Study on Philosophy Job Market Underway
A team of researchers is conducting a study of the job market in academic philosophy and is currently seeking participants.
[Donald Judd, untitled works]
If you are planning to apply for academic jobs this year, we hope you’ll consider participating in our study “The Philosophy Job Market: Applicant Profiles and CV Review.”
Our aim in this first phase is to get a picture of job market applicants: what journals are they publishing in (if at all)? for how many classes are they the instructor of record (if at all)? There has been, as far as we know, no systematic attempt to collect information about people going into the job market. In the second phase of this study, we will the collected information to randomly generate CVs and investigate the behavior of potential search committee members in reviewing them.
In consultation with the Gonzaga University IRB, it was determined that IRB approval was not needed for this phase project. IRB approval will be obtained for the second phase. Nonetheless, all responses will be kept private and information secured on the hard drives of the investigators.
The survey should take about 10 minutes to complete. Participants will be entered into a drawing for two $25 Amazon gift cards. If you are willing to participate, you can find the survey here. Please email Charles Lassiter at [email protected] if you have any questions.
The results of the study will be made available here and at other sites, including The Philosophers’ Cocoon, Leiter Reports, and Phil Archive.
Again, the link is here (https://gonzaga.az1.
Interesting project. I want to flag a couple of potential worries. 1. Representation: there is a potential for bias if the only way the researchers are getting participants is through advertising on philosophy blogs. The type of job candidate who reads these blogs might be quite different from the ones who do not. Are the researchers also advertising the study in other ways? Perhaps an alternative option in the future would be to partner with a few universities who are hiring and seeing whether they would be willing to participate. This would involve a lot more leg work. Preparing an informed consent release, making sure participation information is kept separate from the search committee.. etc. However, the results of the project would be quite interesting, especially if a number of different types of universities participated.
2. If the goal is (in part) to generate CVs, wouldn’t the best approach be for volunteers to upload their CV? There is a lot of information on a CV that can’t be captured by the open text fields in this survey. Also, there are other ways of listing AOS and AOC than the pre-defined categories that are on the survey. Granted I did not finish the survey, so maybe the researchers do ask for the participates to upload a CV.
3. What about the other aspects of a candidate’s profile that make them stand out besides what can be easily found on a CV? The narratives that people supply in a cover letter/research statement help to give the CV context. Have the researchers thought of doing NLP on cover letters/ research statements?
Some quick responses to questions raised above. (Charlie Lassiter may chime in later with additional thoughts.)
General remark: To my knowledge, there aren’t any empirical studies of the academic job market in philosophy. We don’t really have much of a budget for this project, but hopefully we’ll learn enough from our study to get folks interested in going further in the future. We’re not really looking to say the last word on the topic, at this point.
Re: (1) We’re aware of the limitations of soliciting participants through blogs, but the same or similar worries seemed applicable to other potential data gathering mechanisms that we could realistically do without any signifiant budget. If we chose a random sample of schools and then manually researched everyone on the job market from those schools, then we’d necessarily have to choose a pretty small sample because of limited manpower, and then we’re back to wondering whether the sample is really representative. You could try to assuage the worry with a stratified sampling approach, but then you’ve got to justify your stratification, etc. . . . So, we decided to just start with the surveys that are advertised here. Obviously the more people who take the survey, the better, **so if you’re planning on being on the job market this year, please help us get the most accurate picture possible by taking the survey.**
Re: (2) & (3) We’re trying to limit the information we’re gathering to the questions about teaching, research, and the other items currently on the survey because those are the most straightforwardly quantifiable pieces of data we’d see on the CV. The kind of “fit” information you’d find in a cover letter or the professional service items on the standard CV like serving on the grad student council, etc. are really hard to categorize in a way that would make them easy to quantify. That’s not to say that these items aren’t important, just that they’re not easy to quantify.
But the fact that we know we’re not capturing every possible variable isn’t too concerning for two reasons. First, as George Box said, “all models are wrong” (https://en.wikipedia.org/wiki/All_models_are_wrong). Second, one can can still calculate the “goodness of fit” of a model, i.e. how much of the variation in the data our model *does* predict, even if one knows there are other potentially important variables not being considered. Also, although one could always add variables to a model, but the more variables one adds to the model, the more one increases the risk of the model overfitting the data, i.e. “predicting” what is really only random noise in the data.
The APA should maintain a database for everyone on the job market, with their CVs. That way when any school has an opening they can consult the database. So much easier than mailing one application after another to schools.
I appreciate the idea, but at least two worries come up.
First, departments might waste their time looking at people who don’t want that job. Second, a lot of important information is contained in a cover letter–especially for jobs at less elite schools–so an application would still be required if the department wanted cover letters.
It might be good if the survey (and DN) would specify that this survey is clearly geared toward the American (US/Canadian) job market. It begins with the question, “Are you planning on applying for a job with an[y] academic institution during the 2019-2020 academic year?” But the questions about CVs that follow seem tailored specifically to the American liberal arts university setup. Many international universities – especially those that don’t operate on the liberal arts model – don’t have “gen-ed/core” classes. They may also not employ the “AOS/AOC” approach to declaring specialization. And certain kinds of publications (e.g., contributions to edited volumes) matter more/less, depending where you are.
Job markets and hiring processes are not the same everywhere, and many of us applying internationally have different CVs for different markets. To call this a study of “the” job market in academic philosophy seems to ignore all the candidates and readers of this blog that come from and are applying to non-American countries. It’s fine (and laudable) if the authors of the survey want to study the American job market. I’d just like them to make it clear to those filling out the survey that this is what they’re doing. Otherwise some of the prompts need to be modified and/or questions added (e.g., “In which countries are you applying for jobs?).
Point well taken! Yes, we’re looking specifically at the *American* job market. Candidates from other countries are of course welcome to take the survey, if they intend to apply to jobs in the US as well. But specifically we’re trying to find out what the population of people applying to US-based colleges and universities looks like.
FWIW, elements of this survey don’t really capture the Canadian side of things, either (e.g. gen ed, the race categories). As Shane says, it really seems most appropriate to the *American* side of the market.
The APA should maintain a database for everyone on the job market, with their CVs. That way when any school has an opening they can consult the database. So much easier than mailing one application after another to schools.
Why not appeal to (or use) open record laws to request the application packets of all job seekers to state school positions? There would be more than enough in those packets (even with confidential information redacted) to put together this picture without having to rely on self-reporting.
My suspicion is that it wouldn’t work. HR depts (as we learned in when we first started thinking about getting the data) are *very* reluctant to let go of applicant info without it being stated in the job application that the info might be used for research purposes. Otherwise HR depts open themselves up to lawsuits for releasing information without the info’s owner’s explicit consent. I don’t think such statements are common on apps at most unis.
It doesn’t matter what they’re reluctant to do. Unless the relevant open record laws in the state exempt that information, they’ll have to turn those documents over. My guess is that they’ll redact the names, but, again, unless that’s something they’re allowed to do by law they won’t be able. If and when they redact information, they’ll also need to give you a written explanation as to why the redacted it. Happy to talk with y’all further about this if you’d like. But if you approach these discussions armed with knowledge of the relevant state open records laws, and you make that knowledge known when you make the request, you’ll probably get what you want without having to make the formal ask via the records request. Good luck! Seems like a good project.
Hi Chris,
Looking at https://www.rcfp.org/open-government-sections/3-applications/ it seems like the states are kind of all over the place on whether job applications constitute “public records” which are requestable by the public. On a quick skim, it looks like in a lot of states the application materials of the successful candidate are requestable, but the application materials of unsuccessful candidates are not. (on the theory, perhaps, that if you didn’t get the job, you aren’t “personnel” and so these materials don’t constitute “personnel records”.)
Still, it’s an intriguing idea and at least some US states do make all job applications a matter of public record (which came as a surprise to me!)
Shane,
Nice find! Looks like it’s about 50/50. You’d probably only need a handful of successes to put together a reasonably clear picture of what the market looks like. That picture would probably be more clear than what you’ll get from open surveying. Good luck!
Record requests are slow, can be expensive relative to what you obtain. And have to individualized to each state and, depending on state law, each institution you are interested in.
This is why I suggested above that the researchers partner with search committees and get the applicants to sign informed consent forms. It should be voluntary for the applicants to participate. Furthermore, it should be set up in such a way that the search committee does not know whether the applicant is participating in the project. This doesn’t cost money, just time, and probably IRB approval.
This looks like it will suffer from serious sampling errors. Survey methods are already problematic when you simply cold-call people, but when you request responses in this method, you often get non-representative samples. Further, people often lie in surveys.
A better method might be to go through each of the major graduate programs and examine the public CVs of their job candidates. I realize that not every program lists their candidates. (This is too bad. In economics, departments routinely list candidates, their CVs, and their job market papers.) I realize you’ll miss people who aren’t in programs but are on the market, but you’ll at least get a proper sample of current grad students on the market.
@ King Monster
We considered this method and decided against it, for two reasons. First, doing this would take a lot of time and manual effort we don’t realistically have. Second, not every current graduate student is a job seeker and not every current job seeker is a graduate student.
The second point is the more important one. Even supposing you could tell from the department websites which graduate students were about to finish and intended to go on the market, you would still be ignoring lots of other subgroups of job seekers who are on the market as well, such as adjuncts, post-docs, and folks who already have TT jobs but are looking for an upgrade.
@King Monster,
On a re-read, I’m not sure my previous comment was fully responsive to yours. Perhaps a clarification will help: we’re explicitly studying the job market, not graduate students who happen to be on the job market. Obviously graduate students are a big part of our population, but they aren’t the only group we’re interested in.