History PhDs: a Cambridge cohort study

Several years ago I carried out a couple of basic studies that looked at medieval history researchers and their careers:

A Prosopography of Cambridge Medievalists, which explored the careers of my cohort of MPhil students

The Over-Production of Medieval Historians, which used historic data from the IHR Making History project to ask whether there is an over-production of PhDs now as compared to previous generations.

There’s been a lot of recent discussion about the problems of early career researchers and the related problem about whether there are too many PhDs in history being produced. Brodie Wadell has done two very useful posts giving statistics for historians generally.

I want to approach the subject in a slightly different way, doing a cohort study of the careers of a group of PhDs from a particular year. What follows, therefore, is an analysis of what happened to people who got a historical PhD from Cambridge in 2005.

I chose Cambridge for this analysis for three reasons. Firstly, there’s a large enough cohort (66 people) that chance is less likely to be significant in the overall outcomes, but it’s not so large that tracking everyone would take impossible lengths of time. Secondly, I had two different sources of data to use to select the initial cohort (as I discuss below). Finally, Cambridge PhDs make up a relatively homogeneous and prestigious group. They’re all going to be studying full-time, they’re likely to be funded and I’d expect them to be among the most successful on the academic job market. As such, their results can give us a likely upper bound on how many PhDs get academic jobs.

I chose 2005 to study because I presumed that after 10 years anyone who would ever have a permanent academic job would finally have got one (though as it turned out I was wrong). Those studying for it would also have started their PhD in 2002 or before and so been in a more favourable position than later cohorts.

Data collection

The full data is available as an Excel spreadsheet if anyone wants to work with it further. My two sources for identifying the cohort were Newton Library catalogue of University Library Theses where I searched for all PhD theses from the Department of History and the IHR History Theses database. I combined the data from these and did some cleaning up. The biggest problem was that the year the two sources gave for the thesis sometimes differed: in these cases I preferred the Newton catalogue date, since it would have been based on the hard copy of the final thesis. The IHR list included some theses I’d classify as ancient history, so I removed these. But it also identified historical theses produced other than in the Department of History. The final tally by department was:

Anglo-Saxon, Norse and Celtic 3
Architecture 1
Education 1
English 6
German 1
History 41
History of Art 1
History and Philosophy of Science 6
Music 2
Oriental Studies 2
Slavonic 1
Social & Political Sciences 1

This is a useful reminder that the boundaries of a ‘history’ PhD are blurred. I’m sure some people would come to different decisions than the IHR about exactly which ones we should include, but I don’t think that they substantially alter the overall thrust of my results.

Having got details of my cohort and their theses topics, I then attempted to locate them (using publicly available internet resources). My strategy was as follows:

1) Check the IHR database of teachers of history by name

2) For those not found on that, check Academia.edu by name

3) For those still not found, check LinkedIn by name

4) Google the person’s name

5) Check a history database: Historical Abstracts (post 1500), Regesta Imperii for pre-1500 theses.

All these measures obviously work better for those with more unusual names (and may fail completely if someone changed their name soon after completing the thesis). Out of the 66 people there are 11 who I have not been able to trace to a recent job. But with this combination of searches I can at least be fairly confident that these 11 (5 men and 6 women) are probably not in academic posts. My spreadsheet gives the current job title for people for whom I could discover it; otherwise it gives information on their most recent known academic involvement, e.g. books published.

Data analysis

Since the cohort is relatively small, I only tried a small amount of statistical analysis on it, as follows:

1) Employment status.  I split this into 4 categories:

Y = UK academic post

O = Overseas academic post

N = No current academic post

NU = Untraced (including those where I had some previous information about a career, but not current data)

“No current academic post” was where either I knew the person’s job or knew that they had recently had an academic post, but no longer appeared to be in it.

I didn’t attempt to distinguish between permanent and temporary posts, since this was not always clear from university websites, especially foreign ones. Nevertheless, it is clear from job titles and career histories that some of the academic posts were temporary (e.g. Adjunct, Research Associate).

2) Gender. I did a simple F/M split, based on name, pronouns used and photographs. If there are inaccuracies in gender identity that you are aware of, I am happy to correct these.

3) Period. I split the periods covered by the thesis into 3 categories:  medieval (pre-1500), early modern (1500-1789) and modern (1789-). Where a thesis covered parts of more than one period I put it into the one that seemed to predominate.


