Reader query on choosing a PhD program

In the comments section of our most recent "how can we help you?" thread, AnonMA writes:

When choosing which PhD program to attend, is it more advantageous to choose a program based on the reputation of one's (presumed) advisor, or based on the reputation of the program itself? That is, when considering future employment prospects, is it more important that one's advisor/letter writer is a "star" scholar in their specialty, or is it more important that the overall rank of the program where one receives one's PhD is high? I am a recent MA who is interested in some PhD programs in philosophy that boast "star" scholars but not high overall rank.

I think this is a great query, one many prospective PhD students are likely to have--and I am curious to hear what readers think. My own initial reaction, to be frank, is to question a presupposition the questions seem to be based on: namely, that reputation (of a program or advisor) is the best way for prospective students to think about which PhD program to enter or what their "future employment prospects" are. Allow me to explain, and then give a more direct answer to AnonMA's query.

In order to properly approach the question of which PhD program to attend, I think one has to first specify what one's aims are, or far better yet, what one's aims are likely to be years down the road...after one has spent years in a PhD program (after all, it is your future self's life your decisions are going to be affecting). So, then, what is it one is going to want years down the road? For instance, will your primary aim be to finish and get a permanent academic job (i.e. any TT job)? Will your primary aim be to get an R1 job (rather than a job with a heavy teaching load)? Or, will you be happy just becoming the best possible philosopher? 

For what it is worth, here is my own experience. Given that a PhD program can take anywhere from 5-10 years of your life, chances are the most important thing to you those many years later will be for your time in the program to not have been a horrible waste of many years of your life. As I explained not long ago on my mid-career reflections piece on life and luck, it took me about 8 years (spanning two different programs) to finish my PhD...and then another seven years to get a tenure-track job. It was an incredibly trying experience. There were many times I was so close to the end of my rope--without any clear hope of finishing my degree, and then without any clear hope of getting a TT job--that I wished I could go back in time and make a different decision. Finally, I know more than a few people who either didn't finish their program or who did but were never able to get what they wanted: a TT job. 

So, then, what are you likely to want most after spending 5-10 years in a PhD program? In my experience, it's pretty clear: you're probably going to want a tenure-track academic job. That is likely to be your #1 priority. Then, subsumed under that priority, you may be likely to have preferences (i.e. a preference for an R1 job, or a teaching job, whatever). So, then, bearing all of this in mind, how should someone like AnonMA make their decision in the present? Answer: they should probably pick a program most likely to get them a tenure-track job (as that is what their future self is most likely to care about).

The question then is, which considerations are the most reliable predictor of their future job-prospects as such? Conventional wisdom in the discipline has long been that one should "choose the best program", where this is commonly understood in terms of programs with the best Leiter-ranking, which just is a reputational survey ranking for PhD programs. AnonMA's question then seems to be whether that's a better measure of their future employment prospects than the reputation of an individual advisor. However, the cold hard facts here are that neither of these things seem to be the best predictors of whether a particular choice of grad program will be best for one's future job-prospects. For, as I detailed here, the 2017 APDA job-placement report indicates that many lower-ranked and unranked programs have far better TT-job placement rates than the most high-ranking departments in the discipline. 

I personally know people from both sorts of positions. I have friends who came out of top-ranked programs who never got jobs, and people who came out of unranked programs who got jobs and whose PhD programs place about 70% of their students in TT jobs. Guess which group of people appear, in my experience, more satisfied with their choices. Hint: it's the people who actually finished their programs and got jobs. 

So, then, how should one choose a PhD program?

The first thing I would find out is the program's attrition rate--that is, how many of their students actually finish the degree. If I recall correctly, some programs have something like a 50% completion rate. I've known people who spent 8+ years in a PhD program but who never finished. They weren't too happy about that, to put it mildly (and indeed, when I was in danger of not finishing my program, I was terrified at what that might mean for my life).

After looking at attrition rates and avoiding programs (both highly-ranked and otherwise) with low completion rates, I would look at the program's overall permanent academic-job placement rate. I would then think carefully about my tolerance for risk, and how my future self might think about my tolerance for risk. If I thought my future self be most concerned with being the best possible philosopher, maybe I would choose a high-ranked program with a good placement record (like Berkeley, which the ADPA report indicates has a 59% placement rate). What I definitely wouldn't do is join a highly-ranked program--or a program with a super-famous advisor--that has an abysmal placement rate. Finally, if I thought the most important thing for my future happiness would be to just get a TT job, I would probably forget about reputation altogether and join a PhD program with the best overall placement rate (like Virginia, Cincinnati, Baylor, or Florida).

