Enjoyed the episode? Want to listen later? Subscribe by searching 80,000 Hours wherever you get your podcasts, or click one of the buttons below:

… professors say, “Of course you’re only talking about college Bryan because we all know that K-12, that stuff’s all useful.” I’m just like… all useful? Do you remember at all what was done? There is this weird Stockholm Syndrome that people seem to get, where as long as it’s far enough back in time, they fill in useful stuff every minute of every day.

Prof Bryan Caplan

Bryan Caplan’s claim in The Case Against Education is striking: education doesn’t teach people much, we use little of what we learn, and college is mostly about trying to seem smarter than other people – so the government should slash education funding.

It’s a dismaying – almost profane – idea, and one most are inclined to dismiss out of hand. But having read the book, I have to admit that Bryan can point to a surprising amount of evidence in his favour.

After all, imagine this dilemma: you can have either a Princeton education without a diploma, or a Princeton diploma without an education. Which is the bigger benefit of going to Princeton – learning, or convincing people you’re smart? It’s not so easy to say.

For this interview, I searched for the best counterarguments I could find and challenged Bryan on what seem like the book’s weakest or most controversial claims.

Wouldn’t defunding education be especially bad for capable but low income students? Shouldn’t we just make incremental rather than radical changes to policy? If you reduced funding for education, wouldn’t that just lower prices, and not actually change the number of years people study? Is it really true that students who drop out in their final year of college earn about the same as people who never go to college at all?

And while we’re at it, don’t Bryan and I actually use what we learned at college every day? What about studies that show that extra years of education boost IQ scores? And surely the early years of primary school, when you learn reading and arithmetic, are useful even if college isn’t.

I then get his advice on who should study, what they should study, and where they should study, if he’s right that college is mostly about separating yourself from the pack.

We then venture into some of Bryan’s other unorthodox views – like that immigration restrictions are a human rights violation, or that we should worry about the risk of global totalitarianism.

Bryan is a Professor of Economics at George Mason University and blogger at EconLog. He’s the author of three books: The Case Against Education: Why The Education System is a Waste of Time and Money, Selfish Reasons to Have More Kids: Why Being a Great Parent is Less Work and More Fun Than You Think, and The Myth of the Rational Voter: Why Democracies Choose Bad Policies.

Get this episode by subscribing to our podcast on the world’s most pressing problems and how to solve them: type 80,000 Hours into your podcasting app.

In this lengthy interview, Rob and Bryan cover:

  • How worried should we be about China’s new citizen ranking system as a means of authoritarian rule?
  • How will advances in surveillance technology impact a government’s ability to rule absolutely?
  • Does more global coordination make us safer, or more at risk?
  • Should the push for open borders be a major cause area for effective altruism?
  • Are immigration restrictions a human rights violation?
  • Why aren’t libertarian-minded people more focused on modern slavery?
  • Should altruists work on criminal justice reform or reducing land use regulations?
  • What’s the greatest art form: opera, or Nicki Minaj?
  • What are the main implications of Bryan’s thesis for society?
  • Is elementary school more valuable than university?
  • What does Bryan think are the best arguments against his view?
  • The specific effects of defunding education on low income students
  • Is it possible that we wouldn’t want success in education to correlate with worker productivity?
  • Do years of education affect political affiliation?
  • How do people really improve themselves and their circumstances?
  • Who should and who shouldn’t do a masters or PhD?
  • The value of teaching foreign languages in school
  • Are there some skills people can develop that have wide applicability?
  • Are those that use their training every day just exceptions?

The 80,000 Hours podcast is produced by Keiran Harris.


If you want to get the best education in the world for free, you can just move to Princeton and start attending classes unofficially. There’s almost no effort made to stop you. You just won’t get a diploma, which makes it near pointless, because college is more about impressing people than learning useful info.

And again, Washington State from what I understand, now allows kids to use a computer language in place of foreign language. Like, why not do that? People say, “no, no we need to do both.” People don’t have an unlimited amount of time! Shouldn’t teenagers be able to have a frigging childhood! Like how much of their childhood do you want to destroy with jumping through these stupid hoops?

If you wanna do almost anything in social science or humanities, and you are dismayed by the crummy job prospects… Can you do math? If so, why not just go and get an econ PhD and call what you’re doing economics X? It is literally true that you will probably have tenure as an economist, before you would have your first assistant professor job as a historian.

So when would you rather be a high school dropout looking for a job, today or 1950? So today there’s a very harsh stigma against high school dropout so you’re very limited. Now in 1950, there would have been a lot of good people that did not finish high school and so the stigma would have been a lot less. So we can see in the kinds of jobs that you can get as a high school dropout back in those days. You know it would not have been impossible to be a secretary and a high school dropout in 1950, you know, far from it. And many other jobs that would’ve been middle income jobs, would have still been open to you.

So what I say, when we’re thinking about the effects on the disadvantaged, you shouldn’t just think about how it might be worse for the really talented kid from the poor family, you should think about the average kid from the poor family. That’s one where I say at minimum, it’s just a lot more complicated, because there’s a big difference between changing the funding for one individual and changing the funding for a generation. Changing the funding for one individual, your intuition is totally fine. But if you change the funding for a whole generation, it changes the meaning of the education itself. And means that there are a lot of opportunities that the poor have lost in the moderate economy that they can have back again.

There are many graduate programs where almost the only thing you do with is to become a professor of that subject. Or you just don’t use it. So in that case, well you wanna be an English professor, well look at the jobs prospects for the people who are currently coming out with English PhDs and see how they’re doing.

Don’t ask yourself are you as good as those people. Ask what someone who didn’t know you, only knew what you were like on paper, would think whether you’re better than those people. Because the world doesn’t tell you … We don’t even need to go and get into overconfidence and self-centered bias. Let’s just say the world’s not fair and even if you’re awesome the world rewards being awesome on paper, not intrinsic awesomeness. Just accept this as a flaw in the world, and then consider that when you’re deciding whether or not you wanna try what you’re gonna do.

Totalitarian regimes are generally very uncreative. At best they can maintain the living standards before, usually can’t even do that. So this means that over time especially as long as there any awareness of what’s going on outside of the totalitarian country, they know things are getting better and better in a non-totalitarian world. And staying bad in the totalitarian word. So that’s one thing, is just the comparison group. You know this comparison group you and look at and see.

Another thing that’s going on is of course military competition, where if the non-totalitarian world keeps growing, they will have a big military edge over the totalitarian world and that’s another way the totalitarian may not be able to compete. Then there also just the question of maintaining the morale of inner elite, of the inner circle of leaders. Which I say in practice is the real problem for totalitarian regimes. It’s not like the Soviet Union couldn’t have just stayed the Soviet Union, but the problem was that the people, for its durability was that the people at the top lost their faith in their own system.

Which again has a lot to do with, again, there’s another system that you can look at and see it’s better. So anyway, put all of this together and said, if there ever were to be a time that a totalitarian regime basically took over the whole surface of the Earth. Then whether it was a bunch of them, or just one, you’ve removed most main reasons why it isn’t stable. In which case, maybe it could go one for 1000 years, 10,000 years. I guess compared to some of the other global catastrophic risks of permanent extinction of humanity is not bad. But still seems like a pretty bad outcome, nonetheless.

Articles, books, and other media discussed in the show


Robert Wiblin: Hi listeners, this is the 80,000 Hours Podcast, the show about the world’s most pressing problems and how you can use your career to solve them. I’m Rob Wiblin, Director of Research at 80,000 Hours.

I found today’s episode really fun and expect most subscribers will be interested even if it’s not directly related to their career.

If you find the conversation entertaining or informative, think about forwarding it to someone with a strong view about the value of education so they can enjoy it too.

Just a quick reminder that if you’re in Australia or the Netherlands and want to go to local effective altruism conferences this year, you’ll need to apply to go soon at eaglobal.org.

Here’s Bryan.

Robert Wiblin: Today, I’m speaking with Bryan Caplan. Bryan is a Professor of Economics at George Mason University, and a blogger for EconLog. He’s the author of The Case Against Education: Why The Education System is a Waste of Time and Money, Selfish Reasons to Have More Kids: Why Being a Great Parent is Less Work and More Fun Than You Think, and The Myth of the Rational Voter: Why Democracies Choose Bad Policies.

He’s been published in more newspapers and magazines than I can be bothered to name. He earned his B.A. in Economics from UC Berkeley and got a PhD. in Economics from Princeton. Though, I guess he’s gonna tell us that, that doesn’t matter too much.

Bryan Caplan: Yes and no.

Robert Wiblin: Will Wilkinson described him as playful, ebullient, sportingly-argumentative, and dressed unfashionably for comfort. So, thanks for coming on the podcast, Bryan.

Bryan Caplan: Great to be here again.

Robert Wiblin: So, we planned to talk about the implications of your work for listener’s career plans. And an essay you wrote a while back, about the risk of global totalitarianism. But first, what is the core argument that you’d make in The Case Against Education?

Bryan Caplan: A core argument is that, a lot of the payoff from education comes not from learning useless skills in school, but rather, just from showing off. Convincing employers that you’re worthwhile, basically jumping through hoops, and getting lots of stickers on your forehead. While, selfishly speaking, it really doesn’t make very much difference why exactly your education pays. But, from a social point of view, or from an effective altruism point of view, it matters tremendously, because if you go to school, if you were actually learning how to do new productive things, then your gain is society’s gain. But, if what you’re doing is looking better than other people, then your gain is really, the rest of society’s loss.

Essentially, I just go through the case for why it seems like so much of education is signaling and then try to recalculate. What is the social return to education versus the selfish return, once we have factored in this theory of what the new mechanism is? Well, alternate theory. Not really new.

Robert Wiblin: Right, right. What are the top few pieces of supporting evidence for this signaling theory of education?

Bryan Caplan: Ultimately, I just start with the curriculum and just go back to all the classes you’ve taken and just ask yourself, when do you use this stuff in real life, if ever? There’s just so many courses where almost everyone when they aren’t trying to defend a theory, they are supposed to believe, they just admit, yeah I just studied a ton of stuff, that not only never happened to come up, but was foreseeably ultra unlikely to ever come up. And you know what? It didn’t.

You know? Like, trigonometry. You spend a year studying trigonometry and if you ever dare to raise your hand and say, “When are we gonna use this in real life?”, you’d get a sneer from the teacher. “Oh you’ll see”, and I ask, “Well, have you seen?” How much is real life about trigonometry? Yeah, I guess it’s not at all, for almost anybody. So, we just, looking at the curriculum and just going through it and just seeing how much of the coursework that you’ve done, just very foreseeably, you’re never going to use again in real life. Looking at college majors, also. Just seeing how many of them seemed to be tied to almost no job, other than being a teacher of that very subject. That’s really where I start.

Then, there’s a lot of just observations that people have from their time in school. Things like, it seems like people are much more focused on getting good grades than actually learning anything.

Robert Wiblin: Right.

Bryan Caplan: So, how many times have you deliberately sought out the easy A? The teacher who goes and will give you an A, in exchange for doing no work and learning nothing. Seems like students are very interested in tracking those people down. Very unusual that someone says, “Yeah, I’ll only take him if he teaches a ton of useful material.”

One fun piece of support for this, on the Rate My Professor website, at least in the past they had a rating for easiness, but no rating for job relevance. It seems like, wow, people seem a lot more interested in the easiness than the job relevance. Also, like another primordial fact is that, if you want to get the best education in the world for free, you can. Just say, suppose you think it’s Princeton. Just move to Princeton and start attending classes unofficially. Don’t enroll, don’t pay any money, and just start learning. You’ll see there’s almost no effort made to prevent people from learning, from getting a Princeton education for free.

Unfortunately, just one little thing that you won’t have at the end of four years, and that is of course, grades and diploma. So, it seems like because of this, almost no one bothers to take advantage of this seemingly incredible opportunity. You know just another fact is the way that people forget so much of what they learn in school, and in fact you would often forget so much, you would no longer be able of passing a class you got an A in years ago. And yet employers regard failing and forgetting as two very different things.

If you’re being paid for the skills that you acquired, failing and forgetting are the same. But on the other hand if you’re getting paid for impressing people, then failing and forgetting are very different. Because almost everybody forgets but, if you did well in a class, and then forgot it, that says something about your willingness and ability to learn. So it convinces employers that they should hire you.

Then I also go into four sub-bodies of research. All which I say confirm the signaling model, and we can go into all those if you want.

Robert Wiblin: So, listeners won’t be surprised to hear The Case Against Education is a bit controversial, even among people who study education.

Bryan Caplan: That … I would say especially, actually.

Robert Wiblin: Especially, yeah, right. Well it’s very sacred for them. So if you were right, what would be the main implications for society?

Bryan Caplan: The implication for society is spend less on education. The idea that because education raises individual earnings, that it is a good social investment, is just wrong. If I’m right, that idea that is so prevalent, the idea that you go take a look, and you see more people with better education get better jobs. So if we all get more education, we all get better jobs, right? And the signaling model says, “No, wrong.” This is just a case like standing up at a concert to see better, which works fine for an individual, but is totally counterproductive for society.

So I mean essentially, this means a totally new view on what is a good way to invest taxpayer dollars, right? It’s a totally new theory of economic development. It’s so popular in development to just pour money into education and then modernize. And signaling model says, “No, that’s not a good way to modernize at all.” It’s actually going to burn up a lot of useful tax dollars and time, in what is really a mostly zero sum. So yeah, those are the big implications. I also talk about, the value of vocational education, and how … You know, I mean there’s good evidence that selfishly we actually underrate it. Especially for kids who really do not like school, and are not likely to finish, and are likely to end up in jail, vocational education is a lot better for them, selfishly speaking.

But then from a social point of view, even more important because vocational education and whatever else it does, what it does for private earnings, at least it’s teaching tangible skills, that people go out and use in the real world.

Robert Wiblin: Okay, so you’ve explained … You’ve gone into a lot of detail on the case in the book and various other interviews and a whole lot of articles online. So I’ll stick up a link to one of those. I think that the EconTalk episode is perhaps the best source for people who are really interested in hearing, kind of you describe the evidence at length.

Bryan Caplan: Yeah, yeah.

Robert Wiblin: Because I’ve only got you for so long, I was hoping to sustain a bit more of the time on objections.

Bryan Caplan: Great.

Robert Wiblin: So, in estimating the magnitude of signaling’s effect in education, you rely quite a bit on the sheepskin premium. Do you want to explain what that is for the listeners?

Bryan Caplan: Yeah sure. Alright yeah, so sheepskin effect … In a pure human capital world, that is if the only reason why education raised your earnings was that it raised your skills. Then if you happen to be one class short of graduation, that should have only a very tiny effect upon your earnings. Because there’s only a tiny difference in the amount you know. Whether you have all your classes for a degree, or if your one class short. On the other hand, if part of what you’re doing with your education is signaling, and especially if one of the things your signaling is conformity, that you’re willing to just go along with social expectations. Then there could actually be a discrete jump, where finishing that last class could give you a very big pay off.

So in the book, I go over all the studies that I could find on this, and essentially every single one done in the last 25 years finds that there is a much bigger payoff for graduation from both high school and the bachelor’s degree, than there is for the earlier years. Right. And in the book I go and actually average out all the results, and I find that for high school on average … Again, almost all these are from the U.S., maybe one or two studies from Canada. But almost all the studies find that senior year is worth about as much as the first three years combined. And then, for college … That’s in percentage terms. So it’s even bigger than it seems. And then for college, in percentage terms, it looks like finishing senior year is worth about twice as much as the first three years combined. Which again, is a very large difference.

Robert Wiblin: So that seems like pretty good evidence, on its face, that signaling is a large fraction of what’s going on. And I certainly don’t disagree that signaling is an important function that education provides, but I can’t help but feel there’s something not entirely convincing about that specific number that’s generated by just saying, well the sheepskin effect is signaling, and the rest mostly isn’t, or at least partially isn’t. So it seems like you end up comparing someone who just graduates against someone who drops out just before graduation to estimate this like premium from that final year. From getting the piece of paper. Which is where the term sheepskin comes from. Cause originally diploma were printed on sheepskins–

Bryan Caplan: Yes, yes. Your animal rights people may not be happy, but I have nothing to do with this, not my fault.

Robert Wiblin: Usually not printed on sheep anymore I imagine. Two things that occur to me, won’t the sample of people who drop out just before graduating, be fairly small, making the estimate of incomes of people who dropped out last minute kind of imprecise. And also, isn’t the kind of group that drops out just before the final class, or before final year, kind of an odd group, right? Because they were so close to getting this enormous award, and they decided not to. So possibly they got sick, possibly there were serious problems in their life that prevented them from graduating. Or even if not, I mean at that point they’re signaling that despite the enormous amount of money that they might expect to make by graduating, they decided not to do it. Which suggest perhaps a lack of conscientiousness, or poor judgement on their part.

