What love and marriage taught us about market design : Planet Money They say true love is hard to find. Whoever says that isn't an economist. | Subscribe to our weekly newsletter here.

The Marriage Pact

  • Download
  • <iframe src="https://www.npr.org/player/embed/972943944/1198960806" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript




There's this thing that happens every year on the Stanford campus.


Every year around finals week, undergrads all wait for this one email. They're all going to get it at the same time. It usually comes out late in the evening before midnight. And if you live in the dorm rooms, you all kind of wait for this email to come out together.

ALDEN O'RAFFERTY: All of the doors are open. Everyone's doors open. Everyone's standing in their doorway, just refresh, refresh, refresh.

MAYYASI: This is Alden O'Rafferty.

GONZALEZ: And the email that Alden and everyone else on campus are waiting for will have a name on it, the name of their perfect match, the name of the person on campus that they should one day marry. First, everyone gets this person's initials - an hour later, the full name.

O'RAFFERTY: And it's very much, like, shout down the hall. Someone goes like, oh, I got a football player. I run down to that side of the hall. And I'm like, oh, my gosh, let me look. Let me look. And we're Googling them. But then I'm still refreshing. Like, oh, I got mine. And then other people run over to me. Someone's like, oh, I follow her on Instagram. I'm like, oh, my gosh, show me, show me. You know? Like, I need to see.

GONZALEZ: At 11:25 p.m. on November 20, 2019, Alden gets her match - Kyra Dorado Teigen. And this matching thing that's happening, it is not like a Stanford thing in particular. It happens at a bunch of colleges and universities in the U.S. It's called the Marriage Pact. It's not a dating app. There aren't any pictures or a profile.

MAYYASI: You don't get a bunch of matches to scroll through and stress over and ignore. You just get one match, one name.

GONZALEZ: Everyone calls it a backup plan. Like, if we're both still single when we're 30, we'll marry each other kind of thing - real Hollywood stuff.

O'RAFFERTY: I took it pretty seriously. I was like, if I'm going to find my soulmate, it better be an accurate soulmate, you know?

MAYYASI: An accurate soulmate - I love the idea of an accurate soulmate.

O'RAFFERTY: Yeah, an accurate soulmate. Like, I don't want to fill out the survey wrong and then get the wrong soulmate. That would be such a bummer.

MAYYASI: To get her match, Alden had to answer 50 questions chosen by these two econ undergrads on campus.

GONZALEZ: And these questions are not your typical Tinder, Bumble, dating app stuff, like your height or your interests. These questions are supposed to remove the incentive to lie. There are questions like, how many kids do you want?

O'RAFFERTY: I was like, oh, we're really playing the long game here.


O'RAFFERTY: It's like, I'm 18.

MAYYASI: Other questions, they get a little deeper.

O'RAFFERTY: Is it important to me that I make more money than my peers?

MAYYASI: Is it important to you that you make more money than your peers?

O'RAFFERTY: Absolutely [expletive] not, no.

MAYYASI: (Laughter).

O'RAFFERTY: Do you consider yourself smarter than your peers?

KYRA DORADO TEIGEN: The one that's like, do you think you're smarter than, like, the average person at Stanford?

MAYYASI: This is Alden's match, Kyra.

DORADO TEIGEN: I was like, I've never really thought about that. Probably not.

GONZALEZ: All right, the way the Marriage Pact works is you get matched up based on how similarly you answer questions and how compatible your answers are, meaning how certain questions kind of speak to each other. Like, if I say I like exercise and another person says they're a super healthy eater, then maybe that means we're a little compatible in that area.

MAYYASI: And when you get your match, the email says how strong the match is, your compatibility score. Kyra and Alden, their score was in the 99th percentile.

GONZALEZ: Kyra makes the first move on Instagram, of course.

DORADO TEIGEN: Oh, my gosh. I actually sent her a really cringey DM...

O'RAFFERTY: It's horrible.

DORADO TEIGEN: ...That we just don't need to look into what it said because it's so embarrassing.

