Skip to main content
Loading Explore guides…
Behavioral Economics & Paternalism · Acedra
Home
Explore
Economics
Behavioral Economics & Paternalism
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
131 segments
Copy
Hide video
0:00
[JINGLE PLAYING]
0:07
MARK: So I'll get started with some-- what I hope is enlightening and relatively amusing preamble, as to this one. So here we go. So I'll introduce Mario. And then I'm going to tell you why we're starting the semester with Mario. Right? So there is a particular story behind this, which I hope you find quite enjoyable. It's a bit confessional as well. So we'll see where this goes. So Mario Rizzo-- lovely to have him here for the first talk of the semester for the Rhodes Institute.
0:33
He's the co-director at the Classical Liberal Institute and an associate professor of economics at NYU. He's also the director of the program on the foundations of the market economy in the Department of Economics and is chairman of the Colloquium on Market Institutions and Economic Processes. He's the author of Austrian Economics Reexamined-- the Economics of Time and Ignorance-- which is a great title, by the way-- and the book under discussion today, Escaping Paternalism--
0:59
Rationality, Behavioral Economics, and Public Policy, which came out with Cambridge just this very year, which focuses on the new, or soft paternalism, behavioral economics and the economic theory of rationality. So why start here? Ah, great, Andre, you're here. Fabulous. Why start here? Because a f-- about a year ago-- I really hoped that Steve Sloman from cog sci was here, because I exploded in his house. And it was one-- you know when you have that visceral outpouring of something
1:29
and you don't know where it came from? So we're talking about nudges. And I didn't know that I actually had this problem, right? But the problem was-- I was around at his house. We were just hanging out. And I just exploded. And he said, what's the problem? What's the beef? I said, to tell you, this is how I think about nudges. So imagine a conversation between me and somebody who is absolutely not me-- a normal person, right? So hi, I'm an Ivy League professor.
1:54
We live very different lifestyles. We'll probably never cross the same social circles. My kids will definitely never go to the same school as yours. I get a little bit worried about your behavior. Most of you are obese. Some of you are drug-addicted. And eventually, you're going to be a tax liability on me. And frankly, I don't want that to happen. So I'm going to basically not treat you as my fellow citizen, whom I should reason with. I'm going to treat you as a social problem to be managed.
2:19
But I'm not even going to be honest about that. I'm going to use these behavioral tricks to basically get you to do what I want without actually being honest about it in the first place. And Steve basically said, that's ridiculous, that's not what we do, whatever. And I thought, well, maybe it is; and maybe it's not. Then Doug from cog sci contacted me, and said, hey, we hear that you think this stuff sucks. We want you to do a debate with one of our professors.
2:42
Now, rule number one of debating professors is don't do it outside your field unless you're really, really sure about what you're doing. But I did want to do it because I wanted to figure out what I really thought about this stuff. So I thought, I need to read something. So you go online. And I found Mario's book. And I was like, yes. And it's brilliant. It's a really fantastic book. The amount of work that he has done on this is truly incredible. It's encyclopedic. It's encompassing.
3:09
And it's very even-handed, and very fair, and judicious in what it's saying. So that's the setup for the talk. A couple of things that I'll just say in closing. If I take my little rant about sort of people who are not like this, and why don't they all just show up at Whole Foods, and put this into more formal logic, there's three things that I think are deeply problematic with this research program. And I'm willing to be persuaded that they're not. Let's put that on the table.
3:38
But here's what they are just for the sake of it, because I think this will be a nice launch for where Mario goes with this. One of the things that he points out in the book which I think is quite brilliant is that behavioral economics, to the extent that it is a field, is exactly a collection of error terms. So you'll have sort of the classical model of rationality-- consistent preferences, whatever. And then you go, oh, look, people don't really do this.
3:59
Well, first of all, my grandmother knew that. But we've now got 175 different reasons, called biases, as to why they're not like that. But then the weird thing about the program is now that I've collected these biases, I can use that to make you more like the model. That's very weird when you think about it, right? The second thing that falls from that is the problem of true preferences. Now, one of the reasons I could never be a Marxist was because a Marxist, every time that people don't vote
4:26
the way they want, says, well, people don't know their true preferences. Really, all workers want socialism, but they just don't know it, you know. And I thought there's a bit of a Marxist tinge to this whole thing in the sense that there are proper preferences that people should have. They just don't know what they are. And if we nudge them, they'll find them. All right? And that's deeply unsettling in exactly the same way that a false consciousness claim is, because I
4:49
don't think you can do that. And also, I think it's ethically dubious to think that you could do this. One of the key things that this body of work gets into and likes to claim credit for is defaulting people into pensions on the grounds that people under-save and it's a good thing that they'll have pensions. Well, a couple of problems with this right away is if you look at ESRA accounts in the United States, only 22% of them have any money in them. The vast majority of ESRA assets are held
5:15
in a tiny minority of accounts. So you can give all the people who work for UPS, or whatever company you like, ESRA accounts. But the problem is low wages. It's not their behavior and saving. So this kind of puts the cart before the horse in a way that excuses the fundamental problem. And finally, to return to my little rant at the beginning, there's the problem with democracy. We live in a highly polarized age. And the politicization of expertise has, I think,
5:42
quite rightly become an object of [INAUDIBLE].. So you better be sure of yourself before you go out there and tell everybody that it's robust. Because we used to think that banks were the best judge of risk, that all Iraqis really wanted to be liberal democrats. And maybe it turns out that this is just as fragile as well. It's very elitist. It divides the world into those who know how to nudge and those who only deserve to be nudged. And I think that's a very, very good way
6:10
of making populism even worse. And with that, I would like to introduce Mario Rizzo. [APPLAUSE]
6:26
MARIO RIZZO: All right. Thank you. Thank you, Mark. And thank you to the Watson Institute for having me here. It's been many decades since I've been to Providence. And I was welcomed by your excellent weather.