In other words, given the preferences one's future self is likely to have (though there may of course be outliers), my first suggestion to AnonMA is that they may be asking the wrong question. At least with respect to their stated goal (viz. what is advantageous "considering future employment prospects"), reputational matters--viz. program reputations and individual advisor reputations--are probably the last thing they should be thinking about, since again both of these things appear to be relatively poor predictors of job-market success. If you want to maximize your job-prospects, look at the hard-data: at attrition rates and TT placement rates...end of story. If you want to balance concern with job-prospects against "receiving the best philosophical training", then sure, take into account program and/or advisor reputation, but again bearing in mind the hard data (as again, some highly reputed programs and advisors have good attrition and job-placement rates, whereas other ones have abysmal ones).

Finally, a few thoughts on making decisions based on individual advisors. Some advisors are very well regarded in the profession have much more successful students than others. I know some famous advisors who churn out PhDs left and right, whose students almost all end up getting jobs. I know other famous advisors--often in the same program--whose PhD students tend to never finish the degree, and even if they do, rarely get jobs. This is another kind of hard data to try to collect and bear in mind. If AnonMA is thinking of joining a program on the basis of a particular advisor, they should try to find out a lot more information, such as how the person is to work with, how quickly their students make it through the program, whether they tend to make it through the program, and how they do on the market. Finally, one further complication (which I alluded to in my life and luck piece) is that I think there is a great deal of risk "banking on one person." That famous advisor you joined the program to work with? Yeah, they might get hired away by another program. What then? Or what if your philosophical interests unexpectedly change (as did mine - a person who started out my PhD studies doing metaphysics and epistemology but who ended up doing moral and political philosophy)? What then? Or what if you join the program for Person X and you just don't get along that well or they don't end up thinking you're great? What then?

I guess, then, I'd suggest that someone in AnonMA's position should probably rank relevant considerations as follows in making a decision:

• Highest priority: choosing a program with low attrition rate and high TT job-placement rate (regardless of ranking).
• 2nd priority: choosing a program likely to give them the best philosophical training, subsumed under priority #1.
• Lowest priority: choosing a program with a particular person to work with (subsumed under #1 & #2).

Again, this is just how I would suggest AnonMA think about these matters. It may not be the only or best way...but I've tried to justify why I think it's probably (if not universally) a good way. What do you all think?

Comments

[Amanda]

I know famous advisers who place almost all of their students, and others that place almost no one. So yeah, that question by itself doesn't mean much. Look into all the details. If you are only after an R1 job - then I would never turn down a top 5-10 school. But if you are after something more than an R1, then I would consider everything Marcus said.

[anon]

"I am a recent MA who is interested in some PhD programs in philosophy that boast "star" scholars but not high overall rank."

Just to emphasize what Amanda already said, and in light of this part of the original post in particular: if you're really thinking of going somewhere just because of one particular star scholar you'll want to investigate that person. In depth. How do their students do on the market? What is attrition/time to completion like with their students? How timely are they with getting feedback to students? Are they nice or are they cruel?

One benefit of choosing a program rather than a person is that it's hard to figure out all that stuff in advance. These details are often sensitive, too: people who have been harmed by the bad advisors aren't necessarily rushing to tell their story to everyone they've just met. But when you're at a good program with multiple possible advisors, you can figure this stuff out over your coursework years and then choose a good advisor.

[Marcus Arvan]

I could hardly agree more with Amanda and anon.

(1) If you want to choose a department because of a particular person, it behooves you to do a great amount of due diligence on finding out about what they are actually like as an advisor, their student attrition rates, and placement rates (though I disagree with Amanda that it "never" makes sense to turn down a top 5-10 school. Again, I think that should depend on attitition and placement rates - as some top-ranked schools seem to fare poorly here irrespective of their ranking/reputation).

(2) Given the risks associated with banking on one person, it seems to me sensible to choose primarily on the quality of an overall program and subsume concerns about working with a single individual to that.

(3) If you are *confident* your future self would only want a job at an R1, then do what Amanda suggests: shoot for a high-ranking school with a good placement rate. Otherwise, if you just think your future self would be happy with any TT job, I'd advise following the advice in the OP. (Side note: it's not clear how anyone can be reasonably confident about what their future self would prefer here 5-10 years later, given the situation they will find themselves in. Can you really be *confident* that your future self would rather have no academic job at all than a teaching job? On what basis?).

