hmm, what is that one for ?” on one or more instances, there is a pretty good chance that those committees and those titles are, in fact, phony or useless, and that the chair of your department has created them just to produce some evidence of service work done by faculty (thereby making bean counters happy).
One may object: “What is the problem, man ? You make up some committees that never meet, no skin off anyone’s nose, the facade is saved, and everybody is happy.” Well, not so simple (Incidentally: what’s so great about that “facade” that is always worth saving ? Just wondering…).
First off, it has been my experience that, once a committee is created, it almost inevitably ends up holding lengthy meetings, generate vacuous discussions and produce long documents, in order to look respectable and justify the credit eventually given to its members.
Secondly, faculty can and will use the fact that they are, at least on paper, already serving on one committee, even if their contribution is perfunctory at best, as an excuse to avoid other kind of work for which there is more pressing need.
But most importantly, the principle that service work can be regarded, to some extent, as an acceptable substitute for teaching and research, puts a research university on a slippery slope. It implies, among other things, that a system of evaluation is set in place, that is not conducive to the type of scholarly excellence which all university administrations claim to be wishing to promote. Rather, it conforms to the objective (stated or implied) of many administrators, to redefine the academic profession, rendering the job of a professor more similar to that of an employee in the public or private sector.
It would make as much sense, for a soccer team, to hold on to a striker who has not scored a single goal through an entire season, but has taken it upon himself to ensure that all his team mates show up on time for training sessions, doing a pretty good job at that. Valuable as that contribution might be, it is simply not what fans expect of a striker of a soccer team, who will be booed loudly at the end of each game if he does not start scoring soon.
In order to remind ourselves of that, all of us in the sciences, for example, might want to keep a logbook of how many times, in a conversation with a friend or a person outside academia, we are asked the following three questions:
1) What courses are you teaching ?
2) What do you do research on ?
3) What committees are you serving on this term ?
I am thinking that that should give us a pretty good idea of what it is that society expects of us.
 There is also service of a different type, consisting of tasks that are extramural in nature (e.g., involvement with some nation-wide professional organization). That kind of service is, of course, no less important and/or time-consuming than the one based on campus, but typically affects but a small fraction of faculty, and is not what I am not focusing on here.
 It has happened to me personally, and have heard similar experiences recounted by many a colleague at other institution, to be asked by the chair to serve on a departmental committee during my probationary years, and to be criticized for serving on that committee in my annual evaluation… an example of wise mentoring of a junior colleague by senior faculty ? A “set-up” ? Inquiring minds … probably don’t want to know.
 Do not even get me started on the notion that graduate students should be participating to service too…