O'RAFFERTY: No, we do. I think that would be helpful. I think it would help them tell the story.

GONZALEZ: We got them both on the phone together.

DORADO TEIGEN: It starts with the word heyo, which is like I just - oh, God. Anyway...

GONZALEZ: They decide to meet up for a picnic on campus. Alden planned it.

DORADO TEIGEN: There is, like, little thermoses, and there's Pop-Tarts. And there is little squeezy applesauces. Oh, my God. I love those squeezy applesauces.

O'RAFFERTY: You're telling it wrong.

DORADO TEIGEN: I'm not telling it wrong.

O'RAFFERTY: You're telling it wrong.

DORADO TEIGEN: OK, fine. You can tell it.


MAYYASI: Hello, and welcome to PLANET MONEY. I'm Alex Mayyasi.

GONZALEZ: And I'm Sarah Gonzalez. Economists are not known for being experts on love, but they are experts at markets. And marriage is a market. There just aren't prices. And when you can't put a price tag on something, like love, economists have found ways to match people.

MAYYASI: They've been thinking about how to match people better and more efficiently since the 1960s.

GONZALEZ: Today on the show, an island of three weddings, econ class 136, and what happens when economists try to get you a date.


GONZALEZ: Three years before Alden and Kyra went on their picnic with little squeezy applesauces, this undergrad at Stanford, Liam McGregor, was enrolling in this class ECON 136: Market Design.

LIAM MCGREGOR: Yeah, my mom was looking at the course registry at Stanford and said, Liam, you'd like this class. I was like, I guess.

GONZALEZ: Your mom looked for classes for you?

MCGREGOR: She's a hero. My mom's a hero. I love my mom.


MAYYASI: This is Liam, econ and computer science major, surfer, ton of energy.

GONZALEZ: Liam, do you, like, sway back and forth when you talk all the time?

MCGREGOR: I hadn't noticed it, but it is entirely possible.


GONZALEZ: You're just, like, bouncing around over there.

MCGREGOR: I think it's just the energy I have to be here today.

MAYYASI: Liam co-created the Marriage Pact with a classmate, Sophia Sterling-Angus. And it all came out of this class, market design.

GONZALEZ: In this class, Liam was learning about all these types of markets - markets with prices, markets with complicated pricing situations, markets without prices.

MAYYASI: And, of course, in a market where there's a price tag on something, market designers, economists, they don't really need to get involved. We can just leave people alone to buy and sell, and the market will figure out the best price.

GONZALEZ: But some markets are a little more complicated, like auctions. So economists, market designers, they step in. They can set the rules for an auction, determine how many people should be in the auction, what the starting price on a painting or an oil well should be, et cetera.

MAYYASI: But what do you do in markets where there are no prices at all, like the market for human organs? You can't charge for human organs. It's illegal in almost every country.

MCGREGOR: So we've all just sort of decided that we don't want to do prices for some markets. And so economics here said, OK, well, if we can't use prices, how do we allocate things? And it turns out that matching is how you do it, and romance and dating is one of the most salient examples.

MAYYASI: Matching - when you don't have a market with prices, you can have a matching market.

GONZALEZ: Matching people to organs - that's a famous one; matching doctors to hospitals - you know, medical residency programs; matching refugees to cities - all of these are based on matching algorithms thought up by economists. But before they came up with these fancy, important matching situations, economists started a lot smaller. They started with marriage.

MCGREGOR: There's a sort of contrived example that's sort of very foundational to the course of market design and matching, and it's called the stable marriage problem.

GONZALEZ: The stable marriage problem?

MCGREGOR: Yeah, the stable marriage problem.

MAYYASI: In 1962, economist David Gale and Lloyd Shapley published the first paper on matching theory, and it sets on an island.

MCGREGOR: Imagine an island where you have three men and three women. And by the way, it's the '60s, so we assume they're all heterosexual. And we also assume that they all want to get married to each other, which is a very strong assumption. But we'll just go with it for now.

