Ethics and Experiments in Economic Research

Would you sell a kidney? A country? A child? Where is the line? And how can economists inform where it should be? This week we're asking, should economics have fewer taboos ? This is the Economics Show with Samea Keynes. I'm joined today by Alvin Roth, who is speaking to me from Stamford. Al, hello. Hi . So we always start this show with a silly question. So on a scale of one to ten , how relaxed are you about market ization? Um, so you know, ten , you're extremely relaxed about having transactions in literally anything, and maybe five is the average person . So I'm probably a 7.5, maybe 7.52. I love the specificity. Well, human beings are living a lot longer than they did a hundred years ago and certainly a thousand years ago, and a lot of that has to do with markets and cooperation and coordination and competition. So so I'm a a big fan of markets. Also I I think of markets more broadly than many people do, including many economists. We talk about the marriage market sometimes and people think of that as a metaphor, but I think there's a marriage market whose market like nature has been clarified by all these dating and matrimonial apps. Right? Marriage is a matching market, one of those many markets like labor markets, in which you can't just choose your spouse, you also have to be chosen. You can't just choose your job. You also have to be hired. So I think there are lots of markets in which prices are much less important than than we think they are when we think about commodity markets. Okay. And uh how do you think the world would be different if everyone agreed with you at that 7.5? Well, I think we would have more well-regulated markets. We would probably have a clearer set of discussions about which markets we want to have and which markets we don't want to have, and which markets we can have and which markets we can avoid. Okay, well I want to ask for some examples now. So let's think about the US. If you could pick one market that doesn't exist in the U S, that should exist, what would you pick? At this point, I would like to see us be more generous to kidney donors. So there's a big shortage of kidneys for transplantation. There are lots of issues you would have to deal with if you wanted to be able to compensate living kidney donors for donating a kidney. I think that those are man ageable, that we could design an ethical, legal, equitable market that would not exploit potential donors and would save many thousands of lives every year in the U.S. Aaron Powell Okay, but just stepping back, what do you think is the best case against allowing that marketing in kidneys? Aaron Powell Well, I think we wouldn't at all like a market in which only rich people could buy kidneys from poor people and therefore only rich people could get transplant. So that's not the market design that I think we should adopt. I think we like the idea that people can receive transpl ants regardless of their income. So a kind of design that people often talk about is that only the federal government would be able to pay kidney donors and they would allocate the kidneys much much as we do deceased donor kidneys now without regard to the income of the recipient. Kidney disease is disproportionately a disease of poor people, although anyone can get it. So that so that's one argument that you wouldn't want it to to cause poor people who need it the most to not be able to get kidneys. Another argument is that that somehow by paying donors you might be coercing them and so you might want to have very robust informed consent. And it's not that you wake up one morning feeling in desperate need of cash and can sell a kidney that afternoon. What you'd like to see in a well working market is that the waiting list was not for transplants, but for being able to donate a kid ney. So I think those are the two main objections that it might be inequitable to take away the option of transplants from the poor who need them, and that it might be coercive, you might be exploiting vulnerable people. Do you have a sense of what the price would be for a kidney? Like an order of magnitude? So I would think we could be very generous because it would save the American healthcare system or any country's healthcare system a lot of money because di alysis, which is a a medically inferior way to keep people alive when they have kidney failure, is also much more expensive than transplants. In the United States, a transplant costs is around the same roughly as a year of dialysis. But after a year of dialysis, you need another year of dialysis. So Medicare spends $55 billion a year on kidney failure. Almost all of that is on dialysis. So you could really cut the bill to the American taxpayer and still pay eighty thousand or a hundred thousand dollars for a kidney, which I bet is above the clearing price. Yeah, I mean honestly kidney donations is one of those ones where I suppose I'm not that squeamish. But let's talk now about a harder one. So surrogacy say. Where do you stand on that one? Well, let's first describe the landscape of current surrogacy because Britain is in the middle between Western Europe and the United States. In much of Western Europe, certainly in Germany and France and Spain, surrogacy is illegal. They don't acknowledge parental rights to the parents of a surrogate birth. In Britain, surrogacy is legal, but you can't pay the surrogate. It's like kidney donation. And in the United States, commercial surrogacy I think maybe the only exception is Louisiana. So one of the things that one notices in Britain is that there perhaps aren't as many surrogates as there might be if they could be paid. But British people of sufficient means can come to California where I live and have their names on the California birth certificate when the surrogate gives birth. So so the the banning commercial surrogacy is is not an iron wall. It doesn't stop every British person from having a baby in California and bringing her back to Britain. In Germany and in Spain and and France, uh there is an obstacle to bringing the baby back. They have laws against