Human capital adjustment attributes lots of GDP growth to extra education, but if this education is mostly signaling, then we're getting more growth from new tech than we thought.
I meant to comment on your last post about this, but where you say "Graduate degree holders dropped by 13-17 points in a decade", it's not like we are measuring the same groups of people. If you allow more people to get graduate degrees, then that could bring down the average scores since you were previously very selective and now relatively lesser. If anything, though, that lines up with your broader point about credentialism.
I feel like you might be over-indexing on just a few results when you claim that education doesn't cause any increase in skills. I'm theoretically open to the idea that it's nothing but signaling, but I'm not convinced just on the strength of the results of a few math and reading tests. For one thing, there's clearly a huge variety of different environments that all go by the name of "higher education," from online two-year bachelor's programs, to community colleges, to state universities, to ivy-level institutions. That's to say nothing of the range of disciplines available — a master's in Post-Colonial Studies and a PhD in Optical Physics both count as having a graduate degree, for instance.
I also think you're too quick to dismiss the possibility that education increases more specialized skills even though it doesn't improve basic math and reading — after all, classes aren't focused on teaching those basic skills anymore once you get to the higher ed level, and it seems plausible that many students only bring those skills up to a "good enough" level before focusing more narrowly on skills specific to their field of study. If that's the case, you could explain the drop in average scores with selection effects while preserving the possibility that more specialized skills could still be getting taught.
My prior here is that just as in many debates, reality lies somewhere between the "it's all signaling" and "it's all useful skills" positions. Common sense would suggest that the lowering of standards to allow more people to graduate from high school or with a bachelor's degree really are just credential inflation, and it also seems likely that a lot of the "studies" degrees and maybe the accelerated online programs don't teach people that much. But I still think that a lot of formal education does have value, and I would be interested in seeing more work trying to disaggregate useful vs. useless programs.
In its most basic form, is the argument that I didn’t learn anything in college, or that what I learned in college was something I would have learned anyways without going?
The reason I ask is that I clearly recall learning stuff in college.
More the second, also that one learns things which help you pass the tests of skill presented in college, but do not actually make you a more productive worker.
Perhaps also "you learned less than you would've learned if you'd been one of the smaller number of people who went 20 years earlier" is a possibility.
I think there is a tendency to "learn" in college in the sense that you memorize some facts or ways of solving problems that allow you to regurgitate on a test and pass a class without actually "learning" much of anything.
I saw this has someone who took a STEM major at a college which is decently well known for it - I can count on one hand the number of classes I took that I feel like actually taught me meaningful concepts that I retained and use in my life/career after college.
Excellent point. One of the few econ blog posts that needs to be bookmarked for regular, long-term reconsideration.
Note that this does then splinter the question about the break in TFP growth in 1970 into what changed to cause TFP slowdown and also why literacy rates also hit a ceiling in 1970.
As last time and as Caplan you use literacy rate to say that graduate studies are useless. But those are different people. Let's say that some people will never go above math and literacy rates (that also change but who cares) because of IQ. Still all the people who go to graduate studies are above those levels so you are not saying anything about them. Also a few studies are coming out on the negative effect of lockdowns on graduate knowledge on income, so there is an effect. You and Caplan compare two different sets of people, it makes no sense. Unless you wanted to say that some people should go to trade school at 14, and in that I agree
Great post, reminds me of 'Self-Selection and the Diminishing Returns of Research' by Ekerdt & Wu. They show that a lot of the diminishing research productivity effect from Bloom's 'ideas are getting harder to find' paper is just a transitional effect of decreasing average researcher ability as the pool of college graduates and researchers expands while the ability distribution stays the same. Two positive results: the Great Stagnation wasn't as big and new ideas aren't quite so hard to find as we thought.
My theory is that the advent of computers made regulation easier to write and enforce, so we get a vast increase in regulation, which slows down innovation.
I meant to comment on your last post about this, but where you say "Graduate degree holders dropped by 13-17 points in a decade", it's not like we are measuring the same groups of people. If you allow more people to get graduate degrees, then that could bring down the average scores since you were previously very selective and now relatively lesser. If anything, though, that lines up with your broader point about credentialism.
