I am going to dispute this bit: "Obviously, many people leave. We're talking on the order of 90%. That's how it's supposed to work." This is _not_ how it is supposed to work. The publication half life of astronomers who produced their first paper 35 years ago was 30 years, i.e. about 50% of those who completed a Ma/PhD were still in academia 30 years later. That same number today is 5 years. The figures are similar for other fields.
Nobody decided it was a good idea to filter out so many people. This is something that has happened in large part because of the funding environment and the financial stresses on universities. It's also a result of success. We now educate a lot more people to a high standard and many of them aspire to do research/academia. Yet, we have not catered for them.
The thing is, because the drop-out rate was so much lower in the past, academia is set-up so that most of the career rewards come at the end. Salaries are heavily skewed to senior academics. Your ability to supervise your own research students comes with a permanent position. As a senior academic you become more successful at winning grants (the Mathew Effect) and with that comes the ability to travel to conferences and do the kind of exciting work you want to. You also get have the kind of stability that lets your start a family. In later life emeritus status gives you the freedom to pursue the work you want to without other distractions.
To do any of that though, you have to get past the postdoc stage. In the past one postdoc was common. A friend of mine is on his sixth postdoc. Less than two before getting a tenure track job is now unusual. This is _not_ how it is supposed to work. This is the system not being updated for new realities.
You might be interested to know that the lifetime earning potential of a person with a bachelor's degree is greater than someone with a PhD. If you stick with an academic career rewards come in the form of travel and the things I indicated above. But if you drop out, for every year you stayed in your earning potential is lower. This is a cruel situation. It's also a waste; training a scientist takes a lot of state resources, if we're not using them for the work they are best suited that's suboptimal.
I liked much of your post. The truth is somewhere in the middle, the rewards are worth it for anyone who can make it through. I just want you to know that the system was not designed to be this way. The current situation is a recent development
On the point you disagree with: it's impossible to determine at age 20 who will be a good PhD student, who of those will be good independent researchers, and who of those will be good faculty. Across all disciplines and sub-fields, I would be most confident in those that have a huge pool of talented students from which to eventually choose a small number of the very best to be faculty. It doesn't work as nicely as that in practice (see all caveats in the article, and the previous one), but I seriously doubt that any field where 50% of PhD students become faculty is a healthy one. (Cue jokes about astronomers.)
I sympathise with some of the other points, except any talk of lifetime earning potential. We academics have gone to a lot of trouble to convince ourselves that we don't care about it, and I would rather you didn't bring it up.
Of the people who earned their degrees 40 years ago It was just about every field where 50% were still publishing after 30 years. So, those are the people who taught you and I. If you think they did a decent job that counters your rebuttal. The reality is that half of any PhD cohort would be good faculty with the right opportunities.
Academics don't care about earning potential because they get to do what they want to for a job. But if you force them back into the workforce by drying up opportunities after two postdocs, then lifetime earning potential matters because they've sacrificed a great deal of it, without having gotten the pay off.
Thanks for the link. I'm suspicious. If total numbers of research faculty roughly double over a professor's career, then, for a 50% survival rate, on average each professor trains 4 PhD students and 2 of them get a faculty job. That seems very few PhD students per professor. I also don't know anyone in my field, no matter how old, who would claim that there was a time when 50% of PhD students got faculty jobs. So I'm sticking with my view that in a profession that aims to select the very best practitioners across a range of highly specialist skills, it's natural that only a minority "make it".
If hard data and a refereed analysis of it in PNAS can't convince you, then I don't know what will. For what it's worth, I've heard plenty of anecdotes from senior faculty that support it.
I didn't say it's wrong (what do I know), I just find it very surprising, for the reasons I gave. I can certainly believe (and the data in their Fig. 4 bear this out) that the number varies hugely within fields, and I would expect also in sub-fields. I'm tempted to write more about this sort of thing, but, as you see, it's getting way outside my zone of expertise.
I saw that video too. I’ve seen some of her others, notably the “nuclear power will not save us” one. As a layperson, I don’t really have the capacity to judge, but she always seems pretty reasonable.
I loved the life we had in graduate school.
It's amazing how many people complain (on social media, of course) what a grind and a horror grad school is. I guess they chose poorly.
I am going to dispute this bit: "Obviously, many people leave. We're talking on the order of 90%. That's how it's supposed to work." This is _not_ how it is supposed to work. The publication half life of astronomers who produced their first paper 35 years ago was 30 years, i.e. about 50% of those who completed a Ma/PhD were still in academia 30 years later. That same number today is 5 years. The figures are similar for other fields.
Nobody decided it was a good idea to filter out so many people. This is something that has happened in large part because of the funding environment and the financial stresses on universities. It's also a result of success. We now educate a lot more people to a high standard and many of them aspire to do research/academia. Yet, we have not catered for them.
The thing is, because the drop-out rate was so much lower in the past, academia is set-up so that most of the career rewards come at the end. Salaries are heavily skewed to senior academics. Your ability to supervise your own research students comes with a permanent position. As a senior academic you become more successful at winning grants (the Mathew Effect) and with that comes the ability to travel to conferences and do the kind of exciting work you want to. You also get have the kind of stability that lets your start a family. In later life emeritus status gives you the freedom to pursue the work you want to without other distractions.
To do any of that though, you have to get past the postdoc stage. In the past one postdoc was common. A friend of mine is on his sixth postdoc. Less than two before getting a tenure track job is now unusual. This is _not_ how it is supposed to work. This is the system not being updated for new realities.
You might be interested to know that the lifetime earning potential of a person with a bachelor's degree is greater than someone with a PhD. If you stick with an academic career rewards come in the form of travel and the things I indicated above. But if you drop out, for every year you stayed in your earning potential is lower. This is a cruel situation. It's also a waste; training a scientist takes a lot of state resources, if we're not using them for the work they are best suited that's suboptimal.
I liked much of your post. The truth is somewhere in the middle, the rewards are worth it for anyone who can make it through. I just want you to know that the system was not designed to be this way. The current situation is a recent development
On the point you disagree with: it's impossible to determine at age 20 who will be a good PhD student, who of those will be good independent researchers, and who of those will be good faculty. Across all disciplines and sub-fields, I would be most confident in those that have a huge pool of talented students from which to eventually choose a small number of the very best to be faculty. It doesn't work as nicely as that in practice (see all caveats in the article, and the previous one), but I seriously doubt that any field where 50% of PhD students become faculty is a healthy one. (Cue jokes about astronomers.)
I sympathise with some of the other points, except any talk of lifetime earning potential. We academics have gone to a lot of trouble to convince ourselves that we don't care about it, and I would rather you didn't bring it up.
Of the people who earned their degrees 40 years ago It was just about every field where 50% were still publishing after 30 years. So, those are the people who taught you and I. If you think they did a decent job that counters your rebuttal. The reality is that half of any PhD cohort would be good faculty with the right opportunities.
Academics don't care about earning potential because they get to do what they want to for a job. But if you force them back into the workforce by drying up opportunities after two postdocs, then lifetime earning potential matters because they've sacrificed a great deal of it, without having gotten the pay off.
Here is the reference on cohort survival, in case you think I'm making it up: https://www.pnas.org/doi/epdf/10.1073/pnas.1800478115
Thanks for the link. I'm suspicious. If total numbers of research faculty roughly double over a professor's career, then, for a 50% survival rate, on average each professor trains 4 PhD students and 2 of them get a faculty job. That seems very few PhD students per professor. I also don't know anyone in my field, no matter how old, who would claim that there was a time when 50% of PhD students got faculty jobs. So I'm sticking with my view that in a profession that aims to select the very best practitioners across a range of highly specialist skills, it's natural that only a minority "make it".
If hard data and a refereed analysis of it in PNAS can't convince you, then I don't know what will. For what it's worth, I've heard plenty of anecdotes from senior faculty that support it.
I didn't say it's wrong (what do I know), I just find it very surprising, for the reasons I gave. I can certainly believe (and the data in their Fig. 4 bear this out) that the number varies hugely within fields, and I would expect also in sub-fields. I'm tempted to write more about this sort of thing, but, as you see, it's getting way outside my zone of expertise.
I saw that video too. I’ve seen some of her others, notably the “nuclear power will not save us” one. As a layperson, I don’t really have the capacity to judge, but she always seems pretty reasonable.
You could probably say the same thing about a lot of conspiracy theorists. :-)
My guess is that most of it is reasonable, but not all, and that's what makes it tricky. Keep your grains of salt handy.