A while ago Eve Marder's article on the number of PhD admissions stirred up the natives. It's been incubating or brining or gestating (pick your favorite metaphor) in my mind for a while.
I've read that it's a lot like r- and K- selection. To increase your fitness in an r-selection scheme, and the number of your genes in the next generation, do you produce lots of offspring, invest very little in any of them, but hope enough survive to make it (think insects, or fish). Or, as a K-selection type, do you produce a smaller number, invest heavily in them (think whales and elephants)? Of course the point of fitness is leaving your genes to the next generation and any concern/investment/strategy for the offspring beyond survival is still about leaving your genes and your fitness. Needless to say trainees are not genes, and improving one's academic fitness is not necessarily the goal of Good People (though as is true of BSD's, YMMV).
One perspective of the olders is that the youngers are actually living this, and therefore in the worst place to judge. If it was true that the youngers were saying "don't limit admissions", one would be tempted to argue that they perceive themselves as being at risk. If the youngers were saying "do it to Julia", then one would be tempted to accuse them of arrogance (yes, limit admissions, but not me). Instead what I am hearing is the rational concern that verges on anxiety about the future. Their future. Their hopes. And no K-selected parent can ignore that. And, no thinking scientist who cares about the future of the field.
Part of Marder's arguments that irritated me were:
Admissions committees are bad at predicting who will end up deciding to stay in science. ....I f we knew how to spot the 25% of our applicant pool ... [who will] become an outstanding scientist, .... might make it sensible to decrease the size of our incoming cohort. But it would be counterproductive and sad to limit our numbers and then effectively lose the creative, determined and possibly unconventional individuals who might not make it through a more restricted gateway.
It's not that I disagree. Admissions committees are political entities with multiple pressures and goals, many of which are not about the candidates. The problem with Marder's logic is that we should then admit everyone who wants to be a grad student. Why limit it to what we have now? There might be some great future scientist that we are ignoring with the current standards. I am guessing Marder would say no to that scheme. The argument against admitting everyone who applies is that admissions committees can make some decisions, and there are some people obviously not suited for a career in science. But are admissions committees, right now, only dealing off the bottom of the deck, only excluding those we know won't make it. Of course not. We are making hard choices. We are drawing a line. The question becomes where do we draw that line?
Getting into grad school (even right now) is a classic supply/demand problem in economics. While the demand for seats in grad school are higher than the available supply of those seats, that supply is fairly elastic. The real supply/demand problem occurs for faculty positions. We've set the supply for grad school at an artificial, predetermined level. This level, the line we draw, is determined by a number of factors, mostly funding, that have little to do with the excellence of the candidates and their potential for being scientists. It certainly doesn't think about the future much (all my grad students get postdocs).
I still think the real issues go back to what happens to those trainees. Marder can argue that we celebrate all those who don't make it. She can say that it takes grad school to figure out if you have what it takes to make it. But that's not really thinking about the people in those roles. I've said before that it is a bogus and somewhat condescending argument that "PhD programs in biomedical sciences train you for many careers". No one enters a PhD program because they think its a good way to become a high school teacher. The question remains: is there a better path to the "alt.career" endpoints?
Yet, there is still validity in the argument it is not obvious who will be a good scientist at the point of admission. Previous research experience, if it is real research experience, is a strong indicator. One of the advantages of the British/Aussie/Kiwi scheme of doing "honors" as the fourth year in a three year degree is that there is a condensed year where it is possible for the student to figure out what its like and what they want.
The discussion about where to draw the admissions line needs to be an ongoing one. It needs to involve people all along the career pathway, not just the Eve Marders of this world. And to those who say that blogs/twitter/social media are a waste of time, I say: feh. It is one way those other career-stage people get a voice.