In “The Ever-Tightening Job Market for Ph.D.s: Why do so many people continue to pursue doctorates?”, Laura McKenna reviews the data on how terrible grad programs and the academic job market are, then goes on to ask: “Why hasn’t all this information helped winnow down the ranks of aspiring professors—why hasn’t it proved to be an effective Ph.D. prophylactic?” Having observed and participated in the mass delusion, I have some possible answers:

1. It’s a way to (pointlessly) delay adulthood.

2. Fear of the job market.

3. Don’t know what else to do.

4. Magical thinking (despite the numerous articles out there, like mine) that attempt to dissuade it). I think this is the biggest issue. In addition, there seems to be a Lake Woebegone effect: Everyone thinks they’re going to be above average.

5. Contrary to what grad students often say, in many disciplines and programs grad school is pretty easy and fun! You get to hang out on campus, think about ideas, take a minimal number of classes, do a bit of teaching, and have copious free time. Also, let me be euphemistic and say that many straight guys spend a lot of time with female undergrads. The problem is that, as time advances and your priorities start changing (want a real life / job, date people who don’t date people whose lives aren’t together, etc.), reality starts to intrude. Many grad students have an unacknowledged Peter Pan complex.

6. For most, academic success has been rewarded every step of the way (thanks to Hacker News reader mathattack for this point). The individuals who’ve gotten the most mileage listening to their teachers are also the ones who most need to stop listening to them. Professors are very keen on producing more professors and reproducing themselves, even though doing so is often not in the best interests of a particular individual.

7. People mistakenly focus on the outliers who accomplish major, important breakthroughs and think that they’ll be like the outliers, not the medians. This is another variant of the Lake Woebegone effect.

Note that a few fields (econ, computer science) appear to have relatively robust job outcomes for PhDs, so some of the above likely doesn’t apply to them.