Preference Falsification at Dartmouth and Beyond
Something interesting happened at Dartmouth a couple of days ago, with lessons that extend far beyond that campus.
At a public meeting of the Student Government Senate, a vote of no confidence in President Sian Beilock passed 13-2 with three abstentions. This was met with supportive cheering and chanting by many of the (nonvoting) students in attendance.
The student body president then decided that there had been insufficient deliberation prior to the vote. She scheduled a second vote on the same motion, this time in closed session with anonymous tabulation. Just hours after the public vote with a supermajority in favor, the no-confidence motion failed by a vote of 8-9 with two abstentions.
Among those who switched their vote was a senator who explained (while remaining anonymous) that his vote at the public meeting was motivated by a fear of intimidation and retaliation.
The term preference falsification was coined by Timur Kuran to describe the phenomenon at work here. The idea is related to Elisabeth Noelle-Neumann’s examination of the spiral of silence, and to Glenn Loury’s analysis of self-censorship in public discourse. And these are all instances of what Erving Goffman described as impression management in interpersonal relations.
There is a connection here to recent events on my own campus.
On April 22, by a vote of 102-0, the Barnard chapter of the American Association for University Professors issued a vote of no confidence in President Laura Rosenbury. The motion was voted on again at a faculty caucus two days later, passing this time by a margin of 98-1. Neither of these votes gave participants the protection of anonymity. A third vote was then held, this time anonymous and online, and passed 228-56 with twelve active abstentions.
My initial reaction to these events was to observe that the lack of dissent at earlier meetings and the absence of debate between the second and third stages meant that even the final tabulation could not be seen as truly representative of faculty sentiment. Nobody had made an argument against the motion at any point that was compelling enough to convince a single person to change their mind. Perhaps there were no minds amenable to change, but my own conversations with colleagues and the dozen active abstentions suggest that many people did indeed struggle with the decision.
It also seemed to me that attendance at the first two meetings involved a non-representative sample of faculty, who were especially inclined to vote for the resolution. I assumed that none of the 56 people who eventually voted against the motion were present at the first meeting, and at most one was present at the second.
But this assumption might have been mistaken. The Darmouth example suggests that some people may have voted differently under the different conditions.
One might think that faculty would be less inclined than students to care about the opinions of their peers when making such consequential decisions, but that may not be the case at all.
Consider, for instance, a pre-tenure professor whose senior colleagues firmly and visibly align themselves with a particular position on an issue such as this, with passions running high. It would take some courage and a great deal of trust to oppose them publicly, or even to abstain. The prudent choice might be to just go along.
More interestingly, consider the case of a senior faculty member whose pre-tenure colleague takes a firm public position on the issue. Even if the former were completely committed to fairness in the subsequent evaluation of the latter’s work, they might worry about the junior colleague being uncertain of this commitment. That is, they might be concerned that a public stance against their colleague’s view might signal a lack of support in the future. The result may be preference falsification by the tenured professor in this case.
These kinds of uncertainties and suspicions about suspicions are pervasive in daily life, especially in interactions among people who don’t know each other all that well. They are sand in the gears of public communication. They give rise to declarations that cannot be taken at face value. They contaminate even anonymous tabulations, because the discourse preceding them can be stunted and insincere.
These are not problems that can be easily solved, but there is something to be gained by at least acknowledging them, and being a bit more careful in the interpretation of collective decisions.