6:43
OK. I feel that Mark said the whole thing [INAUDIBLE].. So any questions? No. Let me-- I will try to speak for only about 20-25 minutes. Let me start out by saying that I want to be clear that I'm not getting into any ethical issues about nudging or behavioral paternalism more generally. I'm leaving that to other people. There have been a number of books on that. What I'm trying to do is stick closely to what are generally recognized as economic issues. And in particular, in this talk, I
7:29
wanted to emphasize what I call the knowledge problem. A little background-- I became interested in this area shortly before Thaler and Sunstein came out with the book Nudge partly because of the policy implications, but also partly because I thought that economics pri-- had become a bit boring, at least microeconomics. And the behavioral paternalists were drawing on behavioral economics, which I thought was a very interesting and important development within economics.
8:20
So Glenn and I wrote a couple of articles that were published around 2008 or so criticizing behavioral paternalism, or what we called then new paternalism. But I felt that we were doing what other people were doing-- sort of picking at individual points. And the more I thought about it, the more I thought it was necessary to really deal with this in a more comprehensive way. And so the subtitle of the book-- Rationality, Behavioral Economics, and Public Policy--
8:52
really gives you a sense of the three pillars of analysis. Behavioral economics say that we are not rational. We're-- they used to say we were irrational. But now they're saying we're boundedly rational. OK. So what's rationality? And why should we be concerned about it? And then behavioral economics itself-- well, all of these biases-- there are 175 of them listed by Wikipedia. What's the evidence behind them? Is it good evidence, or is it really flimsy evidence?
9:24
And then, of course, the implications for public policy. So I'm going to be talking mostly about public policy. Though if there are questions during the question period, I'll get into some of the other issues as well. OK. So this is the cover art from Getty Images. And there you have it-- self explanatory. All right. So what's paternalism?
9:52
We define paternalism in this way-- deliberate act or policy to cause the behavior in an individual to be different than otherwise for the purposes of making that individual better off. Now, that leaves a few things undefined, namely the type of policy, the extent to which the policy can cause an individual to behave differently, and how you determine "better off." In the-- this literature, "better off" is supposedly determined by the individuals themselves, at least at some deep level.
10:36
So let me get to this little diagram here. All right. So paternalism is not a new thing. You can read about it in Plato's Republic and many other places. So there are initially a couple of types of paternalism. There's informal or legal. Informal is your friend is about to commit suicide. You try to stop them or her. All right? There, you-- it's sort of ad hoc. You have intimate knowledge-- presumably, if it's your friend-- about the person. And you want them to pause and reconsider.
11:21
I'm not dealing with that. What I'm dealing with is what I would call legal paternalism-- getting the state somehow involved in this activity. And in that category, there are two types-- old and new. Old is what you will read, say, in Plato. It's a kind of moral paternalism, right? And to the extent that-- it's not a moral paternalism-- that is to say, you'll be a morally better person if you behave in this way. If it's not that, then it's based on some notion of objective welfare--
12:00
that there are certain kinds of things that well-being objectively requires. And so there, again, it's not linked to the desires of human beings, either their true moral desires or their true welfare desires. And the policy prescriptions are usually hard, like prohibition or maybe just certain-- less hard might be sin taxes. Although sin taxes are also used by behavioral paternalists as well. They're recommended by them. All right. So then there's new paternalism.
12:40
It's based on behavioral economics. And so it has-- there's a whole formal apparatus of research and findings upon which it was based. It's not just your grandmother's insight into human beings. Although to have research that tells you that people procrastinate, your grandmother knows that. And you know that too. I'm kind of a big fan, by the way, of procrastination in this respect, especially when you're in a university and people bombard you with administrative things.
13:20
Sometimes, if you ignore them, they go a way. So procrastination could be actually a good thing. But that's another issue. But behavioral paternalism, new paternalism is based on the fundamental idea of true preferences-- not objective welfare, not objective moral criteria, but the true preferences of the individual. And policies are from soft to hard, hard to soft. Nudge is the soft end, but there are many behavioral paternalists who have said that nudging is not
14:04
sufficient, it's not-- it doesn't work, and harder policies, like taxes, sin taxes, or other things, are necessary. OK. So the central problem is the determination of true preferences. We have to understand that true preferences are not identical to choices. Initially, there's some confusion, because in economics revealed preference and choices are really the same thing. But here, there's this concept that we have preferences that are not adequately expressed in our actions.
14:41
So what are true preferences? Well, Thaler and Sunstein early on described or defined true preferences as the preferences you would have if you had complete information, unlimited cognitive abilities, and no lack of willpower-- choices that people would make under these ideal conditions. So this is a fairly expansive notion in one respect. It's restrictive in another respect. It's restrictive in the respect that I'm not sure if you took this definition literally-- which I'm not going to-- there
15:13
has ever been a expression of true preferences in the history of the world because don't-- complete information, unlimited cognitive abilities, no lack of willpower are pretty strong. Harsanyi has a better definition and one that I'm sympathetic to actually-- the preferences an individual would have if he were better informed about the relevant facts. Now, of course, there's the question of, what are the relevant facts, and how best to convey knowledge about the relevant facts?
15:47
We'll talk about that. All right. So operationally true preferences-- so the idea-- there's a postulation. And it's a postulation that there are fully-- that we have a fully rational inner agent somewhere inside who is continually thrown off course by preference biases and cognitive biases. One author described it as Chicago Man. And Chicago Man is purely hyper-rational. But Chicago Man is sort of inside us. He's struggling to get out.