surrogacy, often expressed in terms of trying to protect the vulnerable who they take to be women who might agree to be paid surrogates. Um but of course if you have a ban and and it's not perfectly tight, if if there are exceptions, a ban on surrogacy, the exceptions are going to be babies. And there's no one more vulnerable than a newborn baby. So the German family courts and the French family courts and the Spanish family courts have had to create avenues for parents to adopt their own child because the problem facing the family court is not what's right and what's wrong, but with whom should this baby go home? And a natural answer is the baby should go home with her parents. So that's the landscape of Saragosy. So what do I think about that? I think that the urge to have children is very strong. So I think it's inevitable that bans on surrogacy are going to have important workarounds. People are going to find ways to have children if if they need surrog ets to have children, and that a regulated market is a better way to deal with that than to have these stateless children potentially being born into the world. Just still on this question of what there should be markets in. So so recently I wrote a column about President Trump trying to buy Greenland, which was a transaction that he didn't see a problem with, um, but was quite controversial, I would say in, many settings. Um and I was trying to draw parallels between, you know, the attempt to buy a country and and other, you know, repugnant transactions and explain why people might have such a problem with it. What did you make of that particular transaction? Well, so one thing is there are people who live in Greenland and we no longer think that they can be bought and sold. Uh there are people who live in the Falklands Island s and and you know there's a question of should they have a say in in what country they're part of, even though they're much nearer Argentina than they are to Great Britain. So in the past there were land sales. There's the Louisiana Purchase that we we purchased a big part of what's now the United States from France. There's the Alaska Purchase we purchased Alaska from the Russians. Um I think that um you know, the the people living there didn't have much say on those things. And today we view that a little differently. But to be clear, I think the idea was that you know the inhabitants of Greenland could be paid some sum of money. And you know, there were folks saying, well, just, you know, name the price. Um, there is there is a sum for which most of the population of Greenland would accept to become another state of of the US. So to be clear, I was horrified as I am by many things that President Trump does. I was horrified that he thought we could either buy or seize Greenland, you know, from our from our allies. Makes it very hard to remember why we voted for him. But the question of could the people of Greenland be given conditions in which they would like to join the United States is not a crazy question. The people of Greenland have a domestic government and it makes a voluntary association with Denmark. So it's not inconceivable that they could want instead to have a voluntary association with the United States. That doesn't seem to be what President Trump was talking about. But but Greenland has uh you know a a parliament of its own which which has has agreed that Denmark will take care of its foreign policy and some other things. And and one imagines they could change the terms of that if they wanted to. I mean the argument at the time was that if you create this market, right, essentially what you're doing is you're opening the door to other actors out there in the world to claim like, oh well we've we've paid the Ukrainians, we've just given them a sum of money and now they're quite happy to join Russia. Um and so once you make borders changeable through cash transfers, you're essentially breaking down a norm that territories are fixed and opening the door to seizures, essentially. So so again, you know, the market design is about how to avoid the the bad outcomes and and promote the good outcomes. I live in the United States and and at the moment we have fifty states. It was sort of appalling to hear President Trump talk about making Canada the fifty-first state. But we didn't always have fifty states. Hawaii and Alaska, I think, are the two most recent states. Puerto Rico is part of the United States. Puerto Ricans are American citizens, but Puerto Rico is not a state. So it turns out people can vote to change the nature of their affiliation with with other states. So so I don't think it's obviously always a terrible idea. That doesn't mean that if you have a terrible government, they might not put forward the idea in terrible ways. Can I ask you about the gambling market, which is obviously a a market that has exploded recently in the US? That's an example where many people had moral associations with it. Do you think that's gone well ? I think that we have a new market in gambling, particularly on sports, particularly on smartphones, that has a lot of problems. And we probably are going to need to regulate it in ways that we weren't ready to do when the Supreme Court struck down laws against gambling. So one way in which gambling on smartphones during a sporting event is different from previous kinds of gambling, legal or illegal, was you used to, if you wanted to gamble on a sporting event, place a bet before the game began about who would win and who would lose, maybe involving a point spread. And now you can bet during the game, not just on the final outcome of the game, but on whether the next the next basket will go in in a basketball game. And that opens up two difficulties that we're starting to live So to the extent that gambling is addictive for some people, I think we're seeing more addiction. And the other is that it causes a sinister interaction between gamblers and athletes. You could bet a lot that I'm going to miss my next free throw, and you could pay me to miss my next free throw. You could say if you get two shots and you make the