John Leahy delivers an entertaining and insightful lecture celebrating his installment as the Allen Sinai Professor of Macroeconomics and Public Policy. October, 2016.
>> Good afternoon, everyone, on this beautiful rainy October afternoon. My name is Andrew Martin, and I'm Dean of the College of Literature, Science, and the Arts. I'm here to welcome everyone to the Inaugural Lecture of the Allen Sinai Professor of Macroeconomics and Public Policy. We are here today to pay tribute to individuals who have made transformational differences for the fields of macroeconomics and public policy and in its teaching. In addition to celebrating the accomplishment of Professor John Leahy, we are also paying tribute to Economics alumnus Dr. Allen Sinai, who along with his wife Lee, established this Professorship. It would be hard to overstate the importance of endowed professorships to any college or university, public or private. Endowed chairs have long been used by institutions of higher learning to honor faculty members that have shown considerable academic excellence in both scholarship and education. Such chairs are typically reserved for scholars of international stature and endowed chairs have proven to be effective methods for recruiting and retaining the best and brightest faculty to our university. An endowed professorship denotes reverence, both on the scholar holding the chair, and on the university. The very presence of faculty of such renown also goes a long way in bringing top graduate students to the university as the best and brightest quite simply attract the best and brightest. This is precisely the type of honor the Sinais have bestowed upon us with their gift. Through dedication to both economics and public policy, and to the University of Michigan, Dr. Sinai sought to create a chair in macroeconomics and public policy that would serve to ignite the interest of our students. Allen Sinai has a history of engagement with the University that stretches back decades. And he is generous in his volunteer of service. He is an Emeritus member of the LSA Dean's Advisory Council, and currently serves on the LSA Economics Leadership Council, and the Ford School Committee. The scheduling of this lecture aligns with the annual meetings of both groups, and you will see many of the members here with us today. Dr. Sinai participated in the first New York City-Michigan Economic Summit in 2003. And also in our second, Wolverines on Wall Street, that was held in 2009. Beyond his distinguished service to the University of Michigan, and in addition to his career as an economist, Allen Sinai has taught at a number of universities and colleges over his career. He has published over 40 articles in academic and public policy journals and books, and over the years, he has testified before the U.S. Congress dozens of times. As a highly respected advisor to policy makers over many decades, Dr. Sinai is known for forecast accuracy and for pioneering use of the National Economic Information and Financial Systems approach to monitoring analysis and forecasting and identification of big wave trends in the U.S. and abroad. Dr. Sinai has contributed tremendously to the fields of macroeconomics and public policy, and this professorship pays tribute to his lifetime of achievement. It is an honor to have Dr. and Mrs. Sinai and their children, Todd and Lauren, with us today. Please join me in welcoming them home to the University of Michigan.
[ Applause ]
And now I would like to welcome to the podium my colleague, Susan Collins, Dean of the Gerald R. Ford School of Public Policy.
>> Well, thank you very much, Andrew. And I must say it's really a pleasure to be here this afternoon with all of you. I am Susan Collins, the Joan and Sanford Weill Dean here at the Gerald R. Ford School of Public Policy. And it really is a very special occasion both for the Ford School, for the university and I must say for me personally. And before I begin, and I have the pleasing task of introducing our special speaker, I really would like to give a personal thanks to Lee and Allen Sinai as well as Todd and Lauren. I must say that today's occasion for making this occasion possible. Today's occasion is particularly meaningful for me personally. And I know that many others in the room very much feel the same way. Allen, a professorship in your honor is really a very fitting legacy to all that you have contributed over the years in macroeconomics, in accurate forecasting, and in public policy. And so thank you so much for all that you have done. Well, as Dean, there are many things you're expected to do, and then there's some things that it's truly an honor to do. And I have to say that among the latter, one of the most rewarding is to recognize a colleague. And I have the great pleasure of doing that. Oops!
[ Laughter ]
Oh, let's see.
>> Don't worry about it.
>> Don't worry about it because you're going to be doing, working on that anyway. So let me move over here.
[ Laughter ]
And a little humor is always a good thing as we're talking about economics. And so I have the great pleasure of introducing my colleague, John Leahy, who is the Allen Sinai Professor of Macroeconomics and Public Policy. Here with us today are many of John's friends and also some of his family members. And a number of colleagues in the Economics Department and in the Ford School, and we're all very proud of John as well. John is an extremely accomplished macroeconomist, and he currently serves as co-editor of the American Economic Review. I first met him many years ago when he was a brand-new assistant professor at Harvard, and I have to say that his high energy and extremely wide-ranging interests made him a really wonderful colleague. And so last year I was particularly pleased that the Sinai Professorship enabled us to recruit him to Michigan. His scholarship blends theory with both empirical and policy analysis. And he's well known for research that develops theoretical models to explain why firms adjust prices infrequently and in large jumps instead of responding to changes in small increments as they occur in markets. And some of the important implications of that work are for monetary policy. Indeed, John has been a Visiting Scholar at the Federal Reserve Banks of New York, Philadelphia, and Kansas City. He's adept at solving models which highlight the very complex role of psychological factors in economic decision making. And so he has much to draw on in crafting a lecture on the economics of wishful thinking. He's also an excellent teacher and mentor, and indeed the current Chair of the Council of Economic Advisors, Jason Furman, was one of John's students. Jason wanted to congratulate you personally. And he sent a note that he asked for me to read today. And so I quote, "John Leahy has been a wonderful, generous, and brilliant teacher, collaborator, and mentor. I am particularly grateful for the fact that I only went to the Council of Economic Advisors for the first time in the 1990s at his recommendation. However, the result of too much time in public policy is that I confess to forgetting a decent fraction of the cutting-edge macro that John so patiently taught me." [Laughter] Well, an endowed professorship really is one of the highest honors we can bestow on a faculty member. The selection is a highly competitive process that draws on numerous recommendations of candidates' peers, mentors, and students. And in John's case, all of those were glowing. But I can't help but reference a student review from online that I think sums things up very well. The student wrote, "Professor Leahy's quirky humor keeps you awake. Everyone liked him so much that when he gave an extra bonus lecture at the end of the term, everyone came." Well, clearly John's students and colleagues have really embraced his sense of humor and his enthusiasm for his field. And so we look forward to hearing from him this afternoon. Following his remarks, John will take questions from the audience. And so beginning around 4:30, staff will start collecting question cards, and Ford School Professor Josh Houseman [assumed spelling], together with students from John's recent macroeconomics class, Cassandra Baxter and Josh Fleming, will facilitate the question and answer session. For those of you who are watching online, please send us your questions via Twitter using the hashtag Sinai Lecture. So now please join me in welcoming to the podium the Allen Sinai Professor of Macroeconomics and Public Policy, John Leahy.
[ Applause ]
>> Thank you, Susan, and Andrew, for the, for the very gracious introduction. I really don't know what to say. It's quite an honor to be here at the University of Michigan. It's an even greater honor to be the recipient of the Allen Sinai Professorship in Macroeconomics and Public Policy. I want to thank Allen and Lee for their very generous gift that made this all possible. I also want to acknowledge my wife, [inaudible], and my mother who came here from Oregon to be here today.
[ Applause ]
Their support has been, been just fabulous. My father, unfortunately, he died last year. And I know that he would really, really have enjoyed being here today for this, this event. I hear that my in-laws in India are watching it on the live stream. This is, this is a definitely a new world which, you know, the, the physical space of the classroom is no longer really a constraint on anything and everyone's connected. I hope I look good in India.
[ Laughter ]
What I want to talk about today are expectations and beliefs. And expectations and beliefs are central to pretty much all economic questions. Think about finance. In finance, we price assets. Pricing assets: what's the price depend on? Expected future price of assets, dividends, returns. The person planning for their retirement needs to think about life expectancy, long-term care insurance, what type of bequests they want to give, taxes. Is Social Security going to be around? Even in monetary policy, and basically modern monetary policy is entirely belief management. It's going out there, giving speeches, trying to get people to think in certain ways. Especially today when there are almost no tools, interest rates are at zero, and there's very little room for maneuver. I recall a few years ago at a conference, I got the honor to sit next to Ben Bernanke. And I asked him what he had learned being on the Federal Reserve Board. And he said, "Inflation is equal to expected inflation." So expectations are, are important. Now my, my benefactor, Allen Sinai, I mean, he's in the expectations managing business. He's a forecaster, so shows you again the importance of expectations in economics. What I want to talk about today though is this problem in the way we think about expectations in my side, in the academic side of economics. And that is there's a tension, and there's a tension between the standard economic view of expectations in which we think of people being very good information processors. That they take the world, they digest it, they understand how the world works, and we put this under the rubric of rational expectations. At the same time, there's a lot of psychological research that people aren't very good information processors. That there are a lot of psychological biases in the way people process information. And that they don't make always, or they'll always make good decisions. And the idea I want to play with in the next half-hour or so is to think about choosing beliefs in the sense that treating beliefs is another economic variable with costs and benefits and subject to choice. And thinking about, thinking about beliefs as an economist who says there's some maximization problem, and there's something people do and they don't believe random things. But they believe things for reasons. And I want to play with this and chat a little bit about how this, how this might marry these two worldviews and bring it together. So a sketch of what I'm going to say. I'm going to first start with the standard model, lay out how we usually think about economic decisions, the role of beliefs and so forth. I'll then present a theory of belief choice: what are the costs? What are the benefits? And then I'll argue that this kind of reconciles some economics with some psychological evidence leads to some testable implications. And then I'm going to speculate and say some stuff that's totally unfounded. But, you know, you guys can -- that will allow me to have a little fun, too. Okay, along the way, I'm going to touch on some papers. The first two papers, these are work I've done with Andrew Caplin. He's at NYU, and Mark Dean who's at Columbia. The first two papers are about putting beliefs -- about why people care about beliefs. And the second two papers are about learning models. And in learning models you essentially try to choose beliefs, but subject to certain constraints. I also want to acknowledge some other people whose work I'm not going to cite, but they kind of sit in the background, so we might as well just mention who they are: Akerlof and Dickens. They were kind of the first people to think about belief choice in economic models. There's also papers by Brunnermeier and Parker. And [inaudible] and [inaudible]. And these guys are all very important, and we'll go uncited into everything that goes, that comes, comes, goes from now on. So trigger warning: there'll be a few equations. I can't do things without equations. I'm an economist. So think about a concrete decision problem. Think about a person buying a house. A person buying a house, there's certain states, future states of the world they care about. Will house prices go up? Will house prices go down? Will the family get bigger? Will the family get smaller? Will they get divorced? Will they get married? These are the things I'm going to collect in this set S of possible future states of the world. And then there's got to be, we have to make -- have expectations and beliefs. So there'll be some probability distribution. What's the probability of a certain state happening? House prices go up, you're married, five kids. And a dog. Then there's a set of choices. A, think about the housing example. Buy the house. Don't buy the house. Simple economic decision. And then the way we usually think about this is we assign some sort of payoff called utility to each choice: buy the house. Each state of the world: house prices go up. Family of three. Dog. Married. And then we have a payoff. And then the goal is to make the choice that maximizes expected utility. So we weigh these payoffs by the probability of event. That tells us how much we get from doing that choice. You then look at the different choices and pick the one that does the best. That's the standard economic model of decision making. The one we teach to undergraduates in like in principle. Well, in this model, the only role of beliefs is to weigh the future outcomes. The only reason we care about the probabilities is the, is the weighing the different states of the world. And this is in some sense the rationale for rational expectations. Nobody would willing want to believe something that's bad. All that's going to do is mean they make a mistake in putting a weight on the, on the certain choice. They're just going to make bad choices. It's just going to look, I mean, no one would want to make bad choices. Everyone wants to do the best that they can. So that's why in economics in the standard model, we tend to say people don't make systematic mistakes. They use the information. If there was an obvious better way to use the information, they would actually take advantage of that. Otherwise they just kind of look like, like, like -- why would somebody choose something that doesn't make them happy? But people do appear to make mistakes. In the psychological literature, they will list biases. I just put a few here; there are lots of others. Overconfidence: people seem to be overconfident of their own abilities and of the outcomes of their own actions. Cognitive dissonance: cognitive dissonance is you have some beliefs. You see evidence that contradicts them. How do you then reconcile this evidence that contradicts your beliefs with your current beliefs? And people sometimes just ignore the information. Confirmation bias: people interpret information in light of their priors. They have beliefs, they see information, and they interpret it as confirming what they already believe. So you can get a Democrat and a Republican both reading the same article and coming to very different conclusions about what it means. Because they come to the article with different beliefs. And then that's extrapolation of trends. People seem to think that trends go on. Now the problem, as I said, you put some of these agents, these agents with any of these biases, and you put them into an economic model. And the mistakes are just so glaring that it's just like why would anybody do this? In the standard model, people would choose to be rational. So in some sense, we have to change the standard model if we're going to have a model of belief choice. We can't work with the equation I put up on the board at the beginning. We'd have to do something different. So now the second part of the talk. I'm just sketches -- a theory of choosing beliefs. Two requirements: one is we've got to get away from the standard model where the only role of beliefs is weighing future outcomes. So I'm going to give you another reason why people carry beliefs besides that. And then second, once people have become unhinged and can believe anything, I mean, people would choose to believe, "Well, you know tomorrow I'm going to have a billion dollars. And I'm going to be, you know, have a yacht and all these," anyway. You can go on. I don't want to go there. Okay, so why would people care about beliefs? And this goes to the first couple of papers I put up, some papers about beliefs and their role in utility. And this is a paper I wrote with Andrew in 2001. Now what we were thinking about then is that there are lots of -- There are lots of situations in which people reveal reactions to uncertainty that they feel before the uncertainty develops. So we're thinking about things like fear, hopefulness, suspense. So all of these are emotions you feel today about something that's going to happen to you tomorrow. Anxiety, the American Psychiatric Association defines anxiety as apprehension, tension, or uneasiness that stems from the anticipation of future danger. And this timing difference is very important. And let me explain why. Let me explain to you the difference between risk as normally studied in finance and fear as I will think about it. So here's a standard macro, a standard economic model of risk. So on this axis, I'm going to have some monetary payoffs. On this axis I'm going to have the utility. And this is a curve that says for this monetary payoff, how much utility I get, so forth, so forth. It's bending down, and the idea is that the first dollar you get is worth more than the five-hundred-thousandth dollar you get. When you have more and more money, the additional dollars give you less bang for the buck. Literally, for the buck. Now if we think of an asset that has an average return, and then maybe it's risky. It could be higher; it could be lower. And then what risk is is the idea that the down is more costly than the up. So this variance tends to make you less happy because the losses get more weight than the gains. That's risk. That's the basis for most finance theory pricing assets. This all is in the future because you're looking at the future payoff. So it's your future happiness if the asset goes up. The future happiness if the asset goes down. Fear is something very different. Fear is in the -- oops. Fear is in the present. It's a view of, from the present towards future utility. So you're sitting there on the bridge scared, and that's a totally different thing than risk. What we did in this paper, the Psychological Expected Utility paper, is that we made wellbeing depend on the standard economic things. But we also made wellbeing -- so the weight, beliefs weight future outcomes, like in the standard model. But we also allow beliefs to affect you today. So that you had an added reason for wanting to believe different things. And this led to a lot of interesting outcomes. First of all, if there's positive and negative reactions to uncertainty. In the standard model, risk's just bad, right? There's fear, but risk is just bad. But you can also with this model have suspense and excitement. And so sometimes you might want to hedge, but other times you might want to double down and, and gamble. For example, in a month, Michigan will play Ohio State. If the world was a model of risk, the optimal strategy would be to put money on Ohio State. Because then if Michigan wins, Michigan wins. If Ohio State wins, you've got the money, right? And there's just something wrong about that. I mean, I don't know. It's like immoral or something, but in, in this, in this view, it's because maybe that's not what it's about. It's not risk. It's excitement, and for excitement what you want to do is double down on Michigan. You want to make the odds, you want to call your friend up and say, "Okay, you know, I'm going to put 25 cents on Michigan. If you put --", I'm, I'm -- anyway. I need even more money. But, but, you put money on to double down to increase the excitement, to, to, to, to, to -- [inaudible] explain gambling. Second interesting thing you get is that fear is a multiplier. Notice back here we had the standard risk story, and then we had the fear story. Much of finance is trying to explain why the people price risk so highly, and this gives you a reason for that. I mean, lots of