The thesis committee is there to stop the natural tyranny of the PI

Gerty-z of Balanced Instability blog posed an age-old problem in post-graduate education.

I was talking to a graduate student the other day. It was a hallway interaction, she had not searched me out for advice. I have known this grad student for several years, and she is one of the superstars in a highly-ranked graduate program. By every metric, she should be graduating. Now. Turns out, her advisor has been suggesting that she stick around for another year or two.

ruh roh! Conflict of Interest raises its ugly head.
I bring this up because this is not the first time I’ve heard a similar story. In fact I’ve heard of what appears to be at least one entire department that is riddled with this tendency to prolong the graduate school interval as long as possible, seemingly only to extract more value out of productive trainees.

The story continues:

As I pressed further, Pre-doc superstar told me that she is the only person in the lab that knows how to work the Magical Data Machine. She is also, bar far, the most productive person in Dr. Advisor’s lab. If she sticks around for longer, she will probably publish one or two more papers. But she will have a LONG graduate career. I think that Dr. Advisor is thinking more about his own lab than the career of Pre-Doc Superstar.

Gee, you think? We can all see this…from the outside. Are we so able to see these situations, which are often slightly more subtle, in our own laboratories?
Comrade PhysioProf was all over it in the comments.

It is the responsibility of the the thesis committee to rein in this kind of PI behavior.

He’s spot on. This situation is wrong. It is absolutely ridiculous that graduate education has stretched up to a median of 6 years. Now of course there are going to be individual cases. Sometimes folks run into a situation where extra time is really needed. An initial stint in a lab that went poorly and a smart student needs a re-start. Fine.
This is what the doctoral or thesis committee is for. To make those decisions about timeline, when to defend and how much work is required to deserve the PhD.
In this, however, they are supposed to be on the side (so to speak) of the student. Why?
Because the PI is inextricably conflicted by his or her own scientific and career goals. Those goals tend to prefer that trained and data-pumping individuals stick around as long as possible. That papers get finished, even when the reviews come back demanding another 6 mo or 12 mo of work from precisely that model in which the student is expert.
It is not even trivial to understand the motivations that the supervising PI / doctoral advisor has to keep graduate students in the lab as long as possible.
And they have all kinds of tricks, most of them no doubt totally subconscious. Sit on those paper drafts. Insist the paper *has* to go to a higher falutin journal (and of course the longer the trainee has been working at a project, the greater the perceived need to show it was worth it with an exceptionally high IF journal acceptance, right?) Insist the data, while statistically significant, aren’t “pretty” enough.
Then there are the more overt ploys. Plead tenure or Full Prof promotion fears. Appeal to funding concerns (real or not). Play the-science-is-most-important cards. Act the ass about the productivity of someone else’s doctoral student when on her committee so as to keep the departmental average training duration as high as possible.
All of this is natural and inevitable. People have intrinsic motivations and goals that are in conflict. The key is to set up administrative structures that minimize exploitation of others. The doctoral committee is one such structure and it should, in my view, function as the gatekeeper that negotiates the differing interests of student and doctoral supervisor.