The Philosophers' Cocoon

In the comments section of my recent post, "Part 10: More on Standing Out for Teaching Jobs", I had a spirited discussion with 'Postdoc' beginning here on whether candidates should take "risks." Of course, what constitutes a "risk" is open to interpretation. What one person regards as a risk another person might not. Here, however, are what I take to be a few plausible types of risks:

• Pursuing ambitious research projects that fly in the face of received paradigms, authors, etc., that people "like."
• Pursuing research that defends morally and/or politically unpopular conclusions.
• Utilizing unconventional teaching methods.
• Engaging in blogging or public philosophy under your own name.
• Etc.

Here, in contrast, are things candidates might do to "avoid taking risks":

• Work on currently hot authors or topics, working well inside established paradigms that people "like."
• Pursuing research that defends morally or politically popular conclusions.
• Utilizing conventional teaching methods (e.g. "chalk and talk" Socratic dialogue).
• Avoiding online discussion, blogging, etc., under one's own name.
• Etc.

I have little doubt there are many candidates who succeed "taking few risks" (more on this below). Still for all that, my experiences on both sides of the market have led me to wonder whether this is generally the best strategy for individual candidates–particularly those who do not come from the highest Leiter-ranked departments. Allow me to explain.

One of the things I really didn't understand as a candidate (and understandably so!) is what things are like on the other side of the aisle: namely, from the perspective of search-committee members. On the candidate side, all you see are your materials, as well as (perhaps) who ends up getting hired for the jobs you applied for. Things look very different to search committee members: they see the hundreds of applications they receive, as well as what their deliberations about those applications are like. so, try to put yourself in the shoes of a search-committee member for a moment. Chances are, you just received several hundred applications for a position. Chances are, further, that many of those candidates look broadly alike. They have the same areas of specialization, are working on similar projects (on hot topics, authors, or arguments), broadly similar publication numbers (most candidates in my experience have something like a handful of publications), broadly similar teaching statements and classroom practices, and so on. There you are, with hundreds of applications. How do you decide who to interview? Well, it's difficult–so many of the applications look broadly the same! So, you read candidates' writing samples, some strike you as better than others, and so maybe you put them near the top of your interview list. But now ask yourself: given the hundreds of candidates in front of you, what are you plausibly looking for the most? Here's my answer (speaking from some experience): you're looking for candidates who literally "jump out of the pile" at you–a candidate to, as a UK Reader put it here, "fall in love with."  You're not looking to hire "just another person." You want to hire someone unique–someone who stands out. Or, at least this is my experience.

And so I suggested that taking some risks may be a good idea for candidates. UK reader agreed, writing, "I agree: take risks. After all, you only need to get one search committee to fall in love with you. Who cares if all the other ones get rubbed up the wrong way by your application?" Yet, Postdoc was unconvinced, writing, 

Your experience is just different from mine. My experience is that those who don't 'step in line' (those who aren't doormats) suffer notable consequences, which in this job market is likely enough to harm one's chances. My experience is that those who make it are those who keep their heads down, do what they're told, and walk the party-line. Those who make it are just those who don't stray far from the mainstream, probably even philosophically.

As I noted in reply, it's probably hard to settle whether it's good for candidates to take risks without clearer data (which is why, in a few moments, I will ask search committee members to weigh in!). But first, I'd like to venture a hypothesis for how Postdoc's experience may be consistent with my theory that it's advantageous for candidates to take risks. As I will now explain, I believe the law of large numbers may make it look like risk-averse strategies are best for candidates when, in reality, taking risks may be more advantageous.

As 'Recent hire' noted in the course of the conversation, "I am just generally not a very risk taking person, and find that many of my colleagues aren't either." My experience on both sides of the market is that Recent hire is probably right about this: that most candidates are probably risk averse. Suppose that's right then–let's say the "risk-averse" pool makes up 90% of all candidates, whereas risk-takers only make up 10% of the pool. Suppose then that a search committee receives 300 applications for a position. In that case, 270 applications are likely to come from risk-averse candidates–candidates whose files are broadly similar–while only 30 are likely to come from risktakers (candidates whose files plainly "stand out from the crowd" in some way). Suppose next that simply because there are so many risk-averse candidates in the pool, there's a 75% chance that the committee will end up hiring a risk averse person. Notice that still means that, if you are a risk-averse candidate, the probability that you will get hired is really, really small. There's a 75% chance that someone in your pool will get hired for the position, but you only have a 1-in-270 chance of being that person! Now suppose, on the other hand, that the much smaller pool of 30 risk-taking candidates have a 25% chance of someone from their pool being hired for the job. In this case, even though the probability that a risktaking candidate will be hired for the job is smaller than the probability that some risk-averse candidate will, it's still a better strategy for an individual candidate to be a risk-taker–as each individual in that smaller pool has a 1-in-30 chance of being the member from the risk-taking group to be hired.

My sense–having spent quite a bit of time on both sides of the market–is that something like this picture is probably true. Yes, Postdoc may be right: most candidates who are hired may appear to be "risk-averse." But that may simply be because so many candidates are risk averse. It is a statistical fallacy, in other words, to assume that the best strategy is to be risk averse because candidates who are hired appear risk-averse. Finally, things may even be worse than this. When I was discussing this hypothesis with a philosopher friend just last night, he added (I paraphrase), "I think you're right–and I suspect risk-aversion tends to favor people from Leiterific PhD programs the most. They don't need to take risks to stand out, because they already have prestige and professional connections that are to their advantage. The candidates who most need to stand out are those who come from lower-ranked programs."

This struck me as probably correct–and it complicates the proper statistical inference about how to succeed a great deal. For let's suppose again, for a given position, that 270 applicants are "risk-averse", and there's a 75% chance of the hire coming from this risk-averse pool. Now suppose however, in addition, that 9 times out of 10, when a risk-averse candidate is hired, they just happen to come from a Leiterific institution. In that case, any candidate who is risk-averse who is not from a Leiterific school has only a 1-in-10 chance multiplied by the 1-in-270 chance they previously had of being hired out of that group. If something like this is correct, then even if most people who are hired are "risk-averse", it is a spectacularly bad strategy–probabilistically–for any individual candidate from a low-ranked school to be risk averse.

Here is why I think this matters. I think it is easy to look at the market from the outside–at who gets hired–and assume that it's best for candidates to be risk-averse. However, from the standpoint of a search committee member–who has recently seen some incredibly risky candidates get hired for jobs–I worry that the assumption may be based (at least in part) on a statistical illusion. Yes, of course, there may be "dumb" risks to take, such as being awful to others in the profession publicly. Still, for all that, I cannot help but wonder whether many candidates–especially those coming from lower-ranked PhD programs–might improve their odds by taking more risks. I certainly did better on the market the more risks I took, and just this year I saw someone get hired for a tenure-track job who defends moral and political views in his research that appear to be widely considered outrageous in philosophy and the academy more broadly. I cannot help but wonder whether it was the very fact that this person defended such unpopular views that made them stand out, and stand out enough to get a job when so many other candidates defending more popular positions do not.

But, this is all–admittedly–just me speculating on the basis of my own impressions and statistical possibilities. Which brings me to my question for the search committee members out there: what do you all think? Is it good for candidates to "take risks"? If so, which kinds?