16:29
And he's constantly thrown off by these psychological shocks. So the psychological shocks are kind of inessential. What's essential is Chicago Man down deep. So preference biases are things like present bias; inconsistency-- we will talk more about that-- influence of hot states when you're deciding; status quo bias; cognitive biases, or overestimating low risk, underestimating high risk; optimism bias, which is kind of magical thinking-- I'll talk more about that--
17:05
and the half-empty, half-full fallacy, which is really framing. Wikipedia has now 175 biases. But operationally, what rationality is in this framework is really a residual-- it's what's left when there are no biases. OK? And what's left when there are no biases is the traditional rational economic man, which is strange. And we'll see that. For behavioral economists to spend a lot of time saying how absurd it is to believe that people behave as the rational economic man, and then
17:53
to say that true preferences are what we want to nudge people to, and true preferences are the inner Chicago Man is really kind of odd. People-- it's absurd to think that people behave like Chicago Man, but people should behave like Chicago Man is the theme. All right. So true preferences have to be distinguished between-- with-- from simple desires, pro attitudes, or stated preferences, right? We all, for example, would like to save more.
18:30
At the same, we'd like to spend more, that's consume more. It's not simply your desire that's relevant, right? It's a preference. That is a willingness to sacrifice and make the trade-offs. So simply to survey people, and say, do you want to save more for your retirement, and they say yes, but don't do it isn't really to discover their true preferences. And stated preferences in general have a-- are a big problem. There's research indicating, in areas
19:04
where you're able to test to see if people are-- actually have certain preferences, stated preferences ex ante are usually very misleading. Secondly, knowing the relevant facts, as Harsanyi would say, often requires knowledge of true preferences. So there's a kind of circularity there. People don't only have necessarily one simple goal. And for example, you want to go from point A to point B. Well-- and you're not taking the shortest route. What's wrong?
19:41
Well, maybe you don't know what the shortest route is. OK. Or maybe you want to also get from point A to point B, but also you want to do it in the scenic way, right? So simply providing a person with the information and then seeing that they don't take the shortest route is not to say that there is something irrational or even boundedly rational about their behavior because you don't really know what their preferences are. So knowledge and preferences really go together.
20:13
[? Secondly, ?] we have to distinguish-- and there's-- this is extremely important. And as I think about it more and more, it becomes even more important-- the distinction between knowledge of facts and salience of facts, vividness and intention. And let me skip for a moment to Orwell because I want to keep going. But there's an interesting quotation in the book about the details of a heat operation-- how do you describe it? And-- open-heart surgery.
20:49
There's a-- you can describe it in a sort of very clinical way that maybe doctors would describe it if a patient asked, or you can describe it in extremely vivid detail, about blood spurting out and a crack being made in the chest. The quotation gives you all the gory details. And depending upon how you describe it, people's reaction may be very different to open-heart surgery. But the problem is there really is no unique way of describing things. And the frame, or the description,
21:34
and the people's reaction to the description is not something which you can say is rational or irrational. Simply to look at things one way or to look at things another way is not a matter of rationality. To look at things in one way-- given that you look at things in that particular way, rationality is something which follows the frame and not precedes it. OK. What I'm going to do is essentially what I'm calling an internal criticism of the policy. So let me just move on. All right.
22:10
So overweight issue-- first of all, from an economic perspective, there's no uniquely optimal weight. All right? The costs and benefits of achieving a particular weight-- even from a medical perspective, the only evidence there is of overweight being medically harmful is when you get to the category of obesity. Before the category of obesity, a simple overweight-- the evidence that this is bad for you is actually quite weak. However, it doesn't matter from my major point.
22:54
The behavioral claim is that people do not adequately take account of the true cost, the future cost when making present eating or exercise decisions. They over-weigh-- pun intended-- the present benefits. So how so? This is the idea. The idea is-- suppose you're in the doctor's office. And you are-- your doctor is telling you about the virtues of dieting, and why you should diet [INAUDIBLE]. So now you go home. And that evening at home, you're sitting down at your table.
23:26
And having just been to the doctor, you decide on a dieting plan. So you set it all up. OK? And you say, OK, I'm going to start this diet next week. Monday morning, I'm starting the diet. So presumably, the behavioral economist says, at that moment, when you decided that you were going to diet, you carefully balanced the costs and benefits, and you discerned that this was optimal. All right? And you're going to go on to this diet. Monday comes.
24:01
It's time now not to have the croissant for breakfast. All right? And you yield-- you eat it anyway. And immediately, the diet is broken. So this is viewed as-- well, lack of willpower, or more technically an instance of present bias. Why? Because now that the decision is right now, you over-weigh the value of eating the croissant. When all of this was in the future, that future croissant wasn't as delicious as the actual croissant before you. So you were willing to give it up. All right?
24:49
So this is the inconsistency of your behavior. The long-term perspective is paternalistically privileged over the immediate perspective. A lot of things could be said about that.
25:05
The first thing that's obvious to say about it is that the cost of giving up the croissant or other similar things, from the perspective of just having come from the doctor's office, is not obvious. Right? Because this is this croissant in the future, the [INAUDIBLE] croissant. Well, I like [? croissants, but anyway. ?] Right? But then when you're really faced with the decision, you know what the cost is. So why is it s-- why are we accepting the view that somehow
25:43
the estimation of the cost when you've just come home from the doctor's office is somehow more valid than the estimation of the cost at the moment of decision. I don't know. But let's forget about that for a moment, or at least go on beyond that. There have been some policies that [? had been ?] championed by paternalists. All right. In 2015-- or actually, in 2010. We're quite early. New York, under Bloomberg, instituted this calorie labeling requirement on menus for chain restaurants.
