5 ms·

Suppose that the skill of the elites you produce is a normal distribution. There are only jobs for the best 0.001% of people in the field, but to get enough can

Suppose that the skill of the elites you produce is a normal distribution. There are only jobs for the best 0.001% of people in the field, but to get enough candidates that reach that 0.001% bar, you have to produce the other 99.999%. You're sifting through sand to try to get the flakes of gold. There might not be a need for the grains of sand that are produced by the system, but without filling the system with them, you wouldn't have been able to extract the gold that's mixed in. This is why elite overproduction occurs. Many of them may be perfectly competent individuals, but as long as employers are competing to hire only the best of the best, the system will have to massively overproduce to satisfy that demand.

stdbrouw 5y ago
Except that's not what is going on. Usually, when the time comes to look for a new postdoc or a new professor, there are plenty of candidates of very high talent and future potential and committees end up deciding based on low-information signals such as publication count because otherwise they'd have no clue who to prefer. Of course you want there to be more PhDs than professors to allow for an additional selection for the very top talent at that point, but the current situation is that we're producing too much sand as well as too much gold.
- philipov 5y ago
  The place where the sand/gold analogy breaks down is that the goalposts are constantly shifting due to competition. If I've hired the best person in the world, and they're the best out of 100,000 candidates, then if someone else also wants to find the best person in the world, they'll need to be better than the current best person in the world. It may take 200,000 candidates to find that person. To keep finding someone better than the current best person in the world, you constantly need to increase the amount of sand. To find the new best person in the world, you'll produce geniuses who are as good as the old best person in the world, but they're no longer good enough to be the NEW best person in the world. So yes, we're overproducing on gold, but now we're looking for something even better than gold.
  - stdbrouw 5y ago
    This presumes (a) a highly reliable ability to identify the "best person", which we simply don't have, (b) the idea that "best" is unidimensional and unrelated to the context one finds oneself in and (c) the idea that such a "best" person will also be the one who will do most to advance science. Just take a look at a century of Nobel prize winners and you'll see that none of those assumptions hold.
    - philipov 5y ago
      The assumptions don't hold in reality, but the myth of them holds in the minds of people doing the hiring, and that is enough.
  - ipaddr 5y ago
    They would cost too much. You take the second best and claim they are the best person on the market.
    - lupire 5y ago
      It doesn't cost too much, because the employer doesn't pay for it; the candidates do.

stdbrouw 5y ago

Except that's not what is going on. Usually, when the time comes to look for a new postdoc or a new professor, there are plenty of candidates of very high talent and future potential and committees end up deciding based on low-information signals such as publication count because otherwise they'd have no clue who to prefer. Of course you want there to be more PhDs than professors to allow for an additional selection for the very top talent at that point, but the current situation is that we're producing too much sand as well as too much gold.

philipov 5y ago
The place where the sand/gold analogy breaks down is that the goalposts are constantly shifting due to competition. If I've hired the best person in the world, and they're the best out of 100,000 candidates, then if someone else also wants to find the best person in the world, they'll need to be better than the current best person in the world. It may take 200,000 candidates to find that person. To keep finding someone better than the current best person in the world, you constantly need to increase the amount of sand. To find the new best person in the world, you'll produce geniuses who are as good as the old best person in the world, but they're no longer good enough to be the NEW best person in the world. So yes, we're overproducing on gold, but now we're looking for something even better than gold.
- stdbrouw 5y ago
  This presumes (a) a highly reliable ability to identify the "best person", which we simply don't have, (b) the idea that "best" is unidimensional and unrelated to the context one finds oneself in and (c) the idea that such a "best" person will also be the one who will do most to advance science. Just take a look at a century of Nobel prize winners and you'll see that none of those assumptions hold.
  - philipov 5y ago
    The assumptions don't hold in reality, but the myth of them holds in the minds of people doing the hiring, and that is enough.
- ipaddr 5y ago
  They would cost too much. You take the second best and claim they are the best person on the market.
  - lupire 5y ago
    It doesn't cost too much, because the employer doesn't pay for it; the candidates do.