[AnonMA]

Thank you to all who have taken the time to respond! I am prioritizing placement rate into permanent jobs (of almost any kind), but I posted my query in part because I also have concerns about this. For instance, it is totally opaque to me why certain programs have higher placement rates than others. It occurred to me that it may be because they have just one or a few dedicated faculty whose deep support of students raises the placement rate. But what if that/those faculty leave? That is, how stable is the placement rate? (I may be able to get some insight into this from the APDA report.) If supportive faculty did leave, would I have to/could I (in part) fall back on the reputation of the program and/or my advisor? Rhetorical question--what I've gathered from the responses is that I'd have to do this research myself by reaching out to current students, etc., in various departments.

[Marcus Arvan]

AnonMA: You're very welcome!

Here's my general sense from talking to people: programs with better placement rates tend to have them because of their departmental culture.

Programs with good placement rates appear to me to tend to (1) have a culture that prepares their students well for the market (with a lot of resources explicitly devoted to mentoring), and (2) have an approach to placement that "fits" their candidates.

For example, my friend who comes out of an unranked program with a 70% placement rate reports that his program explicitly prepares its graduates for "teaching jobs": a strategy based on the (almost certainly correct) assumption that its graduates will not be competitive for research jobs. By a similar token, a friend of mine who recently graduated from a mid-ranked program reports that her program dramatically raised its placement rate by adopting a similar strategy.

Conversely, my sense is that programs with poor placement rates tend to (1) have a culture that prepares their students poorly for the market (with few resources dedicated to mentoring), and/or (2) an approach to the job-market that is poorly fitted to their candidates.

For example, I know people who come out of a program that provides its students with NO official job-market mentoring. When I met a student from this program at a conference a few years ago, he was going on the market for the first time without *any* idea of what it entailed. His program had not familiarized its students at all with (let alone mentored them) with how to write cover letters, research statements, etc. Suffice it to say, his program has a terrible placement record. Then, on the other hand, my sense is that there are a lot of lower or midranked schools who primarily prepare their students as researchers--something which I have hypothesized is a terrible job-market strategy (as students from lower or mid-ranked schools will generally lose out on research jobs to candidates from top-ranked schools and not be competitive for teaching jobs due to their PhD program's research heavy priorities). See http://philosopherscocoon.typepad.com/blog/2018/01/grad-program-rank-publications-and-job-market-a-hypothesis.html

Long story short, it is indeed an opaque and empirically unsettled question what accounts for dramatically different attrition and placement rates at different PhD programs. However, my own sense from talking with people is that the most likely explanations are systemic and thus fairly likely to be pretty stable. Case in point: my friend from the unranked program with a 70% placement rate tells me they've had an excellent placement rate for many years going. A converse whose students apparently rarely finish or get jobs--and in that case too it appears to be a matter of longstanding departmental culture. More generally, I've heard the amount of job-market mentoring that programs provide varies immensely, with (again) some programs basically giving none and other programs giving tons of mentoring from the time grad students first step into the building. It is hard to believe that these differences in mentoring wouldn't make a substantial difference, and in a relatively stable way at that.

To answer your rhetorical question, yes, I would think you would have to do that kind of research yourself by asking around. On another note, I believe the APA has a grad program report that lists attrition rates for programs willing to report their numbers.

https://gradguide.apaonline.org/

[Trevor Hedberg]

One often neglected aspect of deciding on a PhD program is the financial support you'll get. I don't just mean whether you have a position as a teaching assistant. I mean the combination of your current aid, the cost of living in the area, and the opportunities for additional aid (e.g., fellowships, summer teaching) in the future. If you incur a lot of debt during graduate school, it makes the job search that much more stressful because it puts more pressure on you to find a stable, tenure-track job more quickly. Additionally, if you have to adjunct at other places to increase your income, that will likely increase your time to degree or decrease the amount of time that you can devote to your own research (which could hurt your marketability in the long run).

[Amanda]

while obviously one can spend too much time adjuncting, I would say overall it is a huge help on the job market.

[Amanda]

As for a few top places not having good placement records...I'm not so sure about that. I remember reading something that explained it like this. It is common for a number of the top 10 places to put their students in very fancy research post docs that last 2-5 years. So for these people it showed as not placing in 3 years, even though they had a very competitive job and would get a R1 job once the postdoc is complete. I am not positive about this, but I think all the top 10 schools have very good placement records once you look 4-6 years out.