26:23
And so five years later, there was a study-- well, did it do any good? Are people consuming less? So they did research with receipts and other things of that sort from various fast food restaurants. And they found that there was no change in the consumption of calories. Now, why is that? Well, there could be a number of reasons for it. But one reason is that sometimes people mis-estimate the calories in food. They don't always think that there are fewer calories.
27:03
Sometimes, they think there were more calories. So they would abstain, those who think there were more calories, right? When they find out there were fewer calories, they can indulge. Right? And so all sorts of perversities crept into the whole thing. Also, it was the case that in some areas, it was noted that sometimes people were trying to get the most caloric bang for the buck. So you've got $5 to spend. Can I get more c-- why? Because calories mean energy, at least immediately, right?
27:41
So if that's what you want-- you want an energy boost-- well, they're telling you how to get the energy boost. In any event, there was no effect. Fat tax-- tax on fat, sugar, salt. Let's talk about a tax on sugar. Philadelphia instituted one recently. Interestingly enough, it did so in a way that was directed toward raising money for the city. They didn't push the paternalistic aspect for whatever reason. However, in 2017, there was a study by the New Zealand Institute for Economic Research--
28:27
actually, a friend of mine used to work for the New Zealand Institute for Economic Research. In any event, what they did was they studied 47 cases of taxes on sugar content of soft drinks throughout the world. And not one of them did they find any affect on health outcomes or obesity. Why might that be? One of the reasons that we talk about in the book, among a lot of other reasons, is that non-caloric sweeteners do not satisfy your hunger as well as sugar drinks.
29:15
So the tendency then is to actually eat more when you're drinking the non-caloric drink than the caloric drink.
29:29
But in any event, either due to that or many of the other reasons that we talk about, there was no effect, not in one case. So there you have it. But theoretically, the issues are that, first of all, there's no objectively correct degree of discounting future harms of junk food, right? Oh, it's in the future. Well, how deeply should you discount that? There's no objectively correct manner. Some people will value their future health more. And other people will value it less.
30:07
Economics is not supposed to be able to say anything about that. So what are the behavioral economists talking about? The longer the future period, the more uncertain the health risk because of health care advances. In 1960, before statins, before cholesterol-reducing drugs, before many other things, the health of obese people was markedly poorer than obese people today. So what are the technological advances in health that are going to occur in the future that
30:43
will most likely reduce the cost of eating fat, or sugar, and all of that? We don't know. And because we don't know, there's no objectively correct decision that can be made today. And finally, a tax on these substances would need to account for self-regulation in order to avoid excessive reduction in demand. [INAUDIBLE] people are already engaged in self-regulation. And if they already engage in self-regulation and you put a tax assuming that they're not
31:17
engaged in self-regulation, you're overdoing it. This is-- was taken into consideration theoretically by [INAUDIBLE] in 2006. And basically, what they said was assume that half the population is self-controlling and the other half has a small present bias. The optimal tax is 5.15%. [INAUDIBLE] precise. Assume half the population has self-control and the other half has a somewhat larger present bias. The optimal tax is now 63.71%. Just a small difference in assumptions-- and we
31:57
have really no way of gathering this kind of data. But a small change in the assumptions makes for a big change. So either you're way overdoing it or you're way under-doing it-- you don't really know. This is a variance in estimated discount rates-- what should people discount in terms of the future harms? What rate should they use? Well, discount rates are all over the place. Savings-- people say that they don't save enough. Is this a desire or an actual preference?
32:33
Are they willing to forgo present consumption in order to save enough? Well, how much should they save? Well, we've seen that the estimated discount rates are all over the place. So what is the correct level of saving? The long-- if you assume that there is an unambiguous long-term rate of discount-- which there isn't. But if you assume that, you have to assume that that longer perspective, long-term perspective is necessarily better. But really, if you're going to take this behavioral thing
33:15
seriously and talk about people's estimates and knowledge being contaminated, then it's not only contaminated in the short-term. But when people think about the farther future, they're less certain about the utility of expenditure at that time than when they're thinking about the nearer future. What are you going to be spending things on in 30 years? So the shorter perspective may actually be more informed because you actually know what it is that you-- know what you want to do.
33:44
And you know the utility of what you want to do. Farther future is quite uncertain. I believe Keynes had something to say about that. All right. Finally, I want to mention about cases where information is not enough. All right. Talk about cigarettes. Well, first, there was a information campaign-- cigarette smoking may be hazardous to your health, was the-- the surgeon general has determined that cigarette smoking may be hazardous to your health. That was the original warning.
34:19
Then it came stronger and stronger and stronger as time went on. So we get to the point where the warnings are as strong as they're going to get. And so a behavioral economist would say, well, it's still insufficient. Why? Because there's something called optimism bias. It's kind of a magical thinking. You say, well, it's true that cigarette smoking causes cancer, but it doesn't apply to me because I've been unusually lucky in my life-- bad things don't happen to me. All right?
34:51
This is a species of magical thinking called optimism bias. So they say, what's the solution to that? Well, we need to have risk narratives, or scary stories or scary pictures, to make the risks more salient. So you probably have all seen these commercials where they get somebody-- I saw one which didn't last very long because I guess it was considered over the top. It was this 27-year-old woman who says that somehow smoking caused her to get diabetes. And then she had to have her feet amputated
35:29
because of the diabetes. And she-- and this was all due to smoking. That somehow disappeared. I don't know whether there is any truth to that. But the point is that these scary stories are by their very nature biased, right? They take the extreme case and present it as if it's the modal case. All right. But it makes the risks more salient. However, we know that cigarette smokers overestimate the risks of smoking. And there have been a number of surveys of that,
36:02
both in the United States and abroad. They overestimate the probability of getting ill, various diseases due to smoking. Well, if that's the case, then the optimal policy has to balance the subjective risk-lowering effects of optimism-- the magical thinking, right, that kind of lowers it for you-- with the subjective risk-raising effects of misinformation-- with scary stories, right? You're already at the case where people have overestimated
36:43
the risks of smoking, but they lower them by optimism bias. But that has to be counteracted by risk-raising effects of misinformation or biased information. The upshot is unclear. Who knows-- we don't really have data on all of this. So we may be overkilling or not-- no pun intended there. We don't know what the behavioral optimum is. The fact of the matter is-- and this is important-- for behavioral economists, the optimum amount of all these excesses
37:23
is not zero because they recognize, and explicitly so, that all of these bad habits have their counterpart in benefits to the individual. So the optimum is not zero. So you can't just say, well, the optimum will be when we get it down to zero because that's not what the behavioral economists are all about. All right. Finally-- and then I'll end with this slide.
