Search the Site

Episode Transcript

My guest today, Michael D. Smith, is a professor of information technology and marketing at Carnegie Mellon University. His recent book, The Abundant University, is a scathing critique of the U.S. higher education system.

SMITH: Everybody involved in the system is doing what their incentives tell them to do, even though we know it’s leading to really bad outcomes.

Welcome to People I (Mostly) Admire, with Steve Levitt.

Michael Smith doesn’t only criticize education, he’s also got a bunch of ideas for transforming the system. And I’ve got some radical ideas of my own, and I want to put those out there to see what Michael — and you listeners — think about them.

*      *      *

LEVITT: Michael, you’ve got a new book. It’s called The Abundant University, and it’s a forceful indictment of our current system of higher education. Of course, you’re right in the middle of that system. You’re a tenured professor at Carnegie Mellon University, and this is a book that the powers that be will not like. I hope you weren’t banking on any salary increases any time soon.

SMITH: Yeah, exactly. Yeah, I’ve now pissed off everybody at Carnegie Mellon. I wrote a piece in The Atlantic in 2020 and the basic premise was higher education in 2020 looks a lot like the entertainment industry did in 2015. We’re fat and happy and so darn pleased with ourselves and completely unaware that technology might change our business. And the feedback I got from that really opened my eyes to how much we are trying to protect the status quo.

LEVITT: So let’s talk about that status quo. You argue that our current system of higher education is built on three scarcities. Can you explain what you mean by that?

SMITH: When we talk about market power, you get market power by being able to do something or controlling some sort of resource that your competitors don’t control or can’t do. And at least when I looked at that in the context of higher education, I think higher education power derives from being able to control access to the scarce seats in the classroom, access to the scarce faculty experts, and access to the valuable credentials that help you get good high paying jobs.

LEVITT: Let’s cover those one by one. You’re saying that there’s a scarcity of access to seats in the classroom. What do you mean by that?

SMITH: What I mean by that is — you know, my wife and I have had three kids apply to college, and if you type into Google almost any college’s name, the autofill will be “acceptance rate.” One of the key ways we measure quality is proportionally how few students get admitted. Controlling access to those scarce seats becomes a key source of power.

LEVITT: That makes perfect sense. Now, your second scarcity was the scarcity in instruction. What do you mean by that?

SMITH: What I mean by that is access to faculty experts. So in the book, I talk about my advisor, Eric Brynjolfsson, Catherine Tucker, Ed McFowland, Natalia Levina, and the point I make is: if you want to learn from Eric Brynjolfsson, you’ve got to get admitted to Stanford. And if you want to learn from Catherine Tucker, you’ve got to get admitted to M.I.T. And on down the line. And I think it’s the powerful universities who control access to those experts, however we define that.

LEVITT: That makes sense. So the faculty teach the students at the university primarily. They don’t teach the general public. If you want me to lecture to you, then you got to come to the University of Chicago. Okay. Now the third scarcity you mentioned is credentials. What do you mean by that?

SMITH: What I mean by that is controlling access to the really high value credential. And here I’m going to say university brand name — that how people are perceived in the job market, at least early in their career, is heavily dependent on the brand name of the university they graduated from. Getting a degree from the University of Chicago is going to get you a whole lot farther in the marketplace than University X.

LEVITT: I’m glad you said X. I wondered what university you were going to disparage.

SMITH: I’ve already pissed off enough people. I’m trying to minimize it as much as possible.

LEVITT: So the elite schools admit very few students. They have big name faculty that people would like to learn from, but can’t because they don’t go to that school. And then when you graduate with this degree, it gives you a big reward. And the bottom line of all of that is that universities have a ton of market power, especially the elite ones, and they charge really high sticker prices as a result of that. And in your own words, you described the current situation as financially and morally unsustainable. And that’s why you’ve written this book is because you think that higher education is completely and totally broken.

SMITH: We are unintentionally trying to protect a status quo that I think if we stopped for a second, we would recognize needs to change.

LEVITT: And when you say “we,” you mean academics? You mean the universities? You mean the U.S.? Who are you talking about when you say “we”?

SMITH: “We,” I mean academics and university leaders. What we know from Raj Chetty’s research is that kids born into the top one percent of income have a 1 in 4 chance of getting access to a top selective school. Kids born in the bottom 20 percent of the income distribution have a 1 in 300 chance of gaining access to the same university.  

