Dean Search Procedures
May 7th, 2008Though no decisions were made at Monday’s meeting, I think it was the sense of the faculty members at that meeting that we should try to decide on procedures to be followed for indicating faculty preferences on dean candidates before we get into a substantive discussion of the candidates themselves. In the agenda she sent out Monday afternoon Donna proposed 10:00-11:30 a.m. as the time for the procedural discussion.
I would like to suggest that we consider two things during that part of Thursday’s meeting.
First, I would suggest the following procedures for determining the faculty’s preferences:
1. We would hold an initial screening vote.
2. If there is more than one acceptable candidate, and we decide to rank them, we would do so using (a) a single transferable voting system (STV) if we have at least three acceptable candidates, or (b) a straight vote if we have two.
3. After the determination of the ranking, the law school members of the Committee or the Committee will report to the faculty at a follow-up meeting on the report the Committee plans to give the president, in order to get faculty feedback.
These procedures would not be implemented at Thursday’s meeting; we would just be deciding on the procedures we would use later — after we’ve heard the committee’s report on all the candidates (e.g., check-outs, reports on student, staff, and alumni views) and had a full discussion of all the candidates. While the plan is to begin a discussion of the three current candidates on Thursday, there’s no way we can complete the discussion even on those three candidates this Thursday; among other things, I don’t think we’ll even have a report from the committee on Thursday. And there may be additional candidates to be brought through; I gather that the Committee’s plan is to let us know one way or the other about that later today.
I’ve put the details below. The most complicated issue may be the last one, concerning the report to the president.
Second, I think we need to address the mechanics of how we will conduct any balloting:
1. Should any voting we do be conducted immediately at the conclusion of the substantive discussion of the candidates (once again, I don’t think this would be on Thursday), or should we break for a day or so to give people time to digest or consider? If the latter, we would reconvene again in a day or two to conduct the actual voting. This may be something we could decide at the end of the substantive discussion, but it’s useful to consider it now because of the second point.
2. Should we use absentee ballots? I would urge that we not have absentee ballots, for two reasons. First, discussion is crucial to voting. Second, and equally important, the procedures I’m suggesting would contemplate more than one ballot taken. The first is a screening ballot; then there would be one or more ranking votes. Trying to accomplish this with absentee ballots makes no sense. If we are going to have absentee ballots we’ll need to design our procedures around accommodating absentee ballots, and that strikes me as backwards.
At the same time, I don’t think it’s necessary or wise to cut out faculty members who can’t be physically present for the meeting or meetings. We should have telephone conferencing for any faculty member is who is not in town on the day or days of the meeting, with a system for allowing them to get on the queue to speak, and also a way for them to indicate to a designated person how their written ballots should be cast in the course of the meeting. I realize some people may be in different times zones with all the inconvenience that entails, but no system is going to be perfect.
**************************
Details:
1. Initial screening vote: We would begin by voting “acceptable” or “not acceptable” on each of the candidates (however many there are). We would not do this in a sequence. We would it on each of the candidates all at the same time, and announce the results at the same time. For each candidate we’d add up the acceptable and unacceptable votes, with the aim of screening out candidates who had a majority of unacceptable votes.
Comment:
This vote would function as the faculty counterpart to the screening the Dean Search Committee engaged in; potential candidates who did not pass that screening test were not brought before the faculty at all. In part because the Committee quite properly erred on the side of bringing someone in who might or not prove acceptable to the faculty (in order to minimize the risk that a candidate who could prove exciting to the faculty would never get before it), the faculty needs to make an initial determination of minimum acceptability. Equally important, it will be helpful to any dean candidate who may be ultimately chosen to have had at least a straight-out vote of “acceptability” from the faculty. This would be lacking if we used only a a single transferable vote (STV) scheme without an initial screening vote.
Michael’s initial memo, along with Terry’s e-mail and the lengthy exchange on the faculty blog between Michael and Patrick (see http://plan.law.miami.edu/?p=45#comments), all persuade me that all voting systems present complexities and difficulties. In this case, it’s possible that someone who thought that as a matter of principle we should send over only one name might mark all but his or her favorite as unacceptable, even if they thought that another candidate was also acceptable. But this represents a substantive choice concerning the way to handle the selection of a dean, not dishonesty or strategizing. And of course, someone might mark all but one candidate as unacceptable if they thought there was only one acceptable candidate. Once again, this is a substantive choice. If someone concludes, after our discussion, that (a) more than one candidate is acceptable and (b) we should send more than one candidate, there would be no real incentive to mark all but his or her favorite candidate as unacceptable.
It’s worth emphasizing that before we take the initial screening vote, one thing we will discuss is whether it would be a serious problem to send only one candidate to the president — and equally whether we should send any candidate at all. If someone is persuaded, for example, that even if we were to ultimately strongly prefer a candidate it would be highly impolitic to send over only one, there’s no rational reason why that person would then, under those circumstances, mark only one candidate as acceptable. Of course, this assumes that the Dean Search Committee effectively reports to the president that there was a very wide gap between the candidates in terms of faculty preference. Because of the importance of this aspect, it’s crucial for any voting system to work properly that the faculty know, while it’s voting (and discussing the candidates) that it will see the Dean Search Committee’s report with meaningful time to comment. (Stage 5, below).
An alternative approach would be to vote in stage 1 solely on the number of candidates to send over. One concern with this approach is that it would mean that no candidate would have a majority of the faculty saying on a straight up or down vote that they’re at least acceptable. There’s a virtue of having some simplicity at some point in the process. Further, by the time we get to the actual voting, we will have already had the discussion about the candidates on the merits. I don’t see how a vote on the number of candidates would be anything other than a vote on acceptability of particular candidates, but with the results obscured by the absence of names. We can know what an “acceptable/not acceptable” vote on candidates mean about the candidates; an outcome of “1″ or “2″ or “3″ doesn’t give us that information. In general, I think a voting system that gives us that information would be better.