people try to build in models of imperceptible long-run risks that somehow affects current prices. Or they build special utility functions to jack up the price of risk. Here you can jack up the price of risk through fear. Give you an example, another example of fear as a multiplier. So take the right after the financial crisis, consumption fell by a lot. Across the board, people saved. Well, the standard model would try to explain that through an increased probability of unemployment, an increased uncertainty about the future. And therefore people saved for precautionary reasons. To build a hedge against the possible future risks. Quantitatively, that doesn't give you a lot. Because unemployment, while high, was still 10 percent, just one in nine chance. The thing about fear is everyone can fear being unemployed. And unemployed is a very scary state of the world. And so even though it's a small probability, your gut reaction to it could lead to quite a lot of saving. One of the things that we did in the next paper was we looked at the implications for policy advice. We have two roles for beliefs. Decision making and current happiness. Now for an individual, these kind of have to be the same thing. Unless you're like, have multiple selves and multiple beliefs inside your head. But a policymaker can know things that the individual doesn't know. So a policymaker, for example, could know that, something about your future that you don't know. And they can, they then have a question: do they tell you or not? For example, think of genetic testing. You take a test when you are, say, in your 20s. They find a gene that says something about how your likelihood for disease 50 years later. Do you tell that person the result? Or do you let them be blissfully unaware and live their lives for 50 years? A lot of it would depend on their preferences for knowing. Some people want to know; some people don't want to know. And a lot would depend upon whether they could make any decisions that would affect the 50-year outcome. But it's totally possible to think of a world in which you just let the person live with the illusion that life was normal for 50 years. Or take parents and children. A lot of parenting is belief management. We tell our kids everyone can be president. We don't tell them it's probably the worst job in the world. And that, that no one would want, in their right mind would want to be president. But you know, it's a motivational thing, and you want them to think that there are possibilities and that they'll study hard. And so you manage their beliefs. You don't tell them the truth about the future states of the world. That might do damage.
[ Laughter ]
For the current talk, though, what this opens up is now a role for choosing beliefs. Because beliefs don't just affect future outcomes. They also affect current happiness. So now you have a reason why people might want to believe things that are different from the true outcomes. Why they might want to be optimistic. Why when buying a house they might want to believe house prices are going to go up. It's more comforting, less anxiety. You put your whole financial wealth into this asset. You don't really want to think that that asset is on shaky grounds. You want to think that it's, it's potentially going to do quite well. Now with this additional role for beliefs, I mean, things can easily become unhinged. I just believe that tomorrow is going to be wonderful while today is terrible. And there's no real constraint on beliefs. And so, you know, what we do with, with, with the choice of goods is we put budget constraints. And the budget constraint says, "Everybody would like to have more," but your income constrains you to have less, and you have to live within your means and make choices. Well, with beliefs, there's no budget constraint. So it quickly becomes unmanageable. So what are the limits on choice of beliefs? And I can think of two. And one is we've already mentioned is that -- bad, inaccurate beliefs might lead to inaccurate choices. So you might trade off optimism against the downside of optimism. You want to believe the house prices are going to go up. But then maybe you buy an asset, and they don't go up, and then you're in trouble. So that's the first. The second is it's actually kind of hard to alter your beliefs. I mean, you don't just wake up and say, "Okay, today I'm going to believe this." I mean, you have to, you have to work at it. When you go to school, you spend years changing what you think about things. It's not easy to fool yourself. So I think here, one must also have some sort of a cost of changing what one believes. And what's going to happen is it's going to be -- in my view, cheaper or easier to change your beliefs when the world is muddled. When there's less actual hard data in which to base beliefs on. But when there's multiple interpretations out there, multiple things one can believe, multiple potential, explanations, or there's just no data out there. People are going to find it easier to change their beliefs. So in summary, we have costs and benefits. The costs are some beliefs make you happier. The benefits are some beliefs make you happier. The costs are two-fold. Biased beliefs, you're going to behave less rationally, a tradeoff between happy beliefs and happy outcomes. And this not easy to fool yourself. I mean, coming up with inaccurate beliefs is not like a trivial task. So third part of the talk: what do we get out of this? The first thing we get is probably a general bias towards optimism. I mean, people would like to believe that they're better, that their actions are better, you know, and so forth. So a general bias stuff. And this, basically a lot of psychological support for optimistic beliefs. I'm going to pick two studies that I find interesting. They're not, they're just, they're kind of random selections, but I think they give you some sense of how this works. This is a paper by some Stanford guys, [inaudible], [inaudible] and Ross. And their goal was to unpackage confirmation bias. Confirmation bias is normally thought of as confirming your priors. I believe something, I see evidence, then I interpret it through what I already believe. And it makes me more confident in what I already believe. They were wondering what would happen if people want to believe something different than they already believe. Would they go with their beliefs, or would they go with their situation? And so what they did is they create a situation -- a group, look for a group where this conflict might, might, might develop. They picked a group of people who all were surveyed to believe that homecare was superior to daycare. And then within this group though, there were people whose situation was their kids were in daycare for reasons of their job or their life situation. But they were all priors where the homecare was better. They then presented both with research on the costs and benefits of the daycare and homecare, and then they gauged their response to this research. It's kind of a standard psychological confirmation bias type of study. Usually though with confirmation bias, you would have their beliefs being confirmed by the evidence. But now there's a tension between their beliefs that homecare is better and their situation for some of these people that daycare is better -- it is, they're, they're stuck with daycare. And it turned out their situation won, that those who had the kids in daycare turned, they believed what they wanted to. They wanted to feel more comfortable with their situation. They interpreted the evidence as favoring daycare, whereas the other group that was in homecare had the standard confirmation bias. Here's another study. I picked this one because Dunning [assumed spelling] is a, is at the University of Michigan. And this is in perception. And you think of perception, I mean, what you see, I mean, how can that be biased? I mean, you see it. But in these studies, they look at, they flash images that are ambiguous onto a screen. But they primed the people about thoughts about certain images being associated with positive money rewards and other images being associated with negative money reward. For example, we might prime you that numbers are associated with winning and letters are associated with losing. And then people see a 13. When you prime them with letters being associated with winning, and numbers with losing, they see B. When you prime people with sea animals being associated with a monetary gain, and farm animals with monetary loss, they see a seal. And when we do the other way, they see a donkey. I didn't see the seal for quite some time either. So don't, I don't know how -- this priming is fairly strong. So a lot of this might be sight subconscious, you know? It's just the way you process the information underneath. So when should we see wishful thinking? Go through our costs and benefits. The benefits are you're going to win the desire to believe is great. At the beginning of the football season, every team can win the national championship. So lot of wishful thinking. When the stakes are low, I mean, it's not very costly to think that your football team is going to win the national championship, so that's the costs are low. Unless you bet on -- anyway. When the cost of fooling one's self is low. So this is when the data's ambiguous, muddled, or lacking. So think of retirement. We don't get to retire twice. Like we make all these choices for retirement saving before we have any clue what it's like to be older. Right? Housing. We don't sell houses very often. It's almost impossible to price a house. I mean, the, every house is unique. Every appraisal is -- appraisals can be -- automatic appraisal models tend to be off by 15 to 30%. I mean, it's very hard to have any concrete data on housing which leaves a lot of room for people to believe whatever they want to believe. And think about extrapolation of trends. When house prices go up, there's two alternative theories there. They're going to continue going up, or they're too high. And if you want to believe they're going to continue to go up, that's actually quite easy to do. More speculatively, we can think about the macro-economy -- I've been talking about individuals, but put them into a group, right? And you put a whole bunch of wishful thinkers together and you get, you know, optimum, like bubbles. Your -- Greenspan's irrational exuberance. It's easier to believe something if other people believe it. The cost of fooling yourself goes way down. Actually, it's harder to believe the truth if everyone else believes something that's not true. And so you can get that sort of thing. One also could imagine the policymakers themselves engage in wishful thinking. Think of the pension crisis. If your actuary says it's okay to use a seven percent rate of return, it's quite easy to think that your pension is fully funded. And, and, and to go with it. It's definitely, definitely a problem. Climate change: no data. We -- nobody has any clue what the world's going to look like in 100 years. Pretty much free range to believe whatever you want. In a data-free environment, theories can run amok. And that, self-worth. Monetary policy: as I said, monetary policy's all about belief management. It used to be that monetary policymakers used the tools of monetary policy to manage beliefs. But now those tool aren't doing a whole lot. Beliefs are being chosen in some sense as a tool in themselves. And policymakers are saying things trying to directly manage beliefs. Well, what about their beliefs and their wishful thinking? It's -- one can, this is the speculative part of the talk. No data. I want to close with one final thought. So why hire Allen Sinai?
[ Laughter ]
>> Hey, I didn't put you up to this.
[ Laughter ]
>> Why hire Allen Sinai? Well, the traditional view is that he's going to tell you something about the states of the world. What are the possible risks you're going to face in the future? And something about the possibilities of these future states. What are the probabilities? So you can make better decisions. But in some sense, you don't need Allen for that. You need his computer program. Maybe his research assistant because his research assistant might help you run the programs and interpret the output. But you don't need Allen himself. Why do you need Allen? Oops. In -- why do you need a technical assistant? The new wrinkle is to help you manage your own interpretations of the data, to tell you, to keep you from engaging in wishful thinking. And from, and to interpret the data correctly, you need somebody sober mind, lots of experience who can kind of keep you on the right track. And so in some sense, that's, that's why you hire Allen. Thank you.
[ Applause ]
Thank you. I'm not very good at the applause thing. You can stop now.
[ Applause ]
I think, I think we have to take this question first.
[ Laughter ]
But I told everyone to direct questions towards you.
>> Well, on the one hand, they might. And on the other hand, they might not.
[ Laughter ]
So wishful thinking. So -- so, I'm, I'm looking for the data that inform wishful thinking, and I'm looking for the data that wishful thinking informs. So what can you tell me about that?
>> I mean, it's mostly indirect, right? I mean, things like irrational exuberance, bubbles. I mean, how do you explain -- I mean, the housing bubble is in some sense one of the hardest things to explain without people believing that house prices were going to go up. But how long? And the rational economic model would have said this unravels. There has to be some date at which house prices don't go up. And at that date, the -- cuts back to the beginning, and then you don't have a bubble. So in more like indirect, I mean, except for the psychological studies which are more micro, that show bias in individual decision makings. But in economics, I mean, we don't have a lot of data on individual expectations. I guess you can look at some that's popped up recently. But those --
>> That's a great example. Because the data that [inaudible] rising prices were rising prices. And the hope that prices would keep rising. So agents in the system kept buying houses on the expectation, and then derivatives were created on that asset. In turn, on the expectation that [inaudible] more value. And on the wishful thinking informed by the data.
>> So you have the hard data to do that. That wishful thinking then gave us higher prices.
>> And then it -- but at some point it has to unravel.
>> The question is, and I'll stop here. Is there any examples of data that will take wishful thinking and how wishful thinking [inaudible] acts on real live [inaudible] data? And you just may not have gone that far in your research.
>> No, no, no, no, no. Much more on the individual stuff. But the situations in which you would expect, like it's harder to engage in wishful thinking if there is feedback from the world that grounds you in the world. So after that first loss or second loss, it's very hard to believe your football team is actually that good, right? Because there's feedback. It keeps you tethered to the world. Asset prices though don't have that aspect because asset prices are inherently forward-looking and are formed with those beliefs. And they tend to come more unhinged than other sorts of economic variables. But more of when and why than like specific killer examples. Cassandra?
>> Okay, so I'm Cassandra Baxter. I'm a Master's student in public policy and economics, and I have the pleasure of being a student of Professor Leahy's last semester. So the first question we have here is you said that beliefs are the same in the present state as in the future state. But is that always true? For instance, many people believe that if Trump or Clinton wins, it will be, the economy will fall apart. But you don't see them changing their portfolios to align with the winning probabilities of their preferred candidate.
>> I plan on changing my portfolio. But I don't. I mean I think that, that, that to the extent -- I mean, they should. But no, I don't have, I don't, I mean I don't have a good, good, politician's answer for that one. I could have answered a different question which is I guess what I should have done.
>> Hi everyone. My name is Josh Fleming. I'm a second year Masters of Public Policy student also at the Ford School. And I also had the pleasure of being in Professors Leahy's macroeconomics class last winter. The next question: how do you address people who adamantly insist something is true when all, and I underscore all the evidence shows it's false? Such as, for example, the earth is only 6000 years old, or Obama was born in Kenya.
>> Well, Obama wasn't born in Kenya. I think, I mean, okay. So we're not going to explain everything with a theory. Okay? So what you're trying to do is explain -- you know, there are people who, the data doesn't matter, who hold their beliefs very strong. So I guess I would say what we're trying to capture is generally, what economics tends to try to capture is more the middle of the road, the average belief and not focus on the extremes. There may be people, there will always be people who believe very disparate things. And we can't, we're not trying to explain everyone in the world, but to find a framework for thinking about lots of situations and lots of people. I think that pretty much goes for all of microeconomic theory. It's not everyone's demand curve. There's always some people who, when you know, prices go up, buy more of the stuff. There are also some people when supply goes up, supply, you know, less of the stuff. I mean, there are always going to be cases that sound counterintuitive. Sometimes they're interesting. Sometimes they're just not this model. They're some other model. And I think, I think that, like, like confirmation bias. No, no cognitive, cognitive dissonance I think is very different than what I talked about. In the sense of taking information, seeing other information, and just ignoring it completely. That's very hard to put into an economic model, but it's out there, and people do that. And so I don't think we capture -- I was very careful to put some of the psychological evidence. Not all of the psychological evidence.
>> So how do you external factors influence the wishful thinking and beliefs of policymakers like the Federal Reserve, and how much of a person's belief is under their own control?
>> That's a great question! Actually, I don't know the answer. That's where we want -- we're just starting on this, this, this road. And I, I, I, I think it's difficult to fool yourself, so I don't think it's completely under your control. I don't think you can wake up and just believe anything, but I think it's more nudges at the corners and little things. I don't think they can do -- you can't just wake up and think, "Okay, tomorrow inflation is going to be zero. Tomorrow output's going to be a trend. And so, you know, basically my job's done. I'm playing golf." I mean, that would be wishful thinking. In the extreme.
>> Alright, the next question. Is there room in your theory for the fact/belief that the longer and adjustment is delayed, the longer it will be?
>> It will be what?
>> The larger.
>> The larger it will be, sorry.
>> The longer an adjustment delayed, the larger it will be. I think it depends on the information process, like what people know about the, about how. I mean, the word "Adjustment" implies that your beliefs have gotten off track, and then they come back. Now, I mean, you could imagine an adjustment never happening, going to the end of your days and never realizing your beliefs were wrong. Or you could imagine if there's a lot of data, the adjustment's quite quick. I think with some things, they happen. So some examples fit this paradigm perfectly: the housing market. The longer people, it took people to adjust their beliefs, the longer the bubble took over. The easier it is to believe about a bubble, and then the disequilibrium was larger, and the spring-back was, was, was, was bigger. I can imagine other cases where it was very different, where you got off, and then you got a little information, and you came back shortly. I think it depends upon -- with the house prices, they were going up. And when they didn't go up, that triggered a whole bunch of reactions. And so then things collapsed very quickly. So in a crisis and boom-bust cycle, it might be very fast, but for other sort of situations, it might be very slow. I can't -- I'm a two-handed economist.
>> The Federal Reserve has a belief system that --
>> People need to work on their handwriting.
>> That assumes lower interest rates will boost economic activity, and then the limit, but negative rates will work when low rates don't. Given that low, given the low rates in the U.S. has not boosted economic activity, should the Fed change its belief systems, i.e. normalize interest rates?
>> I think the Fed is kind of doing that in the sense that Janet Yellen just, I think last week, said that she didn't understand what the mechanism behind inflation and the working of the economy was. And that they were taking a close look at all of these types of options. But and in some sense the Fed is -- has -- don't know how to say this because they said they were going to raise interest rates. And it seemed like they were going to, but then they didn't. So they, there was talk a year ago about raising interest rates, and it seemed like they were going against their earlier orthodoxy of that we were only going to wait until we saw inflation before we do this. But then they actually ended up not doing it. I think they're confused.
>> They're what?
>> They're, they, they don't, they, they're, they're right now trying to figure -- they're trying to figure out what they believe.
>> The same person would also like to know by how much will Williams beat Amherst in football this year?
>> Oh, by a lot.
[ Laughter ]
And they'll beat Wesleyan, too.
>> I have a question, personally. Is there any input for the role that --
>> By the way, Williams, Amherst, and [inaudible] is "the little three." So it's like, it's like this, like, group of schools that play each other. I went to Williams. And, and it's like the biggest little game out there. You know, 500 people in the stands, and you know, it's very important for a very small group of people.
>> What role, if any, does asymmetric information play into your models? For example, let's say somebody has --
>> Is this for you?
>> Has information regarding actuarial risk regarding their health insurance premiums, and due to wishful thinking, they don't believe it's going to go up. Does the, the role of asymmetric information play into that as all?
>> Who's asymmetric information? Who's the? Asymmetric says there's two people. One person believes something. Who's the other person?
>> The health insurance company.
>> I mean, if people are engaged in wishful thinking, then one could imagine a firm taking advantage of that quite, quite easily. I don't think that a lot of loan officers disabused homebuyers of their belief that housing prices were going to go up. In the run-up to the housing crisis. I mean, it wouldn't be in their interest to do so. The question is though, and so yes, I could imagine, I could imagine a game between a firm and an individual in which the firm tried to take advantage. And pretty much you write down any model with a psychological bias, and then there's some way that a firm could take advantage of that.
>> So is wishful thinking preventing us from talking about Social Security and Medicare and the data that's telling us where it's headed?
>> Well, that's a really good question. Maybe. It's easy to push, I think there is some sense that it's easy to push future problems off in the future. These were very big topics before the Great Recession. Then with the Great Recession, we had more pressing concerns: the present. And we pushed off thinking about the future for quite some time. And that was probably rightly so. I mean, people's lives need to get back on track. People were underwater in their homes, they had lost their jobs, the banking system needed to be, be, be fixed. And so we focused on the immediate as opposed to the long term. My guess would be, the hope, I guess my wishful, my wish would be that we get back to thinking about the long term soon.
>> Okay, the next question. Is there evidence that participation in prediction or betting markets is more rational?
>> Justin? By the way, prediction markets still have Trump at 10% to win the presidency. I said I was going to avoid -- I promised myself I was going to avoid politics. So I'm not going to comment on the 10%, the Trump 10%. I mean, so what's the question? The question was prediction markets prevent you from wishful thinking?
>> Yeah. Is there evidence that participation --
>> I don't think there's any evidence. I don't know. I mean, one would have to do that study at this point. One would think that what -- it's not obvious that a prediction market. So the hopeful prediction markets would, by aggregating organically from the bottom up information, you aggregate information in a really good way. But if you did a prediction market on the housing bubble, it's not obvious what you would have gotten. I mean, maybe a lot of people would have said housing prices were going to go up. Then you would have gotten the madness of crowds rather than the, you know, wisdom of crowds.
>> How does this idea differ from the ideas in game theory?
>> Completely different. Game theory almost always assumes rational expectations. But the closest thing would be the papers by [inaudible] and [inaudible], who have interpersonal games in which they try to fool their future selves. And that's kind of a game-theory theory topic. But they also have -- so they also have, I mean, it's very hard to do this in a public --. They have dynamic inconsistency, I can define that. But, but it's I think very different. You don't choose your beliefs in most game theory models.
>> Is it wishful thinking to assume that one day we will come out of a 20-trillion-dollar deficit and pay back all the money the Treasury owes to foreign counties?
>> It was not very long ago when Clinton was president when the Federal Reserve, one of the big topics of conversation was that we were going to have no government debt. The Clinton surpluses, Bill Clinton, the Clinton surpluses were large enough that if you tracked them into the future, the T-bill was going to disappear in 2010, and the Fed was trying to figure out how we're going to do, you know, trades without a government deficit. Optimal fiscal policy is to keep the debt at fairly constant levels, so I wouldn't think -- I mean we've gone, we've whip-sawed from a period in which the debt was going to disappear. To a period in which the debt's going to explode. I don't know if it's wishful thinking to think we're going to pay it off. I don't think we will pay it off. There's actually no reason to pay it off. You only have to pay -- a person has to pay off a loan because, because they have a finite life, and no one wants to lend to dead people.
[ Laughter ]
But a government lasts forever. Banks are in the business of lending money to people, not getting it back. So we can hold it on for as long as we want.
>> Can you say a little more about how this theory might resolve the equity premium puzzle or the general implications for asset prices?
