Exhibit A: Professor at Ohio U has got a good record of teaching and research, gets a narrow but positive vote from his department and then is rejected at every level above the department. The prof's reactions after the fact seem to justify the prof's rejection as he goes ape-sh!t. The more you read, the more convinced you become that this guy is nuts. However, there is no record of his dubious collegiality before the tenure decision where the majority saw his record as meeting Ohio University's standards.
On the one hand, tenure means lifetime employment, so giving someone like this tenure means being tied to this guy for the rest of your careers. And, depending on how the department runs, dysfunctional people can make a department an awful place to work. So, denying tenure to a guy who seems likely to make life miserable for everyone might not be a bad idea.
BUT how does one measure collegiality?
“When you make [collegiality] an independent criteria, then it’s subject to all kinds of subjective, political or biased principles that really have no place in a tenure evaluation,” he said.Precisely, it was one of the criteria in my old job, and it seemed to be quite a dangerous clause. It is hard enough to measure good teaching, good research and adequate service (see post a, b and c). At least those can be documented, compared and evaluated, even if there are differences about what counts for how much. But collegiality? How much friendliness is required? Or how much disruptiveness is ok?
With the stakes this high, caution ought be required. That is, collegiality can be used by anyone with an agenda to vote against someone without much in the way of repercussions. While one can still vote against someone just on their record of teaching/research/service because of ill-feelings, it seems to me that the collegiality clause gives someone carte blance to vote negatively regardless of the merits of the rest of the case.
There is no perfect solution to this and some difficult tradeoffs to face. Despite the fact that I have encountered tenured folks at previous jobs who would fail miserably on the collegiality clause (and yes, one can hide one's nasty personality for six years so collegiality before tenure is not such a good predictor ...), I still think it should not be considered as part of the tenure process in departments that are sufficiently large that dysfunctional individuals can be ignored. In very small departments, then collegiality might make more sense.