37:49
Information can be too much in the behavioral mindset. Example-- snus. This is a Swedish smokeless tobacco product. Almost all scientific authorities say that it is safer than smoking. Actually, I went to the Snuff Museum. There's a Snuff museum in Stockholm. It's actually kind of boring. There's nothing to see there really-- a few labels. Anyway, so the company, Swedish Match AB, wanted the label in the US to read that their product carries substantially lower risk to health than cigarettes.
38:28
This is true. The FDA rejected the request even though they agreed with the dominant scientific opinion. Why? Well, the New York Times explains it this way-- the label may, quote, "tempt people to use tobacco products they may not have tried otherwise, or some smokers may take up the nicotine habit and progress on to cigarette smoking." So now the implication is you can't-- not only do you need to have bias or exaggerated commercials about the dangers
39:01
of cigarette smoking, or eating sugar, or et cetera, et cetera, but also it may be inadvisable to tell people the truth about something in many cases where you believe that this may tempt them to use tobacco, and get addicted to nicotine, and then move on to cigarettes when obviously they shouldn't. So this opens the door to even more intrusive policies, to suppression of true information, et cetera, et cetera. So there is more or less-- I'm running out of time.
39:45
So that's more or less what I have to say. If you want to talk about the evidence about biases or-- I can talk about that. If you want to talk about what economic rationality is, I could talk about that. So let's go to questions. Thank you. [APPLAUSE] AUDIENCE: I'll let you field your own questions [INAUDIBLE].. MARIO RIZZO: Yeah, yeah. Yeah sure. Hi. AUDIENCE: Yeah. Could you say a little bit more about these biases? And what are some of these 100-- MARIO RIZZO: 75. AUDIENCE: --175 biases?
40:18
MARIO RIZZO: Sure.
40:27
All right.
40:32
There are 175. Sometimes it's hard to tell the difference between them-- the subtle distinctions. But we'll have-- some are cognitive. So they are biases like people overestimate low-probability events. And they underestimate high-probability events. So for example-- and this gets tied up with, say, something called availability bias. So let me give you an example. Suppose there are a number of crashes in airplanes in recent days. Nevertheless, the probability of a fatal airplane crash
41:24
is very, very low. These are sort commercial airlines, right? But people will get all excited or upset about that because they were given a picture or have in their minds a very easily accessible picture of airplane crashes due to the recent news reports. OK? And so what they-- in the old days, what people did is you'd go to a machine at the airport. And you buy for some fantastic price, given the probabilities, airplane-- airline insurance, if you die on the airplane.
42:05
I remember doing it myself when I was in college. [INAUDIBLE] So you put in some money. You get a-- and it expires after your trip. Right? And that would be considered irrational from the point of view of a-- all right? Of this. Let me give another example. And then I'll-- I want to say something more in general. The other example is-- and this is a study which Gary Becker and Yona Rubenstein did a number of years ago. It has to do with terrorist bombings at cafes in Israel.
42:44
So what they studied was, immediately after, was there an affect on people going to the cafes? And what they found was a difference depending upon how important it was for people to go to cafes. Now, you say, well, how do you measure how important it is to go to cafes? What they said was-- and this makes sense-- single people go to cafes to meet people. This is really important. Married people-- well, you can go to the cafe or not. They did a similar thing with buses, right?
43:33
People who had long-term-- who bought ticket by the day or people who buy monthly tickets-- monthly tickets, you have to go to work. It's very important. Daily tickets, shopping-- maybe not so important. What they found was, in both of these cases, there was some effect for people who did not value the activity very highly. So they stayed away because it really wasn't very important for them to go to the cafes or to the buses. But for people who really found it--
44:11
for whom it was important, there was no effect. Even in the cases where there was an effect, it only lasted about a week. OK? So they found some bias-- overestimation-- but not where the activity was considered important. All right? So that tells you something about whether-- if there's a real cost attached to the bias, , people are less likely to indulge it. More generally, I want to say that biases are very context-dependent. And so you have to be careful if a bias exists,
44:54
as Cagan says in this quotation.
45:00
It's-- these psychological traits are very context-specific. So you-- they're not general tendencies. And that's the problem with the literature. The problem with the literature is what we have come up with is these biases as general tendencies in human behavior. But in fact, Cagan is saying psychology and psychologists are good at discovering particular contexts in which there is a certain behavior and they're generally not widely generalizable. All right. So I don't know if I should say more.
45:38
But let me say this-- that these are the four major problems. Right? A lot of these biases are discovered in lab experiments. But generalizing the quantitative results, whether they're significant quantitatively or not to the real world, is unreliable. So that's a problem. Most research doesn't account for incentives in learning, right? Incentives would be either positive or negative from indulging the bias. In many cases, when you provide incentives, the biases disappear.