I feel like you might be over-indexing on just a few results when you claim that education doesn't cause any increase in skills. I'm theoretically open to the idea that it's nothing but signaling, but I'm not convinced just on the strength of the results of a few math and reading tests. For one thing, there's clearly a huge variety of different environments that all go by the name of "higher education," from online two-year bachelor's programs, to community colleges, to state universities, to ivy-level institutions. That's to say nothing of the range of disciplines available — a master's in Post-Colonial Studies and a PhD in Optical Physics both count as having a graduate degree, for instance.
I also think you're too quick to dismiss the possibility that education increases more specialized skills even though it doesn't improve basic math and reading — after all, classes aren't focused on teaching those basic skills anymore once you get to the higher ed level, and it seems plausible that many students only bring those skills up to a "good enough" level before focusing more narrowly on skills specific to their field of study. If that's the case, you could explain the drop in average scores with selection effects while preserving the possibility that more specialized skills could still be getting taught.
My prior here is that just as in many debates, reality lies somewhere between the "it's all signaling" and "it's all useful skills" positions. Common sense would suggest that the lowering of standards to allow more people to graduate from high school or with a bachelor's degree really are just credential inflation, and it also seems likely that a lot of the "studies" degrees and maybe the accelerated online programs don't teach people that much. But I still think that a lot of formal education does have value, and I would be interested in seeing more work trying to disaggregate useful vs. useless programs.
Good knowledge workers are constantly learning new useful skills... on the job. The vast majority of people retain no useful skills from university
+1 on this.
Personalizing the issue for clarification…
In its most basic form, is the argument that I didn’t learn anything in college, or that what I learned in college was something I would have learned anyways without going?
The reason I ask is that I clearly recall learning stuff in college.
More the second, also that one learns things which help you pass the tests of skill presented in college, but do not actually make you a more productive worker.
And because of that, I have forgotten 99% of what I learned in college. I re-learned statistics after not using it for two decades.
Perhaps also "you learned less than you would've learned if you'd been one of the smaller number of people who went 20 years earlier" is a possibility.
I think there is a tendency to "learn" in college in the sense that you memorize some facts or ways of solving problems that allow you to regurgitate on a test and pass a class without actually "learning" much of anything.
I saw this has someone who took a STEM major at a college which is decently well known for it - I can count on one hand the number of classes I took that I feel like actually taught me meaningful concepts that I retained and use in my life/career after college.
Excellent point. One of the few econ blog posts that needs to be bookmarked for regular, long-term reconsideration.
Note that this does then splinter the question about the break in TFP growth in 1970 into what changed to cause TFP slowdown and also why literacy rates also hit a ceiling in 1970.
As last time and as Caplan you use literacy rate to say that graduate studies are useless. But those are different people. Let's say that some people will never go above math and literacy rates (that also change but who cares) because of IQ. Still all the people who go to graduate studies are above those levels so you are not saying anything about them. Also a few studies are coming out on the negative effect of lockdowns on graduate knowledge on income, so there is an effect. You and Caplan compare two different sets of people, it makes no sense. Unless you wanted to say that some people should go to trade school at 14, and in that I agree
That is exactly what they are saying!!
Great post, reminds me of 'Self-Selection and the Diminishing Returns of Research' by Ekerdt & Wu. They show that a lot of the diminishing research productivity effect from Bloom's 'ideas are getting harder to find' paper is just a transitional effect of decreasing average researcher ability as the pool of college graduates and researchers expands while the ability distribution stays the same. Two positive results: the Great Stagnation wasn't as big and new ideas aren't quite so hard to find as we thought.
My theory is that the advent of computers made regulation easier to write and enforce, so we get a vast increase in regulation, which slows down innovation.
Anecdotal story
When I was about twelve or thirteen I read the tarzan series by edgar rice burroughs
Earlier this year, I gave the first book a reread
I was struck by the difficulty level of the reading versus the newer stuff I normally read
I do think there's been a bit of dumbing down in society amd in what we expect