GONZALEZ: On this metaphorical island, there are going to be three marriages. Everyone is going to get matched with someone, and everyone will probably have some strong preferences over who on the island they want to marry. Like, I might want to marry Woman No. 2.

MAYYASI: Oh, I also prefer Woman No. 2, Sarah.

GONZALEZ: Liam says this is the core of the problem.

MCGREGOR: If, say, two men prefer the same woman or two women prefer the same man, how do you make the matches? How do you find who people marry in this case?

GONZALEZ: Yeah. How do you create a situation where everyone is happy with their match, meaning if there are three couples, there are no two people on the island who can or want to run off with someone else and make their own match?

MAYYASI: So to get us to this happy outcome, these stable marriages, these two economists, Gale and Shapley, say everyone should rank their preferences.

GONZALEZ: So my top choice is Woman No. 2, right? But I am not Woman No. 2's top choice. She wants Alex. But my second choice, Woman No. 3, she does have me at the top of her list. So we get matched. And it's kind of a bummer 'cause, you know, she's my second choice, but I actually do know that there's no one else on the island who I prefer who also prefers me back. This is the best I can do, so I'm probably not going to leave my marriage.

MCGREGOR: And so a stable match is one where people have gotten the best they can get. And if they tried to leave and leave behind the person that we gave them, they couldn't do better because everybody they liked better has someone that they, in turn, prefer.

GONZALEZ: Brutal - but OK, this island of three marriages from the '60s gave us what is called the Gale-Shapley algorithm, also known as the propose-and-reject algorithm, where you propose your preferences, you rank them, you reject the people or things you don't want, and eventually you arrive at a stable match.

MAYYASI: And the professor teaching Liam about all of this - economist Paul Milgrom.



MILGROM: (Laughter) Hello.

GONZALEZ: Paul Milgrom just won the Nobel Prize in economics for designing auction markets. But today, we're talking to him about why economists got into the marriage business.

MAYYASI: Is it fair to say the civil marriage problem is not - economists are not just entirely focused on trying to figure out a good way to get people married, but actually, the application (inaudible) much wider?

MILGROM: Yeah, it's much, much wider. This is - you know, it's almost a parable. And when Gale and Shapley themselves did the stable marriage problem, they didn't just do the stable marriage problem. They did that because it's the simplest case. It's one-to-one matching. They also, in the same paper, studied what they called the college admissions problem.

MAYYASI: The college admission problem is about matching students to colleges. Like, what if instead of students stressing about how many colleges to apply to and colleges accepting more students than they have spots for, you just have students in colleges rank each other? And the Gale-Shapley algorithm will match every student to a college, like those couples on the island.

GONZALEZ: Some cities use a version of this for matching students to high schools even. So yeah, the Gale-Shapley college and marriage algorithm has gone on to inform all kinds of things - matching doctors, donors, et cetera. But there have always been some flaws with it.

MILGROM: So - you know, we've been aware for a long time that there was something wrong with just assuming people had preferences.

MAYYASI: Right. The algorithm only works if you know exactly who or what you prefer.

GONZALEZ: But what happens when the island of six people gets bigger - when the island is Manhattan with a million people on it? How could you possibly rank every single person in Manhattan? And when you do have to rank things, you also start to strategize a little. Like, maybe you really want to get into the best high school in Manhattan, but you don't think you'll get in, so you just leave it off your list completely. You don't actually list your real preferences. You're strategizing.

MAYYASI: And sure, in the dating or marriage market, the strategizing is a little more amateur. Like, maybe you're trying to seem cooler than you actually are. Your profile is full of photos of you hiking when you're on the couch most of the time. But Liam says this is still strategizing.

MCGREGOR: For a lot of people, age is something that you want to strategize, where you're like, OK, well, actually, maybe I'm going to say something slightly different than what my age actually is. Or height is a salient example for a number of guys, where it's like, OK, do I report my real height, or do I say - if I'm 5'11", do I say I'm 6 foot?

