Richard Thaler | Nudge

0:06Welcome to the Talks at Google podcast, where great minds meet. I'm Emma, bringing you this episode with behavioral economist Richard Thaler. Talks at Google brings the world's most influential thinkers, creators, makers, and doers all to one place. You can watch every episode at youtube.com slash talks at Google. We make a countless number of decisions every day, but unfortunately, we often choose unwisely. Richard Thaler has dedicated his life's work to understanding why that is. In 2017, Thaler received the Nobel Memorial Prize in Economic Sciences for his contributions to the field of behavioral economics.

0:49His book, Nudge, co-authored by Cass R. Sunstein, shows that it's not possible for choices to be presented to us in a neutral way. The book demonstrates how to best nudge us in the right directions without compromising our freedom of choice. Richard Thaler is a professor of behavioral science and economics at the University of Chicago Booth School of Business. He's a member of the National Academy of Science and the American Academy of Arts and Sciences. He's been published in many prominent journals, and he's also the author of Misbehaving, The Making of Behavioral Economics. Originally published in December 2021, here is Richard Thaler, Nudge.

1:39So I think you're on now. So please, let's hear your current views on behavior economics and the challenges it faces. Wow. You know, I think we've reached a stage, kind of a stage I was hoping to reach, where behavioral economics is part of the repertoire. So it used to be if you were a behavioral economist, you were a professional troublemaker. And it was a small group of troublemakers. Now, lots of economists consider a behavioral approach just part of their toolkit. And I think the biggest challenges are to apply principles to fields where it hasn't become commonplace. I think one area where it started, but there's a lot of work to do, is macro.

2:46So if you think about, say, the questions that the Fed is dealing with now about inflationary expectations, that is essentially a behavioral problem. It's what do people think and why do they think it? And the Fed is involved in not only taking actions about buying or not buying bonds, but persuading people about what's going to happen in the future. And that's an interesting behavioral problem. ED HARRISON There is a recent macro paper that's getting a lot of attention about whether inflationary expectations are worthwhile in the sense that they really help predict actual inflation or not. There's a technical issue there, of course, whether it's a good predictor or not, but it seems to be raising quite a bit of interest in that particular area.

3:49ED HARRISON Well, I think one thing the profession isn't very clear about is whose expectations are we talking about? So you can look at the yield on inflation index bonds and back out an expectation of inflation of traders of those securities. But if you walk down Michigan Avenue in Chicago or Main Street anywhere and ask people what their expectations are about inflation over the next year, they are not likely to have a very strong opinion. They're more likely to have an opinion about whether the Bears are going to win their next football game than they are about what inflation will be in 2022.

4:43So whose expectations is a big question to me. And how do they form those expectations? And this is very much a behavioral topic. Because one of the huge drivers is gasoline prices. Why? Because once a week, you see the price of gasoline when you fill up your car. And that's going to be a big driver, things that are frequently purchased of that sort. So here's an interesting thing. When we're all driving EVs, how will inflation expectations change? Because when you plug in, you don't see $80 or whatever. Well, it wouldn't be anyway. But you don't see a price. And you're right, that gasoline prices are probably the most salient price to most people.

5:33And then there are some grocery store items like milk and bread. But everything else, people don't really know or see. They've got an inflation index called the Sticky Price Index. And what that does is it focuses on goods that rarely change prices. So we're seeing there's something anomalous, to use the term here. There's something anomalous going on in terms of prices that usually are pretty stable are now going up. What does that mean? Yeah. And you know, one thing I've been thinking about recently, so we've talked about sticky wages in economics going back to Keynes. So in the Great Depression, there was a question, how can it be that a third of the workforce is unemployed?

6:22Why don't wages fall to clear the market? And the sort of tautological answer was, well, wages are sticky, meaning they don't fall enough to clear the market. It's not an explanation. but I think there's a question now whether wages are sticky upwards because we hear about labor shortages in all kinds of domains every restaurant you go into has a help on its side and there are all these supply chain problems some of which are because they can't get enough truck drivers or something. And my question is, are wage setters too reluctant to raise wages for reasons that I don't know, but I suspect that there is some of that.