The basic statistics are the most revealing ones. Out of 66 students with PhDs, 36 had academic jobs (22 in the UK, 14 overseas). 30 didn’t (of whom 11 are untraced and 19 had either a non-academic job or no current academic job). Even with this most privileged of cohorts, only just over a half had academic jobs 10 years on. Admittedly, not all those who did a PhD necessarily wanted to go on and have an academic career. I’ve listed the other jobs people have ended up with in the spreadsheet and I might discuss them more in a future post.

It’s also worth noting two other splits. One is looking at who obtained academic jobs by period studied:

Period & status
Med EM Mod
Y 5 2 15
O 1 4 9
N 5 3 11
NU 4 3 4

The other is who obtained academic jobs by gender:

Status & Gender M F
 Y 11 11
 O 9 5
 N 4 15
 NU 5 6

From this, it’s clear: students who researched history after 1789 are substantially more likely to have academic careers than those who don’t and medieval historians have a substantially poorer chance of academic employment than other groups: only 40% of them have academic jobs, as opposed to over 60% of those who studied modern history. (Again, all this is discussing an elite cohort, one unusually likely to be hired for any academic jobs going).

The other is that there’s a bias against women getting jobs. Overall there are 37 women and 29 men in the cohort: 16 women and 20 men now have academic jobs ten years on. It’s possible that more of the women I haven’t been able to trace than the men have academic jobs (because of issues about changes of name), but that’s unlikely to bring them up to parity.

Before assuming this is directly due to sexism, however, it’s important to look at another table, which matches gender to period studied:

Period & gender M F
Med 5 10
EM 5 7
Mod 19 20

In this cohort, women were substantially more likely to research medieval topics than men were and since medievalists are less likely to get jobs, that may have influenced the overall outcome. It would be possible to compare employment outcomes by period and gender, but then you’re getting down to such small numbers that the differences might well occur by chance.

It’s also possible that we’re missing more subtle aspects to sexism. One possibility is that female PhD students may prefer to have female PhD supervisors (or at least cluster in subfields of history which seem less prone to sexism). If women are going into areas of research which have worse career outcomes because of this preference, then that’s a more difficult problem to tackle. But that would be quite hard to demonstrate conclusively without studying a much larger cohort in some detail.


If we want accurate information about the career prospects of historians, I suspect we in the profession will have to start doing some fresh data analysis ourselves: the published information isn’t adequate. This study shows one way of doing such an analysis; what I’d really like to do is compare this with a cohort from a less prestigious university and see the career outcomes from there. But even with this atypical cohort I think there are two important lessons visible: 1) if you’re hoping for an academic career, you do need to have a Plan B and 2) don’t be a medievalist.


8 thoughts on “History PhDs: a Cambridge cohort study

  1. Absolutely fascinating, Rachel! Thanks immensely for doing this. Although, as you say, it is only a tiny and untypical sample of History PhDs, its still provides a very useful sense of what is happening at the top of the job market. I’m very tempted to do this for my own cohort (Warwick 2009), though I don’t have the time at the moment. What would be even better would be for some sociologist (or contemporary historian) to do this for a series of cohorts (several different years, several different institutions) and follow up with interviews to get a sense of how much of these outcomes is choice and how much is necessity, and their general satisfaction with where they’ve ended up.


    • I’m currently trying to work out a way to do something similar for a cohort of non-Oxbridge PhD students, though that’ll be trickier. If you do have time to something about your cohort, that’d be very interesting as well, though the smaller the group you start from, the harder it is to know what’s noise and what’s useful data. I also agree that we really need some more detailed study to look at career paths etc. Did you know that the AHRC did a couple of such studies for the students they had funded? The reports are On the Right Track from 2007 and Career Paths of AHRC funded PhD students from 2012.


  2. I finished my PhD at Cambridge in 2008, and have kept a similar spreadsheet for those graduating in the 2007-8 year. I only considered those graduating from the history faculty, so didn’t include those from HPS etc as you did, who worked on historical topics. My outcomes were similar: currently (7 years out), fewer than 50% have a permanent academic job.


    • It’s very helpful to know that my figures aren’t out of line with others. I suspect including people from other Cambridge departments doesn’t make much difference to the overall balance of outcomes (apart possibly from including ASNaC, since medievalists have worse career prospects).

      Would it be possible to put your spreadsheet (minus any private information, obviously) online, as additional data for this discussion, because we do really need more hard evidence when discussing the topic?


    • It’d be very interesting to find out in more details about career paths etc, but I didn’t include details of that because it’s much harder to get that consistently from academics’ webpages. Some give full details, some are very vague. You’d probably need to contact them directly to ensure you got accurate information, Because of limits of time and resources, I’ve stuck to desk research for the moment.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s