In the end, I think an initial screening will have significant benefits and we should rely on the good judgment of the faculty in executing it in a way that avoids theoretical problems.
2. Ranking the candidates.
a. If there’s more than one acceptable candidate, we vote on whether to rank candidates at all or just discuss strengths/weaknesses. I tend to think that we should rank them, but I don’t know whether everyone agrees, and it’s something we haven’t really discussed, so we should have a point at which we decide this. It doesn’t seem to make sense to take this up before an initial screening, so this would be an appropriate point to determine this issue. Obviously, if we vote “no” on this, the rest of this second stage is irrelevant; only the third stage would remain.
b. If there are more than two acceptable candidates, we use the single transferable vote (STV) system to rank the candidates.
i. Exception: If there are only two acceptable candidates, we can just vote on those two for top rank.
ii. If there are three or more acceptable candidates, we hold a vote using the STV system to pick the top ranked candidate.
iii. We hold a second vote using the STV system to pick the second choice, and so on if necessary until everyone’s ranked.
Comment:
Obviously stage 2 is irrelevant unless we have at least two candidates acceptable to a majority. Assuming we reach that point, I’m persuaded that STV can be a useful system for ranking where there are multiple candidates. But in my view, this conclusion is subject to one major caveat: I believe it’s useful only if used for a pool of candidates who have been determined to be acceptable. Without stage 1, we could have candidates in the mix who aren’t acceptable even to a majority of the faculty. Those faculty members would have no way to so indicate; they could only vote “2″ or “3″ for the unacceptable candidate. This would be misleading; voting “3″ on someone who you think is unacceptable or way below the top candidate or candidates, and should really be a “9″ or “10,” doesn’t capture people’s views. I realize the voting system is supposed to be ordinal, but mere ranking in a pool of people who could be unacceptable to a majority of the faculty looks to me intrinsically misleading. On the other hand, while ranking individuals in a pool of candidates who are all at least acceptable to a majority may not fully convey intensity of preferences, it’s still workable.
3. After the determination of the ranking, the law school members of the Committee or the Committee will report to the faculty at a follow-up meeting on the report the Committee plans to give the president, in order to get feedback.
Comment:
In some ways this may be the hardest procedural issue to be decided at Thursday’s meeting, perhaps more complicated than the question of specific voting procedures. The report to the president is a key part of the dean selection procedure. If we’re going to decide Thursday how voting procedures will be handled, we should also face this particular issue at that meeting. It would seem incomplete to decide on voting procedures but dispense with consideration of the end product.
Procedurally, I realize that the law school faculty as such probably can’t formally bind the Committee. But the law faculty members of the committee form a majority on the committee and would presumably follow the expressed wishes of the majority of their colleagues on the faculty. I don’t know fully how the committee operates; if there were some problem in its procedures with following what a majority of the committee wanted in this regard, that problem would have be confronted, but it’s important to have the faculty’s views on this key question made clear.
Figuring out the substance of what to say to the president and how to say it is going to be a challenging task, to say the least. I have great confidence in the competence and good faith of my colleagues on the committee (not to mention strong appreciation for all the work they’ve been putting themselves through), but experience in a variety of areas shows that it is very hard for any committee to get things right in a first draft or a first take on faculty sentiment, even after a discussion. Feedback from the faculty on what the committee or law school members of the committee believe are the faculty’s views is important.
In theory, we could postpone decision on this question until after we see the faculty’s views on the candidates. But even if the faculty were to decide that there were no acceptable candidates, for example, some follow-up meeting could be useful. There would still need to be a report to the president, and it would still help the law school committee members to get feedback from the faculty on what the committee plans to report.
As for the form of the report, it’s not clear to me how the committee plans to report to the president — just orally or with a written report, too. I thought I recalled the Provost remarking that there would be a written report, but I don’t think we’ve heard definitively from the Committee as to what it plans to do.
If there is to be a written report the substance of the report should be shared with the faculty with meaningful time to comment. At the same time, any written report poses real dangers. If it says anything meaningful, and if the faculty’s views are anything less than “these are 2 or 3 wonderful candidates and we can barely decide who would be more excellent,” having the report leak could be harmful to whoever became dean. The faculty owes it to the committee (as well as to itself and whoever the next dean is) to make sure that procedures adopted respect the need for confidentiality. E-mailing a draft report to regfac01 would be in serious tension with that aim. One possibility would be for the committee to give the report to the faculty on paper at a follow-up faculty meeting with numbered copies and collect them all back at the end of the meeting. A second possibility might be to provide a detailed oral report of the written report at a follow-up faculty meeting.
If the committee’s plan is simply to have a meeting with the president with no written report, or if there is to be a brief written report with the bulk of the report being oral, we would have a follow-up faculty meeting for an oral report to the faculty on what that report is to be, with feedback from the faculty.
I mention “the law school members of the Committee or the Committee” simply to indicate that there could be a question as to who on the committee should attend the follow-up faculty meeting.
Finally, a basic alternative to what I’ve laid out would be simply to have the discussion and voting on the candidates and leave the rest to the committee. We haven’t had follow-up meetings in the past, to my knowledge. But neither have we sent over more than one candidate (let alone no candidates), nor have we had any “strengths and weaknesses” reports to my mind. Once again, I think our experience in other areas shows that some opportunity for feedback from the faculty as to what the committee understands the faculty to think about the candidates should be of invaluable assistance to the committee. Only if we think that the confidentiality aspects of trying to give the committee this assistance pose such high risks should we dispense with it.