46:18
Small-group de-biasing-- there's a very interesting new literature that shows that when people make decisions in small groups or when they discuss their decisions with other people, biases tend to disappear. And most research does not adequately assess people's own efforts at overcoming their biases. So-- all right. Big answer-- sorry. AUDIENCE: Thanks. I-- MARK: You got to turn it on. AUDIENCE: Turn it on? Oh OK. Am I on? Yeah? Thanks. I really enjoyed this.
46:51
And I guess I want to ask three sequential questions that-- I'm going to start by being devil's advocate, defender of behavioral economics, and then maybe pivot to be even more critical than you. But my devil's advocate says, is it the case that a non-nudge world is impossible? We already live in a nudge world, whether we acknowledge it or not. And the example I'm going to give here is your croissant and diet example. Wouldn't a defender of behavioral economics say, all
47:19
we're doing with these nudges is offsetting the biases introduced by the Department of Agriculture in Congress in subsidizing junk food and subsidizing industrial agriculture, and this is just a way of offsetting them? And that leads to my second point, which is I think there's an odd that I'd never thought of before between two economic fads-- one is behavioral economics, but the other is the counter-factual view of causality because you can't actually implement anything
47:41
based upon behavioralism if you don't understand what the counter-factual is. And I think your snus example gives that point really clearly. [? It's ?] the alternative that you're smoking cigarettes that you're not taking on tobacco at all. And that leads to my third point, which is I find it kind of interesting that you bracket the normative and ethical questions that [INAUDIBLE] sat with. Because if I'm reading you correctly, what you're really saying is if we could just
48:00
get the math right, there would be nothing wrong with this. And I actually think what's wrong with this isn't just that we don't have the math down. I think it's that there are deep moral and ethical problems here that you're kind of bracketing. And I think they're actually the best argument against behavioral economics and its deployment in the realm of public policy. MARIO RIZZO: OK. As to the last question, the only reason that we didn't do that is because, first of all,
48:22
the book was fat enough, but more importantly because we felt that our expertise was in these economic and more technical questions, and wanted to leave the rest to somebody else who's more-- has a better comparative advantage. Kind of-- we were thinking sort of the Mises Hayek analysis of economic calculation into socialism. The issue was not whether socialism was good or bad, but really whether it could efficiently calculate the people--
49:06
agents-- we could sufficiently calculate under that system. So stick to the economics questions. About offsetting-- well-- OK. First, that presupposes that the behavior is offset by the nudge or behavioral policy so that--
49:31
what's the goal? Is the goal to have people consume less sugar, or is the goal to have people go to better health status? So the Department of Agriculture subsidizes sugar. The price of sugar is lower than it otherwise would be. So the idea is, OK, you raise the price of sugar. But the paternalists are saying you raise the price of sugar to increase health, not to offset the subsidy. In fact, if you raise the price of sugar in that way and you in effect offset the subsidy, I'm
50:10
not sure what the next stage is. Do the agricultural interests come back and ask for something else? So-- and that presupposes that it's effective at achieving its goal. And we were discussing why it's not-- unintended consequences, this, that, and the other. I forget the second one though. AUDIENCE: Counter-fact. AUDIENCE: Counter-factual. MARIO RIZZO: Yeah. Well, the counter-factuals are more unintended consequences. The current dispute about vaping, right? So you don't want kids vaping.
50:41
So you raise the vaping age, or the cigarette smoking age, now to 21. I don't know if it's happened already, but it's about to happen. Right? So yeah, it'll be just as effective as the drinking age as 21. Everybody-- all the students at NYU drink. Regardless of their age, right, they drink. And the problem with vaping was that it was the non-standard capsules that were contaminated, right? So now you're driving people-- and then with the prohibition of certain flavors,
51:24
you're driving more people into the black market with these contaminated-- where these contaminated capsules are being sold. So unintended consequences are a problem. And it's difficult [INAUDIBLE] people are clever. If you try to force them to do something they don't want to do, there are a lot of different ways around. So you've got to know what they're going to do in response. Now, it's true that-- your ultimate criticism here is that if all of these problems
51:56
were overcome, the book sort of doesn't have anything to say. But at least recognize these. I don't know. MARK: Gentleman in the front. AUDIENCE: Yeah. MARK: Green light. AUDIENCE: Yeah. Sorry. I just want to continue with some of the points that were raised. But I have two concrete questions. And they're related to the fat tax. So of course, you mentioned that it's a complex matter just instituting the tax. But [INAUDIBLE] or any such scheme is a hierarchy.
52:27
So you have a tax, but then you have incentives for long-term health benefits in terms of a health incentive, or a health subsidy, or whatever. So that's one question-- how-- where do you draw the line? Does the hierarchy just continue or-- I'd just like to hear your thoughts on that. And number two, how do you account for macroscopic systematic effects, like cultural reasons? Obesity in, say, state x in the United States-- not to get political-- may not be viewed the cert--
52:55
in the same way as it's viewed in, say, California, where culturally it's viewed as a desirable state. So how does behavioral economics or even microeconomics account for that in their model? MARIO RIZZO: All right. On the last question, there is a recognition of individual differences in this behavioral literature. It's troubling because many of the policies cannot be finely tuned so as to accommodate individual differences. By individual differences here, I'm
53:36
referring to different attitudes toward gaining weight and the benefits of dieting. Some people may-- for some people, their optimum weight, costs and benefits considered, is going to be very different from other people. And so that's-- and if you put a tax that's uniform on sugar-sweetened soda, you are not recognizing that there are these different valuations. Right? You're just doing it according to some average valuations. So some are going to have mis-incentives.