So perhaps, that’s what employers are looking at when they say, “Well, you did three years, really? And then you stopped? What is wrong with you?”

Bryan Caplan: Right, although, hmm. So I guess there are a few things that are getting packaged together. One thing is that as you might expect, people have tried putting in different controls for ability to see whether it’s just the abler people finish. And out of the studies that do this, usually they find that controlling for ability reduces both the graduation gain and the annual gain. But again, if you think of sheepskin just as basically the ratio of those two, than the ratio seems very stable actually to … The main ability control they actually use is, measures intelligence, although that’s also the main one that actually seems to make difference when you’re doing the estimate.

Meaning you could back away, and say that there’s just some unmeasured thing that we just don’t have any measure for and that would destroy the results. Honestly, for all observational kind of metrics, you can never rule that out. But I think it is fair to say, that look the main thing we know works, doesn’t change the result, at all. You know the idea that this is all artifactual, or even primarily artifactual, seems hard to believe. It seemed like you were actually moving into a specific signaling story which, would actually make sense. Which is that what do employers think of someone who does three years and then gives up? They think, “My god, what’s wrong with you that you would do this?”

Robert Wiblin: Right.

Bryan Caplan: Although I think that, you can just flip it around and they say when someone does graduate, they say “Oh, look at what’s right with this person.” So it seems that in terms of being quantitatively wrong, I don’t see why that comparison is off. It’s just sort of saying, like how much of a difference do people see between the three year and the four year versus in terms, not of what they have of the skills they’ve acquired, but in terms of what kind of person that they are.

Robert Wiblin: So that is still a signaling story, but it’s a signaling story in which the sheepskin effect perhaps exaggerates the impact of graduation relative to learning. Because the group that’s just short of graduating is peculiar in this particular way.

Bryan Caplan: Okay, so the other thing is, usually what I’m actually doing because, most studies don’t actually give you return per year. They don’t say freshman return , sophomore, junior, senior. Usually what they’ll do is give you a sheepskin and they’ll give you annual. So we’re not basing this comparison upon just graduaters versus people who do three years. It’s based upon graduaters versus people who do zero … Zero two or three years.

Robert Wiblin: I see.

Bryan Caplan: Basically you’re lumping the one, two and three year people all together into one category. So again, it’s not nearly as heavy selected… There are a few papers that actually do exactly what you’re talking about. Those actually find that every year is its own return. So every year is special. Senior year is still by far the most special. Then actually, there’s a paper too that finds that freshman year is special, it’s like the second most important year. Basically this, you can think of this as the signal between someone who at least tries to do a year, and someone who doesn’t.

Robert Wiblin: I thought that could be bad. That they go to university for one year and then dropping out would look worse than someone who just tries to do something else, and doesn’t even like try and fail.

Bryan Caplan: At least in modern U.S., I think the stigma against someone who doesn’t even try college is pretty great. So that makes sense to me. In the second year, so-so. In the third year, then, that’s normally found to be worthless. Again I think that is … You’re right, it’s a small number of people, maybe 3% of the population, something like that. But it’s a small number of people that is operating under a very negative stigma, where they do a whole extra year, and then employers are so baffled by this decision to give up, that it seems that third year pays zip.

But again, most studies don’t break it down with that level of granularity. So comparisons done or reporting is essentially ones that compare people at one, two, or three years to people who actually finish. It’s not so sensitive to just those people with exactly three years.

Robert Wiblin: I see, so you’re saying you could look at the annual gain from doing the first, second, and third years and kind of average those–

Bryan Caplan: Yes, yeah, yeah that’s what most people actually do. I mean, really what most people … If you wanna get econometric about it, most people do the regression of earnings on years of education completed and degree completion. You just have those two things separate. That means, that in percentage terms, every year is automatically the same. Then since we have a special variable for degree, then we can get a bonus or a bump. Turns out for college especially, the bump seems to be the whole thing.

Robert Wiblin: Okay. Well that would do pretty well with my objection. Do you find it surprising that these ability controls don’t turn up anything peculiar about the people dropping out in third year?

Bryan Caplan: Right. Well let’s see … The main thing to remember is just that, even today, strange as it seems, there aren’t really good ability controls for non-cognitive ability. There’s a lot of different pieces of evidence that employers really value some kind of character or attitude. But none of the measures of character or attitude that are cognitively used, seem to be very good at predicting anything. This is one where you can either–

Robert Wiblin: What about the marshmallow test? I’ve heard that, that one’s predictive, but I wonder if that really is–

Bryan Caplan: Well hardly anyone’s been given the marshmallow test.

Robert Wiblin: I see, okay

Bryan Caplan: So this can’t really be commonly used. I mean honestly I actually did try tracking down studies that supposedly used the marshmallow test, and my memory is I couldn’t actually find them. And in the end, it was like is this vaporware or what is this stuff?

Robert Wiblin: I’ve heard that it’s bit of a myth actually. But I haven’t looked into it very well.

Bryan Caplan: So there’s something the marshmallow test is being used for. But my memory, at least, I could be wrong, is that I couldn’t find an actual study that used the marshmallow test along with income.

Robert Wiblin: I’ll do a quick Google after this, and we can put up a link to a study if we can find. Or otherwise a debunking if it doesn’t exist.

Bryan Caplan: That’s worthwhile. I had a graduate student, in a chapter in his dissertation … He had a very clever new measure of character. And what he did is he got a big data set, and then used the percentage of questions that the respondent didn’t answer as a measure of the conscientiousness. On the theory that a conscientious person fills out every single question, and then the lazy person just goes yada, yada, yada. He found that this measure was quite predictive for some health measures, and accidents and things. But it still didn’t work for income. It seemed like it could have been … What was really neat about this dissertation was that every data set, you can add a measure of non-responses. So what he found is there’s a secret measure of conscientiousness contained in every new data set. All you need to do is just re-visualize what it’s saying, instead of treating non-responses as lack of evidence, treat it as evidence.

Robert Wiblin: Interesting–

Bryan Caplan: So it was really cool idea, which works for some stuff, but again doesn’t work for what we’re looking at, unfortunately. I mean ultimately I can’t believe attitude and character aren’t really important in the labor market, but no common measure seems to be very good. I don’t know what to make of that.

Robert Wiblin: I was gonna say, people who fill out a survey that doesn’t really offer any advantage to them. Maybe they’re signaling that they’re too compliant, or their not independent or rebellious enough. Maybe that’s lowering their income. It’s possible it’s picking up … An ambiguously good aspect of the character.

Bryan Caplan: Yeah, I mean also, you can tell the story both ways. If employers find out about the people who do what they’re told, and then compete for those people, then they could make a lot of money. Just like someone who is a back talker, you might think that this would lead them to do better. But then, like well, once people find out you’re a back-talker, maybe they just wanna get rid of you.

Robert Wiblin: So it sounds like you’re dissing the ability controls that you had to rely on, that you were forced to rely on in the book. You’re just somebody who is very admirable because you’re willing to venture out and say something about areas where the evidence just isn’t that good. Just make your best guess. Is this one possible area you could be getting the wrong result because the evidence isn’t so great?

Bryan Caplan: Well, so … It’s possible, if we could find the magic bullet of character or attitude that does work. This may require a major rethinking of a ton of stuff. In all these cases, I would say you’ve gotta work with what you got. I mean it is true that these intelligence controls do matter quite a bit. And you know, there’s a few other controls that matter marginally. My main answer to this is just that if … You know the intelligence controls, these are the abilities where we know there are other quicker ways of assessing them, than just education. Where the character and ability, these are things that are more fakable.

Imagining going to an interview and someone says, “Here’s how you get a good job, pretend you’re smart.” It’s like wow that’s really helpful, I’ll just pretend I’m smart. On the other hand, if someone says “Pretend you’re hard-working, pretend that you’re conformist.”

Robert Wiblin: Pretend that you’re friendly–

Bryan Caplan: That I can do. So basically if the measures that are very observable, even those leave a big effect on education. That it seems hard to believe. That measures of things that are very fakable, at least over the short or medium term, would be much more potent.

Robert Wiblin: So sticking with the sheepskin premium. You did a debate a couple months ago at the American Enterprise Institute with Eric Hanushek, who’s an education researcher. Who is generally well respected.

Bryan Caplan: Oh yes–

Robert Wiblin: And he, in his presentation–

Bryan Caplan: By me too.

Robert Wiblin: And he put up figures in his presentation for the sheepskin premium which suggested it was a lot smaller than the estimates in your book. I think your response was just, “I’m a bit baffled about where these numbers could be coming from.” Have you had a chance to look into those and figure out why they differ from the numbers in your book and figure out which ones are more trustworthy?

Bryan Caplan: So I have a co-author who is helping me out on that actually, Nathaniel Bechhofer, who is a PhD student at UC San Diego. There’s one glaring problem with it, which is that Hanushek’s data set doesn’t actually have years of college. So you can’t really do a normal sheepskin effect and then, when I was inquiring … What he did was so different from what everyone else has done, that it didn’t seem very comparable. For high school his numbers just look like they were on the very low end of existing research. So you know, I’d been inclined to average them in. Admittedly,he does have an unusually large data set, so I’d been inclined to put more weight on it.

So that’smost of what I can say about it. I haven’t actually played with the numbers myself, I’ve been thinking of getting to it when I get through a queue of other things. I mean I’d really like him to actually go and publish that result, because the sheepskin effect. Considering how big and robust it is, the number of actual publications of top journals people have gotten with it is surprisingly small. It’s one, like where, if you read this review article in Handbook of the Economics of Education. The authors who were very prestigious to say, well its well known that this is so big. Then giving the idea that among econometric insiders, they’ve actually run a lot more regressions than have been published. But I’m a little unclear, I mean I’m just always concerned with what I think of as the autodidacts curse, which is you read a ton of stuff, but what if there’s stuff that’s known that isn’t written.

I mean I always try to reach out to people that are active researchers to see whether they will say, “Oh by the way Bryan, these are some facts that are just … It’s folk knowledge, but it’s genuine.” It’s just that no one ever managed to squeeze a publication out of it, so you don’t find it in the journals. There are quite few facts like that, that people will tell you, if they’re active researchers like, “Well it’s just commonly known.” Alright, thanks you know. I mean I tried to go collect of much of that as I could but there is always the problem that if you’re not an active researcher in any subfield people may just not tell you.

Robert Wiblin: Right, so if we took your estimates of the sheepskin effect from the book, what fraction of the return from an undergraduate degree looks like signaling then?

Bryan Caplan: Let’s see. So for the, for the bachelor’s degree, that comes out at something like about 60%.

Robert Wiblin: Okay so almost two thirds–

Bryan Caplan: Yeah, exactly. 60% for that. So the way that I use the sheepskin effect in the book is I have two different base estimates, one I call the cautious view, one I call the reasonable. The sheepskin effect is what I use for the cautious estimate. Then I have the preferred reasonable where the sheepskin effect doesn’t play any special role. That’s where I say, based on a few pieces of evidence, 80% seems reasonable. Then we get of course an even lowered social valued education. I think that the cautious one is the lowest a reasonable person will go for the signaling, and then the reasonable one is exactly where a reasonable person would go and stay.

Robert Wiblin: So it seems like, kind of an assumption in the book, is that as you move from primary school, to high school, to undergraduate degrees, to post-graduate degrees, that the signaling component is becoming larger. So kind of chopping off the later years of education is more effective than getting people to not go to primary school. Is that correct and kind of, do we have a sense of what the signaling component of early years of education is?

Bryan Caplan: So that’s only correct on the cautious approach. So the cautious approach definitely backs you up. I mean I don’t have anything for elementary school, because again there’s so little … There’s so few Americans that wouldn’t have done K-8 that we really can’t measure that anymore. So what I get for high school, I get something like 37% so, a bit over a third. And then for the master’s degree, something close to like 75% being signaling, using the one approach. On the other approach it just sat at a flat 80% cause I don’t have, other than the sheepskin effect, I don’t really have any particular evidence that says that earlier years are less signaling than later years.

And again, I will say, as a parent of four kids who have been in the K-12 system, very much inclines me to see tons of signaling there because there’s so much of the day that is spent on stuff that is so not academic and just seems so unlikely to ever yield any actual career fruits. I think this is something where I can very easily see someone disagreeing with me. So there are a few bodies of evidence that seem to back up more of just 80% across the board. We can talk about those if you’re curious.

Robert Wiblin: I guess I was expecting you would say that it does get gradually worse. Because if the signaling component is kind of constant across all years of education, and do you think that it implies people should study a lot less. You can always imagine that you’re saying, people shouldn’t even bother going to primary school if that’s also 80% signaling. Maybe it wouldn’t even be worth the cost.

Bryan Caplan: So actually, if you really wanna get into the math, the reason why I wind up getting higher returns for earlier years is because completion rates are so much higher. So the thing that’s really bringing down the rate of return for higher education is the low-completion rate. And that of course is bringing down the selfish return as well as the social return. Then the earlier years, K-8 there’s almost full completion, even for high school. You know people like James Heckman have put a lot of work into how bad our high school graduation rates are, but they’re still awesome compared to college graduation rates. So that’s really, if you really crack open the spreadsheet and try to figure out what’s going on, it’s actually that completion rate that’s crucial.

Robert Wiblin: But you can’t really signal anything from finishing primary school if everyone finishes it right?

Bryan Caplan: In a sense that’s right, of course if you have a system where you’re not even allowed to proceed to the next level until you finish the earlier level, then there’s sort of this indirect signal of if you don’t do that, that you don’t get the later signal. So there’s sort of a domino effect. There is something odd, I’ll admit, about saying that a second graders signaling, because it’s like, well no one knows what they did. Alright, are kids … Its true that employers have no idea what you’re doing in second grade, but they’re looking at what you did in college, and college is looking at what you did high school, and high school is looking at what you did in middle school, and middle schools look at what you did in elementary school. So there is actually this chain reaction … And you can see parents being very cognizant of this chain reaction when they’re desperately trying to get their kids into the honors classes in third grade.

Because the way lot of systems work is once you’re in, you’re in for life. So just get them in, in third grade, and then they get to be in the good track forever, no matter what their performance is practically.

Robert Wiblin: Yeah, that makes some sense. I’ve gotta say though, it’s kind of beggars belief for me, to say that the amount of actual useful information that you learn in primary school, and in an undergraduate degree or postgraduate degree is the same. Because it feels like, in primary school, I learned things that I use every day. I learned how to write, I learn to do basic math. I learned my vocabulary. I learned basic social skills. It feels intuitively to me like you start learning the useful stuff, and then they kind of run out of that, so like in later years it just becomes a larger, and larger fraction of separating the smarter people from the less smart people.

Bryan Caplan: Right. Or again, the more employable people. I like to think of them as more employable, rather than smarter because there’s so much that goes into that.

Robert Wiblin: Sure.

Bryan Caplan: So that may be right, although the main thing that I would encourage you to do is to find out what’s actually being taught in American schools these days, and I think you might be kind of shocked at how unacademic so much of it is. My older kids were in a school where they had three mandatory music classes. All were required, everyone has to do … They had to do choir, and dance, and music appreciation, no exceptions. This is burning up so much of the time, let’s just do music. I mean again, you can say they’re learning some kind of social skills in music. I’m like, alright. It doesn’t seem like that … like there’s much going on there anyway. I mean especially given this is a class where kids are goofing off especially. You may wonder is it just teaching rebellion, or making kids so resentful, but you know I would just say that … Here’s the main thing that’s striking to me, is I think many people you know professors, who will say, “I mean of course you’re only talking about college Bryan because we all know that K-12, that stuff’s all useful.”

I’m just like, all useful? Do you remember at all what was done? I know you’re not saying that, but there is this weird amnesia/Stockholm Syndrome that people seem to get, where as long as it’s far enough back in time, then they fill in useful stuff every minute of every day. Whereas when the kids are there, it seems like it’s a very different experience. They’re just like, why are you teaching us this and why do we have to do this stuff. So–

Robert Wiblin: It’s possible you remember the classes that were useful, because they’re the ones that are memorable–

Bryan Caplan: Yeah, yeah. Yes, yes. So you know I mean, someone could fairly say, Bryan remembers stuff that wasn’t useful because that’s the kind of person he is. Again I do try to control the bias just by making a list of… You know subdividing it. So what were the classes that I did in tenth grade. My memory is actually shockingly good for some of this stuff. I have almost a perfect memory for my high school curriculum anyway. I’m a little hazy on the name of the teacher, I still can remember the teacher, but just to go and pay attention to … You know crack open your old transcript and just go through class by class. I think people will say, “Wow, I just forgot that I had to do all those classes.”

Robert Wiblin: So you say throughout the book that you do think learning to read and write and do basic arithmetic is pretty valuable. Do you think that we could get big economic gains from teaching those things properly in the early years of education and really focusing on those completely?