LEVITT: So let me say, I know a lot about this issue, but I was unaware of that Raj Chetty paper and that fact, which is incredible — that people who grow up in the top 1 percent family income wise are 77 times more likely than people in the bottom 20 percent to go to an elite school. I mean, I knew there was some bias, but I would have never guessed it was so big. But I want to stop you there because having just heard that fact, why do you think that is?

SMITH: We use selection criteria that we know are heavily influenced by wealth. So SAT scores. Were it an unbiased measure of your intelligence, I’d have no problem with, but my wife and I had someone sitting in our living room explaining that universities use SAT and other standardized test scores because they’re the only objective measure that the university has of someone’s preparation for college, and then in the next sentence saying, “And I can increase that objective measure in the next two weeks if you give me $3,000.” It can’t be objective if I can pay you $3,000 to increase it over the course of the next two weeks. 

LEVITT: So for sure, the SAT has this feature that if you study for it, you can improve your scores. And for sure, affluent people, especially historically, have done a lot more of that. And maybe I’m too defensive about the SAT because I’m a huge personal fan of David Coleman, who runs The College Board. I think he’s done amazing things in education, and I had him as a guest on the show. But one thing that’s interesting is that David Coleman worked with the Khan Academy to make sure that really top notch SAT prep materials are available for free. Of course, that’s only a start because how do the kids in the bottom 20 percent even find out about that because their counseling is so bad. An important question about the SAT is: is it actually a good predictor of how kids do in college and beyond? And my sense is it is. And that the SAT is maybe more like the tip of the iceberg, or more an outcome than a cause? Obviously the piece you’re talking about where affluent kids can manipulate their scores on the edge by doing prep courses, that’s a clear problem with the SAT that they’re trying to deal with. But often there’s a broader indictment to say something like, “Well, the SAT is biased against poor kids.” But you would know better than me. I don’t know the evidence. Is there actually any evidence that the SAT mispredicts success in college or in life for affluent versus non affluent kids?

SMITH: The studies I’ve seen is that it’s not all that well correlated with success. And this is not my area of research, but I’ve seen a bunch of people make an argument that high school G.P.A., regardless of what high school you went to, is a much better predictor of success in college.

LEVITT: So to explain the big gap between the affluent and the people who have low incomes. Your argument implicitly is if we got rid of the SAT this gap would shrink and I don’t even think that’s true. I don’t think getting rid of the SAT would do a lot to help the people in the bottom 20 percent.

SMITH: That’s a totally fair point. Again, seeing three of my kids get admitted to college — as someone not in the 1 percent, but above the average, let’s just say — you just see all these ways in which the finger of wealth gets put on the scale of admissions. So my wife and I live in the city of Pittsburgh and we pay to send our kids to elite high schools. We were sitting down with the college admissions counselor in a one on one meeting and we were going through one of our kid’s list of schools, and he had a reach school on his list. And the guidance counselor looks at it and he goes, “Oh, I know the head of admissions at this school. I’ll make sure that they pull your child’s application.” And walking out of that, you realize, my guess is the kids who go to the local public school don’t have that connection to get their applications pulled.

LEVITT: I was surprised at the moral outrage around the Varsity Blues scandal, around the attempts to cheat and lie to get into college. Because obviously it was wrong. But to me, it was different only in degree from what you just described you were doing. And you’re not embarrassed about it. You wish the world were different, but in the world we live in, it makes a ton of sense for parents and kids to angle to get into good schools.

SMITH: You’re exactly right. These were the logical extension of practices we all already accept. And I think that was the part that really infuriated a lot of people outside of higher education is just seeing that up close and personal.

LEVITT: There’s a deeper thread that runs through American society between the affluent and the people on the bottom 20 percent, which is about the information they have. If you’ve never been to college, you have a hard time guiding your kids to apply to college. And I think as a society, we have not figured out a good way of getting around that. The K through 12 schools would be the obvious vehicle for doing that, but they haven’t been able to. Now, I don’t have the answer, but I think that’s a really good problem for people to be thinking about, and I don’t think enough thought is going into it.