54:22
Now, the hierarchy issue is interesting because we do discuss that somewhere in the book, the maybe in the chapter on slippery slopes-- the idea that actually these-- what are considered the mildest policies are ineffective. And so either you have to say, OK, well, we're just going give up on this, or you go to increasingly more invasive policies, right? So-- or policies with wider scope. You say, well, OK, not only do we put a tax on sugar, and salt, and all the nasty things
55:01
that people could consume, but maybe we should also subsidize vegetables and other things as that, or maybe we should put a-- a funny one. Put a tax on computers. Why? Because people sit around with their computers and don't get up and exercise. Right? So I'd like to see how that works-- an extra tax on computers, and cell phones, and things of that sort of to get people to exercise more, or maybe we have an income tax subsidy for a gym membership or something like that.
55:38
We'd make that deductible in your income tax. Well, the thing spreads out. So Sunstein and Thaler start out by saying, oh, just a little nudge here. Well, when the little nudge doesn't work, then, if you're in this frame of mind, the natural thing is to spread out the things, right? And that's what we deal with in the slippery slope chapter. MARK: Gentleman to your left. Not to you, the [INAUDIBLE] and then over. AUDIENCE: So I was going to ask-- MARK: Green light [INAUDIBLE].
56:10
Just hold it until it's on. AUDIENCE: In your book, do you go over certain treatments that can do a better job in research studies of kind of capturing all these I guess externalities you talk about? MARIO RIZZO: Well, we do a pretty comprehensive examination. We do a pretty comprehensive examination of the research.
56:41
And I don't know of any research so far that has tried to take a look at several policies and the cumulative effect of all these policies. That's difficult, especially if the policies don't exist yet. So I don't really know. One of the problems, which I didn't mention, of the comprehensive approach is that the analysis that took-- takes place, generally speaking, looks at one bias at a time. Now, suppose the 175 biases listed in Wikipedia--
57:37
but let's just say that some of them are really the same thing, some of them don't exist in a particular context. So I'm going to be real generous and say in a particular context-- let's say eating, overeating, or overweight-- there are only 10 biases that are appropriate. Well, the problem is the biases move in different directions. Some things cause you to be more present-oriented. Some things cause you to be more future-oriented. Some things that cause you to exaggerate risk;
58:06
some things to underestimate risks. So they move in all different directions. Now, once you get into this frame of mind, where you have to deal with all these error terms at once, it's a very difficult thing to figure out. Right? Now, it's one thing to say-- if this were-- had to do with the survival of the human race on the Earth, you'd say, well, OK, keep-- we'll keep trying. But you might want to say, you know what, is it worth it to accomplish these goals in this way
58:39
when an alternative might be just more information and cultural adaptation, which in fact has taken place over time? People are much more aware of weight problems, of cigarette problems, and all that than they have ever been. So some of this thing is just sort of natural evolution. And it takes time. MARK: Before an-- AUDIENCE: [INAUDIBLE]. MARK: Am I on? Yes. Before an-- this one's dead, but whatever. Before anyone takes off, 21% off the book. Flyer right? So if you want it, I've got them here.
59:12
All right? Before we take I think what might be the last question, just one I want to point out from the book as one of my absolute favorites, right, is alcohol. So alcohol is bad, right? It's become really bad for you. It's the worst thing ever. And the weird thing is about you have to think about the distribution. And that's something that gets washed out in this, right-- the frequency and the distribution. So it turns out that 85% of hard liquor drinking is done by 10% of drinkers.
59:36
So what that means is your standard-- you should have two drinks a day maximum or whatever, right-- already applies to 90% of the distribution. They're there. They're fine. Now, the 10% who are booze-heads, right-- they're going to be completely price-insensitive to whatever you do to them because they're going to drink. So if you then start applying to this, what you're doing is disincentivizing the people who are already in a good space while having no affect on the people who have actually
1:00:02
got a drink problem. And the frequency and distribution of a problem is absolutely huge in this. So. MARIO RIZZO: And by the way, if you take seriously the evidence that says that moderate drinking is actually good for your heart, then what you're doing is you're causing these people to drink less, right? And then we've got to counteract the-- well, fewer heart-- more heart attacks here. Fewer cancer over there. And figure that out. But I don't think we should be doing that. AUDIENCE: Thanks.
1:00:33
I have one question regarding the policy implications that follow from your work in a very general sense. So how should a policy be designed in order to avoid being paternalistic? Did I get this correctly, that the policy should first be honest about what its-- what it wants to do, what it wants to achieve? MARIO RIZZO: OK. There is a literature existing which talks about boost versus nudge. And what "boost" is-- and I'm very sympathetic to "boos."
1:01:07
Boost is to concentrate on improving the cognitive skills of people, or simplifying certain presentations with respect to, say, investment, vehicles, and things of that sort. I have no problem with "unbiased" meaning undirected toward a particular end provision of information. Now, that can be done privately, or it can be by government. I'm not getting into that. But-- so boost would be not having in mind that there should be a certain outcome to the decisions
1:01:48
process, but simply that there should be a simpli-- or a decomplexifying of the decision process. Information should be provided in a way that people can understand and appreciate. Now, Sunstein and Thaler talk about that, but their perspective is that should be done with a few of nudging people into a particular decision. My idea is simply-- or the boost idea is simply to provide it. And then whatever happens, happens. So it's-- there are some subtleties here
1:02:27
because providing a simplification of information simplifies certain aspects, but also then maybe de-emphasizes other aspects. So hopefully there can be some competitive provision-- you can go through different websites and see presentations in a different form. But-- similarly with doctors and presentation of information to patients. How can doctors present information in a way that's unbiased, that doesn't lead the patient to do what the doctor thinks he or she should do?
