Welcome to Behavioral Economics

This site is the course reader. Lectures are written and re-written here in public, in front of you, over the course of the term. If you click into a lecture and notice typos or phrasing that doesn’t quite work, that’s because most of these are still drafts — and your reactions, in the right margin of the page, are how I find out.

Here’s how this will work. I’ll record each lecture, turn the recording into the written notes you see here, publish them more or less immediately, and then revise them as your questions and marginal comments come in. None of that will be hidden from you. The page you read in week ten will be better than the page someone read in week two, and the difference will be partly your doing.

As noted in the syllabus, there is no textbook for this course. The slides are the closest thing to an official text, and your own handwritten notes from class are likely better than anything I can post here. These pages are intended as a supplement to your own work and a place to catch something you missed, re-read an argument at your own pace, or possibly ask a question you didn’t ask in class. They are not a substitute for attendance and in-class engagement.

One warning in the same spirit of honesty: some of what happens in class does not survive the trip to prose. Long stretches of algebra worked at the board, derivations that unfold in real time — some of that will show up here compressed, and some of it won’t show up at all. If a step you saw in class is missing or harder to follow on this page, that isn’t a signal that it doesn’t matter. It’s a signal to check your notes and the slides — or to ask in the margin, which is exactly what it’s for.

Welcome to EC 404, which I lovingly call “Economics Not Found,” or behavioral economics.

As I said before we started, I am trying to encourage humans to do human things and talk to one another. So put your phone away when you come to class, and make a friend. If you have no friends, I will be your friend, though not a very reliable one; I won’t call you.

I love teaching this class, and at least seventeen to twenty other people would dearly like to be sitting where you are. If you decide this course isn’t for you, that is perfectly all right: drop quietly, and somebody will take your place. That is not me saying I don’t like you; I probably will like you. But use this class period to judge honestly whether the course is a good fit.

A few notes before I launch into things. Not counting today, I begin talking two minutes after the hour: class is at 2:40, so 2:42. I am easily distracted, so if you walk in late I will stare at you, and then, very likely, try to strike up a conversation. Consider yourself warned.

You may also have noticed that I am recording. This is a bold attempt to turn the lectures into written notes on the course website. What I say aloud gets transcribed; I clean it up by hand; and it gets posted under notes for that day, which is what you are reading now. It is no substitute for being here. I say things that are funny to me and not to you, and the cleanup is a little rough. But the notes exist, and the website has a pleasant feature: you can comment on them, publicly, right in the margin, and you need not use your real name. Give yourself an alias. I encourage it; it’s fun. If a question about the content occurs to you after it is posted, ask it there.

The course website is your guide to everything: the syllabus, the schedule, the problem sets, these notes. In general I lecture from the slides, and the slides are incomplete. This is a class you simply have to attend. Most of you are juniors and seniors, leaning senior, and it is time to be honest with yourself: are you the sort who comes to class, or not? If you are in the “not” category, again, no harm done. But this is not the class for you. Learning, here, is something you do with your brain switched on, in the room.

That said: have fun, relax, enjoy. I love teaching this class and I enjoy talking with you.

Here is the thing about actual learning: it sucks. I have done a great deal of it, and it is hard, because by definition it means you do not know something now and you will know it at the end. And in this course the thing you will know is not a fact. It is a way of thinking about a process, about the world we all jointly inhabit. That requires an open mind and a variety of tools.

That makes me the most important tool, which means you should come to office hours. They are right after class, Tuesdays and Thursdays, 4:15 to 5:30. If that does not suit you, no matter; send me an email and I will make time. I am very accessible, so do not let the scheduling become an obstacle in your mind. My office is about a three-minute walk from here, and I will happily walk with you and chit-chat. I can probably solve a few of your problems in the chit-chat portion alone. Remember that I am a human being.

I got here in 2017, and before that I was at Harvard.