Bryan Caplan: So I think we get good economic gains, not great. So this is part of my debate with Hanushek, Hanushek does have this idea, which … And he does have some work that I can explain if you’re curious. Saying that not only does the individual get a good pay off, but society gets a much better pay off than the individual. Like for him it’s more math and science rather than literacy and numeracy. But same basic idea. So I think that … To me it’s kind of crazy that you’re going and making kids who don’t know how to read and write properly to take three music classes. If I were the principal of that school, I would say, “Stop all this stuff for the kids who can’t read or write.” So until they can read and write, we’re not gonna go and have them do this other stuff. That’s it.

That would be my reaction if I were the principal in terms of improving individual employability, I think they’re would be good gains for that. So, like the Hanushek results implying many tens of trillions of dollars for boosting these scores. That seems very hard to believe. But the idea that you could go and reduce … Rather increase labor force participation of weak students by getting their literacy and numeracy up to employable levels. That seems pretty plausible to me. For that, like for many other common sense improvements of the school, my reaction is that I don’t trust them to just do the right thing if you give them more money.

Just the fact that they are willing to set up a system, where they’ll teach four other classes that they don’t really need to know before they’ve done they’re basic job tells me that they are not the kind of people that I really trust.

Robert Wiblin: What are the best arguments against your view, in your own view.

Bryan Caplan: Sure, so like the Hanushek stuff, since you asked. Hanushek has quite a bit of international work, which using very standard econometric methods says that if you can get a country’s math and science scores up, there’ll be an enormous gain for the country as a whole. Which is far beyond the gain for any one individual of getting this up. And it’s some kind of like social multiplier story or something like. The numbers that he gets are so huge that anyone in effective altruism, ought to get to go and look at them. If there’s a 1% chance he’s right, it’s worth looking into

He’s talking like tens of trillions of dollars of present value, rated. Again, to me, what’s very hard to believe about this is that the typical job in the … Even in the first world uses little math and almost no science. So I just don’t see how there can be these incredible multiplier effects of teaching something that almost no job actually uses. Again, the better story is that the result is spurious and he’s really picking up a social multiplier from intelligence. Which is much harder to teach than just math and science, math and science just put more into math and science. More time and more time on task.

Making people genuinely smarter on the other hand, much harder to. Of course there’s a lot of classes that can raise your score on an IQ test. But it seems basically like teaching the test for the most part. So I’m just not so optimistic there. Let’s see, and then honestly the Hanushek result, the sheepskin effect, with a really low sheepskin effect that was fairly disturbing to me. A guy who is so good with the numbers and knows the data so well coming up with a result that, at least for high school is very much of the lower bound of all the other numbers and it’s big a data set. That bothered me a bit.

I guess in terms of doing anything radical. Not exactly the precautionary principle, but just saying, “Look, you’re talking about moving far out of the range of anything that we actually have evidence for, so you shouldn’t do that,” right?

Robert Wiblin: Cross the river by feeling the stones.

Bryan Caplan: So, you know, that’s another one. Which again, I mean I admit in terms the level of uncertainty, what would happen if things were very different from what they are. I mean, I just think that’s a fair point. For me, like I say in the book it all depends upon what your starting point, who your starting point is, the status quo is justified unless you show a big problem. That gives you a very different starting point from … A very different conclusion if you say no government spending is justified, unless you show that’s justified. Whatever you consider the benchmark is winds up having a big effect on the result.

And you know the way I tried to write the book is just to be transparent about that and say, these are the things where I don’t think it really matters where your starting point is. And I say these are some others results where matters very much, here’s my starting point, so this is where I end up. But if you have a different one, you’re gonna end up with a different conclusion.

Robert Wiblin: So after reading the book, I spent a bunch of time looking for rebuttals to your view basically. And it does seem like there’s a lot of people who disagree with you, but I couldn’t find almost any very sophisticated engagement with the claims that you’re making and the empirical evidence that you’re delivering. I think one thing that I do agree with is what you were just saying, that it’s a lot easier to make the case that we should make the case that we should stop increasing the amount of education, or reduce it slightly, than to embrace some of the more radical ideas that you have in the book. Because we just can’t say what would happen if you reduced average education by several years, cause it’s just such a large change.

Bryan Caplan: I mean, I don’t think it’s such a large change to reduce education spending by 25%, something like that. There’s plenty of countries that exist that do. Meaning just for the U.S. to have the educational level of Switzerland, that doesn’t seem that weird to me. You know, that would be a big fall.

Robert Wiblin: The streets probably won’t run red with blood.

Bryan Caplan: I don’t think so.

Robert Wiblin: I guess you could also just look at how much education did people have in the 80s or the 70s.

Bryan Caplan: Yeah–

Robert Wiblin: And what did that look like? So you’d just be–

Bryan Caplan: So early 70s would be about a little less than 2 years, little less than what we have now. So yeah.

Robert Wiblin: If you reduced funding for education from the government, how much would that just reduce prices and not actually change the number of years people study?

Bryan Caplan: Yeah, great question.So, there’s whole literature on the effect of government subsidies on educational attainment, and the way they’re doing it should actually factor in the price effect and the quantity effect all together. So that’s really what I rely upon, so essentially when someone says increasing government subsidies by $1000 per student raises educational attainment by, fraction of people that finish college, by 1% point. That is already factoring in these effects you’re talking about. So I mean essentially, say if government spending subsidized by 1000 more you’d expect that to raise price by something, but of course not a full thousand. And the result of that … When someone says that that intervention goes and raises the fraction that finish college by 1% point, that’s already including that. So when I go and reverse that, it’s essentially the same thing.

So in terms of how much would it fall, then again it’s already factored in. I mean it is true that when people are doing thought experiment of the-n it wouldn’t be affordable anymore, and that is a separate issue. Then it’s better on those grounds, so it means that, the price isgoing down and the quantity is going down, but at the same time if you’re someone who wants to get the education anyway, still cheaper, but … I mean, the existing approach was already handling that.

Robert Wiblin: So, I was looking around for rebuttals and actually the best objections I could find to the signaling view of education was from a 1999 paper by John Quiggin, called Human Capital of Theory and Education Policy in Australia. Perhaps that’s a failure on my part, but it certainly had some interesting objections.

Bryan Caplan: I don’t recall reading that one actually. Let’s see… Did I cite it?

Robert Wiblin: No, no actually John sent it to me, so–

Bryan Caplan: Oh, okay. So what’d he say?

Robert Wiblin: I think you’ll be familiar with a lot of the ideas in them. So one is, if the signal about learning is imperfect, then all else equal, , students will under invest in education since they’re unable to provide a fully informative signal of their achievement to employers or their actual education to employers. If so, a correctly specified signaling model might predict under-investment in human education. Does that sound plausible to you?

Bryan Caplan: It sounds like if you know how to do good economic modeling, you can get any result. But no, it does not sound plausible to me at that if some of the payoff from education is acquiring useful skills and some is impressing employers, that it turns out that it’s socially optimal to get even more education than if it were all building skills–

Robert Wiblin: Well I think the idea isif it’s very important to be able to demonstrate the skills but you can’t do that very well, then people will tend to not develop as many skills as they ought tobecause they can’t communicate what they’ve actually learned.

Bryan Caplan: Yeah, so I mean that’s a reason to develop fewer skills but its not a reason to develop … To actually spend less time in school. So again you might say that it’s incentive to focus on the most impressive things regardless of the most useful things, and you know … So that hardly seems like much of a defense of the status quo. I mean, I was in a debate with Miguel Urquiola of Columbia, so he’s got a model which I say is already implicitly in the book, although it’s possible that he’s the one who actually put it down on paper.

So in the book I admit there is some social benefit to signaling mainly to get ranking, but then in the book I say once you’ve got ranking then at that point there’s no marginal gain to raising the average level of education. Basically what he’s got is, basically it’s model where there’s some extra value in having the really good people in the really hardest jobs. So in which case, I mean basically you’ve got redistributive element of getting more education where you look better, but also there’s an efficiency element where you’re improving the match a bit. Essentially with his model, you can always improve the match more and more. So as the level goes up you’re always getting closer and closer to that. Again, that theoretically is possible, it just seems so unreal to me. Again, the idea today that we’re getting a more meritocratic match of people and jobs, than decades earlier. It’s not at all clear to me that’s true. At least if you … At least tabling issues of discrimination that were around in the past that really had very little to do with education.

You know the idea that in 1970 we had a much less meritocratic system than we do now. Like really? Cause it seems, you know–

Robert Wiblin: Setting aside the racial issues, you’re saying.

Bryan Caplan: Yes. Yes. Exactly. Or gender actually, probably is maybe the bigger one since it’s so many more people.

Robert Wiblin: Okay, so another objection that Quiggin raised is pointing to these studies showing when the number of years of compulsory education has increased in some countries, that you can look at the people who just got the extra year of compulsory education and the ones who didn’t, who just missed out because of their date of birth. That seems to show a small, but statistically significant gain in earnings. What do you think about that line of research?

Bryan Caplan: So you said this paper’s 1999?

Robert Wiblin: Yeah.

Bryan Caplan: So that’s interesting. So basically, since that paper was published, the use of these compulsory attendance measures to measure the true causal effect of education has exploded, and it’s become probably the main way of doing it. And then in 2015 there was a critique in the AER of the entire literature, saying that’s almost all wrong.

Robert Wiblin: Why’s that?

Bryan Caplan: Well so, specifically for the … Almost all this stuff is for the U.S. So the paper did cite a bunch of other stuff from other countries just finding that it didn’t work in other countries. But for the U.S., basically the south was the least educated part of the country. Essentially this paper said, just in or at compulsory attendance with region, then everything’s fragile and all the main results go away. I mean again, this is a case where normally I’d say, “Well, that’s just one paper and you shouldn’t really rely upon that.” But there is a general rule that the AER almost never publishes critiques of any kind. And to get a critique published in the AER, you normally have to get it past the people’s whose work is being destroyed.

Robert Wiblin: It has to be pretty devastating.

Bryan Caplan: Yes, it’s very hard to get these critiques published. So, Tyler Cowen has a rule that any time a critique’s published in the AER, the critique’s right. He said this before this paper came out. But I was like alright, so there’s that. Now the other thing is actually, even if the result had been solid, there’s a general issue about how granular you think the education signal is. So for example, suppose there is one county in the U.S. that has a higher compulsory attendance, or stricter compulsory attendance law than all the other places. Now, if employers had very finely grained understanding of the education system, then they might say, “Oh, he’s the kid who went to the Granada Hills High school district. And we all know that there they push the kids to do more. So his degree doesn’t mean as much.”

In the real world, employers do not act anything like that fine grain thing. At most they’re gonna say, well he’s from California. Or again, maybe they’ll just sort of average it out for the whole country. So in terms of how detailed the measurements that employers are making are. I just don’t think employers are likely to be distinguishing between people in one particular area who are born before or after a given cut point in the schedule. So that to me is pretty implausible too.

The last thing worth pointing out is that, I mean this was an actual earlier problem with all this compulsory attendance work for the U.S. is this is not one of these instrumental variables that immediately works. It’s one where you actually have to add a bunch of control variables even to find the compulsory attendance laws raised attendance. You just sort of look at the raw correlation, there seems to be nothing. So it’s one where you actually have to go and really squeeze the rock to get some blood out of it in the first place, which doesn’t mean that it’s wrong, but it is the kind of thing that is just suspicious to me. So again, like if there’s an instrumental variable that immediately works, great, those are the kinds that I would put more weight on than one where someone has to go and really finesse everything just to get the basic result out of it.

Robert Wiblin: There’s a lot of discretion in the specification that they’re using.

Bryan Caplan: Yeah, yeah. So that. But anyhow again, anything … The 2015 critique, that’s one where I haven’t seen anyone reply to it yet, of course there are lags. Of course, the people’s whose life work has been destroyed, or allegedly destroyed, they’re never gonna go and cede to it. But this actually, there is a pattern of these education instrumental variables catching on, becoming a big fad and then someone publishing a paper saying actually it’s wrong. So there’s a previous debunking of using season of birth, which used to be the really popular one, and then someone said, actually … Basically the first people to say well, season of birth, we all know a priori, that they’d be no pattern based on season of birth based upon socioeconomic status or other stuff. And then someone came a long and said, hmm there is. Oh, so rich people like to have babies in some, one season more than others compared to other people. Yeah.

Not obvious but still, it’s the kind of thing where you kind of say, alright I guess we should have checked that a little bit more. I think the early people did some cursory checks, but later people really checked, and essentially people stopped doing that.

Robert Wiblin: Yeah, in general I’m pretty skeptical of instrumental variables techniques cause it’s just so hard to validate that the conditions are met. Although, the changing of compulsory years of education seemed like a relatively promising one. But I don’t quite understand the objection. You were saying that southern state–or different states have different levels of education to start with. But when you’re doing this–

Bryan Caplan: Particularly cause the change in attendance, basically the south. A bunch of things happened in the south during this period. It got a lot richer, relative to the rest of the country. So the south used to be really poor compared to the rest of the U.S. and during this period it did a lot better. And then it also raised its educational lot, and I think it also raised compulsory attendance a lot. Basically this means, if you sort of separate the south out, and put in this regional dummy, then you’ll see it doesn’t … It just seems to be the south was improving during this time rather than compulsory attendance was improving things during this time.

Robert Wiblin: So you’ve got a time trend on both of these things, and–

Bryan Caplan: Yeah, yeah, yeah. And the south has its own special pattern which is good in all respects, and that’s what supposedly is throwing it off. Which, again, I’m not gonna claim that I went and looked at the numbers myself I’m just relying upon these guys. At the same time, if the result had come out the other way, I think I would be at least a bit more circumspect, because of this special stamp of authenticity which is super hard to get critiques published in the AER. The gatekeepers are the people’s whose life’s work is being destroyed. So like when you combine that, that seems pretty credible to me compared to other empirical work where it’s really just your mafia who has your back and try to push stuff through if it supports what they want.

Robert Wiblin: So, here’s another objection from Quiggin:he points out that in the signaling model, a student from a bad or poorly resourced system should be offered a higher wage than a student from a good system with the same measured level of performance. Since the implied ability level of the first student and higher. And then he refers to two papers that say in fact the opposite is true. What do you think of that?

Bryan Caplan: Again, there’s the granularity issue again. Do employers know what your school district is? Now there is a related result which is very well-known and quite strong, which is that there seems to be a higher rate of return to education for blacks than whites. Which very much fits in with the signaling model. I didn’t ever mention that in the book because, I’m not even sure why I didn’t mention it. Probably I just felt like I’d have to go and add multiple big sections on.

But again … Notice, it’s much easier to tell someone is black than that they come from a bad school district, which is one where a lot of employers might … Like, what, I’m gonna go and look at the quality of everybody’s school district before I decide who we’re gonna give an interview too? This gets in the way of the desire to put applications in the trash as quickly as possible. On the other hand it really does look like the percentage increase in earnings from another year of education is higher if you’re black. Which very much does fit. Like I start off with a negative assumption, but you do something that looks good and actually that’s more impressive, than if you were someone who I was expecting good things all along.

By the way, this result … This does not mean that blacks earn more than white across the board. Essentially this is one where, for black high school dropouts they’re doing way worse than white high school dropouts. But then they have this much steeper rate of growth for year of education. And then roughly speaking you sort of reach the point of equality around 15 years of education. So yeah, if you’re black college graduate, then you do outearn white college graduates. At least controlling for some other stuff.

Robert Wiblin: That’s very interesting.

Bryan Caplan: Yes, but again the granularity, which you know … I know that it does feel a bit like a cop out. I will say that there were a couple of papers that supported me that also seemed to me to be very vulnerable to this granularity critique, and I just didn’t talk about them, I didn’t cite them actually because … When I was reading them, suppose this came out the other way, would I still be citing it, like no. Well then it would not be honest for me to go and cite it as support for me, given that the method would not have been convincing to me at all if it came out the other way.

Robert Wiblin: Right–

Bryan Caplan: There’s even a Journal of Political Economy paper that was pro-signaling that in the end I didn’t cite because, I just looked and said, “This isn’t very convincing.” Like it did seem to hinge upon whether employers knowing some very fine grained details about your educational background, which just seemed pretty hard to believe.

Robert Wiblin: One of the main concerns that I had with your argument, was that I thought defunding education would be bad for low income students who are quite capable. But you argue that in the long run it wouldn’t be, which is quite counterintuitive, and I’m not entirely sure that I’m convinced. So can you explain your argument there?

Bryan Caplan: Yeah, so the argument just begins with an observation. So when would you rather be a high school dropout looking for a job, today or 1950? So today there’s a very harsh stigma against high school dropout so you’re very limited. Now in 1950, there would have been a lot of good people that did not finish high school and so the stigma would have been a lot less. So we can see in the kinds of jobs that you can get as a high school dropout back in those days. You know it would not have been impossible to be a secretary and a high school dropout in 1950, you know, far from it. And many other jobs that would’ve been middle income jobs, would have still been open to you.