SMITH: What I’m trying to argue in the book is that a lot of these problems are systemic, which means everybody involved in the system is doing what their incentives tell them to do, even though we know it’s leading to really bad outcomes. And one of the examples I use is President Mark Becker at Georgia State University, who in 2009 became very intentional about, “let’s deemphasize the role of wealth in how we admit students.” So he deemphasized the SAT, he increased outreach, all these wonderful things. And he increased the number of Pell eligible students from 32 percent to 60 percent of enrolled students. Pell eligible means that you’re eligible for a Pell grant from the government, given to students who come from, from poor backgrounds to help them afford college. He ended up with three times more Pell eligible students than the entire Ivy League combined. So he did all the things you would hope we’d do. And the only problem was by doing that, he fell 30 places in the U.S. News World Report rankings. He got hit in the marketplace for doing what society would, I would argue, want him to do. And you could just see this over and over again. I talked to one senior person who had just retired from a major state university, let’s just say, and what she told me was her state school has a cap on the number of out-of-state students it could admit, and out-of-state students generate three times the revenue of in-state students. And so her president and provost came to her and said, “Let’s make sure we maximize the revenue on the out-of-state students and and let’s do that by just targeting out of state students who live in high income zip codes.” Does it make sense from the perspective of the business model? Absolutely. Does it make sense from what we should be wanting to create in higher education? I don’t think so.

We’ll be right back with more of my conversation with Michael D. Smith after this short break.

*      *      *

LEVITT: So we’ve talked about the morally unsustainable part of your equation. What do you mean when you say that our system is financially unsustainable?

SMITH: It’s financially unsustainable because we know college tuition has increased at four times the rate of inflation over the course of the last 50 years.

LEVITT: Okay, so college prices are truly going through the roof, as our basic impression has been. Why are college tuitions going up so fast?

SMITH: When you do this talk at enough universities, you get side conversations with folks that give you interesting little nuggets. And one of the side conversations I had was with a person who said they heard their president give a talk to a local community group — and I document this in the book — and the title of the talk was, “Why Does It Cost $60,000 to Attend University X?” And what the president said is, “Reason one is technology doesn’t lower my cost, it actually increases it. I still got to maintain all the buildings. I still have to pay the faculty. And now I have to invest in all these new technologies. Number two is I’ve got to keep up with the Joneses. If every other university is building big fancy dorms, I’ve got to build big fancy dorms if I want to be competitive. And number three” — and apparently this is a direct quote — “Why does it cost $60,000 to attend my university? Because I f***ing can.” And then he went on to explain, not using these words, but basically: “Every year I raise tuition, every year I get the same number of applicants, if not more. Clearly I’m not on the elastic point in the demand curve. Why shouldn’t I raise tuition?”

LEVITT: What’s weird is why they didn’t do that 40 years ago, right? Or how they missed the boat. Either something changed, and they couldn’t charge a fortune 40 years ago, or they were just out of step — which might be the real answer is they didn’t realize how much they could charge 40 years ago, and they’ve only slowly been catching on to the fact that their demand is extremely inelastic.

SMITH: I think a college degree has become more important in the job market over the course of the last 40 years. I also think that the U.S. News and World Report rankings have become more important over the last 40 years. And only 25 universities can be in the top 25 at any given time.

LEVITT: I don’t know if this is really true. At least in economics, there’s maybe eight schools in the top three, 25 schools in the top 10, and about 50 in the top 25. I don’t know, I think you got something wrong with the math there.

SMITH: It’s a limited number in the top 25. How about that? Another perversity of the system is, when our kids were applying to college, they took the practice SATs and all of a sudden they got a whole bunch of advertisements from colleges telling them how much they wanted to apply. “We’re so interested in you, we’ll waive the application fee, blah, blah, blah.” And one of the schools in particular, he just didn’t have a chance of getting in. And so we’re just scratching our heads. Why is this elite university recruiting our son? And about a month later, The Wall Street Journal came out with an article that documented that elite schools are buying data from the college board to recruit students they know they won’t admit. Because increasing the number of applicants is a much easier way of reducing your admission ratio than reducing the number of people you admit.

LEVITT: You haven’t talked about college debt. Is there a change in what’s been happening with college debt, or is it just that all of a sudden the media is reporting it and the Biden administration cares about it, so it’s getting a lot more attention? Is there a lot more college debt today than there was 10 or 20 years ago? And are default rates going way up as well?