7:22Yeah, yeah. We're seeing it. And of course, what happens is when you raise the wages to track new hires, you pretty much have to raise the wages for the existing employees as well. If you don't, you run into a lot of problems, again, of a behavioral nature. So that you get - Although you could have signing bonuses and retention bonuses. So this is what we call a framing, right? So if you give every new employee two grand and every existing employee two grand to stick around through next year, maybe you don't have to be as clear about what permanent wages are going to be. Right, and they do that in the trucking industry, for example, that you're getting a wage, but then once you complete a certain number of trips, then you get a bonus on top of that wage.

8:23So in the book, I read the book. Maybe we should mention what book this is, Al. Excellent. Why don't you tell us about the latest book? Well, the latest book is called Nudge, the Final Edition. And of course, Nudge is a book that Cass Sunstein and I wrote back in 2008. And you might wonder why we would rewrite a book that was continuing to sell lots of copies. And part of it is that the paperback contract had expired and no one had noticed and they got around to telling us in April of 2020. And if you remember, there wasn't that much to do in April 2020. So we started tinkering and there was a mention early on in the book about the spiffy iPod.

9:29Remember those? Mm-hmm. So when we were writing this book, we had both just bought our first smartphones. And it just struck us that the world has changed so much in the relatively short time since we had written that book. And I was bored. So we ended up, the new version is about two-thirds new. We call it the final edition as a commitment strategy to make sure that we never make this mistake again. It was a lot of work. What I tell aspiring authors is that when you think you're 90 percent done, you're 50 percent done. And all the dull stuff comes at the end of fiddling around with the order of things and with the index and with all that other stuff.

10:25So it's fun to write the first chapter of the book, but then by the time you get to all this, the minutia is not that much fun. You know, one of the things we wanted to do is introduce a concept we were talking about but had not written about, which is what we call sludge. So since the beginning, I've been signing copies of Nudge, Nudge for Good, which is meant as a plea. I'm sure you have a copy somewhere that says, Hal, nudge for good. And people don't always. So what is sludge? Sludge is the same tools, but the idea of nudge, the basic idea is if you want people to do something, make it easy.

11:13And lots of successful firms like Google have succeeded in large part because they make things easy. You can Google it, right? And, you know, it's true of Apple and Amazon and Netflix are all companies that make things easy. And but lots of things are done that make things hard. One thing that drives me crazy is lots of services make it very easy to join and very difficult to unjoin. You can subscribe to many publications with one click and your credit card. But to unsubscribe, you have to call and go through some torture. and I don't approve of that. I've tried to convince some major publications to stop doing it, but it's obviously profitable.

12:19And so I think one of our goals was to get rid of sludge. There's a lot of inadvertent sludge, which is not, so the unsubscribed trap is intentional because it's profitable to get a few more months of your payments. But a lot of sludge is just inept programming. And I think the world would be a better place if we had sludge exterminators. And I encourage all of you Googlers who are watching this live to, if you find sludge in the process, get rid of it. Well, another example of making it simple and easy to do is, of course, Google Maps. Remember what it was like planning a trip 30 years ago. It was painful.

13:29I totally agree. And I think the smartphone is the single biggest technological advance. I mean, and we have so many, more so than the personal computer. And in part because the computer we carry around in our pocket is more powerful than the first home computer we had. And it has a map. Yes. And I have a terrible sense of direction. So I love Google Maps. And, you know, you mentioned that GPS is our ideal for a nudge. And it's because you get to pick your destination. Suppose I want to come and visit you. I plug in your address. And that was my choice. Nobody's telling me where to go. It will suggest a route.

14:36but if I see something interesting, I'm allowed to depart. And all of life would be great if there were a Google Maps equivalent. And of course, search is like that. If I want to find out who has the record for the most touchdown passes in the season, I can find that out instantly. But there are lots of things that figuring out, take our tax system. The U.S. tax system is sludge ridden. In Sweden, one files your tax return by text message. you get a text saying according to our calculations how we owe you 10 000 kroners if you agree press one and the money appears in your account and if you think of 89 percent of american taxpayers would be eligible for such a service because they take the standard deduction.