So what I say, when we’re thinking about the effects on the disadvantaged, you shouldn’t just think about how it might be worse for the really talented kid from the poor family, you should think about the average kid from the poor family. That’s one where I say at minimum, it’s just a lot more complicated, because there’s a big difference between changing the funding for one individual and changing the funding for a generation. Changing the funding for one individual, your intuition is totally fine. But if you change the funding for a whole generation, it changes the meaning of the education itself. And means that there are a lot of opportunities that the poor have lost in the moderate economy that they can have back again.

So that’s the basic argument. And again, underlying this is all the research that I cite on credential inflation. The way, for wanting the same job you now need more education than you used to. And just thinking about the people today that are from poor families or disadvantaged families, that still that don’t have good credentials. And you know … In many ways they are worse off, like they’re at a competitive disadvantage that they wouldn’t have been decades ago.

Robert Wiblin: This kind of has the perverse implications that it’s good for education to not actually show people’s ability, because that’s one way that rich people can distinguish themselves, just by… Yeah. Is that what we should think, that the more accurate education is as an indication of how capable someone will be in a job that we should be unhappy about that?

Bryan Caplan: Only in so … There’s an equity and there’s an efficiency part here. So if you think as, I think it’s very true that there’s some jobs where you really want capable people on them. So you want capable people building the bridges, it’s not pure redistribution that they get those jobs. They actually are saving lives by being good. Then by having a better way of locating those people, than that’s social gain and not just for the engineers, social gain for everyone who crosses bridges. But at the same time it does mean that the stigma against people with lack of credentials does get worse.

So you know, thinking about the late-19th century, there’s a lot of successful people who had no formal education. And in terms of, who would you rather be, what system would you rather be in. If you are a high ability person than the modern system seems better, right. Cause first of all things work better, and you also get rewarded for your talent. On the other hand, if you’re low ability than it’s a bit less clear. Cause maybe only work a little bit better, but at the same time there’s a harsh stigma.

Now then, sort of … You were kind of talking about it. Imagine one where all that your education said about was that you had money and didn’t say anything about your ability. In that case, there’d really be no reason for employers to worry about it, right. So it would reduce the incentive just to go to school just to show off. There might still be a lot of incentive to go there to impress all the rich people, and mingle with all the rich people. And again if you think about it more, you realize well there’s money in that too, right.

So in the book I have a detailed section on the marital payoffs for education, where … And honestly people ask, is it really worse than sending your kid to a fancy school? I’ll say “Well, is your kid ever gonna get married?” You’re like, “Well probably”. Then probably yes, because the fancy school puts them into the elite dating pool. Was it you, or someone else, that was recently quoting Robin Hanson on this. I think it was you actually, was you that was quoting this?

Robert Wiblin: Yeah. So this is a great quote that I actually got from your book, which is don’t-

Bryan Caplan: Ah, okay, so it’s actually from Robin. Robin himself said he got, it’s folk wisdom.

Robert Wiblin: Yeah, it’s “Don’t marry for money, go where the rich people are and marry for love.”

Bryan Caplan: Yeah, so I think this is one of the best reasons to send your kid to an elite school. And there’s one or two papers that actually do find this. That even though you’re unlikely to find your spouse when you’re at school, it just causes this social segregation where you get to be in the elite dating pool. You say well, the fact that the spouse has money doesn’t mean they’re gonna be happy. Doesn’t mean they’re gonna be sad, so. It seems like a good thing selfishly speaking. Although effective altruists may deplore that people are deliberately putting all this effort in just to meeting other rich people, but depending upon your view about whether you need to be an effective altruist every minute of every day or whether it’s just something that you do for your 10% tithe. Or what.

Robert Wiblin: Okay. I put up at a link to a meta-analysis showing that one year of education raises measured IQ by 1 to 5 points. What do you think about that kind of line of research?

Bryan Caplan: Yeah so, main thing to understand is that, that paper is not at all out of line with earlier work. So I assume, that paper came out after my book was already in the can, but I relied upon an earlier meta-analysis by Steve Ceci, where he found a causal effect of 1 to 3 points, rather than 1 to 5. So assuming … There is a difference, again it’s a modest difference. It’s very well understood actually that there seems to be this causal effect of education on IQ scores. What’s nice about the Ceci paper is that he, and after establishing that this result is very solid, this is using a bunch of different quasi-experimental methodologies. It’s not just doing the dumb thing, it’s really trying to crack the nut of causality open. But then says there are two big problems with the interpretation of this.

So one of them, Ceci says, is that a lot of education is in fact just teaching the test. And they say, yeah, teach the IQ test. Yeah that’s what you might say if you’ve never taken an IQ test. Or you don’t remember what’s on them. But you know there’s a lot of IQ tests that have very factual information, like who wrote Hamlet, what’s the longest river in Africa. Stuff like that appears on bonafide IQ tests.

So when you realize that, it’s like hmm. So when you’re learning a bunch of river’s lengths and authors of books, does that make you genuinely smarter. It doesn’t seem like it. I mean its something where, all else equal, you think smart people would know more of these answers and yet if you were to go and pour on the effort to teach the stuff, you would think you probably just giving them extra factual information without really making them smarter. Even for something like RAVEN’s Progressive Matrices, where you’re just working with geometric puzzles. Kids in school do geometric puzzles.

And in the book I point out there’s a big literature on practice effects. Practice works. In fact when I got a chance to finally meet Steve Ceci, my favorite question to ask him is, “Is there any known test ever, that people cannot improve at through practice?” If you know anything about Steve Ceci he is one of the best read researchers in the history of research. He’s just a veracious and credible reader, he writes, and one of his main products are these incredible meta-analyses and literature views. And he thought about it for quite awhile, he goes “No, there is no known test that you cannot improve at through practice.” Also …

Anyway, so that’s one of my main replies to this. The gains are real but hollow. Where you are improving the measurement but not the thing that you are ultimately trying to measure. Which again, the most obvious case about this is, I know a way to make everyone smarter than Einstein, just give everyone the answer key. Alright. Everyone knows that’s ridiculous, and there’s some scent that if you directly teach the test that doesn’t really make you smarter either. Like the SAT prep classes, make the country smarter. That doesn’t seem … They work, they raise scores, but they don’t increase the thing that, they don’t increasingly treat … When we’re thinking about intelligence, which you know they’re like, it’s the kind of thing where it’s a bit hard to define conclusively, my favorite definition is just learning ability. Ability to learn stuff.

I just did an interview with a philosopher where was the only person who asked me … He asked me for a bunch of definitions. Like define intelligence, define learning. Alright. I mean usually I just take definitions for granted but I will talk about that. Anyway, the other issue with education raising intelligence is the issue of fadeout. So, it’s very well established that you can improve some of things, but then if they stop practicing they at least lose a lot of the gain. I think that the student Ritchie lit review they were talking about does at least say that they’re trying to account for this fadeout. At least in the modern literature it’s definitely a serious issue.

Especially in the most famous experiments of trying to raise people’s IQ scores. Those have enormous fadeouts. Then the last thing, if … at mid-range like education costly raises your IQ by 2 or 3 points, this would still mean that we’re only explaining 20% of the effect of education on earnings. So it’s still, it still would leave an enormous puzzle. So we can basically … So essentially there is literature that tries to separately estimate the effect of intelligence or IQ scores versus educational earnings. And these usually say that a point of IQ raises its earnings by about 1%. So even if we took these estimates as total gospel and ignored all of the worries, it would still leave most of the payoff for education totally unexplained.

So that’s sort of my final point there. I don’t want to rule out that there’s some effect there. I mean especially for really extreme cases. So in the non-fiction graphic novel on immigration that I’m working on, I actually did come up with a way of trying to measure the effect of early adoption from the third world on intelligence. So basically if you get a kid from Mali and move them to Sweden at an early age. Essentially I was able to go and take few different literatures that were ignoring each other and snap them together to get a ballpark estimate. And there, there’s an enormous and longer effect of IQ of growing up in Sweden rather than just getting the average life in Mali.

Again to me, that’s the kind of thing where it’s so dramatic it would be crazy to think there wouldn’t be a big gain on true intelligence. So I mean I’m not saying this thing can’t happen, but I think it just has to be a lot more dramatic that just giving people a few more years of education to really yield these great gains.

Robert Wiblin: The fact that you get these pretty significant cultural effects on IQ tests and practice effects on IQ tests does seem to make it very difficult to do comparison across very different times and places. That they maybe work to compare intelligence between who grow up in the same place, or at the same time. But once you move beyond it becomes very hard to tell what’s real and what’s not. Okay, so a final objection is from a paper by Hanushek, called Do Better Schools Lead to More Growth: Cognitive Skills, Economic Outcomes and Causation, where the abstract points out that cross-country growth regressions generate a close relationship between educational achievement and GDP growth that is remarkably stable across extensive sensitivity analyses and specification time period and country samples. What do you think of that kind of research?

Bryan Caplan: Okay, so two things. One like you said he does not have a control for intelligence. It’s much more plausible to think that a difference in national intelligence leads to very big differences, because you use your intelligence every day. There’s no such thing as a job that doesn’t use intelligence. But there are a lot of jobs that don’t use math and science, right. And certainly, most jobs actually just don’t use any science really. I mean I don’t think there’s anything actually wrong with his numbers, it’s just that he’s got this omitted variable which seems like a much more plausible explanation for the general pattern.

And then, there’s a question … So it’s a lot easier to raise math and science scores than it is to genuinely make people smarter. That’s my main reaction to that. I do have a section in the book where I talked about that. That was my main objection to what he’s doing there. Better science scores overall, it’s just hard to believe that there are these big effects just cause it’s so rarely used. Only a very tiny fraction of people who study science ever have a job involving science. Literacy and numeracy, it’s easier to think that there are good payoffs because they’re used frequently, but still … Then again, kind of like the magic multiplicative payoffs that he gets. That’s what’s harder to believe.

So I’m little surprised you didn’t mention actually is the single most common complaint about the book.

Robert Wiblin: What’s that?

Bryan Caplan: That I’m just a philistine economist who doesn’t realize the point of education is not to increase personal or national income or wealth, or any of the economic effects. Rather the point is to elevate the human personality and just to refine our tastes and make us into intrinsically better human beings. Maybe you’re much of a utilitarian to care about that stuff.

Robert Wiblin: That’s right, I’m too much … I actually think philistinism is in kind of the correct view. So we disagree about that, but in this case it works in your favor.

Bryan Caplan: Ah, okay. My main reply to those people is, come on at least read the table of contents before you say I don’t talk about. I’ve done a whole chapter on this stuff. The other thing is they really do just wanna give the school system credit for good intentions, because they don’t even really show that the school system successfully changes our tastes or preferences or anything else. I actually go and look at the data and try to see what evidence is there that people’s preferences are being transformed away from philistinism. There just isn’t much sign that it’s really happening. So it does seem to be a lot of people just want to give the schools credit for what they say they’re trying to do rather than what they accomplish. Which I think almost any effective altruist should agree is a crummy way of giving an institution credit for anything.

Robert Wiblin: Yeah. I found your chapter on that, pretty persuasive. Almost… I mean I’m not too worried whether people love the opera, but you persuaded me that education isn’t making them like the opera either. Even if that was its goal. One thing that was really quite remarkable that I was very surprised by was the finding that number of years of education doesn’t change political views very much. Doesn’t make them more liberal or left-wing. I almost can’t believe that, ‘cause it’s so counterintuitive.

Bryan Caplan: Right, so that’s one where it’s important to distinguish between overall liberalism and conservatism, versus issue-specific views, and that’s what I do in that section. So I mean, one thing I noticed is this is averaging over decades of data, maybe that the world is changing. It could be that it used to be education was more conservative and now makes them more liberal, we’re averaging out these facts. Although it should make you wonder about what the effect what would usually be, because if it’s changed once, maybe it can change again. So there’s that.

Another thing is, if you go and break it down by views and issues, than you do see these effects of education on specific issue views. In particular, it does seem that more years of education makes people more socially liberal and economically conservative.

Robert Wiblin: You would like that wouldn’t you?

Bryan Caplan: Yeah, yeah actually so to me that sounds good and that fits in with my earlier book The Myth of the Rational Voter, where I find a lot of stuff consistent with that. But that pattern is by itself so odd because there’s so few college professors who want to cause that. I want to cause that, but there are very few professors on campus that want to cause a combination of more social liberalism and more economic conservatism. Which then does make you start wondering, well what is the mechanism that is thriving this? If it’s not the professors enlightening/brainwashing students. That’s where I say the best story is probably peer effects. Then I go onto this general issues of, well if school does change people by peer effects, then the social effect and the private effect are very different. Think about the last election, where you might say “Well college graduates are way less likely to vote for Trump, that’s good.”

So yeah, maybe that’s because all the people that would be disinclined to like Trump all hang out together, physically in college and socially outside of college. And they have basically socially segregated themselves from all the other people, and now they no longer influence those people. And those people then influence each other in a pro-Trump direction. So the net effect is actually quite unclear. Meaning … Think about this as a philanthropist, suppose you want to go an promote effective altruism, would you want to go and create one special school of effective altruism where you draw all the be EA people in one department. And it’s like hmm, well if we did that, then this would mean that nowhere else had any EA people around to go and talk about it and introduce people to the ideas. And maybe actually it would be bad for effective altruism to concentrate them all in a place where you could get an awesome peer effect there, but you’ve destroyed the effect of the voices… the access to effective altruism everywhere else.

To me at least, this is an interesting question of whether there’s value in this. And I can see scenarios where there is a value to concentrating people, but at minimum it’s not just at all clear that it’s worthwhile at all to do that. Maybe we want to do a spread. So basically put a little bit of your ideas everywhere, maybe that’s the best thing to help your ideas.

Robert Wiblin: Just returning to the cross-country growth progressions, isn’t it possible although most people don’t use their math and science education, you know, a small number of geniuses do, and they manage to drive up the country growth rate a lot through innovations that they create and that everyone else copies.

Bryan Caplan: So that story makes a lot more sense, of course … I mean like Hanushek is almost just looking at the average results. I think he does have a few specifications where he sort of looks like 90th percentile and median, right. Then he winds up saying that each is important. I think that probably is one of the better ways of saving it. Although, here’s the thing, even out of people who have degrees in STEM, most of them don’t use STEM. It’s very extreme. The share of people that ever use the stuff, it’s not the top 10%, it’s maybe not even the top 1%.

There was a U.S. Census thing that just looked at, what kinds of jobs to STEM graduates have. I think the result was only about 20% of STEM graduates even have STEM jobs. Engineers, about half of engineers have engineering jobs. But still you really are going and slicing the salami ever, ever thinner. So I guess you can imagine that going and having really good science for the .1% of really innovative people is important. But using this as justification of the system we have seems really odd. It might be a justification for an indie institute of technology system where you have really well funded schools for the very best of the best. But again, that’s not the system that we have, is about at all.

So it seems like a really crummy defense of the status quo if that’s the right story.

Robert Wiblin: Okay so let’s push on to talking about personal implications from your research. So the first one that jumps out at me, is that a lot of people will try to learn generic skills, or they try to get better at things that they might do in the future in an indirect way. A classic one is that parents teach their children a musical instrument, hoping that this will make them generally smarter. But another one might be, people they just do writing in general, hoping this will make them better at their jobs. What’s your view of that kind of indirect training?

Bryan Caplan: Right, so … In terms of what is, you know you’re asking what is selfishly best for you or what is socially best for you to do?

Robert Wiblin: Yeah I’m thinking, so now we’re talking to the listeners, and we’re hoping to give them some advice on how they could actually do better in their lives.

Bryan Caplan: Right, right. But selfishly speaking?

Robert Wiblin: Selfishly speaking, yeah-

Bryan Caplan: Yeah. So right, I mean honestly where I always start is gaming the system. I don’t start with learning, because the system doesn’t seem very interested in learning. I would just start with … Right, so what do you want to do and how can you do that with the least suffering to yourself. Basically sort of going through the inventory and finding out which things you can cut corners on and which things you can’t cut corners on. Some obvious things are, if it’s way outside your nature, if you’re just doing it for requirement than just find the easiest person. Most Americans will never use foreign language on the job, so find the easiest foreign language teachers that are around, go and do that.

Like for college actually for most people my advice is just be an econ major. Because I often tell my students economics is the highest paid of all the easy majors. And there’s a lot … Like these aren’t easy. Like come on people it’s not computer science, it’s not engineering. You know you can do really well in econ major and still have a great social life in college. It doesn’t mean that you’ll be like vitamin D deprived, like you would be if you were a CS major.

Economics is not the highest paid major, but it’s not that far from the winner. So winners are usually electrical engineering, computer science, mechanical engineering, finance, then econ. But it’s not that big of a difference. A lot of it, as I say, major in econ because it is the highest paid of all the easy majors, so do that. I say, whatever else you’re interested in you can do that to, but the econ major will just get you a better job and open more doors for you. So I advise people to do that.

In terms of other numbers. General result, this I not primarily me, I’m just reporting what the researchers find. Your majors more important than the selectivity of your school. Better to go to a state school and be an engineering major than go to Harvard and be a literature major. For most purposes, at least for your career purposes, maybe not for dating purposes, or for marriage purposes. But in least of terms of your career it seems like a hard major at a low cheap state school better deal than going to a private school.

Then I’ve also got stuff on is it worth going to an expensive private school. Like the marriage … Like improving your marital options, that seems like the best argument for. In terms of your career, what I’m saying is unless they’re giving you a lot of scholarship money, probably don’t go. Or, the only other reason I would consider going to a top school is if you have a special career that is even snobbier than other careers. Like say professor. If your career goal is to be a professor, then I encourage you to go to a top school because graduate schools are super snobby. And then schools hire professors who are super snobby. So even though it may not affect your earnings, it may affect whether you’re allowed to enter in to your desired occupation at all. So I think about that.

And the other big thing to think about, this probably won’t matter very much for your listeners specifically, but maybe for their kids, you should think about whether it’s worth going to college at all. I say there, the main underappreciated variable is just completion probability. There were got this classic saying of the best predictor of future performance is past performance. Best predictor of whether you’re gonna graduate college is whether you did well in high school. If you struggle to get through high school, or if you know someone who is struggling to get through high school. Those are the people it is not at all clear that it’s a good idea for them to go to college even selfishly speaking. Because they’re so unlikely to get that big bag of gold over the finish line.

Robert Wiblin: Yeah. The reason this issue of people who are unlikely to finish, it seems like the very most low hanging fruit here is to convince people who are losing out personally because they go to college for a year or two and spend a bunch of money and time, but are very unlikely to ever actually graduate. Just to convince them to stop going to university, I mean that’s something you can maybe even get the government on. You don’t have to tell anyone to sacrifice for the greater good, you just have to tell them “Well you’re gonna lose yourself.”

Bryan Caplan: Yes, I mean, funny thing, the government used to have a whole program to go do this. It was called guidance counselors. So back in the old days, the guidance counselor would bring you in, and they’d say “You are or you are not college material.” This was hard to hear … And of course sometimes they made mistakes. Sometimes there’s someone who is really good and they told them they’re not college material and they’re late bloomers. But you know better than to put too much weight on that scenario. There’s also a lot of people who were told not to go, they would fail, and so they didn’t go, so they didn’t fail.

There’s been a big change in the United States towards the college role model. Now I think one of the best ways to get fired as a high school guidance counselor is start using, throwing around that phrase “You’re not college material” and I think that will get big complaints from parents. Like, “Do you know what your guidance counselor told my kid, said he’s not college material.” So the government used to do this, now there’s more of, whether you tell them to go to community college or four-year college. It’s sort of the main division right now. And you say well of course college is for everyone, everyone’s college material but you should probably go to community college to get your grades up and then that kind of thing.

Then go and tell about how great community college is for those who do well in it. Without ever giving them any idea if they’re ever statistically likely to do well. But yeah, in terms of what I’m trying to I do actually try to give this advice. Honestly I don’t think the kids that are struggling in high school are going to be listening anything I say. So I do try to direct this to the parents. Who I think are the … They’re sort of the last line of defense of people who won’t get fired as parents if they level with their kid.

It’s really hard, especially parent’s pride is at stake. This is where my best appeal is look, what’s more important your kid’s future or your friend’s pity? Like “Oh my god, your kid isn’t going to college that’s so terrible. Oh my god.” Right, and I mean know parents, I think their pride is so strong they’d rather send their kid on this academic suicide mission rather than endure the pity of their friends, but I think there’s a lot of parents who have mixed motives, and I just like go and give them some moral support for advice to kids to go and try something else. I mean, I try to frame it more positively as can we find something else your kid likes and is good at, other than academics.

Like can you do that, have you tried? Why not try? You know if your kid just really doesn’t like school, why keep assuming they’re gonna turn over a new leaf and blossom in college rather than say here’s ten jobs that give you a good life and don’t require college.

Robert Wiblin: So as you say, most listeners on the show aren’t trying to decide whether to try to do an undergraduate degree or not. But many of them are trying to decide whether to go to grad school. So what do you have for advice for students who did well in undergrad and they’re considering doing an economics PhD or some other PhD? Who should do it and who shouldn’t and what courses should they consider and what courses shouldn’t they consider?

Bryan Caplan: Yeah that’s a great question. Just to preface data on graduate education is much mushier than data on high school or college. So here I’m just building on a much shakier ground, and I just think everyone should know that I’m doing that. The main thing to know is that grad school completion is even lower than regular college graduation. Remember that people go to grad to school are generally well above average for college, for undergraduate. So when you’re thinking about whether to go, again, the best predictor is how good were you as an undergraduate.

So if you were doing very well as an undergraduate, then probably you would be about average for grad school, and remember average doesn’t do that well. So were you stellar for an undergraduate, those are the people where they’re likely to actually gain from graduate education. Furthermore, you have to consider what you’re majoring in. I’m not sure that I found a single paper that measured the payoff for graduate programs as a function of major. It certainly seems that the pattern for undergraduate holds up for graduate as well with CS and engineering and economics paying well. History and philosophy and fine arts paying poorly.

So you gotta factor that in, and again as usual if you know what occupation you’re training for than just see what does the job market for that occupation actually look like. There are many graduate programs where almost the only thing you do with is to become a professor of that subject. Or you just don’t use it. So in that case, well you wanna be an English professor, well look at the jobs prospects for the people who are currently coming out with English PhDs and see how they’re doing.

Don’t ask yourself are you as good as those people. Ask what someone who didn’t know you, only knew what you were like on paper, would think whether you’re better than those people. Because the world doesn’t tell you … We don’t even need to go and get into overconfidence and self-centered bias. Let’s just say the world’s not fair and even if you’re awesome the world rewards being awesome on paper, not intrinsic awesomeness. Just accept this as a flaw in the world, and then consider that when you’re deciding whether or not you wanna try what you’re gonna do.

Another sort of general piece of advice that I offer people is if you wanna do almost anything in social science, and a lot of humanities, and you are dismayed but the crummy job prospects of people that major in the subjects. Then, my question … Can you do math? If so, why not just go and get an econ PhD and call what you’re doing economics X. I’m not even being flippant here. So I am baffled by people who like history and can do math, who do a history PhD. Why not just do econ and become a economic historian?

It is literally true that you will probably have tenure as an economist, before you would have your first assistant professor job as a historian. That’s how the world works. And then, once you have that tenure, you can work on anything you want. You never get another raise in your whole career, you probably have a better income stream than a historian would. So this is of course selfishly speaking. This is the kind of strategy that if everybody did it, it wouldn’t work anymore. But everybody’s not gonna do it are they Rob. So why don’t you go and do it?

Robert Wiblin: Well we’ve got the special tricks for you there. Just not too many of you do it-

Bryan Caplan: Economic philosophy of course. You can either go and try to become a philosophy PhD and try to be a philosophy professor, or you can go in economics and then do economic philosophy. Probably get tenure before you get your first job as a philosophy prof. And then, yes you’ll have to go and do some stuff you don’t like to get tenure probably, but then afterwards it’s clear sailing. So why not?

Robert Wiblin: It’s a shame that the evidence on postgraduate courses isn’t better. I recall one thing you said was that Master’s degrees don’t tend to do very well. Is that right?

Bryan Caplan: Yes. So we … Of course they still do have higher earnings, but basically when you combine the low completion probability with the modest gain and the high opportunity cost, because you know … The opportunity costs of high school is really low. Basically it’s like a high school dropouts wage. The further along you go, the bigger your opportunity cost is. And also, you’re starting to cut off some of your peak earning years too. And this it what winds up giving the bad result. But I’m not gonna say the master’s degree data is any good either, it’s very sparse.

In a way, the good news is the people are studying are putting more into studying decisions that more people face. So I suppose that’s good. But at the same time, given the sheer volume, you think there would be 20 good papers on it, and it’s very hard to find anything that I really thought was compelling

Robert Wiblin: I just wanted to return to the issue of transfer learning, which I tried baiting you with earlier-

Bryan Caplan: Ah yes, so I love this subject because, background… If you go a get a PhD in Economics as I did, you will hear the word psychology and psychologist said many times with as much contempt as a human being can muster. So, oh that’s what they say over in the psychology department. Oh well why don’t you go be psychologist. Economists just love saying these words with ultra contempt. Of course, this does reflect a real disdain for the field. And yet, when you start going and telling economists, like look, how can human capital be being built by all these classes when it seems like the course work is totally unrelated to the job. And then suddenly the economists say, “Oh well, basic psychology here.” Like you learn one thing, you learn others. Learning how to learn, critical thinking.

Basically they suddenly reference a bunch of things economists never study and who are the people studying them? They’re psychologists. So suddenly they respect the field. Although not enough of course to actually investigate what they found. I will honestly say I have enormous respect for psychology when economists get all delighted by the replication crisis in psychology. Like yeah, you’re going and finding some problems, and problem papers but they’re outliers. And there’s a lot of genuine knowledge that has still not seeped in. And you’re just deliberate going and searching for … You know, this is lemon picking rather than cherry picking. Finding the worst of the bunch and then saying ah-ha we’re justifying our contempt.

Robert Wiblin: Economist papers don’t replicate at any higher rate anyway, so… Go on

Bryan Caplan: I’ve heard this disputed actually but again though an economist I’m kind of inclined to believe you’re right, actually-

Robert Wiblin: It’s a lot of dodgy stuff in econometrics, and a lot of instrumental variables papers just not — Sorry anyway, go on.

Bryan Caplan: So anyway, I’ve got a lot of respect for psychology. I love talking to them and I love reading them. Honestly I’d rather read one more empirical psych paper than rather an empirical econ paper. I feel like I learn a lot more out of it. You know, they’re just addressing a lot of bigger issues that are relevant to economics. But economists don’t know about it or don’t care.

Anyway, so in writing this book I tried to go and read everything I could find that psychologists have come up with on learning how to learn, learning how to think, critical thinking, like all of this whole general area where you justify studying X because it improves Y. Studying something that you plainly don’t use because it’s allegedly going to improve your skills and things that you will in fact use.

So anyway, I found that educational psychologists have been studying these issues for about 100 years. Overall they’re ultra pessimistic. And again, to me, the reason I think this is especially trustworthy is these are the kind of people that very much seem like they wanted to find the opposite results. They really wanted to find that learning how to learn was huge, and in fact, often when they give their life story, their autobiographies, they’ll start off by saying, “Well, at first I thought there must be a lot of learning how to learn, it just hadn’t been studied properly, but after 20 or 30 years, nothing. I’ve changed my mind, I wish it were otherwise.”

How do people specialize in this, they’re generally very pessimistic about learning how to learn, their general view is that teaching someone X at best teaches them X. And if you want to teach them some other skill, focus on teaching them that thing. So you don’t teach writing skills very well by teaching history. You wanna teach them writing, teach writing. Even there of course they’re kind of pessimistic, because you know people forget so much. And also, what people learn they just tend to compartmentalize it so much. But still, this is there end result.

And then in the book I talk about the main exception to these rules. First I just talk about the main evidence, so there’s a lot of experimental evidence where people are just really bad at taking one lesson and applying it to a logically relevant lesson. When you first read the experiments, you might be inclined to say, well it’s because the questions are too hard. So if you were to go and teach them A, and then you need to go and do a 17-step proof why it’s relevant to question B, and people don’t do that. Well yeah, of course people don’t do that it’s really hard.

The neat experiments are the ones where they contrast teach A, see how well they do on B, with teach A and give them the instruction. Use A to solve problem B. Alright, and then you’ll see that people improve a lot. So it’s not that it’s so hard to see the connection. It’s just that even when the connection is seemingly very obvious, people just don’t tend to think of the world as connected in this way and so they tend to treat every problem as compartmentalized. Unless you’re told. And then of course in real life, you’re not being told. There isn’t a concierge on your shoulder saying, “Now Bryan, now reference what happened to you in 10th grade Spanish to convince Rob that transfer of learning is a myth.”

That’s not how the world is so constructed. So that to me is pretty compelling. And then there are some counterexamples which I do try to cover. So it’s not that there is no such thing as transfer that never happens, but it’s just so much smaller and more meager and disappointing and just more fragile. Then what economists assume, and what psychologists hopes.

Robert Wiblin: I have to say this confirms all of my biases. Because I always thought the idea that learning a musical instrument was gonna make you better at math just struck me as total nonsense. You get that illusion because people who are just more intelligence are gonna be better at music, and they’re also gonna be better at math. And unless you figure out some perfect way of controlling for that it’s always going to look as if these things are going together and you can have lots of transfer, but you’re not.

Bryan Caplan: Yeah so the first guy who got really into this was a psychologist name Thorndike, and he specifically targeted the teaching of Latin as this incredible builder of mental muscles. So, I mean, so it’s funny because Latin was taught for maybe a thousand years after it fell out of use, because you needed it to be a churchman. Alright so it was taught as basically being part of the catholic church. But then of course the Protestant Reformation comes, you don’t need Latin anymore for that. Then do people abandon teaching Latin, no they come with a new rationalization for why you have to teach Latin. And it is this mental muscle story, which lasts for hundreds of years. With no experimental evidence of any kind. And then Thorndike comes up with some of the earliest experiments and finds no, it does not teach mental muscles.

And now, basically people are at the level of just learning Latin to improve the acquisition of other romance languages. So they narrow it down to something where it’s totally plausible that Latin would be helpful. And even there actually they come up with decidedly mixed results because, while studying Latin improves your acquisition of French and Italian vocabulary, it seems to actually hurt your acquisition of French and Spanish grammar. Because the Latin grammar is different, so you get what psychologists call interference. Where the learning two bodies of information that conflict with each other actually the initial learning hinders your ability to acquire the second thing because you’ve basically got to unlearn or maintain two different mental books in your head.

Robert Wiblin: It makes me so mad, if you wanna learn French, learn French. I mean, maybe that’s not even sensible. But the idea of learning Latin just to like learn French a little bit faster seems absolutely farcical to me. Anyway…

Bryan Caplan: In the U.S. I’ve heard so many times – I learned Latin and it really improved my score on the SAT because of all the Latin roots of the English vocabulary words. How about you learn some English vocabulary words, wouldn’t that be a little easier?

Robert Wiblin: I’m just… I’m pulling out my hair here.

Bryan Caplan: Well if you wanna pull out your hair a little bit more. Out of all my ultra moderate reforms that I suggested, the one that I stand behind more strongly than any other is abolishing foreign language requirements in the United States. Because there, we’ve got a bunch of facts, which are: hardly any jobs use foreign languages, it takes a lot of time to get any good. And furthermore, in this book I’m able to go and snap together a bunch of pieces of data to show that virtually zero Americans claim to… even claim to speak a foreign language very well in school.

So I say, look, even if it did have these big payoffs, the system is just a waste of time, and people spend years doing it for nothing. And even here, I just run against a brick wall and people say, well in that case we should just improve the teaching of the foreign language.

Well, how about you do that and then get back to me, but continuing to fund the thing that we have, this is garbage!

And again, Washington state from what I understand, now allows kids to use a computer language in place of foreign language. Like, why not do that? Then it’s like, “No, no we need to do both.” People don’t have an unlimited amount of time, and shouldn’t teenagers be able to have a frigging childhood! Like how much of their childhood do you want to destroy with jumping through these stupid hoops?

Robert Wiblin: Yeah, it makes me incredibly mad. There is some value in learning a foreign language. You get some income gain depending on the language. But it doesn’t matter if no one’s learning it anyway.

Bryan Caplan: Well we just need to teach it better Rob, that’s the answer.

Robert Wiblin: We could fund it once we figure out how to teach it I guess.

Bryan Caplan: Yeah, I mean, when my older kids were in high school, they got three weeks of high school Spanish, I can see why the acquisition’s so poor. They had three weeks of Spanish where they were not taught any Spanish, it as just rambling lectures in English about the history of Spain. You know, I was like, “What is this stuff?”

Robert Wiblin: Yeah. I mean I really feel that there’s two ways you can go. I learned Spanish by living in Spain for a year-

Bryan Caplan: Oh nice-

Robert Wiblin: And that worked, and it was pretty easy because you just get exposed all the time. But I learned more or less nothing at high school before that. I think you either have to go all way, and learn it properly by traveling there, or just give up, don’t even start.

Bryan Caplan: By the way, I think this is one … I think there is pretty solid evidence in the immersion technique at teaching foreign languages, where you just ban English in the classroom and it’s hardcore from day one. But again, almost … Like colleges frequently do it. I think a foreign language acquisition in college seems clearly better to me. Like high school’s just don’t do, and it seems to me like the evidence is all there, and they just have so much inertia, and they just have so little interest in actually even using the time they have valuably. That the knowledge, however … A better pedagogy just stays unused on the shelf.