GONZALEZ: Do you? If you're 5'11", do you say you're 6 feet, guys?

MCGREGOR: (Laughter).

MAYYASI: I think it's safe to assume there's always at least a half inch or an inch of inflation.

GONZALEZ: Lying, strategizing - this makes the market inefficient.

MCGREGOR: The best strategy for everybody involved is to tell the truth about their preferences.

GONZALEZ: This is the kind of stuff Liam is learning about in ECON 136. And then the class gets an assignment to write an essay about market design. But Liam - Liam is like, I want to do something else. I want to design a new market on campus for marriages, technically for backup-plan marriages. And that's after the break.


MAYYASI: So Liam is ready to start on the Marriage Pact. He wants to match every person on campus with their one optimal Marriage Pact partner and do it in the most efficient way possible.

GONZALEZ: And Paul Milgrom, his professor, he's into the idea.

MILGROM: It was exciting, and I loved it.

GONZALEZ: You see, Paul Milgrom, famous economist, is kind of a romantic. Just listen to this story about how he met his wife.

MILGROM: (Laughter) Well, my wife - I met my wife, believe it or not, at the 1996 Nobel Prize dinner.

MAYYASI: There was dinner and a dance, and that's where he saw her.

MILGROM: So I met Eva, and I was extremely attracted to her. And I went home and thought, OK, what am I going to do? So being a game theorist, I decided, OK, I needed to do something that was head-turning. So I wrote to her and said, you know, I really liked you, and I'd really like to see you again. And given that you live in Stockholm and I live in California, I'll send you a plane ticket and meet you anywhere in the world.


MILGROM: Yeah (laughter).

GONZALEZ: That's quite the move, Mr. Milgrom.


MILGROM: It worked anyway. So...

GONZALEZ: Nice. Yeah, I would think it would work (laughter).


MAYYASI: Total game. He approves the project.

And now, the first thing you want to do when you're designing a new market is you want to make it strategy-proof. If you're going to ask survey questions, you want to ask questions where there's an incentive to be truthful - no lying.

GONZALEZ: So Liam, star student, he's coming up with his list. He's looking for good questions online. He's asking his mom what good questions would be. And his mom - true brains of the operation - says stick to questions about values.

MCGREGOR: So like, would you keep a gun in the house or is it OK for your child to be gay?

GONZALEZ: Read me some more.

MCGREGOR: OK - I would consider my friends quiet.

GONZALEZ: That's a good one. So they'd be like, oh, God, he has a bunch of boring friends? No, thank you.

MCGREGOR: (Laughter).

MAYYASI: Sarah, I don't think quiet friends are boring friends. I disagree very strongly on this point.

GONZALEZ: I mean, maybe we're just not an ideal match, Alex. I don't know what to tell you.

OK. Now, in the Gale-Shapley algorithm, you would, of course, rank people. That's part of the matching process. But Liam doesn't do that. He says, I'm going to throw this whole ranking thing out the window. We know it's flawed, so no one is going to rank anyone in the Marriage Pact. You just answer these questions, and I will tell you who you're most compatible with on campus.

MAYYASI: And sure, some dating services like OkCupid, they also ask you questions to try to determine compatibility, so this wasn't groundbreaking stuff. But his professor says it was really different from what economic theory on matching had looked at.

MILGROM: You know, these are not economic ideas. These are - you know, these are things that more normally would have come up in psychology or sociology, and they just melded them together. It was wonderful.

GONZALEZ: But, you know, you can have the most wonderful survey in the world, but if no one takes it, you don't really have a market, right? So Liam makes a flyer for the Marriage Pact, and he starts texting it to his friends on campus, telling them to text it to their friends on campus.

MCGREGOR: OK, here it is. (Reading) Listen, finding a life partner is probably not a priority right now. You hope things will manifest naturally - fine. But like many busy individuals, years from now, you may realize that most viable boos are already hitched. At that point, it's less about finding the one and more about finding the last one left. The Stanford Marriage Pact is here to cover the essentials.

GONZALEZ: And everyone starts sharing it on campus. RAs are sharing it. They're like, eh, who cares if you already have a boyfriend? Just sign up for the Marriage Pact anyway because the more people sign up, the bigger the market, the more likely everyone will end up with a compatible match. By the second day, thousands of students signed up.

MCGREGOR: I think it was about around 2,000 people on the second day that I called my mom and told her I had a school project that was going interestingly (laughter).

MAYYASI: And she had suggested the course.

MCGREGOR: Yeah, yeah.

GONZALEZ: Shoutout to Liam's mom, Lori (ph).

MAYYASI: By the end of the week, 4,111 Stanford undergrads had signed up for the Marriage Pact - 58% of the student body. So Liam decides to do the Marriage Pact the next year, too, and the next. That's when Alden and Kyra sign up.

GONZALEZ: So - OK, wait. So are you guys - did you guys date? Was there, like, the spark? Like, she touched my knee. And like...

DORADO TEIGEN: So - OK, we definitely did things that were very, like - like, she took me to her house, and she's like, it's a surprise. And she, like, leads me inside. And I, like, have my eyes closed. And then she puts kittens all over me because she fosters kittens. And so suddenly...

O'RAFFERTY: OK, just for context, I did not know that any of that had date vibes. Did it?

GONZALEZ: Alden, that is, like, super date vibes.

MAYYASI: (Laughter).

O'RAFFERTY: I thought we were just hanging out.

DORADO TEIGEN: Well, I thought we were - I didn't know what was happening.

GONZALEZ: Kyra and Alden say they did fall for each other, just in a platonic way. Kyra currently has a boyfriend. But this is kind of how marriage pacts work, right? Like, you do your thing. You date. You bounce around. You live your life. And then maybe when you're older and you're ready, maybe you come back to the person that you made a marriage pact with.

MAYYASI: Kyra and Alden, they say they do plan on fulfilling the marriage pact one day.

DORADO TEIGEN: We've talked about our wedding. We're getting married outside. There's going to be lights in the trees. We're not going to be wearing shoes. We're going to have flowers, you know, on our heads.

O'RAFFERTY: I think we're going to be friend-married for sure, you know, and we're going to have a life together...


O'RAFFERTY: ...And mutual friends and game nights and sleepovers. And our kids will be friends. And, like, it's a real marriage, you know? It's just not a romantic one.

MAYYASI: As for Liam, he graduated from Stanford last year, and he and a small team have since brought the Marriage Pact to dozens of other colleges, universities. It feels a little like Facebook in the early days, spreading from campus to campus. Like, in a few years, who knows? The Marriage Pact might just be a standard part of college life.

GONZALEZ: And by the way, the algorithm has evolved. Like, the computer has learned how to match people better. But it does have its shortcomings still. Like, sometimes, siblings get matched up with each other on campus. A pair of twins just got matched together this year at Princeton. And sometimes, people get really low compatibility scores, like 4%, which is a terrible rate. As for Kyra and Alden, they actually took the Marriage Pact a second time and got matched up with each other again.

MAYYASI: The second time, their compatibility score was 100%.


GONZALEZ: If you participated in a college Marriage Pact and are now coupled up, please, please, please tag us in your photos on Twitter and Instagram. We are @PlanetMoney. You can also see pictures of Kyra and Alden there.


GONZALEZ: Today's show was produced by Maria Paz Gutierrez and James Sneed with engineering help from Gilly Moon and Michael Cullen.

MAYYASI: We also want to give a shout-out to Sophia Sterling-Angus who co-created the Marriage Pact. She stopped working on it when she graduated but was a big part of its success.

GONZALEZ: Bryant Urstadt is our show's editor. Alex Goldmark is our supervising producer. And Alex Mayyasi, my co-host today, he joined us from Atlas Obscura. He's a food editor there, and he brought us this story. I'm Sarah Gonzalez.

MAYYASI: And I'm Alex Mayyasi. This is NPR. Thanks for listening.


Copyright © 2021 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.