SMITH: The answer to both questions is yeah. And I don’t know whether these are inflation adjusted, but I do know that college debt has doubled in the last decade. It’s 1.8 trillion now. We’ve got a chart in the book that actually shows the increase in college debt relative to other forms of debt, and basically what it shows is other forms of debt — home debt, vehicle debt — are basically flat from 1989 to 2019, and college debt is going through the roof. College debt’s gone up over 700 percent during a time where every other form of debt was basically flat.

LEVITT: Now per se, that’s not a problem. I think everybody who studies it agrees that getting a college degree seems to, in a causal sense, be related to higher earnings. Young people should be borrowing to finance that excellent return in their own human capital, that wouldn’t be such a bad thing. Why is everyone so outraged about college debt?

SMITH: I think a lot of people are outraged because they don’t feel like they are getting the R.O.I. I would also argue that, yeah, all the numbers tell us that college is a great investment. If we were allocating that investment equally to both rich and poor, I wouldn’t have a problem with that. The problem is if we create a system where the rich get access to college and college is the best way to get rich, then we’ve created this terrible feedback loop in society of increasing socioeconomic stratification.

LEVITT: I probably should have known it, because I’ve been part of the system for a long time, but it wasn’t until I read your book that it really hit home to me how powerful our current system of higher education is as a means of maintaining a high and increasing degree of income and wealth inequality in this country. I always, I think, still subscribe to some mostly false American dream that education is key to opportunity, and that in this country we’ve done a better job than almost anywhere else in making education available. But I guess the problem is that access to elite universities is highly closed off, even access to four-year universities. I couldn’t believe you had a statistic in the book that blew my mind with respect to four-year college graduation. So in 2016, among those who grew up in the top quartile of income, 58 percent of young people were getting bachelor degrees or more. And in the bottom 25 percent, 11 percent were getting bachelor’s degrees. I had no idea. In a world where the returns to education are so huge, that fact is critical to thinking about inequality. And it’s strange that I wasn’t aware of that fact until I read your book.

SMITH: I don’t think we think about it in higher education. I know I didn’t before I started writing the book. You start to play around with these things and you realize we really are contributing to socioeconomic stratification by any measure. And I don’t know how we solve it from within the system. I think these are classic systemic problems, which means we’re not going to solve it by just tweaking the way we do things today. I think we’re going to have to come up with something very different.

LEVITT: Alright, so we’ve spent a lot of time criticizing higher education, and before we get into what you see as some of the solutions, I want to take a little bit of a U-turn because you’ve spent much of your academic career studying a different industry, the entertainment industry. And at first blush, someone might say that higher ed and entertainment — they have nothing to do with each other, they couldn’t be more different. But, I think when you take a step back, all of a sudden, some remarkable parallels emerge. Could you talk about your research on the entertainment industry? How it had been remarkably stable, and then — BAM! — suddenly everything turned upside down.

SMITH: My colleague Rahul Telang and I spent the last 15 years looking at the entertainment industry and how technology was changing it. And we were really blown away by how much the entertainment industry, when they saw new technologies, used them as a way to replicate their existing business model. “I’ve always sold CDs. Let’s use iTunes as a way to sell CDs online.” They weren’t thinking about “How do we create Spotify? How do we create a bundled streaming service where all the music is available?” It’s, “How do I replicate my existing business model?” The questions they weren’t asking were, “How can we use the new technology to do what it does really well, which is create these wonderful bundles of access to information?” And also to gather information about the customer, which might allow you to get the content in front of exactly the right member of the audience. I think this is what Netflix did so brilliantly. They created shows that wouldn’t have worked on traditional broadcast television, where you’re trying to hit the middle 80 percent of the market. They created shows that wouldn’t work in that medium, but work perfectly well on Netflix, where I can use data to target, “This is the show that I know you’re going to be interested in.”

LEVITT: I hadn’t really thought about that. They produce so much content, but they have two advantages over broadcast TV — well, three. One is that people can watch it whenever they want. They don’t have to watch it all at the same time as it’s transmitted over the airwaves. Number two, there’s no limit on slots, right? The networks battle with this limit. “We only have six slots per night. We have to fill those up with the best shows.” And the third one you pointed out about data. I had never thought about the fact that knowing what I like and being able to tell me what other six shows I might like allows me to find content that otherwise I’d never find because there’s such an infinity of new content being produced.