I crib my material from friends at smart people’s schools: this class is a grab bag of graduate courses at Harvard, Berkeley, and UC San Diego, and an undergraduate course at Cornell. So it will be challenging, but I hope fun. In most of the economics courses you have taken at Michigan State, you covered economics up to about 1920. In this course you will get to about 2023. That is only because I have not gotten around to updating it. Perhaps I will.

My research concerns a number of topics in psychology and economics, many of which we will cover this term.

The difficulty of this course is not really the content. It is using your brain in ways that do not feel familiar. And if you are planning to graduate in December, or in the spring, you may already be coasting. I went to undergrad at the University of Oregon (go Ducks), so I have been there. But the kind of learning I am promising, if you engage with this course, demands a little curiosity.

What we are going to discuss is not esoterica. It is not things you have never experienced. It is what people do every day, and people you recognize. In the first section, for instance, we will talk a great deal about choices under uncertainty: gambling, for one.

The most important way to make progress in this class, and outside it, is to know yourself. If you are bad at something, I am quite literally paid to teach it to you, and I am glad to. If you cannot remember high school algebra, that is all right; it means somebody did a poor job of teaching it, and I will happily teach it again. Every term I teach high school algebra to something like 40% of this class. Do not let that hold you back.

But you do have to know yourself, because most of the time you will be sitting quietly, looking at me, and I have no idea what is going on in your head. I am trying. I do not know which muscle to flex. So you must tell me what you struggle with, and to do that you must know what it is. Ask yourself: are you the kind of person who comes to class? Who wants a class that makes you think? If the answer is no, that is fine. Don’t be judgmental about it. Use the Ted Lasso credo:

That is how you succeed in this class. I have no other recipe, and I mean that.

This class will feel foreign in many ways. That is by design. It will feel difficult in ways you will find hard to articulate. Also by design. Let yourself think about it, put your phone down, and when in doubt, ask me questions.

The course in one frightening line: I teach using math. Here is what that means.

Suppose I want to know the area of one of the floor tiles in this room. I could take a sheet of paper whose area I know, lay sheets on the tile, cut them down until I know precisely how many fractions of a sheet fit, and count them up. Then I would know the area of that tile. The trouble is that I would know only that tile. If I want to know how big tiles are in general, it is useful to know that the area of a rectangle is its length times its height: $A = L \times H$. That is mathematics. And we write it that way because now I know it for every tile, not merely this one.

In the same way, we are going to try to describe a great deal of human behavior, and humans are difficult to capture one at a time. So instead I will say that behavior follows an expression like $\lambda x^{\alpha}$, which is a compact way of saying “perhaps people differ in this parameter, and other people differ in that one.” I will explain what the symbols mean as we reach them.

This is an unfamiliar way of doing things, because in most math-heavy courses the game is: what do I jam in for $\lambda$? What do I jam in for $\alpha$? You hunt around the page for where the professor hid the numbers, rearrange things with your algebra, and you are done. In this course I may ask you to set up the equation yourself, from a scenario that is nothing but words. That is a skill we will build over time, but it is a muscle most of you have never used. The key is an open mind, and not recoiling.

At least 30% of you, probably more, have an instinct that when you see mathematical symbols or Greek letters, you recoil and your brain switches off. I do not experience that; I am a nerd. But the key is not to let it happen, and that is a trainable skill. $\lambda x^{\alpha}$ is an idea, just like $A = L \times H$, and the second one frightens nobody. Try to have the same reaction to the first.

Why teach this way? This course is full of ideas from psychology (in another world I would call it Psychology and Economics), but intuitions are loose. They tend to be misleading and imprecise. Our comparative advantage as economists is writing things down as equations that we can then apply, profitably, across situations. A psychologist might propose a “hot stove effect”: you touch the stove once, burn yourself, and never do it again. Plausible. What we would say is: if you lose a certain amount of money, perhaps you become unwilling to take certain kinds of risk. And then we write that down as an equation.

This approach has been tremendously successful. There is a Council of Economic Advisers. There is no presidential council of psychologists, or sociologists, or political scientists, or criminologists. Social science was won by economists, and not because we are right all the time. It is because we did this. Other fields have begun catching up, but they are three decades, five decades behind. Good luck. So the formalism is good: economists have enormous influence on policy, relative to other social scientists. And it is bad: you have to think formally. But you are in a 400-level economics class, so you have already made that trade.

Practically speaking, problem sets are how you learn the material. They are long. I know, because I wrote them; I wrote every single thing in this class. There is little to no reading. You may look at the slides and at these notes, and you may find them useful, but the best thing you can do is come and think. That is a peculiar instruction, I know. What it means is: sit on a park bench for fifteen minutes after class and ask yourself, what just happened in there? That leads either to “oh, I sort of get it” or to “I have no idea,” and the second one means you should ask somebody: another student, or me.

I teach the way I think maximizes success statistically. But you are all different people, and maximizing success statistically is not the same as maximizing success for you. I am genuinely trying my hardest; this is all I have. If I am failing, it is my fault, not yours, so long as you tell me. Usually I can get you there.

A sample comment from fall 2024:

I have done nothing to address this comment. (Other feedback was positive.) In a sense it is a fairly accurate summary of the course. I do not think the difficulty is ridiculous; I was trying to teach this student, I failed, and this student probably never let me know. But yes: this course will require you to take an idea from setting A, the area of this tile, and answer a question about setting B, the area of that table. Different words. Same question. Of course it is. That is what exams in this class ask of you, and exams are the majority of your grade.

If you are concerned about the course, see me soon. I cannot help if I do not know.

Read the syllabus for everything; it is on the course website. The highlights:

Participation is stochastic. It is whatever I do when I am bored, and I get very bored: by definition I know everything in this course, so I am telling you things I already know, which is a curious sort of conversation. Sometimes I will ask who you think will win the football game, and that will count as participation. Why? Because I am bored. It isn’t that deep.

That said, assume attendance is mandatory. If you know now that you will miss more than about 20% of classes, drop the class. That is fine, but the class is not for you. If you miss two or three classes, that 10% will not bite. If you miss seven, it will.

The writing assignment is free points. This class satisfies a writing requirement (the capital W in the registrar). I really do not like it when MSU tells me what to do, so I have made the assignment as minimal as I can. It is a no-generative-AI assignment; everything else is essentially fair game. I will describe it much later in the term. Genuinely, do not worry about it.

Grades diverge in two places. Problem sets are graded by a TA on a check system (the details are on the assignments page): if you try hard and attempt every problem, you will very likely get a check, even if every answer is wrong. Exams are not graded that way. They are graded for correctness, by me (I am very good), in person, by hand: you write things. That is where grades bifurcate into good and not good.

Exams follow from and extend the problem sets. By construction they are easier than the problem sets. A problem set will take you anywhere from six hours, if you are a strong student, to fifteen or twenty if you are struggling. There is no reading, remember; that is what the credit hours are for. If you feel you are beating your head against a wall, talk to somebody else, then talk to me. I am trying to make them doable.

I wrote every problem, and I am happy to tell you why I assigned it: the precise idea I am trying to get from my head into yours. They are not busywork. Sometimes there is a little torture in there: I will remind you of things you used to know, because dredging up knowledge someone taught you long ago is a portable skill, and you will need it after you leave. It is no great matter if you do not remember the chain rule. It is not fun, it will not win you friends, and you can look it up. Exams have no calculus on them. None whatsoever. I promise they are easier than the problem sets; you will agree with me afterward and not during, because they will ask you to use a muscle you did not know you had. People improve a great deal in this class.

As the syllabus states, you are entitled to the raw score under the grading rubric. I will curve upward; I always do. It depends a little on the distribution, but not much: on the order of five percentage points, more or less depending on where you sit. Standard-ish cutoffs apply: 92 and above is a 4.0, 87 to 92 a 3.5, and so on. If you are looking at the low end of that table, aim higher.