15:59And they passed a law saying it's illegal for the IRS to send you a pre-populated tax return. That is sort of encapsules everything that is wrong with Washington and with government. Well, not everything. I say not everything, but it's a good chunk. It's just, I'm not saying it's the only thing wrong. It's just an example of some special interests, and if you guess who they are, you'll be right, have got that law passed, and it makes all the rest of us worse off, and there's a lot of that going on. By the way, you know who we owe the geopositioning systems to? The government. Ronald Reagan. Why was that?

17:03The story goes that the military, of course, was a moving force to build the GPS system, satellite system. And Reagan got a demo. He said, this is wonderful. Can ordinary people use this tool? And they said, oh, no, no, it's all encrypted. And only the military can use it. He says, we don't want that. And so then they changed policy and made it available to everyone. It was a little extra noise in it because they figured that consumers didn't need as quite an accurate position as the military needed. But that was the story. Right. And then they got rid of that. You know, that's actually a good example of another theme we talk about in the book that we call smart disclosure.

17:52There's lots of government rules about disclosures. And if you think of the technology used for almost all disclosures, it's first century, right? I mean, you might as well be doing it with scrolls on parchment. So putting the ingredients on a cereal package on the side of the package, come on. Yeah. And that's true of almost all disclosures. Richard, this site uses cookies. Well, and, you know, I mean, the regulations that they passed in Europe saying you have to warn about cookies. And then so, you know, the first time you got one of those, you say, hmm, if you don't want cookies, press here.

19:04And good luck. Good luck getting out of wherever you go if you start choosing, right? So that was a well-intended bit of regulation that nobody took any steps to think about, okay, well, what else? and so yeah there are two of the big themes of the new version of Nudge smart disclosure and sludge and they try a new form of disclosure and just create more sludge and God knows we can do better So one of the impressive things about the original Nudge book was it inspired several governments to put together nudge groups who would try to improve regulation and practices in ways that you advocated. Can you tell us a little bit about that?

20:11Sure. So the first country to do that was the UK. And the story of how that happened was when the, it was David Cameron, Conservative Party Prime Minister, and there was a young guy working for Cameron named Rohan Silva, who read an early version of Nudge and bought up, I think, the entire UK stock, which was about 10 copies. and piled them up on a prominent desk at the Conservative Party headquarters. And David Cameron picked one of them up. He was nudged and read it. And parties there have manifestos, kind of like platforms here. But unlike our platforms, they take them somewhat seriously. And it was in the Tory manifesto that if they're elected, they will do behavioral science informing policy.

21:23And a week after the election, I get this phone call. OK, we're doing this thing. Can you come over and help? and a guy called David Halpern is the one who created that first unit and is still there. And the White House followed suit after Barack Obama got elected and that unit was created by Maya Shankar, who is now a Googler. And somebody has been counting these government units and got up to 400. Really? So that's governments and NGOs. And, of course, there are countless equivalents in companies like Google that are – so, you know, the first few, I had my fingers in the pie, but they're out on their own now.

22:28And, you know, what are they trying to do? They're not trying to, let's go back to the GPS. They're not trying to tell people what to do. They're trying to help people get where they want to go. So if you think of what's almost certainly the most successful application of behavioral economics, It's been in the redesign of defined contribution retirement plans, what we call 401k plans in the US. And there have been three big innovations that have made those a lot better. The first was changing the default to join, what's now called automatic enrollment. Lots of people just didn't get around to joining, even in companies that were matching contributions.

23:27So they're just throwing money away. And if you just sign up people and tell them they have to fill out a form if they don't want to join, enrollments go to 90 percent, almost everywhere. UK has done this at the national level and less than 10 % opt out. The second is what we called save more tomorrow, which is now called the generic is automatic escalation. So raise people's contributions gradually because they initially signed people up too low and they should be saving 10%, 12%, not three. And then the third was to create good default investment strategies, like these target date funds. And that has worked.