SMITH: Rahul Telang and I wrote a book in the mid 2010s titled Streaming, Sharing, Stealing: Big Data and the Future of Entertainment. And in that book, we started by trying to use some of the standard models of disruption, and what we discovered is they didn’t work that well to explain the phenomenon we were seeing. And so we developed something that I’m going to call “structural disruption,” which basically says, what are the factors that have always defined success in this industry? And in entertainment, it was: the big six studios were powerful because they controlled access to the scarce financial and technical resources you needed to make content, they controlled access to the scarce distribution channels you needed to reach your audience, and they were able to use copyright law to create an artificial scarcity in how people got access to content, right? So it’s still about scarcity. And none of the technological shifts they’d faced over the last hundred years had changed the power of those sources of scarcity. And yet, in the mid 2010s, they were facing a set of simultaneous technological shifts that were making it easier to make content, easier to distribute content, and — because of piracy — much easier to get access to content. What we argued in the book is the new scarce resource is customer data. And guess what? Amazon, Netflix, and Google own your customer, and they own the customer data. That’s the threat you face.

You’re listening to People I (Mostly) Admire with Steve Levitt and his conversation with Michael D. Smith. After this short break, they’ll return to talk about some possible alternatives to the four-year college degree paradigm.

*      *      *

We’ve covered all the background. We’ve talked about the problems with higher education. Now we get into the good stuff. What might we actually do to shake up the status quo?

LEVITT: Arguably higher education has been the most stable industry in America over the last hundred years. I went back trying to find a ranking of the top schools. I found a ranking from 1910 and the top five schools in 1910, so 113 years ago, were Harvard, University of Chicago, Columbia, Yale, and Cornell. MIT was 10th, Princeton was 14th, Stanford was 15th. Really, you know, the same set of schools are there. And just for fun, I went back and I looked at the highest market capitalization companies, and I found from 1917 that the five biggest companies in America by value were: U.S. Steel, AT&T, Standard Oil of New Jersey, Bethlehem Steel, and Armour and Company, which I had to go look it up — they were a meatpacking company. So the economy has been completely transformed. Higher education hasn’t. So what in the world makes you think that we’re at the cusp of a radical change in higher education when everything else points the opposite direction?

SMITH: I started the book by trying to apply that same sort of structural approach, like the entertainment industry. What are the scarcities that define power in the market for higher education? Back in the early 2010s when massively open online courses were available, some really prominent people said these online courses are going to disrupt power in higher education. So here we are; nothing much has changed. Why is that? And the argument I’m trying to make in the book is that online courses created abundance and access to the seats and access to the faculty experts. What it didn’t change was the value of the credential. You can take as many edX courses as you want — it’s not going to give you the same power in the job market as a four-year degree. If that were to change though, I think higher education could be significantly disrupted.

LEVITT: And you’re saying the degree is what gives you the market value. The degree is what is important if you want to get a great job, a well paying job, and that piece up until now has been missing and, consequently, you don’t see a lot of people taking the equivalent of Stanford’s coursework online, even though they could if they wanted to. But what I don’t see is the easy avenue to that last piece changing. Do you see that?

SMITH: This will only change when employers change their practices. So this is all predicated on the idea of: as long as employers still rely on four-year degrees, we in higher education are going to be just fine. Could that change? I think not only could it, but I think it is. If I have an objective signal that you’re a good coder, then I don’t need to rely as much on the Carnegie Mellon degree. If I have an objective signal that you’re a good writer, I don’t care whether you graduated from Columbia’s J-School. I want to know if you can write. Show me whether you’re a good writer or not.

LEVITT: So usually I’m way behind the times, but I want to say that on this one dimension, I’m right on the cutting edge. Now, it’s not because of me. I run a little center at the University of Chicago. It’s called the RISC Center. And I was too busy last year to be involved at all really in the hiring. And so I just delegated it to my team of young people, recent college graduates, all from elite universities. And they decided unilaterally that they wanted to do hiring differently. So what we did last year is the application required carrying out a task, which was not too different than the kinds of tasks that we do at the center, and anyone was invited to apply. I even over this podcast advertised it to anyone. I said, “Look, you don’t need a college degree. You certainly don’t need an elite college degree. Anyone can apply.” And we did the first round of cuts completely blind. And two things really jumped out at me as being interesting. The first is that even though we tried to cast a really wide net, in the end, two thirds of our applications were still from elite schools, from the set of schools that we typically would recruit from. The second thing that was interesting was the success rate. So if you took an application from an elite school, it turned out that, analyzed blind, 26 percent of those applications made it over the hurdle and got an interview. If you take the non-elite applications, so from non elite schools or people who didn’t have college degrees — although very few of those did apply — 13 percent made it through. I actually was surprised at the difference that the elite school applicants were twice as effective at getting through our screen as the non elite applicants. I would have expected maybe there’d be a small difference, but I hadn’t expected that difference to be so big.