This is what I see so much going on with education, and someone says let’s just improve it. It’s like look there’s so many things that are already known to improve but they’re not being used. When someone has that bad of an attitude towards evidence, I think they should have less money.

Robert Wiblin: Yeah, so one place where I think I might disagree with you on the transfer learning is in writing. Or at least my experience is that, basically all of the writing I’ve done I feel just makes me better at writing in other situations. Writing essays makes me better at writing emails, I think. Writing blog posts just makes me better at probably doing this podcast, or at least writing up the blog post with a podcast. Would you agree that there are some generic schools that people can improve that have wide applicability?

Bryan Caplan: So for the writing, I’m very inclined to agree with you, if you’ve got, if you’re putting in a lot of effort or you have a really good teacher. And again, a teacher that actually gives you line by line feedback on your writing. What I’m skeptical about is that normal writing classes are having this broad effects. So again, the typical crummy English class where you just go and write like five essays in the whole year, and then the teacher just puts a grade on it, maybe corrects a couple spelling mistakes, that’s what I don’t think is improving.

On the other hand, if you’re going very hardcore, if you’re writing an essay every single week with detailed feedback than I completely believe that that works. And in fact that is the main way that I invest my time in homeschooling my other sons, is on improving the writing. That’s one way … Like every essay, I sit both of them down, we go over each essay with a fine-toothed comb. It’s very to see the improvements there. Even there, when they switch from history to English, than the general lesson of answer the question was lost at first. Although it was easier to regain the knowledge of remember, answer the question they asked you, don’t just ramble on about the general topic.

Easier to get someone to apply that when they already learned it in another subject. When yet, when you’re asked a history question, answer the history question. Then when you’re asked an English question, answer the English question. Like oh, yeah.

Here’s the thing, none of this research shows that transfer doesn’t happen. Or that it couldn’t happen. It just shows that not much of it happens in the real world. Right. And there are actually a couple of prominent transfer researchers who claim to have discovered the magic bullet. I’ve got the techniques that do work. Generally it’s only one person who believes in each particular technique. So there’s like, each guy has his own magic bullet, and no one else is convinced. Even if they’re right, I just say this just doesn’t explain what’s going on in the current system, and that’s really what I’m trying to understand. But yeah, I mean there’s a lot of area where it seems like transfer could happen.

Out of people, not just that I know, but out of a lot of your listeners I bet, that they transfer a lot. Because a lot of it is more about attitude and curiosity and determination and follow-through, then about logic. So it’s something where, given that the logic of transfer is clear, the failure for it to happen is all about psycho-logic. And again, what it takes to get the psycho-logic into the logic, I think it’s determination and curiosity. But that’s much easier said than done. Be curious, have determination. Like, oh, okay. Well, yes sir.

Robert Wiblin: I didn’t realize. When you’re saying that, a lot of things that people learn, they never use later in life. It doesn’t quite resonate with me, because I feel like I use my economics training every day. I mean I think you can see both of us using our economics training here. I mean, I use it just all the time I feel. Both in my work and my ordinary life. Am I just an exception, and many of the listeners might be an exception?

Bryan Caplan: Yeah, you’re extreme exceptions. So if you go and talk to typical econ undergraduate five years out, they can still do supply and demand. I mean I don’t think they do much else. More depressing, there are a lot of econ majors who will dismiss very basic questions of opportunity costs unless you just retaught them opportunity cost, right? And then, so I mean this is one, you listen to be podcast with Russ Roberts, so there are a lot of economists who don’t walk out of movies. Like econ professors don’t walk out of movies.

And Russ was saying “Well I don’t think it’s so complicated.” It is, it’s not complicated come on. You wouldn’t have paid to be there. Someone would have to pay to watch this movie and no one is paying you to stay in your seat after you’ve already paid the money, so if you don’t like it, just walk out. How hard is this? And yet, again, I doubt that even half of econ professors walk out of movies. And if they were complaining about it, again the transfer, I think you would actually have to go and tell them, You realize how against this … The sunk cost fallacy. And again, a lot of them are like, well yeah, I guess it is. And then they don’t change their life.

So, I mean like a while back I went and blogged about how four behavioral economists, very prominent ones, what their portfolios look like. And all four basically said that they have portfolios that their research says are stupid portfolios. I’m just like, what is wrong with you, like why do you do this. I try to remember … One of them, I can’t remember if it was Akerlof or Kahneman, but he just said “well, I just think that if I thought it about I would just make myself unhappy, so I just go and do a simple rule of thumb. Which yeah, I wouldn’t recommend to anyone else, but that’s what I do.”

To me it’s just so frustrating. This is the way a lot of people approach economics, it’s like a game. So you get publications or maybe you go and write things, but you don’t actually really use it to change behaviour. Whereas I – I mean – I try to use it to change behaviour all the time. All the way down to something as simple as… look this is one that is so mundane, and Rob you’re so grounded you might even have trouble believing this.

When it’s time to go to a kid’s birthday party, my wife would say, “Well we need to go to the store and get that kid a present.” And I’ll say “Why don’t we just give him some money.” Then it’s like, “no we can’t do that.” And I say look, do you think the kid is gonna be mad that he got money? I mean, the parents … you think the parents have so little going on in their lives, that they’re gonna go and hold a grudge against us because we gave their kid money instead of a toy?

I mean… it’s easier for us, and the other people will be totally oblivious to us, or they won’t care, or they’ll be happier. So why don’t we just do it? No, we can’t do that. Ahhgg!

Robert Wiblin: Alright, this has been a lot of fun, but we should move on. Okay, let’s push on and talk about your essay on totalitarianism. This was in a compendium book, global catastrophic risks that I think came out in about 2011. Can you just describe the problem that you were describing there?

Bryan Caplan: Well, so in the 20th century we saw the rise of what are called totalitarian regimes. Which are, governments that basically control all the important aspects of society and human life. If you were a utilitarian you seemed to make people really unhappy. And again if you also just think of, you believe in some of the … If you are not a philistine, you might also say even if the people are happy, they have crushed the true meaning out of human existence. And if you just bred people to like the totalitarianism it would still be a terrible hellish dystopia.

So, anyways. So countries like this have existed. The standard examples are usually the Soviet Union under Stalin, and then Nazi Germany, although only in like the last year. Like people who specialized in it will say it’s really only the last year that the totalitarian ideology actually becomes, they actually start living up to. Of course, Khmer Rouge in Cambodia, like China under Chairman Mao. So these places sort of where every little aspect of human life is being controlled a horrible and oppressive government.

This is an outcome that haunted … The idea that this could actually become the normal human method of government haunted George Orwell. It’s the inspiration for 1984, where 1984 the world is divided between three totalitarian governments with perpetual peace. This book is unbelievably influential and how so many of our concepts about the world actually are straight out of this book. You’re like, double think, double speak.

Robert Wiblin: Memory hole

Bryan Caplan: Yeah. Memory hole. So many of these concepts are … Unperson, the unperson is out of that, or from this book. So when Nick Bostrom was putting together this book on global catastrophic risks I have been very interested in the history of totalitarianism. So I said, why don’t I go and write a chapter on this horrible catastrophe of, humanity remains alive, but we are stuck under the rule of one or more totalitarian governments and they basically crush the meaning out of human existence and turn us into these horrible robots.

So that’s sort of the background. And then in the paper I tried to use what knowledge we have, you know. How likely is this, and what are the ways that it’s likely to occur. So my main story is inspired by, well by Orwell and also others, is that it seems like the main reason why totalitarian regimes don’t last is that there are non-totalitarian regimes that are around, which has a bunch of salutary effects.

One is that, totalitarian regimes are generally very uncreative. At best they can maintain the living standards before, usually can’t even do that. So this means that over time especially as long as there any awareness of what’s going on outside of the totalitarian country, they know things are getting better and better in a non-totalitarian world. And staying bad in the totalitarian word. So that’s one thing, is just the comparison group. You know this comparison group you and look at and see.

Another thing that’s going on is of course military competition, where if the non-totalitarian world keeps growing, they will have a big military edge over the totalitarian world and that’s another way the totalitarian may not be able to compete. Then there also just the question of maintaining the morale of inner elite, of the inner circle of leaders. Which I say in practice is the real problem for totalitarian regimes. It’s not like the Soviet Union couldn’t have just stayed the Soviet Union, but the problem was that the people, for its durability was that the people at the top lost their faith in their own system.

Which again has a lot to do with, again, there’s another system that you can look at and see it’s better. So anyway, put all of this together and said, if there ever were to be a time that a totalitarian regime basically took over the whole surface of the Earth. Then whether it was a bunch of them, or just one, you’ve removed most main reasons why it isn’t stable. In which case, maybe it could go one for 1000 years, 10,000 years. I guess compared to some of the other global catastrophic risks of permanent extinction of humanity is not bad. But still seems like a pretty bad outcome, nonetheless. So that’s my story.

Then I also, this is my most science-fiction-y essay, where I think about different ways in which totalitarianism could have been extended. So, you know, like fun fact is Stalin actually had a life-extension program dedicated to try to make himself immortal. It didn’t work, but my view that is if it worked, than I think the Soviet Union would still be ruled by Joseph Stalin. I really do. Again, 80% probability anyway. He really did crack the nut of how do you gain and maintain absolute power.

And again, Like there’s nothing … He was able to get nuclear weapons. So once he got nuclear weapons, I don’t see there’s anything else that happened in the outside world that would have made him do anything really.

Robert Wiblin: So to me this, is kind of one of the medium magnitude global catastrophic risks. I think we might disagree about some of the top tier ones. I’m worried about risks of artificial intelligence, and new bio-technology. But the risk of a stable totalitarian or authoritarian regime, seems to be reasonably significant. And there’s not a whole lot of people thinking of working on that, especially in the effective altruism community. It’s more or less totally absent.

And I think for me, the reason that I worry, is that it seems like there’s a lot of technologies that could be developed that would make totalitarianism a lot more stable. So we just talked about life-extension. I did an interview with Anders Sandberg where he has written a paper about where the life-extension would actually make totalitarianism more stable. And he was a bit skeptical. So I’ll stick up the link to that.

Bryan Caplan: Interesting. Again lot of this comes down to how much you believe in great man theories in history. I have moderately great man theories of history. So again, there’s a lot of economists and other social scientist who like structural theories where the individual leader doesn’t make much difference. I would just say, I know way too much history of the Soviet Union to think that isn’t all plausible. Like it’s really true. Within weeks of Stalin dying, there are radical policy changes that you would have been shot for. For even suggesting, before him.

The idea that Stalin was not personally, single-handedly crucial for so much of that stuff, seems just completely factually wrong to me.

Robert Wiblin: I think, and this Anders’ main piece of evidence, is just that although it’s true that you’re likely to get regime shifts when the lead dies or become sick. They still do fairly frequently lose power before that. So he was trying to estimate, if he had life-extension, how much would it extend the life of the regime. It was you know, it was only a moderate amount basically. They can still can collapse for other reasons.

Bryan Caplan: Yeah, I guess I would say, for the real hardcore totalitarian ones, that I just don’t see that at all. It’s true that the Eastern European ones collapsed because Gorbachev comes to power, so that’s one where it’s all correlated. Where you have a bunch of leaders that are not the real leaders, they’re not full Soviet puppets, but close to Soviet puppets. The Soviet approach before Gorbachev was, the Brezhnev doctrine, you know, what we have we hold. That seemed again, like super stable.

So the only … Out of the totalitarian regimes that I mentioned, the only one that I know of that actually didn’t just last as long as the life of the leader was Cambodia under Pol Pot. That’s were another, almost totalitarian regime, you know Vietnam invaded and took over. Although, notice Cambodia did not have nuclear weapons. So once you got a well-established totalitarian ruler and you’ve got nuclear weapons, of course … So true that the external pressure got rid of Kruschev, although his main thing was ramping the totalitarians down a lot. He tried the other approach. It’s one where there are these periods when … Before you get a total stranglehold on the system, and then you might still get toppled. But when you do get to the stranglehold point, again it seemed like there’s a pretty nice checklist of how do you know.

When anyone breathes a word of criticism, you expect to die. That’s[crosstalk 01:42:41]

Yeah, so like North Korea. I remember when Kim Jong-un, or rather when Kim Jong-il died, I went home that night to my kids, and I said “This is really exciting.” I said “Look, anything could be happening there right now. The next few days, they could be killing that kid right now. They could be killing Kim Jong-un right now. They could be murdering each other. Like anything … There could be a massive purge, the preemptive attacks. Anything can be happening.”

Then a couple years later. Then Kim Jong-un has his uncle executed. I’m like alright, I think he’s got things pretty well under control now. So I’m not totally … I mean I’m almost ready to go and bet that Kim Jong-un will rule until his death, at this point. Even though he’s only like 31. But hopefully his life won’t be very long.

Robert Wiblin: I was actually gonna go in the direction of talking about other technologies that could extend a totalitarian rule. You’ve probably seen this new Chinese social credit system, that they’re ranking each person based on, well a bunch of things. How nice are they to their families, how good are they at their jobs. And also, just how loyal are they to the state. That determines potentially whether they can get loans, and whether they can get jobs and whether they can move within the country.

And people are a little bit concerned about this, but I got … Is this a ten-alarm fire? Is this gonna be a technology that in China could totally stabilize the regime forever? And could spread to other countries, and make authoritarianism very stable in general. What do you think of that?

Bryan Caplan: Yeah, so you mean … Good that you used the word authoritarianism, because I think that China would no longer be reasonably called totalitarian. Part of it is, I’m no longer nearly as worried when it’s only authoritarian. Humanity has lived under authoritarian regimes for most human history, and you know human life has had meaning for most of this time. I think it’s only under the totalitarian regimes that it’s truly this hell-state. But yeah, I think it’s very sensible to worry. Again, you can see this being pretty stable as long as the government goes and punishes someone for coming up with an alternative credit rating that takes out the loyalty to the state part. In a free system, the government could rank you, but if it someone else were free to give you a different ranking, and ignore the political stuff, than it would make sense for employers to say, yeah, I don’t care about his politics, I just care about whether he will repay the loan.

But yeah, if governments were to do this, yeah that makes quite a bit of sense. I would say again this is probably just formalizing something that’s probably already in existence. North Korea has long had a whole caste system based upon perceived loyalty of the regime. This is just one, where at the margin, it’s easier to move up or down. Where in the North Korean system is much more heavily based upon the caste of your parents and grandparents.

That makes a great deal of sense that this would be amplifying … I guess I would think of this as 25% worse than what these regimes, like what totalitarian regimes, do already. For China, again, since it liberalized quite a bit, I think it’s probably a bigger step in a bad direction.

Robert Wiblin: What about the fact that mass surveillance is getting a lot cheaper. So it’s possible for a government, like the Chinese one to just have cameras everywhere and figure out where everyone’s going and potentially monitor most of the communications. That seems like that would allow intelligences to stabilize the government for a long time.

Bryan Caplan: Yeah, I mean as long as their willing to back it up with some kind of threats. Which could be actual jail time or whatever, or it could just be, as you said, ruining their government credit score. In that case, you don’t even have to have a lot of people locked. I mean, so some are concerned about that, and there is the question about whether China will go and make credit card companies go and base the ratings based on the government rather than do their own. Cause you know, as long as there’s some competing credit rating, that is general informative. And it sort of strips away the pure obedience to authority part in favor of whether you’re a good trading partner, then I wouldn’t be so worried. But if the government is actually going to be serious and try to stamp out these alternative measures, then it does seem pretty scary.

Think about where we would be rated. I’d rate you better than me, I’d give you a better social credit score than me. I don’t think either of us would be doing too well.

Robert Wiblin: No, I don’t imagine. So if a listener wanted to work on this problem, you know, make it less likely that you’d have a stable totalitarian or authoritarian regime, is there anything that you’ve envisaged that people could do that would be helpful?

Bryan Caplan: Now that’s a good question. I mean, there’s the … An answer which is not in spirit with my education book is just saying, learn the history of totalitarian regimes, because those who don’t know the past are doomed to repeat it. I want that to be true and yet, everything I know about transfer learning just says, that even people who know history usually just employ it for their own nefarious purposes, and don’t really use it a fair-minded way. To figure out the best thing to think.

In terms of what could actually be done, of course there is voting on the basis whether someone plausibly is going to be totalitarian. There’s something like that.

Robert Wiblin: I’m sure that would never come up Bryan.

Bryan Caplan: Of course a lot of what this would mean is just voting for people who are very boring. Just vote for the boring person with no new ideas. Assuming that you’re not totalitarian now, then the boring person with no new ideas isn’t going to do that. I mean I will say that, I’m always nervous whenever any political side has a new inspiring leader. I’m like, “Oh great, another one of these guys. Another person’s whose gonna get people excited.” It’s the worst possible thing that could happen is that … Well it’s the worst likely thing, is there gonna get excited and motivated and there probably gonna mess things up.

