I received a few emails in the past couple of days about my previous blog post on the supply of political science Ph.Ds.  I didn’t realize that I have readers others than my mom.  Hi mom.

The content of the emails included:

1. Are these “real” publications?
2. Are you going to write a PS article using this data?
3. What is the take home point?

1.  Publication Accounting

I did a little cleaning of the data I collected.  This data is really just for internal purposes (as the Director of Grad Studies) so I didn’t want to spend any additional resources collecting this data or coding pubs.

But I did quickly eyeball the publication data and decided to go through and see how many of these are “real” publications.  I’ve been on a bunch of searches over the years at WashU (comparative, open field, methods, formal theory, and IPE) so I went with my gut on how committee members would perceive journal publications.

No offense to any journal editors out there, but I thought of journals like PRQ, the Journal of Democracy, International Interactions, etc as the floor. This “floor” might piss off a few of you with pubs in these outlets.  But the point is for us to do some really conservative counting.

Basically, there are about 30 or so journals that could fit my criteria.

How many students have quality articles?  90 candidates (or 30.51% of the sample).  These percentages ranged from over 40% of methods candidates with pubs, and 16.42% of the theorists with pubs.

My gut on these “quality” publications probably greatly understates the quality pubs out there.  I didn’t count pubs outside of poli sci (Nature, Stats and Econ journals) and I know less about some fields than others.  I bet the low theory number is due to my lack of familiarity with theory journals.

A quick personal note.  Talking about “quality” pubs sounds like dickish bean counting.  But my loyal reader wanted my best guess.  I aim to please.

A total of 21 candidates had pubs in the big three (APSR, AJPS, and JOP) with a bunch of students publishing in QJPS, BJPS and top subfield journals.  I would say there are 40-50 publications in “top” journals.

So my 45% figure from the previous post overstates things a bit, but 30% is probably under counting.  We can split hairs on how to count this but this is my blog.

This summary counts solo and co-authored pubs.  Solo authored pubs are less common, but there are some candidates out there with top solo pubs as well.

2.  What to do with this data?  Will you publish a Political Science (PS) piece with this? 

First, I honestly didn’t count PS as one of the “quality” publications in the data.  PS articles are good for the discipline, but I feel like there seems to be little professional incentive for individuals to publish in PS.  I might be wrong, but hey, this is my career.

Second, there really is little “personal” incentive to publish with this exact data.  Much of the rumor mill in political science is glorified gossip that often ends with attacks on grad students at the start of their careers.  This data would mostly be used for this purpose, and not to improve graduate training. I want no part of this.

3.   What are the conclusions? 

The farther I get into my career, the more I realize that I don’t know a whole lot about how this discipline works.  I had a gut feeling that pubs differentiated ABDs, but this story is a bit more complicated.

Given the difficulties with this quantitative data, let’s go back to the search committees I talked to this year.  The word that I am getting is that there are pools of 20-30 candidates in most searches with reasonable records, generally including publications. But only 3 or 4 get interviews.

At least one search committee has tried to break out of the bean counting and simple heuristics of Ph.D., NSF grant, etc.  Ethan Bueno de Mesquita at U of Chicago Harris Schools explained their method for search committees and he gave me the ok to share it here.  The summary is something like this:

For every single application received the committee reads the abstract of the job market paper without looking at any other materials (CV, grad school, letters of rec, etc). If a candidate has an idea, research design, theory, etc that appeals to the committee, then he or she goes in the long list. This generally yields a long list of 15 people.  After the long list is made, the committee looks at the whole file.

This is the first time I’ve heard of this procedure, and it really does put ideas at the front of the search process.  I haven’t thought this through in any detail, but I really like it in principle.

But without search committees making conscious choices like the search that Ethan described at the Harris School, I think simple heuristics are often how candidates are selected. I don’t have anything more profound than that.  Sorry.

But search committees did constantly stress the importance of both “quality” and “fit”.  Giving grad students advice to write “quality” pieces that “fit” with potential departments isn’t what I had hoped to conclude from this exercise.

One final quick point.  Stephen Saideman compares the political science market to the NFL draft.  If we go with this analogy, it really is a draft with unlimited free agency.  Thus the top schools can forego the “draft” and hire those with a longer track record.  Rather than predicting success from some heuristic, just hire those that are already successful.

For the record, I hate the New York Yankees.

[I get a ton of spam and generally leave the comments closed.  But I’ll keep them open in case folks have thoughts.  No Viagra ads please.]