SMITH: I think there’s a correlation between where you got your college degree and your capabilities, but it’s weaker than I think most of us recognize.  

LEVITT: So we hired maybe seven people. And one of the people we hired, he never would have been hired otherwise. His path started in community college and he went through a state school, didn’t have very good grades at all. But he was really good and really impressive. He’s full of ideas and he had talent. We liked this experiment we did in terms of hiring. We’re doing it again this year. But let me just say, I agree with you that you need employers to make a change in order to displace the current system. Being totally realistic, I think the amount of change you would need from employers, I just don’t see it happening. I think you’re a little bit unrealistic in that regard, to think that you could rely on employers acting one by one to radically change what they’re doing when the easiest thing in the world to do in HR is to use college degrees as a credential. And for that reason, I think that the change is not going to come from within the system of HR at big firms.

SMITH: What I would say is the other thing I think we’re seeing is employers vertically integrating upstream into training their own employees. One of the really interesting facts we talk about in the book is there was a 2014 Gallup survey and what they found is that 96 percent of chief academic officers say their university is “very or somewhat effective” at preparing students for the workforce. Fourteen percent of Americans agreed with that same statement, and only 11 percent of business leaders agree with that statement.  

LEVITT: The problem with most training you get is you can’t really tell if it’s going to help you or not. And maybe it doesn’t. You pay the costs up front. And believe it or not, I actually pitched something to one of the big tech platforms a couple of years ago. So this is a big tech company that offers certificate courses. And they charge people a little bit for them, and a fair number of people start them, and a decent number finish. They don’t make money on it. They do it just because it seems like a good thing. It gives them positive publicity and goodwill. But they can’t even tell right now if the program helps people get jobs or pay increases. And that’s why they approached me initially is they wanted to see if I could help them measure what was going on in their program. Okay. But the more I learned, the more convinced I became that their approach was completely backwards. And if they thought about it differently, they could do enormous good. They could make a ton of money. And they could also begin to radically transform higher ed. So the idea was really simple. I suggested to them that they should make the class free, or probably even pay people to complete the class, because the money for them is not in the tuition. The firm offering the course can internalize a good chunk of the value by combining this certificate course with a job matching market, right? So employers would be willing to pay the firm offering the courses a premium to have the first crack at recruiting these students. And the firm offering the courses could even skim off the best performers for itself. And when I tried to work out the numbers I could see this could be hugely profitable. They could really create an alternative job market for any skill that was easily measurable, whether that’s computer programming, maybe data science, project management. There are a whole set of things you could think where you could really make this work. So I sat down with some people from the firm and we wrote up this manifesto where we talked about how this was a once-in-a-generation opportunity for this firm. And we passed this memo around to anyone who would look at it and in the end, nothing actually happened. And that was super disappointing to me.

SMITH: I also totally agree with you about the data. Going back to entertainment, one of the problems with broadcast television is you’ve got to make the middle 80 percent of the market happy. I can’t really serve the people with niche interests. I feel the same thing in the classroom, right? If you’re teaching a class of 50 students, I’m pretty sure that 10 percent of them are bored and 10 percent of them are lost. I just don’t have time for them because I’m trying to keep the middle 80 percent of the classroom happy. Wouldn’t it be cool if we could create a much more self-paced, personalized system of education to train people? I think we’ve got an opportunity to do that. I would have loved your manifesto. I’m sorry that it didn’t get picked up.

LEVITT: It’s yours. If you want it, please take it. You probably hold a lot more sway with the tech companies right now than I do. I still believe, to make this worthwhile for the people who are producing the content, they have to be able to coordinate a job market afterwards. I think Google is the one firm that most easily could just say, “Hey, higher education is broken and we’re Google, so we’re going to fix it.” And they’d have a crack at doing it.

SMITH: Yeah, because they hit both sides of the two-sided market.

LEVITT: Exactly. And they’re great with data, and  they have immense talent, and they’re trusted — so that if Google would essentially guarantee workers and say, “Look, if someone passes our class, we are saying they’re good enough to be hired at Google, and we are hiring huge numbers of graduates of our program,” and then other employers I think would line up to do the same. Let me ask you about another alternative path that could have some disruptive force on higher education, which would be something like a college G.E.D. So just to make clear, in this country, if you don’t graduate from high school, there is a test called the G.E.D., which you can take, and after you take it, it is a credential which is supposed to say that you have acquired the knowledge that a high school graduate has. What do you think of the idea of doing that via, say, College Board, as an example? The same people make the SAT, they could create a college G.E.D. test and a set of materials to study for it that could credential people with the equivalent of a college degree, but you wouldn’t have to pay any money and you wouldn’t have to spend any time in a classroom. Do you think that is a good idea? And do you think it has any legs at all?

SMITH: The challenge is doing it in a way that matches the diversity of the skills employers need. Google’s been very public about the fact that, “We’ve looked at our data and what we discovered in the data is where you graduated from college, what your degree was, what your G.P.A. was isn’t all that predictive of how well you’re going to do at Google relative to our entrance tests.” I don’t know whether you could create enough of a generalized test that would apply to all sorts of different employers, not just Google.

LEVITT: What you just described was an even higher standard. Let’s put that aside. I could at least imagine designing a test that would involve some writing, a little bit of math, maybe a little bit of some foreign language, all the kinds of things that we make students do at colleges. Maybe it’s hard because one feature of college is we let students major in a wild variety of things and what a fine arts major knows and what an economics major knows are very different, but I could at least imagine that you could have a version of this that has a core set of things you’d hope any kid would know after they graduate from college, and then something that corresponds to a major. There might be a fine arts version, there might be an economics version, an engineering version. So I don’t know if it could work, but it feels to me like an idea that if somebody like College Board got behind it, might just make a difference.

SMITH: What again is shocking to me is how much we in higher education are trying to protect the status quo versus trying to use all of our creative skills to figure out how to improve this system.

LEVITT: So you’ve written this book. It’s really a call to arms to change the higher education system. One of the things that I’ve noticed is that our incentives in higher ed are very much to point out problems but then none of the returns come from actually rolling up our sleeves and changing the world. Every incentive you have is to go back to Carnegie Mellon and do exactly what you were doing before. Do you have any inkling to break out of that mold and to actually start to change what you’re doing to be a force for transforming the world?

SMITH: It’s a great question. I wrote the book and invested a whole lot of hours into writing a book that honestly was not designed to be a bestseller. I was told right up front, if you want this to be a bestseller, you should call it The Death of the University. This was really a book written to my colleagues to say, “Hey, I don’t think we can keep doing what we’re doing.” I hope to be a force for change in higher education in terms of, let’s rethink from scratch how we do this. And this comes back to the entertainment industry. They’ve been really good at responding to disruption and the question is why? Why did the entertainment industry figure this out when so many other industries didn’t? And I honestly think the answer is they rediscovered their mission. The entertainment industry initially opposed technological change because they saw it rightly as a threat to their business model. But at some point they realized: my mission is different than my model. My mission isn’t selling shiny plastic discs for 20 bucks a pop. My mission is creating great entertainment and getting that entertainment in front of an audience. And that mission was so compelling to them that they were willing to blow up everything else about their business to go after it. What’s the parallel? What’s our mission in higher education? If our mission is to help rich kids get a leg up in the job market, we’re doing great. But that’s not our mission. If our mission is helping kids from all sorts of backgrounds discover their talents, develop those talents so they can use those talents to the benefit of society, I think we could do a whole lot better if we were to embrace some of these technological changes. That’s what I’m excited about.

Conversations like this get me really excited because I agree wholeheartedly with Michael Smith. We need and deserve a radically transformed education system that works better for far more people. And there’s nothing like talking to someone who agrees with you on every point to make you feel like change is possible. But let’s be realistic. The university system is remarkably resistant to change. I’ve been trying for years, and all my efforts so far in this area have been complete failures. Is there any reason to think that things will be different in the near future? Michael Smith thinks so. I’m not so sure. What do you think? Are any of these ideas we talked about any good? Maybe you have better ideas. Is there any chance, even with a great idea, of breaking the stranglehold that elite universities have? Send us an email. The address is, that’s I’m really curious to hear your take on the situation.

LEVITT: So now is the point in the show where we take a listener question, and as always, I’m joined by my producer, Morgan.

LEVEY: Hi Steve. A listener named Nathan wrote to the show about student labor unions. He’s a Ph.D. candidate at M.I.T. and graduate students at M.I.T. voted to unionize last year. This is happening at other universities as well, the University of Chicago included. Nathan generally supports unions, but he’s found himself increasingly skeptical about the role of traditional labor unions in academic settings. At a university, the laborers are students conducting research and teaching activities, and the management are professors and university administrators. He believes it’s a fundamentally different dynamic than most labor unions encounter, and he’s curious about what you think about the graduate student unionization movement.

LEVITT: So what I think makes the union question so hard, in part, is that there are very different kinds of graduate students. I think there’s a real divergence between the graduate students who have good career prospects and the graduate students who don’t. And the ones who don’t have been, by and large, the ones who’ve been leading the charge for unionization because I also think they’re much more exploited by the university because they don’t have a lot of leverage, and because they often tend to stay longer because there’s no reason to rush out into a job market that’s not going to be very rewarding anyway. The grad students who don’t have good prospects will be the ones who end up running the grad student unions. And the fear is that they’ll put in place a bunch of rules that are beneficial to them, but are costly to the graduate students who have really good career prospects right now.

LEVEY: So let me give you an example. Say I’m a graduate student and I’m working with an advisor who’s very exploitative of my time and energy. Isn’t that the point of the union to protect graduate students who are being exploited by their advisors?

LEVITT: It seems sensible to argue that that’s what the union might be good at, but I don’t actually think that’s true. So certainly when I was a graduate student, I didn’t understand just how important my advisor’s endorsement was on the job market. If you’re not a superstar, you can’t get a decent job unless your advisor stands up for you. So if you have an exploitative advisor, the union can probably help you in making your years in graduate school somewhat less horrible by limiting what that advisor can and can’t do. But what a union can’t do is to force that advisor, when you’re in the job market, to go out and sell you so you actually get a good job. I actually think the union is dangerous in this regard because the right thing to do is to get the hell away from that advisor and go find one who’s not exploitative. I think the logic that union will help you in this case is actually backwards.

LEVEY: But what about for protections that an employer gives its employees, things like parental leave. Wouldn’t a union protect students in that regard?

LEVITT: You’re absolutely right that those are the dimensions in which you think a union could really help. That’s a much more powerful case for the unions than the first scenario we talked about, where the student had the bad advisor. But let me say this. In a situation where the union is actually successful and actually manages to shift the power greatly in favor of the students, that raises the cost of the university to having Ph.D. students and the university’s response almost certainly will be to reduce the number of graduate students, especially the number of graduate students who are seen as primarily a cost to the university and not much of a benefit.

LEVEY: Nathan, thanks for writing in about this complicated topic. If you have a question or a note for us, our email is That’s We read every email that’s sent and we look forward to reading yours.

In two weeks, we’re back with a brand new episode featuring Werner Herzog. People of my generation know him as an incredibly innovative film director. Among the younger generation, he’s earned a following for playing the role of The Client in the Mandalorian series. He has a well-deserved reputation for being completely unpredictable, so who knows what will unfold.

HERZOG: I don’t like to speak about feature films or documentaries. I’m a writer. I’m a poet.  

As always, thanks for listening and we’ll see you back soon.

*      *      *

People I (Mostly) Admire is part of the Freakonomics Radio Network, which also includes Freakonomics Radio, No Stupid Questions, and The Economics of Everyday Things. All our shows are produced by Stitcher and Renbud Radio. This episode was produced by Julie Kanfer with help from Lyric Bowditch, and mixed by Jasmin Klinger. We had research assistance from Daniel Moritz-Rabson. Our theme music was composed by Luis Guerra. We can be reached at, that’s Thanks for listening.

LEVITT: Oh, wait. I’m sorry. Can I throw in one more thing while we’re on this topic? 

SMITH: I love how you run your podcast. This is fun.

Read full Transcript


  • Michael D. Smith, professor of information technology and marketing at Carnegie Mellon University.



Episode Video