24:19And why? It's easy. You don't, I mean, the idea that everybody needs to become their own financial planner and figure out how much they have to save and then form the best possible portfolio is a bit preposterous. There's a lot of interesting work to be done to help out guys our age of the second part. Okay, you've retired. How are you going to take the money that you've saved up and what are you going to do with it? That turns out to be a more difficult mathematical problem than the saving up, because you have to keep, you have a pretty good idea of when you may retire, but you don't know when you're going to die.

25:15You may know you're sick and you can spend a lot, but no one knows they're going to live to 100. So solving that problem is hard, and that's one of the things that behavioral economists can and should be working with. Well, this is why they have annuities, right? Annuities are one solution to that problem. But here's what's interesting. people who are in old defined benefit pension plans, like your former colleagues at the University of California, love them. And those come with annuities. But people in defined contribution plans almost never choose to buy them. So that's an anomaly in and of itself.

26:08that I got involved, there was a labor dispute in Chicago in the Chicago Symphony Orchestra. There was a strike. And the big issue was the management wanted to switch to a defined contribution plan. And the orchestra members didn't want it. And a friend of mine and I looked at it at the request of somebody and it looked to us like what they were being offered was better than the thing they had. But there was a guy in the union whose view was to find benefit, good, to find contribution, bad. And they could have lost one of the great symphonies in the world on this problem. And there were some nudges.

27:04And fortunately, they're still playing. Great. Fantastic. I think one of the troubles is the pricing is not often as attractive as one would like. And here it's a case where it looked like it was better to go to defined contribution. Well, the orchestra was offering a guaranteed minimum return on their investments, which is not an offer you often get. Right. Yeah. And of course, then the question is, well, were they really being liquid enough to pay up or are they outsourcing this to an insurance company or just how it's structured? There's a lot of devil in the details problem. That's true. So let's return to the book.

27:52I noticed as I read the book, this time, there's a lot of lists in the book. And this is a good thing, I think, because these are sort of the items that say, well, you have to look at problems with anchoring, availability, representativeness over contests, loss aversion, status quo bias framing. And it's like you could go down the checklist. Was this policy? Is this showing overconfidence? Is this showing status quo bias, et cetera, et cetera? So it seems to me it's more operational than the first nudge book was. More how to nudge. Yeah. You know, in the first book, we introduced the term choice architecture.

28:37Uh-huh. And that's a term I made up. Now, it's one thing to coin a term. It's another thing to know how to do it, right? So, you know, I coined a term, but I wasn't an architect. And thanks to all these nudge units and lots of academics and economics and management and psychology, We've learned a lot about how to do choice architecture better. And like we were talking about before, lots of technology firms have become very good at choice architecture. If you think about Amazon, they have every book in the world for sale. Now, imagine going into a bookstore that had every book, like the Library of Congress or something.

29:35It would be a nightmare. You know, I want to find a copy of Hal Varian's microbook, and it might take me months to find out where it is on the shelves. It won't take long at a good website. So we've gotten a lot better at choice architecture, and it doesn't take us long to find a movie we want to watch or a route we want to go to. And so much of the reason we felt like it was worthwhile for us to rewrite this book and for people to read a new version is we've learned so much more about how to actually do stuff. Yeah. So sometime when you're bouncing around YouTube, look up the AT &T commercials from 30 years ago.

30:32They ran a series of commercials about someday you'll be able to find any book in the world. Someday you'll be able to find a movie that's perfectly attuned to your taste, et cetera, et cetera. And the amazing thing is how good they were. They were really quite remarkably accurate that they were able to make this forecast 30 years ago. Yeah.