>> The fear as, as a multiplier, that was one of the -- so we have an example in the 2001 paper where we generated an equity premium through, through fear. It's pretty straightforward. I mean, you just have a risk and you have, you basically double the response to the risk. So what happens is the source of the equity premium puzzle is the curvature of that utility function when you look at, you ask people questions about how would they, what would they do when faced with this gamble or this gamble? You find that that curvature's very flat. But they're always doing it on the second, the future, the curvature of the future utility. When you add this other term, fear, in the beginning, you're completely out of that paradigm. So all the parameterizations, the things that go into the equity of premium, they all go out the window. And you can basically explain it all because we don't have any evidence on how big it is. But it could potentially explain it all. By the way, the equity premium puzzle is one of the things in economics where we have ten theories that explain about 300% of the equity premium puzzle. So it's not, you know, we have to whittle it down more than come up with new ones.
>> Can you speak to the potential benefits of an illusion or wishful thought for mental or social survival against the realistic appraisal?
>> I got lost, I'm sorry.
>> I did, too.
[ Inaudible ]
>> I got lost. Could you try that again? We'll try rephrasing it, move the words around.
>> Can you speak to the potential benefits of an illusion or wishful thought for mental or social survival against the realistic appraisal?
>> You mean so they're facing a world which causes mental anguish, I'm guessing. And I mean, again, now we're getting into psychiatry and, and I'm not qualified to talk about that. Again, staying within the average person, not the -- I don't have anything to say about the mentally ill. I don't know anything about the mentally ill.
>> Can you speak a little bit, too, about how people might update their beliefs based on new information? For instance, is there a Bayesian process to do so?
>> So I'm not thinking about this being Bayesian, and that's the big difference. I didn't go into the details here. But I'm thinking about being Bayesian is actually a pretty hard thing to do because you have to come up with a whole bunch of posterior beliefs that add up to a prior. And that's just pretty complicated. So I'm going to allow my guys not to be Bayesian and that then allows them to shift their posteriors in ways that, that, that lead to basically belief choice. But non-Bayesian-ism is entirely -- if they're Bayesian, then they, they, they -- so Bayesian decision maker's one who interprets information accurately so that the, the, there's some mapping from -- you see something. And there's some mapping of what you've seen to the world. And you interpret those probabilities correctly, and you update your prior in light of that new information correctly. And you're deeply inside the rational expectation's paradigm. That's what the standard economic agent does. We want people to like see the information, and have trouble doing this. And that allows them to have all the posteriors shifted off in different directions. Conditional on buying a house, you believe house prices are going to go up more than you should.
>> Okay, and this is our last question. So how can small, separate assumptions and beliefs combine to create sort of a collective flaw policy. So basically is there a multiplier on fear?
>> A multiplier on fear? Multiplier --
>> I think so, totally. But I don't, I mean, people -- so there's this -- once this paper by these guys at, at Northwestern. It was an epidemiology -- logical model of housing bubbles. And people, like you had people who were infected with optimism. And you went to cocktail parties in the model, and if you met a lot of these people, then you became infected. And then eventually the infection spread and everyone believed that house prices were going to go up. Until people could no longer afford to buy houses, and then reality set in. And the crash happened. It's not obvious that you know, that this theory put into groups wouldn't work in very similar ways. In the sense, I'm talking to you. I should be talking to them. In very similar ways in the sense that meeting people have certain beliefs makes it easier for you to believe things, I think. It's easy to believe things if other people believe things. It lowers the cost. You're not fooling yourself, you're taking advice from a friend. Right? And you listen to your friend when you want your friend to be right. You don't listen to your friend when you don't want your friend to be right. And, and so we've worked similarly I think to this, this, this mass contagion.
[ Inaudible ]
Ah, I think it does. Well, it depends. I guess it depends upon whether like I could only believe a little bit. And then you believe a little bit, and then your belief makes me say, "Hey! Gosh, if he only believes, he believes a little bit, then I can believe more." I mean, you could imagine, it's almost like a Keynesian multiplier on beliefs, that other people's beliefs ratchet up mine. Totally possible. Again, I think Allen would like to see the evidence, but, but we're having fun here. So in a data-free environment, we can say anything we want.
[ Laughter ]
>> So on that note, thank you very much!
[ Applause ]
He's not allowed to take a seat because on behalf of [inaudible] Management Department and the Ford School, we have a token of appreciation and something to help him remember the celebratory event here today. And so John, we have a framed poster of the --
>> Of the poster?
>> Of the poster advertising this event today.
[ Laughter ]
I can open it.
>> And we hope that you will keep it in your office and remember our celebration today.
>> Oh, beautiful!
>> So we had a designer who was trying to capture the sentiments of all the synapses and wishful thinking in economics, and so hopefully that will prove to be a nice reminder.
>> Thank you, thank you, Susan.
>> Absolutely. I had some other thank yous. It has been really a pleasure to work with [inaudible] and the Economics Department in putting together this event. And so we very much appreciate that. It is really wonderful to have [inaudible] and Jane, John's mother, here with us. We are thrilled that you are able to be here. And again, a warm thank you to Allen and Lee Sinai and their family for making all of this possible. We are truly grateful to you. Wonderful to have all of you here with us, and all of the questions. I know that the questions were still coming, and the good news is that we'd like to continue the conversation over a reception out in the Great Hall. And so please join us out there, and we hope that you will stay and enjoy. A final thank-you to John Leahy.
[ Applause ]
>> Thank you.