If you are here to satisfy a 400-level requirement and want the easy route, come see me after class and I will tell you which courses are easier than this one. I like this course. You will learn a great deal and we will have some fun. But you will need to use your brain.

The course is hard. I made it; of course I know. It is meant to be hard. If grades are the most important thing to you, tell me honestly, and that includes the good side. If you have had a 4.0 all through MSU and you are determined to keep it, please tell me, so I can make sure you do not get 15% on the first exam, which some people do. The first exam is a genuine wake-up call. And if you are on the other side, with other priorities and a particular grade in mind, that is fine too. Tell me what you are trying to get. I am here as your employee. You are employing me to take ideas out of my head, which I worked very hard to get, and put them into yours, where you will work less hard to get them if I do my job right. Success is achievable, but it will involve doing things that may be new to you, including talking to your professor, which people have historically found terrifying.

There is no text for this course. That is good (no book) and bad (you need to come to class, pay attention, and ask questions).

Engage with the material in nearly all your waking hours. This sounds extreme, but it will not make the top fifteen extreme things I say this term, so keep a scorecard. What I mean is low-level thinking. I am as online as you are; when you are on your phone you are not exactly here, not entirely. So when you leave class, resist the urge to pull it out immediately. Take a couple of minutes to digest. Talk to somebody else in the class. Make at least one friend, or half a friend. They need not be especially good.

Some of the problems are hard, and I can tell you exactly which: problem 2 on Problem Set 2 and problem 4 on Problem Set 5. They are hard because they require you to dredge up an idea you do not remember, so you do not even know where to begin. That means: talk to somebody else, and if you are still stuck, talk to me. There are no secrets. I am not guarding a treasure trove of truths that you extract by answering my riddles three. I am trying to get you to engage with an idea that is unfamiliar to your present mind. Sometimes the material will be confusing, or interesting, or both, and that means you should think about it.

Office hours are the best way to get answers, even to half-formed or silly questions. Email is a poor medium for most questions. Unless your question is “is it okay if I miss Thursday?” (answer: yes; I am not your parent), do not email it. The reason is that the path runs both ways. I am trying to put ideas into your head, but I also need to find the ideas already in your head that are not the ones I want there, and root them out. Over email that is very hard. In front of me, I am rather good at it. Perhaps you looked at $\lambda x^{\alpha}$ and decided that to get $\alpha$ out you simply divide by it and it cancels. That is not how algebra works (you need a logarithm to get rid of an exponent), and it is no big deal; I will dredge the log rules out of you kicking and screaming. But over email? What am I going to do, typeset it? Come to office hours.

Also, ask questions in class, especially once we get into it. Today is not a real day. But later on, the more you can be the person who says “can you repeat that?”, the better. I talk fast; I do not mean to. A life tip from Professor Bushong: “can you repeat exactly what you just said? I didn’t catch it” is a remarkably useful sentence in all sorts of situations. It simply owns the thing. Perhaps they said it fine and it just did not land. I am happy to say it again.

I am a human being doing his best, so: some lectures will be too long or too short. Usually not; I am fairly dialed in at this point, and I know where the beats are. But every once in a while. Some content will not make sense, mostly because I made an algebraic mistake on the board. If you are good at math, shout at me; there is a real chance I slipped. And sometimes I will simply stop talking and stare. It happens.

Perhaps my most important request: comment throughout the course, not merely at the end. You are paying for this. It is a strange thing to go to a restaurant forty-seven times and only then say the food was bad. Perhaps try something else on the menu. I do not know what works for you; I am trying the approach I think is best statistically. If you want a different phrasing or a different approach, say so, whenever. The material improves with time and feedback, and that is especially true of these recorded notes. This is an ongoing experiment in letting you see how the sausage is made. The slides are dialed in after eight years; writing them out is a giant pain, so I record and transcribe. There will be transcription errors. Let me know.

If I call on you and you do not know at once, do not panic. Take your time, or say you do not know. I do not call on people much, but sometimes I get bored. I will try to avoid asking questions you cannot answer.

Ask questions. There are no bad questions. (Well, there are, but I would much rather hear them than not.) If you ask a dumb question, it does not mean you are dumb. I say dumb things all the time; I will write dumb things on the board. One of the essential skills in maturing as an intellectual is divorcing your ego from what you produce. First drafts are bad; mine certainly are, and unless you are James Joyce, nearly every author’s and artist’s first drafts are terrible. Do not hold yourself to that standard. The worst thing that can happen is that I tease you a little, which is nothing much, and you can tease me back. I am trying to relax you by making you confront the scariest scenario: you ask a question, and it turns out to be a dumb one. Okay. Now what? It’s fine. Move on.

I want to learn all your names. I actually do this; it takes time. Statistically, the best way to get your name into my head is to tell me about seven times. The first time, it goes in one ear and out the other; I am merely being polite and smiling. It is not personal; I do it to every human on the planet. Until I know your name, I will call you That Guy, and nobody wants to be That Guy, so just give me your name. Be patient. My memory is bad and there are a lot of you, but only one you. You are a beautiful flower in a garden of intellectual curiosity, which we will cultivate into an even bigger flower. Unless you do not like flowers, in which case you may be a dragon.

Your assignment is to read the syllabus. Things to stress from it:

Despite the hard line in these intro slides, I am here to help. I am genuinely trying to teach you things. When I fail, it is because I am not doing it as intended, not for lack of trying. Every problem has a purpose. If you want to know why something is hard, ask me and I will tell you why I am making you struggle through it: it is either dredging up an old idea or connecting to a larger idea in the course, and I want you to struggle a little so that you understand the texture of the idea. That is an odd thing to say, but we will get there. I care about you as a student and as a person, and I try very hard. I think professors should try to make every course wonderful, and I aim for that.

What is the goal of economics? To understand the economic outcomes we observe in the world, and to design policies to improve them. (Yes, I just used “economic” to define economics. Good pedagogy.)

In this course, “economic outcomes” is deliberately broad. We will study choices that look like standard economics (whether people buy insurance, how they allocate their time across a torturously long problem set) and things that might feel like decidedly-not-economics, such as the psychological costs of poverty.

Our goal in this course is to use and embrace the substance, the techniques, the annoying mathematical notation, and most importantly the goals of standard economic analysis:

We will do all of that, but we will introduce psychological factors that standard analysis tends to neglect. What is a psychological factor? It is harder to define than it appears. You have ideas in your head (citation needed, but I believe you do), and you produce behavior that accords with those ideas. You have a theory of tomorrow’s weather, and your outfit follows the theory. The theory may be grounded in reality, but it exists between your ears. It is a psychological object, and it leads to an economic outcome: say, buying shorts. We will explore psychological factors of that kind, the ones economists tend to under-emphasize.

In brief: behavioral economics attempts to incorporate more realistic assumptions than ordinary economics does, both to understand the behavior we see and to improve our analyses of welfare and policy. For now you will have to take my word for it, but here are some “standard” economic assumptions, and the examples essentially form the outline of the course.

People evaluate outcomes as final levels of wealth. If I am deciding whether to bet $10 on a game, I think: if I win, I will have $10,000,010, and if I lose, $9,999,990. (I am a ten-millionaire, obviously.) That is the standard model, which we will meet Thursday. A more psychologically realistic version of a $10 bet is that the object in my head is plus or minus ten. I dare you to argue that the former is more psychologically resonant than the latter. Plus or minus ten simply feels truer to me. I do not live in your head, though.

People behave exactly as they plan. That does not mean people are never surprised. It means they contemplate even improbable events, decide what they would do, and then, when the event arrives, do it. Now suppose that in one of your other courses you had a problem set due and you planned to work on it an hour a day for four days. Might you deviate from that plan, even with no uncertainty whatsoever? Tuesday, something comes up. Wednesday, something else. By Thursday the plan is gone. Your narrative will vary.

People care only about themselves. This one is less standard than it looks. Adam Smith wrote two books; the other one is The Theory of Moral Sentiments, and it is all about not caring only about yourself. We will talk about it.

People are Bayesian information processors. When you update your beliefs, you apply Bayes’ rule. If you do not remember Bayes’ rule, do not worry; nobody does. And how can you possibly be following it if you do not know what it is?

People act in their own best interest. This is where the course departs most sharply from other economics you have seen, and certainly from a neoclassical perspective in which there is no role for intervention.

Let us return to the problem set. Meet Kevin. (Is anyone here named Kevin? Yes? Sorry.) Kevin is my recurring fictional character, and Kevin is not good at this. Kevin planned to work an hour each day, Tuesday through Friday. Kevin made that plan himself; I did not impose it on him. And if you ask Kevin on Tuesday, working an hour a day is in his best interest. But Kevin deviates predictably: he slacks off the first few days and crams on Friday. If you offered Kevin the menu “an hour a day” versus “all of it on Friday, in a panic,” he would obviously choose the first. The problem is that Kevin cannot bind himself. He cannot be Odysseus lashing himself to the mast.

So there is a role for intervention. I, the benevolent dictator of this class, impose deadlines. Nothing requires me to. I could have you turn in every problem set on the last day of class. Do you want that? Serious question. I will give it to you. It is strictly better: you can still turn things in whenever you like, and you know there is no deadline.

Nobody wanted it. You know you would blow everything off and fail every exam, and once you have failed every exam it does not matter how the problem sets went. That is evidence. Not very scientific evidence, but I asked, and you said no. Behavioral economics challenges these assumptions in precisely that way: not with dogma, but with evidence, from psychology and from our own eyes and intuitions, that casts doubt on these and other standard assumptions. We will probe them all term.

The goal is to incorporate insights from psychology to do better economics. I framed Kevin’s problem as one of allocating time over a week. Kevin also has to allocate money over his life. You have money now; you could spend it all now; perhaps there are reasons not to. Understanding the trade-off between saving and consuming has an obvious parallel to Kevin’s homework: left unbound, the consumption-savings Odysseus fails to save adequately for retirement. That is essentially why 401(k)s exist in their current form, and we will describe it in detail. Why do some people fail to search for jobs as hard as they themselves planned to? Same idea. Why do people consume addictive products, or seemingly innocuous products that become addictive? Many of these would be mysterious if we insisted on dogma about what happens in the six inches between our ears. Guided by data from other social sciences, and by our own eyes and intuitions, we can do better economics.

A closely related field is called decision research, or judgment and decision making. It tends to live in business schools, and it attempts to develop descriptively accurate models of human judgment. Behavioral economics attempts to apply those models, or simplified versions of them, to do better economics. There are no bright lines between the two, and when you are thinking about this class in every waking moment, these are interesting rabbit holes.

It is not psychology taught by an economist. I have the unusual distinction of two PhDs, one of them in neuroscience, so I know a fair amount about how brains work and how people think; that is where the “staring into your eyes and pulling out the wrong ideas” comes from. But we are interested in doing economics. The major focus is economic methodology, so much of this will feel familiar from other courses, and our benchmark will always be the standard economic model. Do not worry; I will always teach it to you. There is no assumed prerequisite in this course whatsoever. Ignore the course requirements. I will teach you anything you need. If you have not done high school or college algebra it will take more work and more office hours, but it is not that hard; they teach it to high schoolers, after all.

It is not behavioral finance. We will talk about financial decision-making, but we will spend no time on the determination of asset prices, which, so far as I can tell, involves spirits descending from another realm. Do not ask me how to invest. Okay, I will tell you: buy low-cost index funds and forget about them for a long time. There is your investment advice.

We will, however, discuss what behavioral phenomena imply for financial decisions, because you cannot separate them out. If you buy a burrito (my recurring example, since I am usually hungry by the time this class meets), you have made a financial decision. Everyone knows the heuristic that you should not go to the grocery store hungry. Why? Part of it is that you are shopping for the whole week and you will overbuy. How do we formalize that? How do we come up with some Greek letters so that the same intuition, which is obviously true, also explains why people buy too many convertibles in the summer and too many pools in May and June?

This is the slide that took me the longest to make and will underwhelm everyone, including me, because of the arrows. Look: we have economics and psychology, and then sometimes you have some experimental economics, and then you get the first wave of behavioral economics, and then you get some other stuff, and then you get some other stuff. Hooray. Fun for me.

The history of thought is underused in the teaching of economics. In a physics class the history is embedded in the names: Newtonian physics; Newton leads to Maxwell, and Maxwell (skipping a great deal) to Einstein and quantum theory. We seldom do that in economics, and I will try to ground things as we go. On Thursday we will begin in the 1400s, where some of these ideas originate. But where we end up is recent. The first wave, Kahneman and Tversky, is the early 1980s. The second wave, the theory, is the 2000s. The modern work is from the last couple of years, or this year.

Some names you will see. The two psychologists most associated with starting the field: Daniel Kahneman and Amos Tversky, Danny and Amos. The first behavioral economist: Richard Thaler. The real first: Herbert Simon. The real real first: Adam Smith. Influential experimental economists: Vernon Smith, Charlie Plott, Al Roth. Early prominent behavioral economists: George Akerlof, Robert Frank, George Loewenstein, Matthew Rabin, Colin Camerer, Ernst Fehr, David Laibson. The current generation: Nava Ashraf, Botond Kőszegi, Stefano DellaVigna, Ulrike Malmendier, Devin Pope, Nick Barberis, Supreet Kaur, Dmitry Taubinsky, Justin Sydnor, Josh Schwartzstein, and many others. It is a small field; I know most of these people personally. Danny and Amos have passed away; Dick is still kicking. Several on the list (Kahneman, Thaler, Vernon Smith, Roth, Akerlof) have Nobel Prizes, and more of them will. When the names come up later, you will have a fleeting “oh, that’s who that is” moment.

The topics are on the schedule page. The first, choice under uncertainty, runs to Exam 1. The second, intertemporal choice, plus a short section on public policy to combat errors, runs to Exam 2. Then errors in statistical reasoning, and social preferences and scarcity, where we end. Four topics that sound small and are large; choice over time alone could fill several courses. The first two exams cover only their own stretch of the course; the final is cumulative, weighted somewhat toward the end.

For each topic:

Two forms of theory. Abstract theory defines a stylized environment and works out what models of decision-making predict in it: the “length times height” level, where I get to ask the philosophical questions, such as what two dimensions really are. Applied theory takes the abstract theory and boils it down to something Kevin is actually doing, such as choosing among insurance policies.

Sometimes I will skip slides, as I just did. That is deliberate. They are there for the curious, and I am never skipping the slide with the real answer on it. This is not a class in which you look up an answer anyway. You might look up a way to think about something, but in the end you are thinking about it.

Two forms of empirical analysis. First, experimental data, sometimes from experiments in class. Those occur at random, according to my whims and boredom. They are always voluntary. They always use real money (mine), and frequently involve me giving you a fair bit of it. Once I gave somebody $1,024; why that exact amount will become clear next class. They are almost always fun (for me). An extra reason to come to class. Second, field data. One of my favorite topics involves the finishing times of marathon runners and what they reveal about targets: income targets, time targets. It will be clear when we get there.

Finally, policy analysis. Behavioral economics is perhaps most important for its welfare implications, and because my ambitions are enormous and my time is finite, this course has relatively little of it: two days, essentially. That is an error. Policy analysis is the beating heart of economics, and by policy I do not mean only what governments do. Amazon setting prices is a policy. The due dates in this course are a policy. Everything here is meant to speak to real-world policy, and I will point it out where I can, and happily wander off on a tangent if you ask. But it will suffer for want of time.

One more note, and then we are done. A perspective we will take in this course, and one economics generally adopts, is that we observe people’s choices and infer their preferences from them. If I see somebody wearing shorts, I infer that he would rather wear shorts in this weather than not. You might ask: this is a class about psychology, so why not simply ask him? Here is the thing. He might have a good answer. But in general, people do not know why they do what they do, and we are going to proceed as if we cannot ask people how they make their choices. I believe, from studying brains, that you do not know the answer, and what comes out is some after-the-fact logic. Here is a curious fact. Suppose there were a clock in this room, spinning rapidly, and I told you to say “stop” whenever you liked. If I were watching your brain, you would begin making that choice about a quarter of a second before you said it. Ask you when you decided, and you would say “right when I said stop.” That is physically untrue. So yes, I could ask why he wore shorts today, and the honest answer is “I don’t know, I felt like wearing shorts.” Better to look at him, note the shorts, and conclude that he preferred shorts. It isn’t that deep.

That has two implicit features. We cannot ask people how they make their choices. And the preferences we infer from choices are used for welfare analysis. Here is where we will differ a little from other classes: we will allow the possibility that you made a mistake. Not with the shorts. But the person who postpones studying until the night before the exam (especially in this class) probably did not prefer that themselves, and if you asked them how it happened they might invent some after-the-fact story. So we will study the choices people make, not the rationalizations they verbalize, and we will allow that the choices do not exactly match what people themselves want. That is the essence of where behavioral economics goes, beginning Thursday.

Thursday is real material. No warm-up; I will be gentle, but be ready to take notes. Questions or comments about the class: office hours, or come see me.

Every paragraph, heading, and blockquote on a lecture page has a little icon ”+ ask” that appears when you hover to the right. Click it (or click the paragraph itself) and a composer opens in the right margin. From there you can type a question, observation, or suggestion, submit, and your comment appears next to that specific paragraph, attributed to your display name. You can reply to other students’ comments and replies thread underneath. There’s a small rate limit — roughly one comment every few seconds, thirty per hour — to keep the conversation reasonable and not break the server, so don’t spam messages.

If a paragraph’s text changes substantively in a later draft, its annotation thread is severed and a new thread starts on the new version. That’s a deliberate choice: a comment that says “I don’t follow line 4” doesn’t make sense once line 4 has been rewritten. If you’ve never used a margin-comments system before, the picture below should help.

A pseudonym and a password hash — that is the whole list. When you sign up you provide a username (which is what you’ll log in with), a password (the only thing you need to keep safe), and a display name (your pseudonym — what other students see next to your comments). The site never asks for your email or your real name; it does not connect to MSU SSO; it stores no NetID, no enrollment record, no campus identifier.

Comments are written next to your display name, so other students see something like QuantumOtter next to your replies — never your real identity, never even your username. The picture below maps out who can see what.

I want to be honest with you about what this is and isn’t.

Because this system runs on my own infrastructure, I — as the operator — can see the row that links your username to your pseudonym. There is no cryptographic blind between us. The privacy guarantee is that I don’t look unless there’s a real reason to, and that no third party — no advertiser, no platform, no analytics service — has any of this data, because there’s nothing to share. Passwords are stored as Argon2id hashes; even if the database file were stolen, your password would not be recoverable.

The other thing no system can protect is you, from yourself. Your prose is recognizable. If you write with a distinctive voice, classmates can probably identify you over the course of a term — no cryptographic scheme will prevent that. If you need true anonymity for something serious — a concern about another student, a question you’re embarrassed to ask in class — write to me on paper or come to office hours. The margin is for engaging with the material, not for confidences.

Hover any paragraph on this page and you should see the + ask appear in the right margin. Click it, sign up with the join code FS26 — pick a username (private; what you’ll log in with), a password, and a display name (public; your pseudonym, like QuantumOtter) — and post something. If it works, your annotation will appear next to whatever paragraph you commented on, attributed to your display name. Welcome.