30:57Well, it's usually not the case. Forecasts of technology are usually hopelessly bad. Well, I think the idea of having, let's say, a portable computer or a mobile phone, something like that, that was in Dick Tracy, right? In the 30s, you had the watch that you spoke into. Right. And now it's a wearing one. Yeah. Yeah, it's amazing. So some of these things have done very well. And when I saw my first mobile phone, what impressed me the most was the fact that it had a camera with it. And think of the implications of having a camera that you're carrying with you all the time. And it's had a huge impact on all sorts of things of having that little device.

31:46Everybody knew what cameras were. Everybody used cameras, but only on special occasions. And with the mobile phones, you had it all the time. It made a huge, huge difference in all sorts of things. Crime, punishment, et cetera. Right. Yeah. So one of the things we're very interested in at Google is how we can be helpful. In fact, that's kind of our theme these days is being helpful in terms of the products we design, like using the search, of course, but other products of other sorts, the maps. The maps, I was going to say earlier when we talked about maps, what's interesting is you went through all this pain and suffering to figure out your route on the map.

32:25And then you had the final insult. You had to fold the thing back up and it never quite worked. You know, there's a whole generation of people listening to this that have no idea what you're talking about. These are two boomers who actually had to deal with folding up maps, you know. So let me ask you one topic I wish you had mentioned. You have this astute observation. Confident speakers are more likely to be correct. That's the belief that people have, maybe unwarranted. Yeah, I don't think we say they are more likely. We say that people believe. Believe there, yeah. Believe. I didn't mean to say it was your opinion.

33:16It was an opinion of one of the common errors that people make. But how do you speak and convey true uncertainty? Because maybe the best thing to say is, well, it depends. Or maybe, maybe not. Or the probability is about either direction. Well, then it makes you look very wimpy compared to the person who says, by God, inflation is going to go through the roof next quarter. Whereas really, it's a very question that, in fact, in reality, there's a lot of uncertainty around it. Yeah. So, I mean, think about the abuse that Donald Rumsfeld got for this line about the known knowns and the unknown knowns.

34:01Now, he had his warts, but that was a very sensible thing to say. Yeah. And, you know, we economists talk about the difference between risk and uncertainty. And we're gradually learning that with COVID that there's a lot of unknown unknowns. Because every time we think we've rounded the corner, there's a new variant or a new problem. And, you know, there's a paper I like that looks at the forecasts of Fortune 500 CFOs about the returns on the stock market over the next year. And they ask them for confidence limits, 80 % confidence limit. So they say, give us a high forecast that will only be exceeded 10 % of the time and a low forecast that the market will only do worse than that 10 % of the time.

35:06Now, of course, the actual number should lie between those two 80 % of the time. And these are highly professional people. the actual came in within their confidence limit one third of the time. And the reason for that is, of course, their confidence limits were way too narrow. Now, it turns out forecasting the stock market is nearly impossible. So we shouldn't be criticizing them for not being able to do that. No one knows how to do that. Well, I mean, this is one of the... But the problem was that they were giving forecasts as if they could. And the reason I wanted to tell that story is I asked the authors, well, what would be a forecast, a rational range, if you knew you don't know?

36:12And it was something like up 30%, down 10%. Now, suppose the Google CFO, Ruth, says, okay, she's at some meeting and she says, my forecast for the market next year is it'll be somewhere between up 30 and down 10. She doesn't look like she knows anything. And, you know, suppose we go to the chief economist and we ask him and he gives that answer. They're going to say we finally need a second chief economist. So it's very hard to give forecasts that are as wide as they should be because it makes you look dumb. and when it comes to things like COVID and forecasting the stock market, we are dumb because we don't know what's going to happen next.

37:19All right, well, on that note, I'm going to switch to some audience questions that they've asked and there's a series of this, five or six of them, I think. So here's the first one. Will you accept cryptocurrency as value for something you're offering, like black hot tickets or corner office booth? What do you think of cryptocurrency? Well, I think that I'm just too old to get it. Yeah. And I think most economists share that view. And it's clear we've been wrong so far. it's clear that cryptocurrencies are not good as currencies because they're just way too volatile. Right. So now, you know, I sometimes play golf with my colleague, Gene Fama, who is the leading exponent of the efficient market hypothesis.