Robert Wiblin: So I guess you were a big Hillary supporter then, Bryan.

Bryan Caplan: Yeah.

Robert Wiblin: Yeah, the most boring possible.

Bryan Caplan: Yeah. I mean I actually, early in the election I just said she’s the lesser evil, cause she inspires very little enthusiasm and would immediately have a hostile congress, so we’ll just have great luck and that’s about the best. I think this is likely to happen.

Robert Wiblin: I’m surprised they didn’t invite you to speak at the DNC conference, Bryan.

Bryan Caplan: Yeah.

Robert Wiblin: So I guess there’s also a reform of like intelligence and surveillance law, might be an option?

Bryan Caplan: Yeah so, especially given that the existing system seems in total violation of a bunch of laws that we already have in the books, and yet nobody seems to care. I don’t know what to do about that than to tell people to care. I mean like, how can … Government is collecting all this information without a warrant, they don’t have probable cause, how do they do it. It’s like well –

Robert Wiblin: They have a secret court, Bryan. That says that it’s fine.

Bryan Caplan: Yeah, so like what? I mean, when I was having arguments with people about Snowden, and there’s saying “He broke the law.” I’m like, “look the government broke the law in the first place, he’s just exposing what they’re doing. How can you say he’s breaking the law?” There like, “No, no, no, the government says he’s breaking and only the government can say whether its breaking the law itself.” I’m like, “Well then how is this any different from any dictatorship or anything else?”

Just the way that it doesn’t seem to bother people at all. Or very few people actually care. I think currently I don’t see its doing very much, but there is this downside risk which kind of is the whole point of thinking about global catastrophic risk.

Robert Wiblin: Yeah, I mean I think for many other global catastrophic risks, having good intelligence and surveillance could be quite important. So for me, the focus would be on regulating collection of intelligence and making sure that laws are followed. Such that you are less likely to get authoritarian backsliding cause abuse isn’t possible. But it’s difficult to do, because just the intelligence services in general aren’t that accountable to the public because there’s so many layers in between them. And voters.

Bryan Caplan: You might remember, in that chapter I did talk about the worse case scenario for totalitarianism being one-world government. Right, and saying … It seems to me that this is most likely the scenario that is most likely to lead to permanent totalitarianism if you’ve eliminated all outside comparison groups, all outside checks. And even to go and grow economic growth for military reasons. I found this interesting because there’s a lot of other global catastrophic risks that people plausibly, a one-world government being the solution. Eventually there is a trade-off where you are risking this other one to a greater extent at the same time that you are perhaps reducing the risk of these others, but at least be aware there is a trade off.

Robert Wiblin: Yeah, that’s where I was about to go next because a lot of people in the effective altruism community are very focused on improving international coordination in order to deal with, I guess issues like climate change, reduce the risk of serious war. And also to regulate dangerous technologies so that you can’t … You kind of have to have most countries willing to regulate them, otherwise someone will find a place to use them. And then, if one person using it is bad, you kind of have to get everyone on board.

But of course there’s a trade-off. If you have international coordination, then you could also just stabilize a bad global system. Remove competition. How do you feel about the trade-off there? Do you think more global coordination makes us safer, or more at risk?

Bryan Caplan: I mean just based upon the principle that the best predictor of future performance is past performance, I don’t see the global coordination has yielded great fruit. If someone were to say it’s given us some improvements on sort of like marginal wars, like are less bloody and are shorter. Countries recover from them more rapidly. Famine’s work out a bit better. Those seem like possible ones, but in terms of has it actually lead to stable world peace. Or actually has even reduced the risk of a really bad outcome, it’s reduced the risk of a serious war between great powers. That to me at least seems pretty dubious, there hasn’t been much gain there.

Robert Wiblin: Interesting. I feel pretty differently. It seems to me like international institutions have reduced the risk of war quite a lot and could have been in essentially in preventing war between the U.S. and Russia. And also just having international norms against conflict between countries. You know anti-nuclear proliferation work, arms control in general.

There’s been some gains I think, in fighting climate change, although we haven’t done a great job of coordinating it.

Bryan Caplan: I mean, so during the Cold War, I can’t … I mean I know the history pretty well, but I can’t think of any case where I thought that the United Nations did much to reduce tensions between the U.S. and Soviet Union. I mean so, the case of the Korean war is bizarre if you know the history because the Soviet’s walked out. So they didn’t exercise their veto. And that’s the only reason why United Nations troops were able to even be in the Korean war. I mean I have read and … What in the world were they thinking?

Robert Wiblin: As I recall they thought it would make America look bad or something like that. I did look this up once.

Bryan Caplan: So that’s a case where you might say … And again, that just lead to war between Soviet back line state and the anomaly of the United Nations, well really the only U.S.-friendly part of the United Nations. Then for the rest of it, I don’t really see there was anything accomplished by the United Nations there. So since the end of the Cold War, that is more plausible that it’s made a difference.

Robert Wiblin: I’ll try to find a quote there from the Soviet ambassador to the UN, because I can’t quite recall what the reason was. What about the World Trade Organization then?

Bryan Caplan: That’s actually a better one. That’s one where it’s I don’t think it’s done much in the way of peace, but yeah in terms of improving trade, then yeah. So I’m not sure … I’m far from sure, and that I haven’t looked at the data very much. People in respect say that it’s been very important, so yeah. In terms of what the counterfactual would be, how much higher tariffs would be. There’s some question. But yeah, you’re right that’s a good counterexample.

Robert Wiblin: What about agreements to reduce nuclear stockpiles, do you think that’s made us safer?

Bryan Caplan: Yeah, so I mean I think those are basically bilateral. Like the ones that are important. So nuclear nonproliferation, those are ones where it’s actually interesting as to what’s occurred. I mean, of course … A common story about why North Korea is going so far in trying to get nukes, is because Libya got rid of this program and then immediately got double-crossed and Gaddafi was taken out.

In terms of whether international institutions are mattering very much there, whether it’s basically just the United States and bilateral pressure. I mean, I incline more to the second. Maybe its making a bit of a difference but. And then by the way, its also worth factoring in earlier failures. So a League of Nations generally considered an enormous failure. Probably rightly so.

Robert Wiblin: What about the Geneva Convention? What about making war illegal?

Bryan Caplan: Yes. Yeah, that is an interesting one. The Geneva Convention is fun because … So I think it’s actually … If you go and read it, the rule are so stringent that it would be very hard to do much of anything actually. Like the rules for the treatment of prisoners and so on, they’re so rigid. How do you ever expect anyone to fight a war under these conditions? The way it worked out is there’s some rules like that poison gas, that they actually mostly followed. There’s other ones, like the treatment of prisoners, where they don’t really do it. So they don’t actually go and take risks to their own people just to go and capture just cause they’re surrendering. Instead if you saw Saving Private Ryan, there are some prisoners and their kind of inconvenient to take, so just kill them.

Totally against the Geneva Convention but it happens anyway.

Robert Wiblin: It sounds like it’s a very difficult trade off to make, and I would be interested in having someone look into this. Whether potentially effective altruism is causing harm by promoting international coordination in a way that it could be dangerous in the future. I guess I think I’m still fairly pro-coordination. But we should consider the downside seriously.

Bryan Caplan: I mean in terms of how far we are from the downside, I don’t see that anything much has happened yet. So I’m not someone who thinks that the United Nations is effectively pushing much of anything. Essentially they still really need the cooperation of all the countries to sort of go a long with it. You know, there’s a few rogue countries where they’re just totally defiant, or almost totally defiant. But again, a lot of it is they just define their goals so loosely that almost every gets to say they’re in compliance with it.

United Nations Declaration of Human Rights, what country’s ever actually been officially declared to be in violation of it. And then face sanctions. I don’t know, maybe someone has. It’s basically just a lot of hot air. It’d be one thing if United Nations were seriously in charged of distributing what human rights on Earth were. Then I think I would be nervous. That they would just do a bunch of terrible things.

Again of course, United Nations makes education a human right. So it’s like does every country have to go and spend as much as South Korea on education or be declared a human rights violator?

Robert Wiblin: Just send them your book, Bryan. So let’s just push onto the last topic which effective deontology, or effective human rights protection. So, personally I’m kind of a consequentialist and kind of focused on welfare.

Bryan Caplan: Kind of, I don’t see any signs that you’re anything less than full-blown 100% die-hard [crosstalk 01:56:52]

Robert Wiblin: You should listen to the episode of Moral Uncertainty, Bryan. All things considered I try to take into account other views that are plausible. Which is kind of why I’m asking you about this. So I’m trying to look for the greatest suffering to prevent. But you’re kind of, you take a more libertarian view. You have more concern of human rights and prevention of coercion. So in light of that, for people who shared your moral outlook. What do you think would be the most effective things for them to work on? That people aren’t currently focused on?

Bryan Caplan: Very good question. So for me, you know like my big personal cause is deregulation of immigration. So right now, I’m working on a non-fiction graphic novel of the ethics and science of open borders. I’ve been a big pusher of open borders for long time. This is one where my view is that whether utilitarian or broader consequentialists, or whether you … You know folks on human rights and other deontological standards. All of them actually might point in the same direction, though I will still say that not nearly enough attention is paid to deregulation of immigration. So that is the cause that I’m most enthusiastic about. The one that I’m most convinced … Taking into account moral uncertainty. There’s so many views that all push very strongly towards at least very strong deregulation of immigration.

And in the book I try to bring together both the evidence and all the philosophical sectors as well. So there’s chapter I go through a bunch of different philosophers and how all of them should respond to the science. For me that’s probably the main one. This is one where in terms of alleviation of harm, alleviation of suffering, alleviation of poverty. But also, equality of opportunity. Just the right of someone to go and take a job from a willing employer and rent from a willing landlord. All these point in the same direction.

I will say a lot of the reason why I’m so enthused is because this is one where someone says “Isn’t it terrible that people are so poor, let’s help them.” And I can say “How about we just let them help themselves.” They’re able to do it. They’re not intrinsically charity cases, they are adults that are perfectly capable of getting a job and going and solving their own problem. If only they were allowed. Which in terms of conversing philanthropic energy I think makes a lot of sense. But also I will say, to me, it just seems more outrageous if there’s someone who’s starving on the street because the law says it’s illegal for them to work, than someone who’s starving on the street because he’s not able to work. The second one is like, well, that’s harder because then who’s gonna have to be responsible for taking care of him.

There’s questions like that. In terms of … Another thing, I am doing a book call Poverty: Who to Blame. And this is one where it’s not so much of a libertarian book as an old-fashion puritanical book. You know there’s some libertarian elements, but a lot of it is just trying to revive a lot between the sanction between the deserving and undeserving poor. Which yes I’m happy to recognize is a continuum. It’s not that you’re in one category or the other.

But still, there’s an intuition that almost everyone has. Which is that if someone born with a genetic ailment, so they’re unable to work, this person is poor through no fault of their own. But there’s someone who just won’t get off the couch and get a job and they’re perfectly able bodied and they don’t. Then two people can be in the same current situation, and yet moral intuition says that you should help the person that couldn’t help themselves. Or at least there’s no reasonable way for them to help themselves. Again to me these are very convincing intuitions.

So in this book, a lot of what I’m gonna talk about is reviving this … Focusing on philanthropic energy on the deserving poor. I think a lot of existing programs already have this intuition in mind. Like there’s a lot of programs that focus on kids. Right, and I think the idea is that kids, look even if a kid could have a job, it’s not really reasonable to expect a 5-year-old to go and work its way out of poverty. Most of them would just have no idea. It’s not reasonable to say, if only were a movie star kid.

So anyway, so think of anytime you’re starting with kids, or you’re starting with people who are severely handicapped or born with severe handicaps. Anytime you have welfare programs or philanthropy, the focus is on people like so you have this intuition in mind. These are people who are suffering through no fault of their own so they at least should be first in line. So basically if we’re doing philanthropic triage, they’re the people to help first.

Now the other side of this, there again, there are standard intuitions about who does not deserve our help. So if you have a cousin who has lost his job, and his wife because he doesn’t show up to work and cheats on her. And then he comes and sleeps on your couch and then instead of trying to get his life together he drinks all day. And then he just says “Woe is me I can’t help it.” Almost everyone at some point gets fed up with this and says “That’s ridiculous of course you can help it.”

My view is the people who are fed up are correct. So yeah, part of what I wanna do is to have a decidedly un-modern view of people with substance abuse problems. And work discipline problems. And say actually, you know that intuition that they are choosing evil on purpose, and so they don’t deserve much sympathy. That’s a good intuition and so let’s not help them. Or at least, let’s help them last. After all the hungry kids are fed, after all the people born without legs have been helped out. Then maybe we’ll consider your cousin, but until that day.

So again, there’s sort of a response that people often have, like how is this my problem. Which again you could say to a starving orphan too. But people are not generally inclined to say to the starving orphan, how is this my problem. Where someone whose wife won’t talk to him anymore because he cheated on her five times, well then why am I supposed to go and intercede on your behalf for you. Like why? Why should I? Aren’t you just getting your just desserts.

So again I would think of that as part of, sort of like another deontological effect of altruism.

Robert Wiblin: Yeah. I think as consequentialist, the reason to help some of those people and not others is because they respond to incentives differently. That you’re not gonna create more orphans potentially by feeding orphans. Whereas you might create more lazy people by giving them lots of money for being lazy.

Bryan Caplan: Of course a classic answer, although often when you really scrape the surface, you find, maybe not. There are people who will actually, you know like in desperate circumstances they’ll just put their kid out on the street as a beggar. When they were able to care for the kid. Or read about some of this stuff that Nicholas Kristof has done in third-world countries. Like dads with kids that don’t have mosquito nets spending money on tobacco, alcohol and prostitutes.

This is horrible, but then again, the idea that … So maybe the dads will respond to incentives, but again it seems like a lot of times when you sort of broaden the idea of all the margins, it’s no longer all that clear of what’s going on. I don’t think the intuition is all that changed. One of my favorite thought experiments is, you’re stuck on an island for the rest of your life in 1945 and then who happens to wash up but Hitler. And you’re both gonna be stuck on the island for the rest of your life, and what should you do?

So do you go and befriend Hitler? Do you say, well he’s not gonna do … everyone thinks he’s dead, it’s not gonna change any incentives at this point? Or do you just avoid him because you don’t want to get contaminated by his evil? He’s and old man, he’s not gonna hurt you anymore. Or do you go and kill him, to balance the moral scales?

Robert Wiblin: Well I would say to create incentives. That retribution helps to create incentives for other people

Bryan Caplan: But the whole thought experiment is designed to say there’s no incentives that will be given. I mean, and yet, if you watch enough movies about revenge, as I have. Almost everyone on some level thinks that you should torture Hitler to death, and in the most horrible way that you can imagine. Then there’s some high level theory of, no, no that will bring you down to Hitler’s level? And like, really, does it really bring you down to Hitler’s level?

There’s also, there’s a bunch of movies where the hero goes and mows down like 200 minions. And then he finally gets to the horrible guy responsible for it all and says, “Killing you would make me as bad as you.” Like well, you killed all the minions who weren’t nearly as bad as this guy, so what are you talking about?

Robert Wiblin: But that’s a case where I feel the intuition is unreliable because it would be so abnormal. Basically it would never happen that you have a case where people will never find out what you’re doing. Basically we’ve involved an intuition about retribution and punishment that applies basically everywhere. And then when you change the situation such that there are no incentives created, the intuition is no longer reliable.

But that would take us down a different line of conversation. I just wanted to ask you, why is that both on the left and the right, very few people … Even people who are concerned about liberty and human rights, view immigration restrictions as a human rights violation. Or just as, on its face, impermissible without very strong reasons to do it.

Bryan Caplan: I mean I think that the ultimate answer is that most people are nationalists first and whatever else they are second. If you are a Christian and democrat and American, the most important one of those three identities if you really scratch the surface is American. Most famously, World War I, there were dogmatic internationalist socialists who said of course all socialists are gonna be against this war. And then what happened is almost all the socialists took the side of their own country. And then there’s a few oddballs saying, what just happened. Generally everyone else is like, well of course German socialists are gonna vote for the German war. French socialists will vote for the war. British socialists will vote for the war. Russian socialists will vote for the war.

Because the most important of identity, whatever they say, is nationality. And again, of course you might say, yes well you could be, consider being an American and whatever your nationality to be your most important identity and still recognize the great benefits of immigration for a country. But it does mean that you’re going to put very little weight upon the welfare of people in other countries, which is the weight that makes it a no-brainer. Right, again I say that there are a lot of economic gains of immigration, a lot of other gains. But it’s really a gain to the immigrants that usually makes it’s completely beyond debate.

So look, once you factor in how much better off the immigrant is, then we can’t really argue about this anymore. If you go and set the biggest term in the equation to zero, and then say still convince me, than at least it’s harder. I think ultimately it still comes down usually, but less strongly with less confidence, and so on. And then furthermore, people sense of nationality is so strong that the idea that you could let someone into your country permanently and not make them a citizen is also anathema?

So it’s like look, if we let them in, then they would become one of us. Then we would have to care about them, and I don’t want to have to care about them. It would be bad for our country if we cared about them, so we shouldn’t let them in at all. I’ve heard this argument from people from a wide variety of views. Sort of the way they phrase it is different, but even … Let’s see, so I’m blanking on his name at the moment, but a Washington University philosopher that I debated on immigration. He said, “Look it’s okay to let people in as guest workers for a limited number of years. But it is morally wrong to let someone for their entire life and not make them a citizen.” Then I say, “Well how about you let them in for like 99 years.” And he says, “No you can’t do that.”

Alright, so I say, so basically this is … So everyone who comes has right after the probationary period to become a citizen. The primary effect of this right is going to be that people are not allowed in at all. And yet, it’s very important that we recognize that right, and he said yes. I guess it was Christopher Wellman.

Robert Wiblin: Yeah, it is perverse to have a human right that makes you so much worse off.

Bryan Caplan: And non-waivable. A completely non-waivable right. I really feel perfectly fine with getting to work in your country and never become a citizen if you’ll allow me to come to your country and work as [crosstalk 02:08:23]

Robert Wiblin: Well you would say that, wouldn’t you?

Bryan Caplan: Yeah, yeah. I mean ultimately the nationalism of most people is very pronounced. You may say this is human nature, although it’s not human nature to, you may say … I believe it’s human nature to be tribal. It’s not human nature to be nationalists because most people throughout history were not nationalists. You know a great, great book by Eugen Weber, Peasants into Frenchmen. Even in the 19th century most people in France barely thought of themselves as French and most didn’t even speak French. It was a whole propaganda campaign to convince them they were French.

Whatever the process was in the modern world, people do generally think of themselves first and foremost being a member of their country. Most politics are so domestic they may not even talk about it very much, but it’s just beneath the surface. Sadly.

Robert Wiblin: Throughout my primary high school education I was told that I was a global citizen rather than an Australian, and maybe that’s gotten to me. Maybe that did influence my moral outlook, cause I really don’t view citizenship as having much moral relevance.

Bryan Caplan: And yet I bet that most of your classmates, that whatever they say about global citizenship, if there were a war between Australia and the world, would take the Australian side.

Robert Wiblin: That’s only because we’d definitely be right, Bryan. So a couple of year ago I wrote a presentation where I tried to brainstorm what should libertarians or classical liberals think of their priorities from an effective altruism point of view. Or from just a general effectiveness point of view. I just wanna throw just a couple of options at you. What about domestic violence? Or sort of this small scale violence that just seems really endemic around the world?

Bryan Caplan: So it’s the kind of thing where its hard to know what the numbers really mean, cause you figure that there’s going to be a whole lot of stuff that’s beneath the surface. But to how much of that, that there is, it’s hard to say. It’s also one where it’s hard to understand exactly what we would do about. Again, sort of like a standard police procedure in the U.S. is, if they’re called to domestic disturbance they must arrest one person. So normally they’ll just arrest the guy, but then they can’t really hold him because they don’t have any real witnesses or anything, he gets released.

I can imagine doing something like that, but it’s hard to imagine what you’d do. My instinct would always just be to rash out the punishments. [crosstalk 02:10:34] Yes, yes, of course those can be costly too. There is a standard list of lower cost punishments we could be using but hardly ever do. There’re fines of course. That’s one where, you might get even fewer police called in, because if the wife knows that there’s gonna be family money that’s attached, there’s that problem.

You may have heard about this half tongue-in-cheek book called In Defense of Flogging.

Robert Wiblin: I have, I think there’s something to it. I’ll still up a link to that.

Bryan Caplan: Yeah, yeah, so if you were to go and flogging for domestic violence, that might actually be a bit more convincing. I don’t know.

Robert Wiblin: I suspect that you could try a broader advocacy campaign to try to shift culture and attitudes rather than try to use the criminal justice system. I agree it’s not entirely clear what would work. What about slavery? I mean there’s still, we believe, millions of slaves in the world. That’s something that I don’t really hear libertarians talk about that much even though it’s kind of the most severe human rights violation.

Bryan Caplan: So I have looked into that. I think a lot of those numbers are very misleading, because they’re including things like debt contracts as being slavery. I mean again, from not only a libertarian but effective altruism standpoint there’s a lot to be said for classic indentured servitude, where someone who otherwise couldn’t get a loan, goes and gets a loan and then they have to work it off.

So it did seem … When I looked at the numbers they did see like they really were advocacy numbers, where there trying to define as many possible things as being slavery as they possibly good. So it looked to me that the actual numbers were a fraction, a small fraction of what the official statistics were. So you know, when I looked at that, still obviously disturbing. As to what can be done, there is the … There either was a kid who tried setting up a charity to buy people out of slavery, I think you can see the problem with them.

Increasing incentive to reduce people in slavery. Again, as for what else can be done. That’s the tough issue. So usually this is happening in countries that are so screwed up in so many ways that it’s not so clear what to say. Again the kinds of things that are actually being targeted actually are things like child labor. Where instead of legal child labor the kids would become prostitutes. Then it’s something [crosstalk 02:12:51]

Robert Wiblin: Not a huge improvement.

Bryan Caplan: Or worse. Yeah. At least one where you can either go, your parents go and say “Okay you’re gonna go work for this guy as an apprentice,” and yeah you’re a kid and you don’t really get to choose and in some sense is involuntary. Although on the other hand, of course, even in this country parents can go make their kids do all kinds of unpleasant things. Or your … and that could be legal. Or you could go and have your parents go and sell you into the black market or the prostitution markets in Thailand, where your parents aren’t even around, you don’t see them anymore. They aren’t there to look out for you anymore. This is some pretty horrible stuff.

Robert Wiblin: What about the U.S. criminal justice and prison system?

Bryan Caplan: Yeah, so that’s one where, back when I believe that the large majority of people in prison for drug defenses, I was much more enthusiastic about. Then I learned that’s not really true, so maybe only … Well, people in federal prison, so there I think it’s a case where the majority are there for drug offenses. But in the overall it’s maybe 20%. Again, probably an underestimate because it’s only factoring in the official sentence, rather than the social causation of what might end up happening. Again I can still believe that 40% of people wouldn’t be there but for the drug war.

But still, that leaves a whole lot in there for offenses that … At least saying they think it is reasonable and just to punish people for. Even then you might say, well they should be punished but not as harshly and there’s not much value to it. It’s one where I will say libertarians have been very versed in it. So including of course the Koch brothers. Noticed they have not specifically talked about getting, reducing sentences for people on nonviolent offenses or just for drug offenses. They focus just in general on trying to reduce fences for people who have already aged out of crime anyway.

Sort of in the background is the idea that, the marginal deterrent effect of going from a 20 year sentence to a life sentence isn’t that much for an 18 year old male. So that seems reasonable in terms of saving tax dollars. In terms of a principal, focus on the people who are in jail who are innocent. So going back to the deontological effect of altruism. Again the first people that I’d get out are all the people there on immigration offenses.

So I was shocked to find out now, in federal prison people there for immigration offense is now a substantial group. I think I remember something like 10%, 10% of people in federal prison are for immigration offenses.

Robert Wiblin: I thought there were just 600,000 people just awaiting trial in immigration court. It’s kind of extraordinary.

Bryan Caplan: Yeah, so those cases are a bit different because I think those are generally cases where if they disagree to be deported than they’d be freed. Which again, I think is still terrible, but it’s not quite as terrible as someone who just is serving a 20 year sentence for a crime and they got nothing … So basically right now you’ve got people waiting trial, where lock up families in a horrible prison. But the lower bound on how bad it is, is you could go home. Of course, maybe they don’t want to go home because they’re worried that there’s gangsters who are going to kill them or something

Robert Wiblin: Right yeah. I think that’s fairly often the case, at least among the legitimate refugees. Okay, I’ll stick up a link to my presentation. I think many libertarians have been quite good on the criminal justice and prison stuff. I think some others seem remarkably unbothered government goons just going around and harassing people on the street. I’m not quite sure what’s going on there.

Bryan Caplan: So I would say that both the unflattering and the flattering stories are all true.

Robert Wiblin: Yeah.

Bryan Caplan: Right. So regardless of people’s philosophy, a lot of what they care about is based not upon any kind of quantitative examinational problem, it’s just based upon social proof and fads. Like what are my friends talking about? And there are some people who try to go and get a new fad going. I’ve been spending years trying to get the open borders fad going. And saying, look this is the main thing we should be worried about. Right now, there’s a growing fad for something that I think is good, which is getting focused on deregulation of housing.

Robert Wiblin: Yeah.

Bryan Caplan: So you’re actually in the Bay area right?

Robert Wiblin: I am, yeah. We have [crosstalk 02:16:44]on this one, yeah.

Bryan Caplan: So I mean this is something there’s enormous gains to be had from that from almost any point of view. And again this is one where libertarians have been pushing this for quite a while, and you know economists from a wide range of political views. Then, I mean in terms of liberals versus conservatives, I say like, sort of tap into both groups for different groups. Again, for liberals I think there’s just like, well someone would make money building these houses right? Ah-ha! Oh yeah, that makes me more charitable, but when I do talk to liberals about this there is this immediate, I think you’re trying to help someone make some money. Like that would occur, yeah, someone would make some money, okay. But houses would be built. And they’re like, no, no they should be built without someone making money.

Well that doesn’t sound very realistic and the houses would probably be pretty crummy. And among conservatives, as to why they don’t care about it … Intrinsically it seems like the kind of thing they might like. Make it really cheap for American families to have their own home. It seems like it’s really easy to make a conservative case for it. And also, isn’t it just like a bunch of liberal cities that are ruining America. And you’ve got that story. But there isn’t any really good, sexy news story about an individual person who can’t build a home. It’s too much about homes that are not built rather than some visible victims. So I think that’s it harder to get off the ground.

And again, it’s also one where it doesn’t have quite the right villain. The ideal conservative villain is like a transgendered illegal immigrant who-

Robert Wiblin: In a union

Bryan Caplan: Yeah, in a union … So put all those things together to get your ideal villain. Where if it’s just like local zoning boards. And maybe there’s actually a little bit of leniency that conservatives have towards local government versus higher levels of government, also. I haven’t actually heard this, but the obvious conservative case against the proposed California state deregulation of zoning is that it’s a state government that is supplanting the wish of local government, grass roots government, close to the people. I haven’t heard that argument be made but it seems like a pretty easy one to make which would tip the scales back the other way.

Robert Wiblin: Big and charitable. I suspect the real reason people are against de-zoning is that they own houses and they don’t want the competition of new houses. It’s just, their voting their pocketbook.

Bryan Caplan: So I mean this is the kind of story that economists love. Although I spent a lot of the first 10 years of my career just looking at the evidence for stories like this. And in general it just seems like, as a rule it does not hold up. As a rule, people don’t throw their pocketbooks. The main thing you can go and look at people with the same income level as people that own homes, of the same wealth, but who rent.

So it seems … Normally it’s more social proof and what do people in my circle believe and what is the ideology say. This kind of thing seems much more important for people’s’ views about almost any political question than, am I personally gonna be losing money. So again, should someone be able to build a house right next door to me, when it gets down to that level, more likely self-interest gets activated. But if it’s like, what about in general. In general should it be harder and easier to buy a house, this is one where there is very little evidence that this makes much difference on people’s political views.

This is one that it is true that sometimes when you talk to people, sometimes they lie and overstate how good their motives are. But people would have to be Oscar-worthy actors to be as angry as they are about deregulation of housing on moral grounds and really self-interest. Cause when I’ve talked to be, like “So you just wanna go and let everyone just build anything they want. They can build your giant pink Disneyland right there. And that’s your world. And there should be no nature and no trees anymore.”

The person just seems to be genuinely incensed by the idea. Developers making money. Like here’s the main I notice, that people who own sub-dividable real-estate don’t seem to be any more interested in this stuff, in the deregulation than anybody else. I mean a lot … When someone has a piece or a plot of land where they could into five homes and make a ton of money. I haven’t seen any sign that they are more pro-development than someone who is basically just gonna see their housing price fall. If they don’t sell out then basically they’ll be losing money.

The social science has a lot of nice angles to pursue, cause there’s always subgroups that will benefit and you can look can compare them. See are they really different from other people that seem to be in their social circles. And my general answer is very rarely actually.

Robert Wiblin: Well it’s good to know that people are stupid rather than evil.

Bryan Caplan: Impulsive. Again, it’s not so much even intelligence as just there’s a knee-jerk answer and people don’t exercise self-control. Which as a deontologist, I say is blameworthy. They all interact together.

Robert Wiblin: I think it’s bad for a different reason. So you’ve been really generous with your time, and I gotta let you go. One final question, in the case against education you suggest that you think opera is objectively one of the best art forms in the world, which strikes me as just insane. Cause not only do I think it’s not objectively, I don’t think it’s subjectively good. So do you wanna make the case for opera?

Bryan Caplan: Uh, sure. In general with … Alright I say there’s trade off between accessibility and long-term enjoyment. There are a lot of songs that are immediately catchy, and yet after people listen to them a few times, they lose their flavor. Often you come to dislike a song that you once liked, right. And so opera at the opposite end of this usually. Opera’s at the end where a lot of people do not like it at first, and yet … Especially out of people who are smart, people who put time into it are unusually unlikely to get an enormous replay value of it, and be able to enjoy it many more times than they would be able to enjoy popular song. Especially because there’s usually so many more levels working where you’ve got … There’s the plot, there are themes, and the music is also just normally much more complicated. So its not immediately pleasing to the ear for a lot of it.

Look at Faulkner, like a lot of people … There’s a line at the time saying “It’s not as bad as it sounds” and that’s true actually. So it’s the kind of thing where you put years of study into it, I think you really do get increased returns. Now again, this is something where if you have never tried to put that kind of effort into something, then I would say just try it for one thing and we’ll see what happens for you. There’s certain pieces where I’d be happy to say, these are things where I think that if you are someone who is capable of getting these increasing returns, then if you just listen to this thing 30 times, you’ll get it.

And then, also, when you get it you will get a pleasure that is greater than what you’ve gotten from the popular songs that you’ve enjoyed up to this point. Again, especially if you take the integral of all the aesthetic pleasure you get over this time. That is the main case that I make for it.

Robert Wiblin: It’s a very consequentialist answer.

Bryan Caplan: Yeah. Maximize the aesthetic control you get out of it. Yeah, sure. You know there are some people I would say they would just never enjoy it, and then I wouldn’t push it on them at all. You’re someone where I think there actually is a good chance that if you invested something into it that you would get a lot out of it. Now I wouldn’t want you to go and start investing thousands of hours, because maybe I’m wrong and maybe you wouldn’t. But if you were to go and invest, say, 50 hours over the next three years, into something like this, I think that would probably be well worth your time in terms of just the option value.

Robert Wiblin: I’ve gotta preserve my self-image those Bryan. I’ve carved out the niche as someone who thinks Nicki Minaj is the greatest art form that’s possible. My guest today has been Bryan Caplan. Thanks for coming on the 80,000 Hours Podcast, Bryan.

Bryan Caplan: My pleasure to have you, or my pleasure to be on it rather. I’m gonna say you can get The Case Against Education for only $20 on Amazon. It is packed full of useful advice, not only for effective altruists, but also for effective egoists. So if you are either of those or a combination of the two, you’ve gotta buy this book. Right, Rob?

Robert Wiblin: Absolutely.

Bryan Caplan: Absolutely.

Robert Wiblin: I read the whole thing. We didn’t have time to discuss your previous books, but I’ll stick up links to those as well and I’ve read them both and very much enjoyed them. So people should go out and give you a dollar with each copy that they buy.

Bryan Caplan: Alright, thanks a lot.

Robert Wiblin: Have a great day.

Bryan Caplan: You too.

Robert Wiblin: If you’re enjoying the show please leave us a review on iTunes. That helps other people find out about us. We’re also looking for quotes to put on our homepage, and we might grab some from there.

The 80,000 Hours Podcasts is produced by Keiran Harris.

Thanks for joining, talk to you next week!

About the show

The 80,000 Hours Podcast features unusually in-depth conversations about the world’s most pressing problems and how you can use your career to solve them. We invite guests pursuing a wide range of career paths - from academics and activists to entrepreneurs and policymakers — to analyse the case for and against working on different issues and which approaches are best for solving them.

The 80,000 Hours Podcast is produced and edited by Keiran Harris. Get in touch with feedback or guest suggestions by emailing [email protected]

What should I listen to first?

We've carefully selected ten episodes we think it could make sense to listen to first, on a separate podcast feed:

Check out 'Effective Altruism: An Introduction'

Subscribe by searching for 80,000 Hours wherever you get podcasts, or click one of the buttons below:

If you're new, see the podcast homepage for ideas on where to start, or browse our full episode archive.