first one, miss the second one, and I'll pay you ten thousand dollars. And that might be a profitable arrangement for the two of us to make, but it undermines the nature of the sporting competition that's supposed to be going on. So so there are problems in both those directions. We're starting to see them and to try to deal with them. Basketball players have been banned from the NBA for for engaging in that kind of transaction. Uh, we probably need more regulation and better regulation in this new gambling environment than we had in the old gambling environment. We've been talking about repugnant markets. What do you make of prediction markets?, R rightight. And so the idea that people could make money from, say, war breaking out. There's an interesting history to that. There was a time some years ago when when DARPA, the defense projects Agency, proposed the prediction market for terrorists, and it was shouted down because it looked like a way for terrorists to make money by planning terrorist attacks. But of course, I think that's a misunderstanding. That's what a prediction market is supposed to do. You're supposed to be able to predict terrorist attacks because the insiders are trying to make money from it. But I don't think it would have worked well because if you were party to the nine-eleven attacks in which airplanes were crashed into theld Wor Trade Center. You could make a lot more money by selling airline stocks short the day before than you could on any prediction market. Right? There's there's a a big thick market in securities that if you have some inside information that's going to affect the fortune of of airline companies would allow you to make a lot of money. And I think the the problems with insider trading on on who knows what the next uh truth social tweet is going to be about are concerning for Americans because they they they might reveal national secrets and because they they seem to be evidence of corruption. But I bet you could make a lot more money if you knew what President Trump was going to say tonight by betting on oil futures, right? By selling Brent crude short or or buying it, right? Because the oil futures is a big thick market in which it's very hard to spot those transactions, unlike the prediction market. So I think the problem with insider trading these days on prediction markets is a problem of corruption, not a problem of the markets. So essentially we shouldn't be so worried about the prediction markets because you can already do plenty of insider trading um through these other markets. I think I'm not worried about insider trading in prediction markets. There are some aspects of prediction markets that that worry me, but some of them have to do i i if if I were concerned with organiz ing military operations for the United States, I wouldn't want the prediction markets to give them away the day befor Right. So strategically they might not be a good idea. Well, on that note, let us throw to a break. And when we come back, I want to ask you about another taboo, which is repugnant economic s We are back from the break . Okay, so Al, you and I last got in touch about a column that I was writing about moral questions in economic research, right? And so in economics, the quest to get good evidence can raise ethical questions as researchers essentially run experiments on people, right? And so the reason we got in touch was you co-wrote this paper describing an experiment during which in job hunting season for economists there was an experiment where junior economists sort of had papers and some of them got retweeted on social media and some of them didn't. And the experiment was to see whether it affected their chances of getting a job. And when the details of this experiment came out, uh it was a little bit controversial, shall we say. Do you want to summarize the reaction ? Well, economists do experiments, and sometimes they do field experiments. That is, experiments that change the rules of a vegetable market in a village in India or change the way anti-malaria mosquito nets are distributed in Zambia. And those experiments help us better understand the implications of different kinds of policies. How should we try to protect people from malaria in Zambia, for instance? And a Nobel Prize was given to Esther Duflot and Michael Kramer and Abajit Banerjee for this work. So I don't think that a lot of people find those experiments repugnant. That is if we're gonna try a new way of distributing mosquito nets, we can't initially distribute them to every village in Zambia, so we might as well randomly choose the ones we 're going to distribute to and see if it works by comparing them with with other randomly chosen not yet getting those mosquito nets. So I don't think that arouses a lot of repugnance. But but some experiments avowed repugnance and, the experiment that you talk about that I did with Yen Chen and some of her colleagues at Michigan, it involved the job market for economists. So there were two objections that that showed up on Twitter. One was experimenting on economists, surely that's not ethical. Economists shouldn't do that. We're special. And the other was maybe these markets are zero sum. And if if the people you randomly have tweeted twice instead of once. That was the difference. Everyone got tweeted. If the people you have tweeted twice instead of once get jobs, surely that just means they stole those jobs from other people. And so we were surprised by both of those things. I was particularly surprised that there were sophisticated economists who thought that markets are zero sum. But but your column caused us to add a section to our paper, because if even you were under that misapprehension, then then we needed to explain better that in many job markets, jobs go unfilled. So there's very little reason to think of labor markets as there are some because the markets are congested, and that makes it hard to know of the many, many applications you get for each position, who you should interview and who you should subsequently make offers to. And if you make wrong decisions, if you're miscoordinated, if