The start of a new school year is a great time to meet new colleagues and catch up with those you haven’t seen in a while, especially after being gone for a year. But every year you come back to graduate school there are inevitably a few faces who aren’t there any more. This is just as much a part of the academic rhythm as anything else.

There’s a statistical side to this. A 2011 study based on responses from 138 history departments found that the median completion rate for students entering PhD programs in history between 1996 and 2005 was 42%. Yes, you read that right. At least at that point in time, any given student beginning a PhD program was more likely not to finish than to earn a degree. The good news is that this number has been improving over time, even dramatically when compared to half a century ago. It is a phenomenon that has received some attention, of course, but, given the normality of leaving a PhD program, it is still a surprisingly taboo subject within the halls of the university.

It seems likely that changes that have lowered the attrition rate have also contributed to failing to treat not finishing as a normal outcome of entering a PhD program. Increasingly, incoming graduate students are given funding packages that, at least in theory, will provide them enough support to carry them through the entirety of their programs. Despite reasonable complaints about the insufficiency of these funding packages, this is a far cry from the situation in programs twenty years ago (or even less). Instead of giving all or even most students guaranteed funding packages, departments would admit large numbers of PhD students without funding and use the first year of the program to weed out large numbers of them, asking those who did not make the cut to leave. Departments could then offer some support to students who had passed this first hurdle.

There is not much nostalgia for those halcyon days of cutthroat competition, although professors who came through this system are sometimes at pains to explain to today’s graduate students how we have it so much better than they did. This undeniable truth notwithstanding, these structural shifts in graduate education actually raise some troubling questions about doggedly low completion rates for PhD’s.

If departments are no longer using the first year or two of the program as a kind of a tryout for the remainder of the PhD and systematically over-admitting and then “culling” cohorts, then this would drive down “attrition” by a significant margin. It would not, however, necessarily say anything about the likelihood of students to complete the program after they had navigated that first trying year or two. Moreover, it does not preclude the possibility that some of these pressures have not been so much eliminated as displaced beyond the PhD program itself. For departments now admitting fewer students but with higher expectations, it is both more possible and desirable (given the substantial investment a multi-year funding package represents) to extend offers of admission to only the most qualified candidates. Anecdotally, this is leading to cohorts comprising an increasingly large number of students who have already completed MA’s. Of course, when these same universities run large MA programs that attract students who are hoping to advance to a PhD and willing to pay tens of thousands of dollars in tuition for a chance to do so, the changes to graduate education described above don’t seem quite as positive as at first glance.

Up to here, it may seem like I’ve been describing not completing a PhD as an inherently bad outcome. I actually don’t believe that’s the case. I’ve seen people leave PhD programs for lots of different reasons. Sometimes students realize early on that this type of work isn’t a good fit for them. In those situations, getting out early is probably the best solution. On the other end of the spectrum, getting a PhD is a long process. Any time someone takes on a project that will take something like seven years to complete, you have to account for the possibility that what people want from life will change along the way. And sometimes when that happens, again, maybe the best thing is to move on.

Maybe this is to acknowledge that I’m also an attrition statistic. I had to leave one program to follow an adviser to another institution. That’s probably as good a reason as any to leave behind a PhD; and even though I wasn’t quitting “the PhD,” I still had kind of mixed feelings – I was leaving a place I had really enjoyed, people who meant a lot to me, and a bright future I had imagined for myself there.

But I’ve also seen very different situations. Students not necessarily kicked out but pushed aside. Who didn’t get the support they needed. And sometimes very much worse.

So here’s my pitch for why we should have open, probing, and honest discussions about “attrition”:

Because not completing a PhD program is a likely outcome for entering students. If we don’t admit this, we are lying to ourselves and prospective students.

Because a completion rate around 40% is probably lower than it should be and there are things we can do about it to make graduate school better for everyone. This includes educating prospective students about graduate school so they can make better decisions about whether it is the right route for them, providing community and guidance for students at all stages of the program, and checking professors who emotionally, verbally, or sexually abuse their students.

Because mental health in graduate school is a serious issue and because acknowledging there are other options will help us be happier and healthier, whether we stay or not.

Because we are increasingly recognizing that even completing a PhD is not a one-way ticket to a job as a professor and because students who do not complete PhD’s will, like many of their colleagues who do, utilize their skills as historians (etc.) in other professional fields. Training at all stages of the PhD should reflect that.

Because, as intellectuals, if we can’t talk honestly about some things things, then we shouldn’t be trusted to talk about others.