“If a prospective student can’t, through no fault of their own, properly evaluate the disvalue of not getting a job, this changes the way we need to assign responsibility for the choice.”
Are there obstacles to rationally deciding whether to go to graduate school in philosophy. If so, what follows from that? These are the questions taken up in the following guest post* by L.A. Paul, professor of philosophy and cognitive science at Yale University, and John Quiggin, ARC Australian Laureate Fellow in the School of Economics at the University of Queensland.
The Transformative Experience of Graduate Study in Philosophy
by L.A. Paul & John Quiggin
In a few months, thousands of students will be starting graduate school. For some, this will seem to be part of an obvious and natural progression. They’ll move naturally from undergraduate to graduate study, without much reflection or deliberation. For others, the choice is more fraught: the stakes are high, tenure track jobs are scarce, and the proposed PhD does not provide an obvious route to nonacademic employment. For those deciding on whether to do graduate work in philosophy, this is often the situation they find themselves in.
Imagine yourself as the student in question. How should you proceed? For an important decision like this, standard decision theory suggests an analysis based on expected utility theory, or one of its generalizations such as rank-dependent expected utility, also known as risk-weighted expected utility theory, explored by Quiggin 1982 and Buchak 2017.
To choose rationally, using standard decision theory, first, determine the set of (relevant) possible outcomes. Assume you know what you would do if you decided not to go to grad school in philosophy (call this outcome ~PhD). Also assume you know what you would do if you completed grad school but didn’t get a tenure track job (call this outcome ~J). Now, attach utilities to the three outcomes of special interest: u(~PhD), u(~J), and u(J), where J is getting a tenure track job—the preferred outcome.
If you decide not to go to graduate school, outcome ~PhD is certain. If you think you want to go, you have two possibilities left. If u(J) > u(~J) ≥ u(~PhD), that is, the utility of getting a PhD but not getting a tenure track job is higher than not getting the PhD, then the option of going to graduate school yields a preferred outcome with certainty, and so dominates. On the other hand, if u(J) > u(~PhD) > u(~J), you need to attach weights w(J) and w(~J) to the outcomes J and ~J. These may be either probabilities (a standard expected utility approach) or decision weights (as in generalized expected utility). Assuming that the weights sum to one, the decision to attend grad school is optimal if and only if w(J)u(J)+ (1-w(J))u(~J) > u(~PhD).
However, all this rests on the assumption that the value of attending graduate school can be anticipated and evaluated. And, unfortunately, this assumption is questionable. Why? Because grad school can be transformative (Paul 2014). That is, you can’t fully anticipate the salient epistemic changes it involves until you undergo it, and your preferences can be changed by the process of epistemic transformation it involves. If so, then, before you go to grad school, at t1, when you are attempting to assess the utilities of the relevant end-of-grad-school outcomes at t2, your utility for outcome ~J (at t2) is not well defined. Thus, you can’t properly anticipate your future preferences.
Put in terms of our analysis, suppose when you decide, ex ante, at t1, to go to grad school, you prefer to get your philosophy PhD, regardless of the outcome: u(J) > u(~J) ≥ u(~PhD). You think it’s worth doing even if you don’t find a tenure track job when you finish. But by the time you’ve finished grad school, at t2, your preferences about the importance of finding that tenure track job have changed. Now, ex post, at t2, you find that u(J) > u(~PhD) > u(~J): that is, now, getting the PhD and failing to find a tenure track job is worse than never going to grad school in the first place!
Our point is that this possibility needs to be recognized when advising prospective students on whether to get a PhD in philosophy (or more generally, when advising students whether to go to grad school in any field where employment prospects are dim). The crucial question is whether students can prospectively assess how much they’ll disvalue not getting that tenure track job when the training is done (or, put differently, how much they’ll disvalue the life they’ll find themselves living as an unsuccessful job candidate). An economist committed to the assumption of unbounded rationality would claim that of course students can assess their future preferences: just as in the case where a person chooses to take an addictive drug, these choices can be made in full knowledge of the changes in preferences they will induce (see, e.g., the theory of rational addiction, Becker and Murphy 1988).
We don’t find this claim convincing. In this case, we think it is plausible that, at t1, students lack the ability to imaginatively project themselves into their future job-candidate “shoes” in the way they’d need to in order to properly assess the disvalue, at t2, of failing to find a tenure track job.
This is, again, because the process of becoming an academic can be transformative: your future utilities for failing to get a tenure track job cannot be properly assessed until you’ve actually become an academic.* If a prospective student can’t, through no fault of their own, properly evaluate the disvalue of not getting a job, this changes the way we need to assign responsibility for the choice. In particular, graduate schools have more responsibility to mitigate the possibility of adverse results, for example, by providing an education that lives up to the claim of being useful and desirable outside the academy as well as within it.
*We observe that the issues described above are not relevant to all students. Some may not find the process transformative. Others will push on the academic path as long as they can, regardless of any evaluation of the consequences. And for some (Botts et al 2014) the issues involved in the choice of whether to go to grad school may be quite different.
Becker, G.S. and Murphy, K.M. (1988), ‘A Theory of Rational Addiction’, Journal of Political Economy, 96(4), 675–700.
Botts, T., Kofi Bright, L., Cherry, M., Mallarangeng, G., and Spencer, Q. (2014), ‘What Is the State of Blacks in Philosophy’, Critical Philosophy of Race, 2, 224-242.
Buchak, L. (2017) Risk and Rationality, Oxford University Press, USA.
Paul, L. (2014) Transformative Experience, Oxford UK.
Quiggin, J. (1982), ‘A Theory of Anticipated Utility’, Journal of Economic Behavior and Organization, 3, 323–43.