you're I used to teach at the University of Pittsburgh, if you're at the University of Pittsburgh, you enjoy the same people who Harvard and Stanford enjoy. But if you only interview those people, you might not get good hires that year. So you have to figure out who you want to interview and you interview many people who you don't make offers to. If you can make that process more efficient, everyone gains. The idea that that that things that economists already do, which is tweet about papers, should not be experimented on, I think doesn't withstand well examination for two reasons. One is we think the market is not zero sum, we can be helping it be more efficient. And the second is that things that economists already do should be studyable by economists. That is, no one is objecting to people saying, I read this paper that I really liked, and I thought I would make a uh post about it on X or on Blue Sky. That seems fine to everyone. But what effect does it have? So our paper uh studies that now there is some general repugnance to experiment. There's a sociologist now at Penn, used to be at Yale, who studies networks and who has some papers that say consider an experiment in a in an operating room trying to get surgeons to wash their hands. And treatment A is you put a sign-up in the operating room saying wash your hands and treatment B is you you have something on everyone's identity badge that says don't forget wash your hands. So neither treatment is obviously better than the other, but apparently people find it repugnant. Some proportion of the people he he surveys find it repugnant to do the experiment because maybe one of those is better than the other and you're randomly assigning surgeries to these two conditions, and you don't know which one is better. They wouldn't find it repugnant for you just to put up the sign or just to put the reminder on the badges, but they don't like the randomization. So we economists who do experiments have to think about that too. Not everyone likes experiment s. I suppose one of the other areas of controversy though was the issue of consent, right? So that the people essentially participating or being studied in your paper didn't realize what what they were signing up to. They didn't realize they were part of this beforehand. So we describe the procedures in detail in the paper. And in many experiments, people who are participants in the experiments don't understand the whole experimental design. When you're allocating mosquito nets in different villages, you might try different procedures for allocating them. And the people in one village don't know that they're being allocated differently in another village. And you might decide that one of those procedures is inferior, and when you roll out to the whole country, you all you'll just use one of them. And so some people got the worst procedure. Similarly, in clinical trials, when you're trying to test COVID vaccines, you enroll 40,000 people in a COVID hotspot and you give twenty thousand of them the vaccine you're testing and twenty thousand of them you give a placebo. And they know the design, but they don't know whether they're getting the vaccine or the placebo. So experiments do not always tell every experimental participant what is happening to them. And that's often important for the for what's going to be learned from the experiment. Yeah. Just to push back a bit on this zero sum idea, right? Because of course it is possible that the labor market for economists is positive sum, right? And just more information is clarifying and you help get better matches, and that's great. But it's also possible that there are some zero sum elements in this market, right? And so, I mean, in a sense, you're making quite a strong claim that there are no zero sum elements, and of course, it's it's kind of only illuminating um these candidates, right? And I suppose the the critique was, well, there is this risk that there there could be some zero sum job allocations. You know, it could be that, you know, supposing you showed two amazing candidates on on social media and, even then it's not like someone else is going to create a new opening or what have you. So that I feel was some of people's concern. So I I think we were very careful not to make the claim you just made that that there's no distributional effect. If Harvard is only hiring one economist this year, then they'll only hire one economist. And if they have better information, they might hire a different economist than they would if they had worse information. It's not clear that that's a bad thing for the market. But if you were the insider who could have gotten the Harvard job if other people weren't so visible on the market, then it's bad for you. Absolutely. Okay. Well, I suppose my last question is just about, you know, when it comes to ethics in economic research, do you think there are examples of economists being too squeamish, or do you think they've basically got it right? I think there are examples of economists being too squeam ish and not squeamish enough. I think we have to understand when markets get social support and when they don't, because markets by and large are very good human inventions, but maybe not in every circumstance. But just as markets need social support, bans on markets also need social support. And when we ban a market, we're sometimes inadvertently designing the black market that will replace it in the face of the ban. And I think we need to understand that better too, because when we have some market that we're squeamish about, it's not always a good idea to relegate its governance to criminals . I'll thank you so much for joining me. Thank you . That is all for this week. You have been listening to the economics show with Samea Keynes. If you enjoyed the show, then I would be eternally grateful if you would rate and review us wherever you listen. This episode was produced by Misha Frankel Duval with original music and sound design from Breen Turner. Special thanks to Stephanie Stacey. Our executive producer is Manuela Saragosa. I'm Samea Keynes. Thanks for listening.