1:03:06
But doctors a little bit-- it's a little bit tricky because sometimes people want the doctor to be the paternalist. They want paternalism from the doctor, which is fine with me. But it should be something that's sensitive to the individual in that case. And doctors are-- try to do that-- at least a lot of them do anyway. MARK: Anyone else? Tony. Hold on for the mic because, remember, we record this. AUDIENCE: I guess I'm listening to this. And I'm wondering whether there is any really clear line
1:03:41
between providing advice or information and a paternalistic moment? It seems to me a matter of degree, not of kind in some way. And I wondered whether you might comment on that.
1:03:59
MARIO RIZZO: All right. The first question is, is the state involved or not? To the extent that's feasible-- and I think it's feasible to a large extent-- I'm happy with the private provision of information-- say, employers providing information about retirement options in their 401(k) plans and things of that sort, which NYU does-- I'm sure Brown does as well. So that's one thing. So I'm less worried about the sometimes blurry line between provision of information,
1:04:40
boosting, and nudging when it's done by private individuals and not done by the state. But even there, I think that--
1:04:54
yeah, we can make that distinction. We convey all sorts of skills in teaching to people which are going to be used for decisions in an unknown future. When you're teaching mathematics, you don't know what your students are going to use mathematics for in their future, right? It's a tool. So similarly, if you teach people or instruct people about certain basic principles-- for example, it's well known that Americans are financially illiterate. There have been a lot of studies.
1:05:26
People don't understand compound interest. They don't know the effects-- they can't conceive of the effects of inflation over a long period of time. They don't know about diversification of-- all of that. So if people learned about these things, they will apply them in unknown contexts. So it's not a question of nudging them toward a particular end. It's a question of giving them the tools to make the decisions according to their own values. I think we would all agree that those three
1:06:00
things that I mentioned-- it's important to understand them all, right? So. But I do agree. I think you could probably come up with some cases in which there is a blurriness. I agree.
1:06:15
MARK: Final question.
1:06:20
AUDIENCE: Thanks. I have a question to do with advertising. And I guess thinking back to the example of alcoholism, one just hypothetical explanation of the distribution of alcoholism is that some people-- MARK: Probably just alcohol consumption, because there's no [INAUDIBLE] alcoholism. AUDIENCE: Oh, OK. Alcohol-- very high, unhealthy alcohol consumption. MARK: According to some. AUDIENCE: According to some. Or substitute in whatever good you want. MARK: Exactly. AUDIENCE: [LAUGHS]
1:06:55
One just plausible explanation is that some people are exposed to advertising or for some other reason are highly susceptible to advertising, right. And that manipulates biases in maybe a direction opposite to that which governments would want to manipulate them in. And so I might just say, sure, this is an imperfect world. There's lots of money spent on getting people to consume more alcohol. If they're consuming it for irrational or biased reason, then we might as well do our best
1:07:28
to counteract those using the state. MARIO RIZZO: OK. Let me say this-- advertising is a form of manipulation. I don't disagree with that. It's also a form of conveyance of information. But the critical thing about advertising done by private companies or private individuals is there's free entry. Let me explain what I mean. So people are all concerned about calories, at least in some areas of the country. Right? So what has happened? Without any government mandates, or incentives, or anything,
1:08:15
I-- we started to see a number of years ago, the cans of soda now-- there are small ones readily available in supermarkets which are about six ounces. Going back to the-- Back to the Future sort of-- the original Coca-Cola bottle was six-and-a-half ounces by the way. And one of the reasons why, as a kid, I preferred Pepsi to Coke because Pepsi was 12 ounces and the other one was six-and-a-half. And that wasn't quite enough. In any event-- so smaller stuff.
1:08:43
Advertising and placement in supermarkets of the smaller cans. Calorie-controlled portions of all sorts of goods-- cookies, salty snacks, et cetera, et cetera-- all calorie-controlled portions. So what you've seen is, in response to people's greater understanding of health consequences of food, and preferences about weight, et cetera, that advertising has in effect appealed to that as well. So in a way, what I'm saying is you don't really need the state to come in and counteract
1:09:20
because the market is counteracting in a certain sense itself or providing the alternative options that some proportion of the market desires. So I'm not sure it should be the role of the state to do that.
1:09:39
Really, the private companies that are trying to sell these smaller cans, et cetera, are much more expert in getting people to be aware of these options than I think the state would be. yeah. AUDIENCE: Could I follow up on that real quick? I can [? also talk about-- ?] MARK: 30 seconds. MARIO RIZZO: Yeah, OK. AUDIENCE: OK. OK. Yeah. I would just respond to that. That's a compelling anecdote in one case. But if you take cigarettes, for example, there are a couple problems.
1:10:08
One is that it might not be very easy for companies to break into certain markets just due to monopoly power. And then the other is that the-- if the product that you're providing and advertising for it are incredibly coercive or biasing, then just the fact that there's also an opportunity to provide products that don't appeal to people in that way doesn't seem to cause the initial problem. MARIO RIZZO: OK. Look, we've got Juul, right, competing with cigarettes.
1:10:47
Now the state's getting involved in this and probably going to screw up the whole thing. But here, you have an alternative being provided, a healthier alternative being provided, and advertised, and promoted. Nothing to do-- the state did not invent Juul. And similarly, with Snuff, the smokeless tobacco product, they're, again, much healthier; the FDA not wanting them to advertise that they're healthier. So I think that the market does respond to people's wishes
1:11:24
to cut down on products that they think are bad. The ideal-- look, you're a cigarette manufacturer, right? So you begin to buy an interest in electronic cigarette production or take over an electronic cigarette company. Why? You want to get people both ways-- those who want to continue smoking, make money from them; people who don't want to continue smoking, make money from them. Free entry into nudging. OK? That's the-- or advertising. Anyway, that's my view. MARK: Thank you.
1:12:01
Fabulous conversation, as ever. MARIO RIZZO: Thank you. Thank you very much. MARK: There we go. [APPLAUSE]
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat