Eric Weinstein on the Application Gauge Theory to Economics (YouTube Content)

Eric Weinstein on the Application Gauge Theory to Economics
Information
Host(s)	Bob Murphy
Guest(s)	Eric Weinstein
Length	02:01:08
Release Date	5 December 2025
Links
YouTube	Watch
Portal Blog	Read
All Appearances << >>

Eric Weinstein on the Application Gauge Theory to Economics was a discussion with Eric Weinstein, hosted by Bob Murphy on InFi.

Description

Eric Weinstein joins Bob Murphy to discuss his work with Pia Malaney on a more economically relevant differential operator, which has important implications for cost-of-living adjustments. Eric Weinstein has a PhD in mathematics from Harvard University, and among other academic appointments was a Research Associate in the Harvard economics department from 1998 - 2001, where he co-founded the Science and Engineering Workforce Project (SEWP) with Richard Freeman, housed at the NBER and funded by the Sloan Foundation.

Transcript

00:00:00
Eric Weinstein: The fact is that we mathematicians pushed on the economists the wrong notion of a derivative because we didn't tell you that it was a bespoke object. You have to tell us more about your problem. You can't just grab the off-the-rack, one-size-fits-all derivative. There is no the derivative. Derivative is a choice, and you're not choosing the right derivatives for the problems that you're facing in economics, which is why, and I hate to say it this way, your entire field is off.

00:00:28
Bob Murphy: [upbeat music] Welcome to InFi, the future of finance. [upbeat music] Hey, everybody. Welcome back to the InFi podcast. I am very pleased to present the following conversation with Eric Weinstein. But allow me to spend more time giving the context of this discussion on the front end than I normally do in these intros. So in case you don't know who he is, Eric Weinstein, among other things, is a very popular podcast guest, appearing frequently on places like Joe Rogan and other enormous podcasts, and he advances ideas such as his theory of everything in terms of physics that he calls geometric unity. Now, Eric has a PhD in math from Harvard. He's been a managing director at Thiel Capital, and he was actually the person who coined the term intellectual dark web, which includes Eric, but also his brother, Bret, who's a biologist. You may have heard of him as well. Okay? But what we're talking about specifically in this episode is the work that Eric did, along with Pia Malaney, in applying what's called gauge theory to economics, and it has implications for cost of living adjustments. And, and some people ask me over the years, like, "Is this just... He's just got a different formula for computing CPI?" And no, it-it's way deeper than that. Okay? I first had a conversation with Eric more than a year ago, and at that point I realized, oh, yes, this guy knows mathematical economics better than I do, and again, economics isn't even his field, as it were. And so he and I decided at that point that, yes, I was gonna go do more research so I could be a better interviewer. Uh, but what pushed me over the edge to reach back out to him recently and say, "Hey, let's go ahead and do this," is Eric on Twitter/X had a conversation with Grok about his work in applying gauge theory to economics, and the reaction I saw there just appalled me, that a lot of professional economists just circled the wagons and their attitude was, "Pah, here comes some outsider strolling in and he's just gonna blow up our field. Okay, dude. Whatever. No, actually, Eric, we have some general impossibility results, things like Arrow's impossibility theorem and some other results from, um, more recently, like a guy, Van Velen, that show the thing you're claiming that you and Pia Malaney found you couldn't possibly have found, 'cause we've proven that it doesn't exist. So get out of here. Go try to, you know, pull the wool over somebody else's eyes, but we economists know better." When I went in and looked at the specifics they were claiming, given what I already knew about Eric's work in this area, I just thought, "No, you guys, that, that's not correct." Right? Now, here I'm not saying that the economics profession is gonna end up using gauge theory in the way that Eric and Pia Malaney suggest 30 years from now, that they're gonna be-- that it's gonna be commonplace. People will look back and say, "Ah, yes, it took a while to break through, but finally they..." I-I don't know if economists are gonna find it useful or not, but what I'm saying is the breezy dismissal that this guy doesn't know what he's talking about and everything's under control, we've investigated our field in social choice theory and price theory and whatnot and cleared ourselves of any wrongdoing, uh, no, that doesn't work. That's what I'm doing here, and so the function of this discussion, I'm not challenging Eric. What I'm doing is giving him a forum and I'm asking clarifying questions to make sure, especially the either professional economists or fans of economics who are really hip deep into this stuff, that you understand what Eric is claiming on behalf of his approach and why he thinks it's better than what the economics profession has adopted, you know, since they brought in the marginal revolution, basically. Okay? So I don't need to get into more specifics here in the intro 'cause we fully cover all that. I'm just explaining here that's what I'm trying to accomplish in this discussion, and again, I think I pulled it off. In fact, I thought I did such a good job that I [laughs] downloaded it onto my computer just to make sure in case something happened, that we didn't lose this discussion. Okay. Beyond what I've said thus far, let me also just provide more evidence that Eric knows what he's talking about. So he got his, uh, PhD in math from Harvard, and I think he got that in 1992. And then from '91 to '93, he was at the Hebrew University of Jerusalem, a postdoctoral fellow at the Racah Institute of Physics and the Einstein Institute of Mathematics. From '93 to '98, he was at MIT, where he was instructor and National Science Foundation postdoctoral fellow in the Department of Mathematics. And then from '98 to 2001, he was at the National Bureau of Economic Research, which is headquartered at Harvard University. And so there he was co-director of the Project on the Economics of Advanced Training in Science, Engineering, and Related Disciplines. He was a research associate in the Harvard Department of Economics. And with Richard Freeman, Eric helped co-launch the Science and Engineering Workforce Project, which is out of the NBER, um, which was funded by the Sloan Foundation. Okay? So And I can just say in talking with Eric, it was immediately clear to me that he wasn't bluffing. That, you know-- For example, just to give you a teaser, when we go over this in minute detail in this conversation, but Arrow of Arrow's impossibility theorem fame one time approached him and said, "What you're claiming can't be true because of my impossibility theorem." And then Eric explained to him, "Well, no, here's why. You know, what we're doing, you know, lies outside the scope of your theory." Okay? So if nothing else, if you're a listener who's familiar with things like Arrow's theorem, you can listen to Eric respond and understand, oh, here's which of Arrow's axioms or, you know, his assumptions is, isn't satisfied, and that's why what Eric is saying goes through. Last thing I'll say before we turn it over to the conversation is, my Austrian listeners might be tuning out at this point and saying, "Okay, yeah, this sounds like squabbles among mathematical economists, and we don't really have a dog in this fight. People can't be reduced to a bunch of equations, so what do we care?" And y-yeah, that's-- you could have that attitude, but I actually think there's more here, even for you. In particular, Eric's approach to, like, how do you account for the problem that, gee, if you wanna compare, like, somebody who earns fifty thousand dollars in the year nineteen ninety, what income would you-- would they need in nominal terms to be equivalent today? There's a lot of different things going on with why that's a difficult question, including stuff like there are goods and services available today that weren't available back then. So how do you deal with that? Okay? Also, in principle, even if it's the same person you're talking about, well, they're, they're different now than they were back then, so their taste would have changed. And if you try to say, "Oh, well, let's talk about somebody who's, like, the typical middle-class breadwinner from nineteen ninety to today," well, again, they're not the same in any meaningful sense. Like, so what does that even mean, right? So that's a difficult question. And on a lot of that stuff, I think the Austrians basically just punt and just say you really can't do that. For example, in The Theory of Money and Credit, Mises critiques some proposals by Irving Fisher for how you could do things like, you know, using declining marginal utility of certain goods in order to try to r-- you know, measure a util or give it operational meaning. And Mises just points out that, well, no, even the thing you're, quote, "using as a measuring rod" itself suffers from or exhibits diminishing marginal utility, so that's not a stable ruler and things like that. And so that's all true, but what Eric is proposing, I haven't seen the Austrians grapple with. Okay, so it may be that still at the end of the day, you're gonna say, yeah, in general, what you're asking is impossible. If people change, you know, you can't compare utils. But what Eric is suggesting, at least procedurally, might be the theoretically best you could do to answer questions like the one I said a minute ago. So I think Austrians even should pay attention to this stuff, and they might have a new way of, of looking at things. Okay, so that's enough preamble. On to my fascinating discussion with Eric Weinstein. Eric Weinstein, welcome to the M5 podcast.

00:09:33
Eric Weinstein: It's great to be here.

00:09:35
Bob Murphy: So there's a lot we could talk about, but I think just to set expectations for ourselves and for the audience, my task that others have asked me is, they said over the years, people have been saying to me, "Bob, you know this Eric Weinstein guy? He's got this thing on gauge theory. What the heck is that? Is that something I should care about? What is it, a big deal?" And he's got all these theories about why they were, you know, suppressing the truth and everything. So I looked into it and yes, there is definitely a lot here. So for sure, I wanna make sure the audience at least understands why, and I agree with you that this is something very important that you've uncovered here and as a way to help neoclassical mathematical economic theory for sure. But then also it's not just a mere, of mere interest to academics. There's a lot of real-world implications on this, including impact on Social Security adjustments and things like that. So there's a lot of material here, and there's various ways we can try to approach this. And also too, just like this idea of gauge theory and what is it and how you took, you know, its application in physics, for example, is what, you know, where you were using it first and then realized, "Oh, we can apply this over here in economics." So there's a lot of different angles here. I guess turn it over to you. Where do you think the best place to start is?

00:10:47
Eric Weinstein: Well, given that we're doing this at a high level, the way I look at it is that this got framed radically incorrectly.

00:10:55
Bob Murphy: Okay.

00:10:56
Eric Weinstein: And this came out in nineteen ninety-six. It had been brewing in the early nineties between Pia Malaney and myself. And it's not about index numbers, it's not about inflation, it's not about growth. It's about marginalism. And if you think about marginalism as the penetration of the differential calculus into economic theory, obviously you don't call something the marginal revolution if it's small potatoes. And any time changes in the differential calculus have penetrated economics and market thinking, you get nothing short of a revolution. So for example, stochastic differential calculus in option pricing in Black-Scholes-Merton is a re-revolution. You know, you could make case for nonlinear programming. Various aspects of differential calculus thinking have completely upended how we think about economics. So the attempt here is to upgrade all of economics to the most modern form of the differential calculus. Now, if we call it gauge theory, it sounds sexy. It goes back to Hermann Weyl calling it gauge theory, I think nineteen seventeen. But it's not anything more than a very simple idea about differential calculus. And the claim is that not only does economics need differential calculus It already is a version of gauge theory. That is, economics as it is understood today, the very concepts that underlie the heart of, let's say, Marshallian demand or theory of the consumer-

00:12:26
Bob Murphy: Yeah

00:12:26
Eric Weinstein: ... are gauge-theoretic concepts not recognized as such. And so just as, you know, Moliere's bourgeois gentleman does not know he's been speaking prose all of his life, I don't think economists understand that they've been doing gauge theory all their life and that they need to recognize that because they're often using the inappropriate choice of a derivative-

00:12:48
Bob Murphy: Mm

00:12:48
Eric Weinstein: ... in the calculus sense. And so in essence, there are multiple concepts of a derivative where we tend to teach our calculus students that you take the derivative as if there's only one, and you have to choose the correct derivative because in all of calculus, we should agree that when something is derivative equals zero, there is some notion of constancy. And in economics, there are different notions of constancy, and they come from different choices of a derivative operator. And somehow we haven't taught economists that we have this marvelous upgrade of differential calculus because we hide it away in differential geometry and in particle theory and in relativity theory where it goes under different names. Sometimes it's called the absolute differential calculus, sometimes it's called gauge theory, sometimes it's called theory of gauge potentials, derivatives. All of these things represent the modern notion of a derivative, and economists needed this a hundred years ago. And in fact, I think it is exactly a hundred years since one of them, an Algerian economist, stumbled on a primitive notion of gauge theory, confusing the theory of measurement of inflation and productivity ever since. That was a man named Francois Divisia. He didn't recognize that that's what he'd done any more than James Clerk Maxwell realized that he had invented gauge theory with Maxwell's equations way back in the eighteen hundreds. So that's sort of where the story is. And because the story debuted in nineteen ninety-six in a department where two of the five Boskin commissioners were engaged in rejiggering the entire US CPI and about half of all federal receipts, there was a war for narrative where two of us were trying to say you need a new version of the differential calculus called gauge theory, so you need a second marginal revolution. And a different collection of five people were trying to say there is exactly a one point one percent overstatement of the US CPI, which was secretly intended to transfer one trillion dollars over ten years, which was not something that the economics profession was advertising openly. This was a-

00:14:57
Bob Murphy: Mm

00:14:57
Eric Weinstein: ... sometimes is known as crimson washing, where you get Harvard economists to say something in an academic sense and solve the problem of getting ought from is pretending to do something technocratic, which is actually the bidding of politics.

00:15:12
Bob Murphy: And that's a play on whitewashing?

00:15:15
Eric Weinstein: It is.

00:15:15
Bob Murphy: Just making sure people get the joke. Okay. All right, well, right there you've given a good commercial for... So here, folks, if you're saying, "Do I wanna listen to this?" By the end of this episode, you will understand everything that Eric just said. [chuckles] So that's... You probably couldn't go prove it on your own and do theorems, but you will understand the meaning of what he just said. So yeah, let's just unpack all that. Let me just establish something here, and I know you don't like to even deal with the nitpickers and haters, but I just wanna say something at the outset because I have, in preparation for this discussion, Eric, I went through and I watched YouTube videos and people quote blowing you up and stuff like just to understand the terrain and figure out like what... And let me say, so Pia Malaney, she got her PhD in economics from Harvard. This was her dissertation. So it's not like this is some nutty thing that doesn't make any sense and that, oh geez, we have impossibility theorems proving that this can't be true, right? So I mean, obviously we can argue about the significance, but this idea that this is just some crank thing that people are spouting off in corners of the internet, I mean, that's nutty. Am I right that her advisor was Eric Maskin?

00:16:16
Eric Weinstein: Yeah. Formally. I mean, this is joint work between the two of us-

00:16:20
Bob Murphy: Uh-huh

00:16:20
Eric Weinstein: ... that Eric was very enthusiastic about.

00:16:23
Bob Murphy: Yeah.

00:16:23
Eric Weinstein: And I'm gonna, I'm gonna be more forthcoming than I usually am.

00:16:25
Bob Murphy: Okay.

00:16:26
Eric Weinstein: And then, um, Dale Jorgenson, representing the Boskin Commission, basically let it be know, known in no uncertain terms that this was not welcome in the Harvard Department. Harvard is basically the collision of sharp minds and sharp elbows, and I would say Eric Maskin originally represented the sharp minds.

00:16:42
Bob Murphy: Okay, great. And again, besides just being an economist at Harvard, he also won the Nobel Prize, right? So I mean, it's not like he's some schlub. And [chuckles] so the fact that he signed off it and she got her degree, I mean, again, this is showing that there's some meat here.

00:16:56
Eric Weinstein: It's, it's much more than... I'm not even the goal. I appreciate what you're doing.

00:16:59
Bob Murphy: Yeah.

00:17:00
Eric Weinstein: But let's assume for the moment that there's some chance that I'm wrong and some chance that I'm right. On the branch of the decision tree where I'm right, it changing out of the back end of all differential calculus to an upgraded version for all of neoclassical economics is no small claim. So if it's right, one would expect profound implications of this. And the fact that it is being discussed as whether it is anything at all has a humorous component.

00:17:29
Bob Murphy: Right. I agree with what you said, but what I'm trying to do is establish the validity-

00:17:33
Eric Weinstein: Yeah. I'm trying to say I, I think this is quite a lot bigger than Eric Maskin.

00:17:37
Bob Murphy: [chuckles] No, I know, but I'm saying that there's not some basic flaw.

00:17:40
Eric Weinstein: Bigger than game-

00:17:40
Bob Murphy: Yeah.

00:17:40
Eric Weinstein: It's bigger than game theory.

00:17:41
Bob Murphy: Yeah. Yeah. I understand the, the-

00:17:44
Eric Weinstein: The claim

00:17:44
Bob Murphy: ... implications and what you're claiming.

00:17:46
Eric Weinstein: Yeah.

00:17:46
Bob Murphy: And so-

00:17:47
Eric Weinstein: But I'm just, I'm just saying I don't want-

00:17:48
Bob Murphy: Yeah

00:17:48
Eric Weinstein: ... I don't wanna derive a dog from a tail. Eric Maskin's work in game theory is legendary, and it should be. This is larger as a claim, and the question is to adjudicate it.

00:17:58
Bob Murphy: Yep. Okay. Why don't we hit this right now 'cause this is something I definitely wanna make sure we cover is, and obviously correct me if I'm not stating this, but economists hearing this are gonna say, "Yeah, there's various indices to correct for changing prices over time." Like we're trying to say, "Hey, the purchasing power of money between the year nineteen fifty and nineteen sixty, how do we..." There's various ways, various types of indices. And yeah, the Divisia is one of them that was, yeah, proposed in the nineteen twenty-five or whatever you said. Sure. But each of them have their pros and cons, and one of the problems with that one is it's path-dependent, like just actually computed, so it takes more numbers in between, and that seems kinda counterintuitive. Like, if I just wanna know between nineteen fifty and sixty, can I just take a snapshot of various prices and quantities? And the Divisia index takes it. And so one of the results in Dr. Mullaney's dissertation that I, I found very compelling was to say, if we do make the switch that you're saying in terms of, like, the differential operator swapping out the one that economists conventionally use for this new one that she proposes, then all these various indices give the same answer.

00:19:06
Eric Weinstein: That's right. But more importantly, what I would say is the big shift-- So Divisia found sort of an echo of gauge theory, and it had the mysterious property of being path dependent, and that was thought to be a negative, and that's bedeviled the field ever since. What gauge theory shows you is why it is path dependent, why that is a positive thing. It is something that you should not try to rid the world of. Effectively, what this is, is curvature, and curvature is what generates path dependence. And so what you're seeing when people try to say, "We don't want a path-dependent measure," is that they are effectively the flat earth society of economics. They're the flat market society. And so you're trying to have a conversation saying the economy is round with a bunch of people saying, "No, it isn't. It's flat. We just can't get this thing to go away." Because if you take any mechanical index, that is anything that takes in, let's say, dual period prices and quantities, and you chain it, and you, you say, "Well, okay, originally I have a year between readings. Now I'm gonna do it every half a year, then quarterly, then every month, then every week, so I now have fifty-two readings." As you increase that frequency, you approximate usually one index, the Divisia. It doesn't matter whether you begin with a Fisher, a Tornqvist, a Posh, a Lespares, they all become the Divisia. And the reason for that has to do with the fact that it is the correct way of thinking. The path dependence is a feature. It's not a bug. You can't get rid of it. And if you don't have path dependence, it's a sign that you're not actually doing economics.

00:20:48
Bob Murphy: Okay. Right. So let me just say that again, is that for people, for especially trained economists listening to this, trying to understand, do I wanna look more into this? To me, I think that's one of the most important points that not only is this a interesting mathematical result that, oh, if we switch and use this new differential operator that is suggested by these folks, out pops this thing where all these different indices that we've been trained to use that in our thinking give different answers depending on the quantities and prices. And oh, you see, if you start with the first bundle and then go to the second one is one thing, but if you start with the second bundle and look backwards, it's a different one. But using this operator, it all gives you the same answer in terms of what happened to the price level over time. So that's interesting, but then again, you're adding that there's economic intuition to it, that actually if you think it through and understand the enterprise of what you're proposing, you realize afterward that, oh yeah, of course, path dependency would be in there. How could it not be?

00:21:44
Eric Weinstein: Exactly. I mean, if, for example, you imagine that you're a portfolio manager and you've got a portfolio on January first that's one-third bonds, two-thirds stocks. And you have two sub-managers, and you give them each half of the portfolio, and you tell them to trade off between stocks and bonds, and you say, "Well, because you guys are interns, we're gonna cover all of your transaction costs so you can transact costlessly. And all we ask is that by next January first, that you're back one-third bonds, two-thirds stocks." Do you have any sense that they should show exactly the same profit if you didn't know what they did during the course of the year, whether it was exactly the same or different, even if they returned to the same mixture on January first? There's no reason that you would expect those answers to be the same. So that's a great example where your intuition is that the skill of the manager is in fact reflected on the path, in the path dependence of measure.

00:22:44
Bob Murphy: Hmm.

00:22:45
Eric Weinstein: And the problem is that we have, and, and as you well know, the thinking around financial markets and the thinking around regular consumer markets and production markets is different. The mathematics is different. The culture of financial economists versus regular economists is different, tools, what have you. And so in a weird way, we've already accepted path dependence in the financial markets. We've accepted the curvature of the financial markets. And something I talked to Soros about years ago, he had this concept of the participating function and the cognitive function. And I said, "You know, you could take John Wheeler's statement about relativity, which is matter tells space how to curve, space tells matter how to move." And I think you could say that traders tells markets how to curve, and markets tell traders how to think. There's a way in which effectively the curvature that's used in both cases is the same mathematics. It's the mathematics of gauge theory of differential geometry and fiber bundles or covariant differentiation. All of these are equivalent ways of saying the same thing.

00:23:51
Bob Murphy: Okay. I know what you're saying. I, I wanna circle back, and I remember just as I was dipping my toe in and, like, looking at your interviews with other people or whatever on this stuff to try to see, like, what's this guy talking about? I remember you said something that when I first heard it, like made no sense to me. Now, having-- Oh, let me mention too, The Physics of Wall Street by James Weatherall does a really good job, I think, of for folks who wanna see this spelled out more and that's accessible to a lay reader, like explaining the physics stuff and, you know, giving good intuition, like what does it mean? Flat versus curved and what are you talking about and path dependence and all that stuff, and then applying it to what Dr. Mullaney and you, and you did in terms of economics. Now, I totally get this, but I remember I saw-- I forget who you were talking to. Maybe you said this multiple times. But you said the way normal economists think about price indices, you know, year nineteen fifty, if all the prices are cert-- you know, vector prices and vectors of quantities, and then look at nineteen sixty, if it just so happened that all the prices in nominal terms were identical and all the quantities were identical, most economists would say, obviously, the purchasing power of money is the same, like the price level is the same. And so we can all be sure about that. And you were saying that as an example to show that they weren't thinking about it the right way. And so at the time, I remember thinking, "What are you talking about?" Like, that's the one thing we could all agree on, right? Of course, they would have to be the same. The, the issue is when the prices and quality of the goods are different or something. And now am I, am I right in saying the reason is because there's all sorts of different paths you could take to end up at the same-

00:25:19
Eric Weinstein: Yeah

00:25:20
Bob Murphy: ... endpoint? Yeah.

00:25:21
Eric Weinstein: And in fact, if you do this in an investment, I think people, sometimes people who don't know much about economic index numbers don't realize that price and quantity index numbers are supposed to be paired, so that if you look at the change in the value of a basket between two different periods, you're supposed to be able to recover that from multiplying the price times the quantity index. And if you happen to have an investment thesis, and it just so happens that all quantities and all prices are the same after a ten-year period, so the size of the fund hasn't grown, you're in exactly the same instruments, prices are exactly where they were. The price index should show the prowess of the investment manager, and the quantity index should show the prowess of the marketer. Because in other words, if the investment manager kept losing money by buying high and selling low, but it was offset by the marketer being able to recruit new funds into the investment vehicle, you could easily have no change in the AUM, no change in the prices, no change in the objects, but the marketer should be given a massive bonus for covering for this terrible investment professional. Likewise, somebody who is killing it in the markets and buying low and selling high should be showing a fantastic performance. And if the marketer is just so offensive making tasteless jokes and chasing away investors, that person should be docked and fired. And in fact, price and quantity indexes would disaggregate the separate effects of whether one is buying high, selling low, buying low, selling high, chasing away funds, attracting funds. And so in a very clear sense, the problem goes back to Irving Fisher. Now, Irving Fisher, back in the twenties and thirties, was on a crazy quest like Ahab after a great white whale to find the perfect axioms to isolate what would be the ideal index. And strangely enough, he came up with a geometric mean that is now known as the Fisher Ideal Index, even though it is not an ideal index. So it's a little bit... It's like imaginary numbers. The marketing somehow lives on. And it was Ragnar Frisch who showed that there was no way of satisfying all of Irving Fisher's axioms. Now, that feels like a very ancient discussion, but the fact is, is that Fisher's axioms were terrible. They were very naive. They were based on bilateral comparisons and in part, the failure of the field to distinguish actually distinct mathematical problems. It feels like the same actor over two instance of time is equivalent to two different actors at the same instance of time. And this was one of the problems with Fisher's intuition. And as we will get into, this is also a huge problem in social choice theory and leads to some false intuition. But the basic problem is that the naivete of Irving Fisher's axiomatic approach doomed the field. And Ragnar Frisch was quite correct mathematically to point out that Fisher's hopes for pony cakes, bread, and circus for everyone were to be dashed. But the implication of the incident was just completely misunderstood within the field.

00:28:38
Bob Murphy: Okay, yeah. So I definitely wanna come, you know, let-- explain a little bit more, people understand more about what you're proposing, then we can kinda come back. But yes, part of the backdrop, folks, here is there are a bunch of standard results, Kenneth Arrow's so-called impossibility theorem, more recently, Van Velen, who I'm not as familiar with, but I've looked around and people say, "Oh, yes, he showed the impossibility." That proposing postulates of, oh, i-you know, ideally a good way of gauging the purchasing power, either among different countries, different currencies, or the same country over time would satisfy these postulates, and then showing that no measure, one measure could satisfy all of them simultaneously. So am I right, Eric, in saying so there's this pessimism among economists thinking that, yeah, we're never gonna have perfection, so we might as well just grope around and do whatever we have the data for. And you're saying, "No, you guys are missing it. If you think about it the way I'm thinking about it, we at least could theoretically imagine perfection," and then, you know, there could be data limitations, but-

00:29:35
Eric Weinstein: Well, it's a great question, Bob. Let, let me take a stab at it. Very often, you have a naive hope, and your naive hope is based on bad intuition. And then something remarkable happens where you're not expecting something to occur. For example, there's a famous thing that happened in the late fifties that taught physicists that they didn't understand the basics of electromagnetism. Now, nobody had this on their radar screen that physicists could be confused about something as basic as E and M. But what they found is that if you took a current in a solenoid and you shielded it so that there was no electric or magnetic field outside of the solenoid, the question became, could you figure out whether current was flowing through the solenoid by doing something outside of the system? And it turns out if you pass an electron beam all around the solenoid, and then you watch the interference pattern, you can see the effect of turning the current on through the solenoid. Now, that didn't make any sense to anybody because the electric and magnetic fields were essentially zero. So the thought was, well, how can you possibly have a zero electromagnetic field interfering with something as concrete as the interference pattern of an electron beam? And it turned out that was because it's not the electric magnetic fields that are primary. It's this more primitive thing called the gauge potential. And that gauge potential, which had been thought to be a mathematical artifact, convenience product, if you will, for figuring out the E and B fields, the electric and magnetic fields, turned out to be the star of the show. And the thing that measured its failure to give a sane answer was the Faraday tensor. So in other words, sometimes when you have a hope for something like I hope that the current in a shielded solenoid has no effect on the electrons outside of it, and you find out that that's wrong, your next instinct should be, let me measure how badly my naive hope crashed and burned. And then you take that object, which would be like hope versus hope minus reality or reality minus hope. You turn it into a mathematical gadget, and you make that new gadget the star of your show. And that's exactly what economics missed as an opportunity. You should have taken the failure of the chained indices to behave as you expected in a Fisherian sort of a way. And instead of just saying that there's an impossibility theorem, you should be constructive about it. You should say, what did we hope for? What did we get? What's the difference? What's the ratio? What object measures how wrong we were, which economists don't like doing. And then make that the centerpiece of the story. You would have invented curved markets. Devisia would have become the James Clerk Maxwell. And whoever did that, rather than it being Milani and myself, would end up as the analogs of the Aronoff-Bohm effect. Right now, you guys call it the cycling problem. It's exactly the same effect as the Aronoff-Bohm effect. You just don't recognize it yet. You think it's an artifact, it's a mistake, it's something you want to get rid of. It's the star of the show.

00:32:47
Bob Murphy: Okay. At this point, unless you disagree, can you explain a little bit more about what's-

00:32:54
Eric Weinstein: What's wrong?

00:32:55
Bob Murphy: Do you mean the differential-

00:32:57
Eric Weinstein: Yeah

00:32:57
Bob Murphy: ... like the operator and like what economists are doing right now versus like what you're proposing they should use instead?

00:33:02
Eric Weinstein: Sure. So let's first of all just get the feel of why there's an issue. So I hope that because economists typically have like a master's degree or sometimes even more in mathematics, the following should be pretty clear. Let us imagine that we're going to take as our definition of constant. Something is constant if there's a derivative of it that is equal to zero. Typical marginal thinking.

00:33:27
Bob Murphy: Mm-hmm.

00:33:29
Eric Weinstein: Now, let's notice that sometimes you have two different notions of constant. So imagine that you are my employee. If I pay you a constant dollar salary, that is one notion of constant. And if we differentiate your salary over time, we spread it out so that it's a stream, you will see that the ordinary derivative detects that you have a constant salary. But what happens if instead you're facing exponentially rising prices in a hyperinflationary economy? You have a second notion of constant, which is constant purchasing power. So into this way of thinking, there should be a second derivative that detects that an exponential salary denoted in the numerair of your currency, whatever it is, should be seen as constant in a more meaningful economic sense. And the question is, well, what is that derivative? So you have to imagine that, well, it's probably your original derivative with some zeroth order term, which I think is something like the derivative of the logarithm of the rising price.

00:34:34
Bob Murphy: Yeah.

00:34:35
Eric Weinstein: And you subtract that off, and then that new derivative operator detects an exponential as a constant purchasing power salary. So now we have two different derivatives, and that shouldn't be very controversial. That should be people could say, "Okay, I can see that would be how a mathematician thinks about such things." It turns out that all derivative operators are pretty much the same at the level of the first order differentiation, and they differ in some sort of multiplicative factor. The way in which you can think about this, if a derivative is thought of as rise over run, that's how we teach this in ordinary calculus. That's how my father was taught at Central High in Philadelphia way back in the '50s. You know? So what was going on back then, he didn't have the following adjustment, rise above reference over run. So it's rise over run versus rise above reference over run. There's an implied reference level, which is that there is a horizontal that you're measuring the rise from.

00:35:46
Bob Murphy: Mm-hmm.

00:35:46
Eric Weinstein: But what if the rise that you should be measuring from is not the horizontal level, which is the amount of money you're taking in. What if it's the price level that you're facing as a consumer when you take home your paycheck. So if I try to defraud you and say, "Look, you're getting a constant salary. Why are you complaining more and more every week? I've been to college, and I can show you that it's constant because the derivative of your salary is zero." And you come back and you say, "Well, actually, I have this gauge theorist friend who's been advising me, and he says you're using the wrong derivative. Let me show you the right derivative to be using, and I'll show you that my salary is negative. It's declining. It's not constant in any meaningful sense." So The fact is, is that we mathematicians pushed on the economists the wrong notion of a derivative because we didn't tell you that it was a bespoke object. You have to tell us more about your problem. You can't just grab the off-the-rack one-size-fits-all derivative. There is no the derivative. Derivative is a choice, and you're not choosing the right derivatives for the problems that you're facing in economics, which is why, and I hate to say it this way, your entire field is off.

00:36:57
Bob Murphy: Okay, great. Yeah. So I think people-- I mean, that's a good way to start and, you know, make-- get people warmed up to the idea that, okay, yeah, the notion of constancy there, it depends what the frame is. So this part, maybe you can err on the side of being more technical, then we can try to give intuition. But so what, what are you proposing? O-okay, so we don't use, you know, what, d of f over dot equals zero. What, what are you proposing instead that we use?

00:37:23
Eric Weinstein: Well, what I propose is that you actually [chuckles] you have all this marvelous gadgetry that is really lying around that you don't think of in the ways that we do, and you can decide that this is technical fluff and complicated things. But once you see it our way, I don't think you'll ever go back. What is the right notion of constancy, right? So we have economic notions of constancy, and we have words that no mathematician has ever heard of unless they hang out in econ departments or have an economist in their family. One of which is called indifference. Another notion of constancy is called substitution. So if you think about income versus substitution effects, substitution is the notion of instantaneous constancy. Indifference in utility space, right? So the idea is welfare is orthogonal in a certain sense to indifference. And this division is what mathematicians call the division into vertical and horizontal subspaces. You can do this in the theory of, for example, evolution, where you could say fitness is akin to growth, is akin to the vertical. And you could say drift is akin to indifference or substitution is akin to the horizontal. That horizontal is the reference level that you have to put into the linguistic formulation of the derivative, rise above reference overrun. And the key idea is something like budget constraints or substitution effects change with the price-- with the collection of prices. Now, technically, if you consider a basket of goods to be a vector in a vector space, the prices are not a vector in the same space. If you had a column vector of prices, you would have a-- I'm sorry, a column vector of quantities, you'd have a row vector of prices, and then you would multiply the row vector against the column vector to get the value. That row vector we would technically call a covector or a dual vector. And that dual vector, as prices move, changes the notion at every instant as to what one means by substitution. And where is the income expansion path in ordinary consumer demand? It is at the exact point where instantaneous indifference, that is linearized indifference of the indifferent surface, exactly coincides with the budget constraint leading to substitution effects. So that point of tangency where you have a linear gadget, which is generated by the covector, is exactly coincident with the differential to-topological or differential geometric concept of the tangent space to the indifferent surface. And so because that is constantly moving, in a certain sense, your derivative is not a static target. It's constantly moving. The notion of economic constancy built into the covector of prices is constantly updating. And the way in which we handle that is they say, okay, let's consider all possible pairs of vectors and covectors, all column vectors from the positive orthant, all covectors from its positive orthant. And so now you're in a space of two end dimensions if they're in goods and services to begin with. And what you do is you say that an economic history is a path of pairs in that Cartesian product. And that object allows you to deal with one overarching derivative, which takes into account the right notion of economic constancy at every pair of points. And so this is an absolutely gorgeous and somewhat standard construction in mathematics from which you get the Divisia index. So you-- the first thing you do with the Divisia index is that you get the intuition about why you're-- why have we been fighting for a hundred years this supposed cycling problem, literally called the cycling problem, when it is nothing of the kind. So you may not get a new index in terms of the Divisia, but you are getting new intuition as to why it's there. It's the fact that you did the economics right. If you did the economics wrong, you could get rid of the cycling problem. The next problem you have is imagine that you have bilateral trade. So for example, you have Erikastan and Bobvakia, okay? We trade with each other. My imports are your exports and conversely. We have two different currencies. You use dinars, and I use lira, okay? Imagine that we have a failure of the law of one price and that your covector and my covector for the same goods and services are not linearly coincident. How do we figure out whether our trade has grown or shrunk if we can't even agree on what the price of goods are? The reason we're trading in part is in order to arbitrage away the vastly different prices in our two economies. In that case, what you have is that you have an N minus two-dimensional space upon which we agree in terms of our prices, and we have a slight amount of discrepancy And so you have to talk about mutually indifferent effects, mutual substitution effects. What you get is you end up with a two-by-two matrix that your field doesn't even know exists. And what's more, if you look at the analog of the Devisia formula, there's a new object in it you've never seen before in your lives called the time-ordered product due to Freeman Dyson. It's not a simple exponential the way the Devisia index is an exponential. It's in fact an integral over simplices, which if you want to make it look as if it's an exponential, you need this time-ordered product. And why is that? Because two-by-two matrices, unlike one-by-one matrices, are not commutative. They're not abelian, as we say in group theory. So it sounds a little pretentious, but the Devisia index that you know and love is really valued in GL one R, one-by-one matrices of determinant not equal to zero. And the reason that sounds pretentious is because it's just a positive number that you should be getting. Okay, but it's the wrong intuition. For bilateral trade, immediately that jumps to two-by-two matrices. So you're in GL two R. And the difference between GL one R and GL two R is night and day. You can't get away-- There's no way Francois Devisia could have stumbled into the Devisia index analog for trade. So you don't even have this in your literature. And for some reason, there's this like really fierce streak of anti-intellectualism inside of economics so that if something isn't blessed by Harvard, Princeton, Stanford, or Chicago, you guys will just sit there not realizing that you don't even have the formulas above the Devisia index. In the case of the Devisia index, you have a GL one R valued gadget, so you know the formula and you're just, you just don't have the intuition because you think you have a cycling problem. So once you get that, you move up the food chain, then you get to trade, you see a new formula you've never seen before in your lives, which has time-ordered products because you have a non-abelian object, and you have real group theory, which essentially doesn't exist. Like it's very strange that groups are almost non-existent in standard economics. They're the bread and butter of mathematics. You guys know a ton of mathematics. You look like you're proving mathematical theorems, but you basically do analysis. You do very little geometry topology, and you do almost no group theory. And one level above this, you're in infinite dimensions. And not only are you non-abelian, where A times B is not equal to B times A, but you're sort of in a realm that's much closer to quantum field theory. In fact, you're some-- closer to something called string theory because their, their main group is called Diff S one, the self-maps of a circle. And your main group is actually the reparameterizations of a circle fixing a point, i.e. the positive reals because what you-- your entire field is built around something called a principal bundle. You just don't know it. Now, what is a principal bundle? If you think about ordinal theory, you take all cardinal utility functions and all ordinal utility functions that are relevant for a standard problem. So let's imagine that the indifferent surfaces are complete and convex to the origin. So you always need a little bit of every commodity. You don't want zero. You don't run into the edges. There are no bliss points. Let's just get rid of all the funky stuff that can go wrong. And imagine that all the cardinal utility functions are surjective onto the positive reals. So they map the positive orthant to all of the positive real numbers going out to infinity, and they're increasing. You can act on that set, that topological space by reparameterizing the real line. In other words, if you have a cardinal utility function, we call it capital C, taking the positive orthant to R plus, and you reparameterize R plus, and you compose the two, you get a second map, C prime, from the positive orthant to all of the positive reals. And in fact, every single cardinal utility function that generates the same ordinal map can be gotten as what we would call the orbit of that group action. That group action also is guaranteed to move every single cardinal utility function. Nothing avoids being hit and moved. And that's what we would call a free action of a group on a topological space. And the quotient of that action, if you replace all the cardinal utility functions that generate the same ordinal map with that ordinal map, that is the quotient of a group action. So in other words, and again, this won't mean anything, and people will say, "Oh my God, this is just mumbledy, mumbledy gook," but when you find out what it is, you could-- you, you'll thank me later. Your entire field is based around a principal bundle between cardinal utility theory [chuckles] and ordinal utility theory that nobody noticed. It's like finding a monolith on the moon.

00:47:43
Bob Murphy: Okay. I definitely want to get into that and unpack that more, the whole cardinal ordinal thing. Let me circle back a bit. I'm gonna s-say some statements. You tell me if it's just totally wrong or if it's, yep, nailed it, or yeah, I get what you're saying and it's fuzzy. So I think the way standard economists think about a price index, if all-- if you had a vector of all the prices for the commod-- Let's assume the commodities all stay the, you know, the set of the commodities stays the same over time. If the vector of the prices from the-- and just make it discrete, from, you know, T one to T two is the same, identical, they would say, "Yep, that's constant. The purchasing power of money is the same constant price level." Are you saying, no, what you need to actually know is for a given income and prices that the welfare of the representative household from T one to T two is what's held constant?

00:48:40
Eric Weinstein: So There is, so first of all, there's a major-- Let's back up and talk a little bit about index numbers as the economists see it, and then we'll get to how Dr. Malani and I see it, which is different.

00:48:53
Bob Murphy: Okay.

00:48:54
Eric Weinstein: Yeah. So the first thing to say about the field, the field thinks that index numbers are a backwater, which is hysterically funny to begin with. It's anything but because anything in the theory of e-index numbers changes everything it touches given how many things are indexed to indicators. So it's prima facie ridiculous to say that this is a backwater. It's-- This is the most beautiful stuff in economics I think I've ever seen. It's right at the heart of the theory of the consumer production functions, what have you. The next thing is they would say that there's a division into mechanical versus economic indices. So in mechanical indices, you take in two kinds of data, price and quantity datas, and then you come up with numbers, and then you mumble things about price levels or changes in production or whatever. And this is completely the wrong way to think. So that's-- we'll get to how-- the right way to think or why I think it's the wrong way to think.

00:49:50
Bob Murphy: Mm.

00:49:50
Eric Weinstein: But the idea that there is a price level is the problem. If I say to you, "What is the temperature in America right now?" You will not understand what I'm saying.

00:50:02
Bob Murphy: Right.

00:50:03
Eric Weinstein: Right? But if I say to you, "What is-- what do you think CPI is gonna be next month?" You're dumb enough to say, "Oh, I think it's going to be..." and then you give a number.

00:50:12
Bob Murphy: Can I stop you for one second, Eric? Just so you know, I'm giving you-- you, the Austrian libertarian types who are listening, when you were going on with the group stuff, they were like, "What the heck is this?" But now you just want them back 'cause they-- that's a standard Austrian point.

00:50:24
Eric Weinstein: Well, wait a second. I wanna say something to my friend.

00:50:27
Bob Murphy: [laughs]

00:50:27
Eric Weinstein: First of all, I'm thrilled to be on a libertarian podcast given all the negative things I've said about libertarianism. I will say one thing about you guys is that you're intellectually pretty honest, and I love that about you, that you will have somebody on who radically disagrees. Good on you for that. And the next thing is you guys tend to be a little bit more aggressive against mathematics because I think you're resistant to the overformalization of a theory before its time. Like mathify it if and when it's warranted, but don't mathify it early. I've always had some sympathies, even as a math guy, with that Austrian perspective. The map is not the territory, and a lot of what you see in economics is that the territory is the real world. The cartography of mapmaking is the province of the economists, and they get carried away forgetting, for example, it's like a m-- if a mapmaker forgets that there's a real planet, the real planet is not the geoid, which is computed as a smooth surface based on gravity wells. The geoid is not a sphere. There's so many layers of distortion, and I think that the Austrians do a better job of saying, "Let's stay our hands and not screw everything up by mathifying too early." So-

00:51:43
Bob Murphy: We've just been hurt before by mathematical economists, and so we're very protective, and we're just, we're cautious.

00:51:48
Eric Weinstein: Look, we've all had bad relationships in the past, and our exes are somehow hanging around like ghosts. Yeah, I believe that... Okay, sorry. We derailed on Austrian and libertarian economics. You were saying I won them back with?

00:52:00
Bob Murphy: But by saying there's no such thing as the price level, and given that analogy of the temperature, everyone would see how goofy that is, and yet, yeah, we talk about, you know, "Oh, CPI under Biden did this, but under Trump it did this," and when like, "What are you talking about?" So.

00:52:12
Eric Weinstein: Well, this is-- R-right. So the idea is that if you say that the elevation in California is, and then you give a number, is that an average weighted by area? Is it weighted by popula-- like what, what are you even saying, man? And so in part, the problem is that for economists to wield power, they want to say that they are druids that can compute the price level. And the short answer is anybody-- Look, if you hear somebody talking to you about computing the price level, clamp your hand over your wallet, and if it's possible, run. They-- That's not what it does. It-- That's not what these ga-gadgets are. You have to understand what you're measuring. And what I want to see is I want to see the government being honest about making maps of inflation. Not just maps geographically because we face different prices in different locations, but because our preferences are different between people. Our preferences are different between age groups. They change over time. Information comes in, and we, we're not the same people we were. We don't want to wear bell-bottoms and have big hair, you know, necessarily. Whatever it is, we're changing, we're dynamic, and somehow we're stuck with this unwanted priestly class of druids who tell us about homo economicus, which doesn't exist, and markets that don't exist, and flat Earth soci... Roughly speaking, I, I've been an economist. I've been in economic departments. We, we've got to recognize that the field has to grow up, and a bunch of petulant people who've been told that the style of an-analysis that they do as mathematical economics is valuable, it's gonna be a bitter pill to find out, look, you're not even good mapmakers. At least understand that you're dealing-- you're trying to map a curved territory.

00:54:09
Bob Murphy: Let me say it right now because as you're going through, I, I realize formally, we haven't officially said this during this discussion so far, Eric, that part of what you're doing and why I understood that all the scope and the audacity of what you're claiming that you're giving us here is that your approach or your framework can easily handle changing preferences over time. Whereas for standard theory, that's, it's kinda like, you know, just, well, yeah, that's why you can't do anything. You know, like from an Austrian point of view, just remain real agnostic and just say, "Yeah, that's why these mathematical economists are nuts 'cause preferences change. We all know that. And so look at their goofy models where they have to assume they stay static, otherwise they can't use calculus." And so you're trying to

00:54:51
Eric Weinstein: Yeah

00:54:51
Bob Murphy: Like combine the best of both or-

00:54:53
Eric Weinstein: I, I sense your-- Yeah. Let me make an analogy. Economics, professional economics, particularly macroeconomics, is like a Death Star from Star Wars. And this thing pulls up, traditionally USAID and the CIA and the State Department dispatch a bunch of economists from Harvard or Stanford or Chicago. Could be the Chicago boys to Latin America, could be Andre Schleifer and Jeff Sachs from Harvard, could be anyone. And we tell some country what they need to do, and sometimes we do things through the IMF or the World Bank, and join or die, and that's how we roll. And my claim is your Death Star has an exhaust vent. And if you don't play nice, and if you continue to pretend that you have a subject that's figured it all out and that you guys have ascended to some mountain, I can show you that your entire Death Star can be killed off by changing preferences. It's very clear in the literature. I think I sent you a great deal of excerpts and quotes that point out that what's happened is, is that more or less Gary Becker taped over the exhaust vent vulnerability with like no exhaust vent here written in red crayon, and said, "Preferences are perfectly stable. They don't differ between people, and they don't differ over time." And he got Stigler to go along with that one. And that is why, in part, we rely on the Chicago School to tell the ultimate lie that no one believes so that the field doesn't collapse. Because once you understand that changing preferences are the Achilles' heel of all neoclassical economics, the, it, the field does not exist in this, in its presence. And that is why from the time of about Alfred Marshall, people have increasingly become fanatical about asserting that somehow preferences are fixed and stable, when it is abundantly clear that there is no backing for this. This is a cult-like belief in a, in an assertion about human beings that is invalidated by everything we know about advertising addiction. You have these crazy articles like rational addiction, where people are being completely reasonable in their heroin habits, and they're not affected by advertising, or they have meta-stable-- me- tastes that are stable in a meta sense. It's such nonsense. You have to recognize that we can fix the vulnerability of the Death Star, but you're gonna have to play nice because I don't want you guys rolling up and destroying the world the way you have been. Or you can feel free to take a swipe and claim that you don't have this vulnerability, and we'll just send a couple of torpedoes right down the exhaust vent and watch the entire neoclassical edifice disappear.

00:57:47
Bob Murphy: Let me, Eric, and presumably I'm gonna s-speak on behalf of the mathematical economists. They're gonna respond the same way in Star Wars when the one lower guy goes up to the admiral or something and says, "Sir, we've analyzed our attack pattern and there is a danger." And he's like, "What? At our time of victory, we won't evacuate." [chuckles] So let me, um-

00:58:05
Eric Weinstein: Can we read some of these quotes, Bob?

00:58:07
Bob Murphy: Yes. Well, let me first-- I mean, I pulled up the Stigler and Becker's paper just for the folks at home, 'cause I was telling you, Eric, before, you know, over email, I reread this just to get ready for talking to you, and I had forgotten how nutty this paper was. It's a classic paper, folks, in the, you know, was it '77? George Stigler, Gary Becker, giants in this field. De gustibus non est disputandum. And they say, um, there's no point arguing o-over taste. So they say the way this is used, the, the way they're starting their paper, the way this is conventionally used is to mean like, hey, once you reduce the differences between people down to their tastes, stop arguing because there's no arguing over, over taste or preferences. And they said, "But we're gonna flip it. We mean it the other way around." And that they say, "Our title seems to us to be capable of another and preferable interpretation that tastes neither change capriciously nor differ importantly between people. On this interpretation, one does not argue over taste for the same reason that one does not argue over the Rocky Mountains. Both are there, will be there next year too, and are the same to all men."

00:59:06
Eric Weinstein: All men.

00:59:07
Bob Murphy: And that's not hyperbole. Like, that's literally what they're doing in that paper, folks. They're just trying to say, it's not just that my tastes today are the same as twenty years from now, and if I develop, you know, an affinity for Beethoven because I listen to classical music every day thinking, "I really ought to do that. I want to be that kind of guy," they're saying, well, basically, your underlying preferences were the same the whole time, and it's just that experience along the way changed the outcome or the behavior we see, but basically you're the same guy. They're not just merely saying that. They're also arguing that Eric and I have the same taste right now, and the reason he's got a quote on and I don't, well, if we looked at the past and the prices and the income and blah, blah, blah, then that would explain it. It's crazy.

00:59:46
Eric Weinstein: You're into, you're into Beethoven, I'm into Darby Crash and the Cramps. It doesn't really matter. Chicago has this brilliance that the rest of us are just not clever enough to understand because both Stigler and Becker were giants of the field. And by the way, they did a lot of good work, but they also lied. This is not a serious paper. And I don't know how to say this politely, but this is like cheating at touch football. When you can't admit the truth, which is, guys, if we don't do something about this vulnerability, we may not have a subject. The reason that they're lying is something I agree with. They understand, and I talked to Becker about this. They understand what's at stake. Becker also gave a definition of the economic method, which perhaps you might share because many of your listeners will be familiar with it, some may not be. But you'll notice, I think it's twenty-two words, and it's a very strange tripartite breakdown. Do you have it?

01:00:47
Bob Murphy: Yes. Okay, so from nineteen seventy-six, Becker says And sorry folks, I'm getting over a cold. My voice is coming and going here. The combined assumptions of maximizing behavior, market equilibrium, and stable preferences used relentlessly and unflinchingly form the heart of the economic approach

01:01:06
Eric Weinstein: Spoken like a guy who knows exactly where the body is buried. Stable tastes in, in the top three things. If you don't have stable tastes, you can't evaluate anything. You can't ask the question, are you better off now than you were four years ago? You can't do revealed preference over time. So the whole thing hinges on a lie, and that's why this is controversial. So when you see me being attacked, remember what's really going on. So what's going on is, is that this problem got solved in the nineteen seventies once and for all by Gary Becker, and Becker together with Stigler. You can tell from reading Memoirs of an Unregulated Economist that this is not predominantly coming from Stigler. He's a much more reasonable fellow. This was Gary Becker claiming that somehow the air in Chicago had a particular feel to it that allowed people to think thoughts that the rest of us could just not even contemplate, like the fact that your tastes and mine are the same. We could be gay, straight, male, female, children, old people, it doesn't matter. Our tastes are exactly the same. It's preposterous.

01:02:13
Bob Murphy: Yeah. And even some of the other stuff too, like Becker famously, folks, if you're not familiar, like was accused of what people call economic imperialism. So like he would come in and like, "Oh, the economics of the family. Well, that's just obviously, you know, income and budget constraints and da, da, da, da, there you go." And you know, how do you pick your mate, and how do you do this? And the economics of crime, and it's just a matter of incentives and da, da, da. Things where, you know, like people can understand, oh, you go into the grocery store and if there's a sale on Pepsi, people who normally buy Coke might buy more Pepsi and whatever. That doesn't offend them too much. But when you say if the police change, like, I don't know, they do three strikes you're out, and then all of a sudden that means people have with two felonies already have an, now have an incentive to murder everybody when they do something because they can't leave witnesses or some- like a lot of people requ- or to say if the government starts paying mothers more if they aren't married, don't be surprised when you see the rise of illegitimacy, that sort of thing. A lot of people are just offended by that. So on the one hand it's like, yeah, yeah, the economic approach and incentives, but a- again, in this, the light of what I think we're talking about here, Eric, is he, he's viewing it as like everybody is basically the same and it's just your circumstances make you behave differently.

01:03:18
Eric Weinstein: Again, he gives it away. He's the unreliable narrator who's secretly confessing to the crime. He says, "Used relentlessly and unflinchingly." And there's a case that where he-

01:03:29
Bob Murphy: Yeah, that is a weird-- I would never write that about something I was doing. Why would he say that? [laughs]

01:03:33
Eric Weinstein: Because he ascended to this, to the top of this pyramid, which was-

01:03:39
Bob Murphy: Mm.

01:03:40
Eric Weinstein: We are going-- It, it-- I think it's also not appreciated that our economics departments, and particularly our top economics departments, were heavily integrated into the State Department, CIA, USAID, and were a tool during the Cold War for the projection of market supremacy over central planning. Now, I happen to be very partial to markets over central planning. I'm not a, I'm not a person who likes communism. But my belief is that we have a product that you don't have to lie about in order to convince people it's better. And instead, what we did is that we chose a path where we imputed mystique into a tiny group of people, some of whom deserved it. I mean, I think we'll get to Ken Arrow and Paul Samuelson, both of whom I was lucky enough to interact with. And I really do think that we have giants in the field. So it's not that I'm negative about all economists. But when it comes to Becker, what Becker did was to try to come up with ingenious ways of covering up the holes in the argument that economics could be an imperialist subject. You have to remember that economists delighted in sticking it to people who couldn't understand that markets were everything. And there's a paper called Economic Imperialism that is for it. It's just, it's going to be open about it. The idea that you can go places where no one else can go. It's very similar to evolutionary theorists. They can think about things-

01:05:09
Bob Murphy: Right. Right. Mm

01:05:09
Eric Weinstein: ... that no one else can think about, and then you treat everyone else as weak because they just don't have the guts to see life as it actually is. And I don't think that's true. I think that there was a time when Gary Becker needed to be rewarded for his insights, and there was also a time that he needed a juice box and a timeout rather than a, than a medal from Sweden.

01:05:30
Bob Murphy: Well, the thing is, you're basically Becker. It's just the incentives he faced were different, and that's why he did what he did, so.

01:05:37
Eric Weinstein: Well, the point was is that he was acting in part on behalf of the United States that had not yet defeated a Soviet menace. And while we can argue about whether or not this was the right way to defeat the Soviet menace, I am reasonably certain that having won the Cold War, it was incumbent upon those who had set up the Chicago School of Economics to debrief the rest of us about the fact that it was not a pure academic movement. It was in part deliberately set up to project through academic sheepskin, "Look, hey, we're just looking at the facts, and we're not pushing an ideology. We're not trying to play geopolitics. This is the benefit of our best minds." And that wasn't true. This was hardball Cold War stuff that didn't get wound down. Now we are thirty-five years approximately into the zombified Cold Warriors. It's very much like the Japanese in the Pacific who weren't told World War II was over until many years later. And a lot of the Chicago professors don't know that they're part of a State Department CIA-USAID funded project.

01:06:47
Bob Murphy: Okay. The conversation is drifting in a good way, and I wanna do focus on this stuff, but I just don't wanna lose the thread

01:06:55
Eric Weinstein: You drive, Bob

01:06:56
Bob Murphy: For people to understand the connection here, is this a true statement that what you're proposing in terms of the differential operator incorporates indifference-- Like that's how you can s-- Like it matters between, oh, between nineteen fifty and nineteen sixty you need to know moving along because you gotta know bundles that are the, the market at any moment in time consider interchangeable with another, and that's the linkage. So it, it brings in preferences and market prices. I think that's part... I'm partly saying this to excite Austrians to realize there's something cool there where-

01:07:29
Eric Weinstein: Okay

01:07:29
Bob Murphy: ... your approach takes in subjective preferences and market prices and like gives more information-

01:07:34
Eric Weinstein: Yeah

01:07:34
Bob Murphy: ... than just the static price and quantity vectors from nineteen fifty and nineteen sixty snapshots.

01:07:41
Eric Weinstein: Yeah. So even though I consider myself an intellectual lefty who cannot stand the modern Democratic Party, I am also a hyper-individualist. I view Ayn Rand as a collectivist because she needed the objectivists in order to push her point, and I think that she should have gone it alone. Anyway, so I love the Austrian emphasis on the individual. What I would say is that these are three different tiers. So at the lowest level, you have GL one R, where you're trying to find a single number to represent price and quantity indices, which is the Devisius story. And what we're adding that's new here is an explanation of why the unwanted aspects of it are in fact desirable, positive, should not be gotten rid of, and why they're there. One level up from that, you have bilateral trade. We're showing you that, in fact, there were simplifications in the one-by-one matrix story that aren't true, and the two-by-two matrix stories. Everything is still finite dimensional, but you're learning about group theory and non-abelian structures, perturbation theory, and the Dyson series, et cetera, et cetera. And then one level above that, you're in the infinite dimensional field theory space. And in that space, I wanna talk about what it looks like from a level of calculus. Not to mystify it, but to sort of give an idea. Imagine that you're looking at some analog of the X, Y, Z space where you're looking at a function of two variables.

01:09:04
Bob Murphy: Hmm.

01:09:05
Eric Weinstein: So you can picture in your mind some sort of a surface floating above the X, Y plane at a height Z, where the Z changes with the two coordinates X and Y. That is a place you can do calculus. The modern idea where you do calculus is in something called a fiber bundle, and in particular, we're gonna talk about a principle bundle. The base space is the analog of the X, Y plane. Okay? So whatever the X, Y plane is, we're gonna rename it the base space. And in the theory of utility, we're gonna call that the space of ordinal utility function. So we're going to-- What we would say is you're gonna foliate the positive orthant with hyper surfaces convex to the origin with complete leaves. But it just means come up with a nice preference map and don't get cute with bliss points and terminating the axes. Above that, you're gonna have the full XY. So if, if the space of all such preferences, ordinal preferences, is the analog of the XY plane, the XYZ space will be the space of cardinal utility functions. And the analog of the translates of the Z-axis are going to be all the cardinal functions that share a single ordinal map.

01:10:19
Bob Murphy: Mm-hmm.

01:10:20
Eric Weinstein: Okay? Now, it turns out that there's a God-given function on this, in this XYZ analog if you also include prices, and you can call that the Samuelson shepherd cost function. What it says is, given a vector, really a covector of prices, tell me the cost of the minimally priced basket on every indifferent surface, and we'll let that number represent an as if utility level. Okay?

01:10:52
Bob Murphy: Mm-hmm.

01:10:52
Eric Weinstein: If that makes sense. So in other words, you have an ordinal map, a covector of prices. You find the point which is the least expensive at every level of utility. If you get the price, and then you say, imagine that that was that number of cardinal utils for somebody's crazy utility function.

01:11:10
Bob Murphy: Okay. I-is that equivalent to saying draw a line for all the bundles among which you're indifferent, and then for each such one, what's the least cost? What's the minimum amount of money to buy a, a bundle in that's-

01:11:21
Eric Weinstein: Yeah

01:11:21
Bob Murphy: ... each of those sets? Okay.

01:11:22
Eric Weinstein: Figure out the angle curve or the income expansion path, whatever you wanna call it.

01:11:26
Bob Murphy: Yeah.

01:11:26
Eric Weinstein: And then ask what the price is at that point and assign that number to every point on that indifference.

01:11:32
Bob Murphy: Yep. Okay.

01:11:33
Eric Weinstein: So that thing we would call a section of a principle bundle. Okay? So it's just a function. It's a function that it's like something, you know, sine of X plus two Y times Y equals Z. That would be a function. This is some gadget that assigns to every ordinal preference map with a price covector a cardinal utility function which recovers that map as its ordinal substructure. So that would be called the Samuelson shepherd section. Okay?

01:12:06
Bob Murphy: Yep.

01:12:07
Eric Weinstein: That thing is not constant in the natural derivative. There is no known derivative looked at in this way, the analog of the XYZ space. If you take all choices of a single cardinal representative and try to figure what the derivative is, there's no derivative operator in the absence of prices. So when you only have ordinal preferences, you're really in social choice theory.

01:12:31
Bob Murphy: Mm-hmm.

01:12:32
Eric Weinstein: You're not in market theory. So you have to add prices, and then you get a distinguished differential operator. And then you can ask yourself, okay, well, we have one function that is God-given to us, the Samuelson shepherd cost function, and the claim that I make with Dr. Mulani is that there is a derivative operator that's economic, and that says That you can only move as you're moving around preference spaces in ways that the cardinal representative is not changing at the point of the income expansion path, right? So in other words, you have a notion that if-- imagine your ordinal preferences stayed the same, but you were manic depressive. I think Herb Gintis and Amartya Sen both have talked about this. They talk about man's efficiency as a pleasure machine. So imagine that your relative assignments don't change, but one day you're happy with your two cups of coffee and one cup of tea, and the next day you're miserable with the same-

01:13:31
Bob Murphy: Mm-hmm.

01:13:31
Eric Weinstein: Exact bundle, even though you don't shift your consumption pattern.

01:13:35
Bob Murphy: Right.

01:13:36
Eric Weinstein: So that would be called a purely vertical change. Man's efficiency as a pleasure machine goes back to that vertical horizontal distinction. So then the question is, what's the horizontal movement? And the horizontal movement is moving by a cardinal function that is not moving in its cardinality at the point of the income expansion path. So that notion exists only when there are income expansion paths, and for that you need prices. So there is a derivative operator. There are no constant functions under that derivative operator, interestingly enough, which is something that happens with these fancy derivatives that don't happen with the derivatives that we teach in college. But there is always a constant function over a particular time path that has to do with existence results for ordinary differential equations. And so when you use any constant function over the same ordinal path in preference price space, it doesn't matter which cardinal representative you choose as long as it's constant in this derivative sense. So let me just say what I mean by that. There's never been an intertemporal comparison problem in cardinal utility theory. It just-- it's never been there because you can always say, well, I'm going to associate the indifferent surface of seven utils today with the indifferent surface of seven utils ten years from now.

01:15:02
Bob Murphy: Mm-hmm.

01:15:02
Eric Weinstein: I have a way of matching up levels of utility-

01:15:05
Bob Murphy: Right.

01:15:05
Eric Weinstein: So I can just price.

01:15:07
Bob Murphy: Yeah.

01:15:08
Eric Weinstein: So if you're willing to-

01:15:08
Bob Murphy: You might have a discount rate, but you're saying what a util today is from the perspective ten years from now, that's what a util means. Like that's a unit of measurement.

01:15:15
Eric Weinstein: A util is a util, and that becomes the means of solving the changing preference problem. So the problem is, if you don't accept cardinality, you want to get the solution that you have in the cardinal world without paying the price of being cardinal.

01:15:30
Bob Murphy: Yep.

01:15:31
Eric Weinstein: So that's what Milani and I did, is we said, look, you can afford to go upstairs into XYZ space, into the space of all cardinal utility functions, because no matter which cardinal representative you choose to begin with at time zero, if you stay in a constant function using the special economic derivative over the path, when you get to your endpoint, the beginning and endpoint will be shifted by the same amount as any other choice. So the same disease occurs twice, and one of them with an inverse sign. The two of them kill each other. And the map between ordinal preference leaves, surfaces from time zero to time one, turns out to have no dependence on the cardinal choices made. You can make any different cardinal choice at time zero. You'll end up with different cardinal choices at time one, but it's the pair of the two that determine the welfare map. So do you have this quote from Carl Christian von Weizsäcker?

01:16:34
Bob Murphy: Yeah.

01:16:34
Eric Weinstein: So Carl Christian von Weizsäcker, who I spoke to, very interesting. He is the son-

01:16:39
Bob Murphy: Go ahead.

01:16:40
Eric Weinstein: Of the German physicist who was active with the Nazis, but a very great physicist. And strangely, his son became interested in something which turns out to be exactly the gauge theory of Hermann Weyl. So there's something running through that family which is recapitulating the search for this abstraction that we now call fiber bundle theory, covariant differentiation or gauge theory. Can you read the quote?

01:17:07
Bob Murphy: Yeah. Traditional neoclassical economics has worked with the assumption that preferences of agents in the economy are fixed. This assumption has always been disputed, and indeed, in the social sciences outside of neoclassical economics, the assumption has never been accepted by anyone. The obstacle is the lack of an answer to the question, how can you do welfare economics if preferences change endogenously? After all, preferences of individual agents are the basic measuring rod of economic welfare of the performance generated in economic system. How can we evaluate an economic system with a measuring rod that itself changes with the system?

01:17:42
Eric Weinstein: That rod is the gauge of gauge theory. I mean, it could not be any more stark. Carl Christian von Weizsäcker is one hundred percent correct. Nobody's ever accepted this outside of the cult of neoclassical economics. And it is also the case that he's phrased it correctly. You have a rod that is varying in time, and this is what gauge theory was exactly designed to do. Now, there's another quote from Travis Nesmith, who is part of the research division of the US Federal Reserve. And if I could ask you to read that, what I'm trying to do is to establish that inside the field, forget Econ Twitter, which doesn't seem to be very smart. Inside of the field, the premier theorists all know that this is a central vulnerability. So he says it quite starkly, if you can find Travis Nesmith.

01:18:30
Bob Murphy: So this is from two thousand and six, from Travis Nesmith, who's on the board of governors of the Fed. A time-varying objective function generally cannot be tracked by an economic index.

01:18:41
Eric Weinstein: Is exactly the case. As of now, without this work, you cannot track a time-changing dynamic ordinal preference map with an economic index, which is an extension, technically, of what would be called the Laspeyres-Paunuis index. After Alexander Conant, who I believe was also writing in nineteen twenty-five.

01:19:02
Bob Murphy: Can I, can I stop you just for a second? I want to make sure. I think some people who like know enough to be dangerous might be thinking, so in standard utility theory that you'd learn, you know, at any decent program, they'll say, "Yeah, we have cardinal utility functions. We don't really believe in it. It's just a mathematical convenience. It helps us to, you know, calculate, solve." But really what it is, is the results show that for any ordinal mapping, given some assumptions you make that are, you know, pretty innocuous, you can construct or exist a cardinal utility function such that, you know, put in any two bundles, and the cardinal utility function will give more utils, uh, you know, higher number to the bundle that's preferred, ordinally preferred to the other one and, you know, base transitive and all that stuff. But they'll stress it. But, you know, you don't need to ascribe any significance real where it's not like it really measures something intrinsic in reality because-

01:19:50
Eric Weinstein: It's sca- it's scaffolding. In other words-

01:19:52
Bob Murphy: Yeah

01:19:52
Eric Weinstein: ... you could choose to put up metal scaffolding, wood scaffolding, plastic scaffolding. The scaffolding comes down at the end. There's no cardinal dependence in the story. You're just going through cardinal. What gauge theory says is that you can afford to dip into cardinality because there will be no dependence on it when the deed is done.

01:20:11
Bob Murphy: Yeah.

01:20:11
Eric Weinstein: And-

01:20:12
Bob Murphy: Let me, let me just... Because I-- Well, one thing I want to check, run this by you, Eric, make sure I'm not getting confused myself. But so there in, in that thing, and then they'll-- if the students say, "But there's no such thing as cardinal. What do you mean ten utils? What the hell does that mean?" They say, "Oh, no, no, don't ascribe it because it's only unique up to a positive affine transformation." So if you had a cardinal utility function that represented these ordinal preferences, you could take it and do two times that plus seven or something, it'd be fine.

01:20:36
Eric Weinstein: It's not even affine. Take any di-

01:20:38
Bob Murphy: Hold on. Yeah, I'm-

01:20:39
Eric Weinstein: Any individualism you like.

01:20:40
Bob Murphy: I'm doing the, the one with the risk. But so yeah. So they say that, and then but you're saying something... But even in that framework though-

01:20:48
Eric Weinstein: Mm

01:20:49
Bob Murphy: ... it's assumed that over time, if you want to talk about intertemporal choice and like saving today and, oh, I've got to reduce my consumption today to have more in T plus one, they're assuming that the underlying ordinal, the indifference space curves, the family of indifference curve is the same from T to T plus one. Otherwise, how could you do anything? Whereas what you're showing is even the ordinal maps taken per, you know, at a snapshot in time can evolve.

01:21:15
Eric Weinstein: This would be-

01:21:16
Bob Murphy: Is that right? Am I-- Is that correct what I said? Okay.

01:21:18
Eric Weinstein: Yep. You have to be allowed to have dynamic ordinal preferences. And this gets into-- I just turned sixty. I'm a little nostalgic. Economists used to be much smarter.

01:21:30
Bob Murphy: Well, yeah.

01:21:31
Eric Weinstein: Yeah. So I want to talk about Ken Arrow and Paul Samuelson because I had the good fortune to be able to talk to both of them, in the case of Arrow face to face, in the case of Samuelson over the phone, about this. And what they had to say was like legitimately fascinating. So Arrow, when he heard about this result, had me initially to breakfast and then out to Stanford. And at breakfast, he says, "I have to tell you, your result can't actually be right." And I said, "Okay, can you explain why?" And he said, "Well, I have this thing called the impossibility theorem." And I chuckled because of course I knew about it. He said, "What it says is that you cannot aggregate preferences in a coherent fashion between many individuals at a single instant in time. Can't be done. It's impossible. And what you're doing is you're fighting a duality. And in that duality, a single individual over many instances of time is directly equivalent to many individuals at a single incident in time. And what you're talking about would be a solution to the impossibility theorem, and I'm positive that the proof holds." And you know, at first I didn't even really grasp what he was saying because I didn't see it in the same way at all.

01:22:47
Bob Murphy: Yeah. Can I, can I stop you just... I don't want to say it, but let me just make sure the list-- because I, I love this story, but I want to make sure the listener gets it. So folks, Arrow's theorem has some postulates, like in social choice theory, it's a, you know, landmark result. Say you got a bunch of different people, they have different ordinal rankings of various states of the world or outcomes. And with some basic postulates about like there should be a dictator. If everybody prefers A to B, then the social welfare ordering should prefer A to B, stuff like that. And he showed there's no way of aggregating all these disparate subjective individual ordinal rankings into one social ordinal ranking that satisfies all these very reasonable criteria. And so what Arrow is saying is this guy, Eric Weinstein, is coming along saying, "I figured out a way to solve the problem of different ordinal rankings by an individual over time." And then Arrow's saying, "Well, no, I mean, mathematically that's equivalent to just treat, you know, myself at T and T plus one plus two plus three as Jim, Bob, and Sally at the same time, and my theorem showed that's impossible, so you must be wrong." Okay, so go ahead.

01:23:51
Eric Weinstein: So what I didn't understand was is that he saw that as equivalent. And I said, "No, no, no. I hadn't understood that this is how you were seeing it. There are two pivotal differences between the two situations. This is a false duality." He said, "What do you mean?" I said, "The first thing is your result is a result in social choice theory. There is no market. You're voting on, let's say, three different candidates, or as you say, three different outcomes of the world. And what you're saying is that you cannot aggregate a heterogeneous group as if it was a single consistent super individual, so that if everybody agrees on a choice, that choice has to hold. There isn't one person who has dictatorial rights, and small changes don't result in violent changes, you know, whatever this list of axioms is." And he said, "And the second?" And I said, "Well, the second is that there is no morphing of Bob into Sally, into Arjun, into Leticia. That just doesn't exist. And you do morph into your future selves over time. So you're also not in possession of a path." Which connects all of the individuals who are being aggregated. And those two things break the apparent duality that you're appealing to, and I'm not going to disagree with you that in their absence you'd be right.

01:25:19
Bob Murphy: Mm-hmm.

01:25:20
Eric Weinstein: You'd be correct. But that's why the intertemporal market situation and the intra-agent social choice situation are not in duality. There was like a long silence, and he said, "Oh." And then, you know, he was good enough to have me out to Stanford.

01:25:36
Bob Murphy: Mm.

01:25:36
Eric Weinstein: And we talked further, and he said, "Look, it's hard for me to understand what a principle bundle is, and groups are not something that..." He said, "I know about groups, but they're not something that we see typically in economics. We don't imagine that there are symmetries in economic data because it's too noisy." And then I said something to the effect of, "Well, the group doesn't act on the data. It's not like a time series of a, you know, a price movement that has a head and shoulders pattern where there's some vague symmetry. It's acting on the target space of cardinal utility itself. You're using a group action to get rid of cardinality, so you're entitling yourself to work with cardinality because you have a means of erasing your dependence, which is group theory and symmetry." And somehow that wasn't easy on him for some reason. Now, the other really interesting-- I don't know if you want to add anything to the arrows.

01:26:31
Bob Murphy: Let, let me just-- Yeah. So that's great. Let me connect that to what we said early on about that issue when I said I f- I saw you on somebody else's show, and you were saying, "Yeah, the way economists think about it, if I told you that this year the prices and quantities were such and such, and then we checked in ten years later and the prices and quantities are identical, they would think it has to be the same." And then, you know, and I was like, "Yeah." And then we've been stressing the path dependency, and so what you just said there, your answer to Arrow, I think should give people a hint. Like you see why the path dependency mattered because yourself today becoming yourself ten years from now, what you do along the way would matter, right? Because otherwise, yeah, if you didn't have that in between, it would be like your future self is just some other guy. Anyway, I'm just trying to connect it.

01:27:12
Eric Weinstein: When you erase someone's history.

01:27:14
Bob Murphy: Yeah.

01:27:14
Eric Weinstein: I mean, you just imagine that I wiped out, you know, I neuralyze you from Men in Black.

01:27:19
Bob Murphy: Yeah.

01:27:19
Eric Weinstein: You, you would probably not feel that that was a fair treatment of whoever it is that you are because y-your history matters to you, and it should in economics as well.

01:27:27
Bob Murphy: Okay, I just want to make that point, and then-

01:27:29
Eric Weinstein: Sure.

01:27:29
Bob Murphy: Did you want to-- Were you going to talk about Paul Samuelson?

01:27:31
Eric Weinstein: Yeah. So Paul Samuelson was mathematically just as interesting.

01:27:36
Bob Murphy: Mm-hmm.

01:27:36
Eric Weinstein: Yeah. So I started telling him about this, and I actually never met him face to face. I feel sad about that. And he got really animated, and he said, "Wow." He said, "I know all about this. This is amazing." He said, "You should go back to my Nobel acceptance speech." And I said, "How so?" And he said, "I think what you're telling me is that all of these notions of instantaneous indifference don't fit together, and that's why we're getting path dependence." I said, "Yeah, that's exactly what I'm saying." He says, "Nobody appreciates that I tried to raise this in a different issue." I said, "How so?" He said, "You'll see in my Nobel acceptance speech." So I looked at it, and in there he's much more radical. So I called him back, and I said, "You're telling me that you don't even necessarily believe in preference maps." He says, "Exactly." He said, "That speaks to a level of coherence in agents that I don't believe was ever realistic." And I said, "So are you telling me that you have an idea of instantaneous situational indifference that does not necessarily knit together into preference-- into a different surface? In other words, you have an instantaneous version." He said, "Yes, but the problem is the integrability of distribution." So he's using really solid mathematics-

01:29:04
Bob Murphy: Mm-hmm

01:29:04
Eric Weinstein: ... and speaking the language. He says, "The integrability of distributions is not guaranteed, and we went wrong by putting so much on indifferent surfaces because you couldn't explain to economists what non-integrability was." Now, this is the Escher staircase problem. So I mentioned that Penrose stairs, the Escher stairs, I said this is the mathematical concept, which is it seems like you're just going up and up and up, but then you're back where you started. Or if you've ever seen like ten people sitting on each other's laps in a circle, or if you've played rock, paper, scissors, you know, these are notions called holonomy or cocycles. And so what Samuelson was really saying is, "I, Paul Samuelson, did not have a way of describing to economists with ordinary levels of mathematical training what the failure of integrability means, and so we ended up with indifferent surfaces." And that blew my mind. And it's right there in his like-- By the way, I've been in a lot of different fields-

01:30:10
Bob Murphy: Mm-hmm

01:30:11
Eric Weinstein: ... so I have a, a unique multidisciplinary view of this. Every single field has very zealous young people who have no idea what choices were made by their elders.

01:30:23
Bob Murphy: Mm-hmm.

01:30:24
Eric Weinstein: They think that somehow the field is on solid ground, and it's always been thus. And if you actually talk to the old people, they say, "Wow, this hardened into dogma."

01:30:33
Bob Murphy: Yep. You obviously have a much better view of it, but yeah, I've seen that too. I, I can just tell-

01:30:38
Eric Weinstein: I've seen it. I've seen it in Donaldson theory, in differential geometry. I've seen it in physics, took a huge wrong turn claiming that the main problem was to make gravity quantum mechanical. Was never the problem in nineteen eighty-three, and suddenly by nineteen eighty-seven, that's all anyone could talk about. There are these cult-like moments, and Samuelson's critique is, "Go back to my Nobel lecture. I was trying to warn the field you don't have preference surfaces. You need to learn what integrability is." And what he was really saying was Learn to love the cycling problem. Learn to love path dependence. Path dependence is not wrong. And I've heard it said that he was really the true student of Gibbs and intellectually at some level. Gibbs really didn't have any errors except for Paul Samuelson. I don't know whether that's true or false, but it is the case that I believe that Paul Samuelson was an early believer that path dependence, failures of integrability, curvature, incoherent distributions in a foliation sense or in a differential equation sense are at the heart of the self-inconsistencies of agents, and that we should embrace this rather than trying to create homo economicus. If you don't believe me, go check out his lecture.

01:31:55
Bob Murphy: Hmm.

01:31:55
Eric Weinstein: Somehow the field doesn't know that the people who wrote it are well aware of these problems.

01:32:01
Bob Murphy: Just real quick, a few times in this discussion, you've referred to the cycling problem. Is that something like A is preferred to B, is preferred to C, is preferred to A? Is that example what you mean?

01:32:12
Eric Weinstein: No.

01:32:13
Bob Murphy: Okay.

01:32:13
Eric Weinstein: Uh, but that is-- it's a related phenomenon.

01:32:16
Bob Murphy: Okay.

01:32:17
Eric Weinstein: In this case of Paul Samuelson's issue about the coherence of distributions, it is very much that sort of a thing. But what I mean is that if you have prices and quantities and you go around a loop, and from time zero to time one, all prices and all quantities come back to their original levels. Any good price index has to have the capacity to show that it is not coming back to unity, because otherwise it's non-economic, right?

01:32:49
Bob Murphy: Hmm.

01:32:49
Eric Weinstein: It's exactly the counter-intuition to the entire field. If I might, I would love to be able to show you a place where this is visualized. So you see a curve at the bottom of the screen, and the-- that is gonna be tracing out a circle. And that's like it's happening in the XY plane. And the height of the curve above it is being depicted in XYZ space. And all of those floating blue squares are these instantaneous notions of constancy that we were discussing before. What Paul Samuelson was saying can't be talked about with economists is the idea that these floating blue planes are in a certain sense incoherent. That means that when you come back, even though you're constantly parallel to the system of blue planes, you don't come back to exactly the same place in the helix above the circle. Those blue squares are like indifference or substitution effects. And so while we are trying to track a change, let's say in ordinal preferences or purchasing patterns, we are locally always constant. But what the path dependence measures is whether or not the system of notions of constancy, the floating blue planes, the distribution, the incoherence, the Penrose steps, all of those concepts that Samuelson were saying you can't really talk to economists about. This is an actual-- So far as I know, it's the only visualization of its type in the world. This shows you what gauge theory actually is. It is a series of local notions of constancy that don't knit together. And if you weren't using those local notions of constancy, you wouldn't be doing economics. And it shows you that for that particular path, you better have a gauge that when you come back to exactly the same point in XY space registers a difference, which is, well, you know, constantly I would-

01:35:03
Bob Murphy: Hmm

01:35:03
Eric Weinstein: ... I traded a gum wrapper for a safety pin. I traded a safety pin for a marble. I traded a marble for this, and suddenly now I have a mansion. And the answer is, yeah, you went around a circle, and you didn't come back above where you started, even though all you did was make trades. So that is the mathematics that economics has to move to, or when the Becker-Stigler stuff falls apart, you will have this exhaust on the death star, and you just-- you don't have a means of staying ordinal. If you move to cardinality, by the way, which I'm very keen to avoid, because cardinality will lead to communism. People don't really understand this. If you have diminishing returns, diminishing utility to money, you will find the argument that says that it-- if we can have aggregates among all individuals, that we have to go after the best off in our society because we have a social welfare function, because we can add up utility among agents. So it's very, very dangerous. I understand very well why we're avoiding cardinality, but I don't think people understand that y- right now the, the only way economics is supported is that you have very few outsiders who know where the bodies are buried. I can tell you that I know that you use cardinality in the theory of risk going back to Bernoulli's solution of the St. Petersburg paradox. Then you've claimed that, oh, we only use von Neumann Morganstern's sub-utility functions. We don't actually have those as the ultimate utility. We integrate out in an intermediate step. It's like, yeah, you don't have a subject, and that's the problem. So we're trying to help economists build a legitimate subject, but they have to understand that picture. Now, we have a second video that will also be helpful to explain what fiber bundles are and why we use fiber bundles when it comes to understanding how calculus has matured. I want to show you two different derivatives based on our earlier discussion about income and what it means to be constant. So what you see right now is that all of the horizontal levels are like the analogs of the floating blue planes that we saw. This is cons-- this is going to be constancy, where we measure rise over run the way you learned it in calculus. And then it will be followed by an economic derivative, and you'll see the difference in the two derivatives. And this is without an XY plane. This is just... Sorry, this is an X, a real line, which is in this case T. So there isn't a two-dimensional base space. There's just a one-dimensional base space. And we're gonna see that even in this simple example, you'll be able to see how violent a change it is when you become freed from the notion that there is one derivative. So now what you're going to see is you're gonna see ten dollars an hour, derivative equals zero. And now you're gonna see what should be viewed as economically constant. The effect of economically constant is that the levels that you measure the rise over run from have to themselves be economically responsive to the change in prices. So that curve in this new system of horizontal levels is in fact constant. At no point is it rising above the reference levels that you've seen. So let's play that one more time. So to begin with, we see the reference levels. This is the standard notion of constancy from calculus, but somebody's facing hyperinflation. So I would now claim that that function we're seeing is the right notion of constancy for somebody facing hyperinflation. So now we have a new notion of a derivative just by changing the reference levels, and the rise above reference over the run is now the right economic notion of constant. And what we've just done is gauge theory. That's gauge theory with-- you don't need fancy formulas. You can see exactly. There's no attempt to obscure anything. What you're looking at is a powerful tool for recognizing that you have to make sure you're using the right derivative, and that you can't have a marginal revolution in which you just use off-the-shelf derivatives without asking for the right notion of constancy. And I think we have one more video.

01:39:21
Bob Murphy: Yep. And also on that one too, just intuitive for people like, "Oh, the economy's not flat, it's curved." Like they-

01:39:28
Eric Weinstein: You don't actually-- you don't see any curvature there in the sense that we mean curvature technically because your base space is just time. It's one dimension. It's not until you have two different dimensions. So imagine, for example, we had Cobb-Douglas functions with three goods. Then we'd have an alpha, a beta, and a gamma exponent, and they all have to add up to one. Alpha-- The gamma is therefore one minus alpha minus beta. So it's-- it can all be parameterized in terms of alpha and beta. That is already a situation in which you have a sense where Paul Samuelson's floating plane, the distribution do not knit together. So if you have changing preferences over guns, butter, and carrots, then the sad fact is that you're already in a situation in which going around a loop will have an effect of changing your aggregate level. And if you don't have a Divisia style index, you will not actually be able to capture the economics. If you do not embrace path dependence, you will wither and die or you will be forced into lying to the public that you supposedly serve. So in the next example, I wanna look at trigonometric functions like sine and cosine. And I want you to ask yourself the question, if you had a periodic function between zero and two pi radians, what would happen if, for example, you were looking at the cosine function and you wanted to only go between zero and pi rather than zero and two pi, and you were trying to put the function on at first a cylinder between zero and two pi, but then on a Möbius band between zero and pi. Let's roll tape. So there's a sine function, there's a cosine function, and you can obviously see that you should be able to do calculus on a cylinder. But if you try to do it with a sine and a cosine and you put them together, you'll notice that they don't-- they're not smooth at zero and pi until you put in a twist. Now, the question is, is there any reason you shouldn't be able to do calculus on this structure? It's smooth everywhere. It's not an XY plane. It's not a familiar place to do calculus. In the same way, cardinal utility functions are like the totality of the Möbius band. Ordinal utility functions are like the circle going through the middle of it. And the Samuelson Shepard cost function is like these sine functions and cosine functions. So what I'm trying to say is you can get visual intuition pumps for what it is that economics needs to do next, even if this is infinite dimensional, if it's group theoretic, et cetera, et cetera. The basic issues that gauge theory was meant to solve was to allow you to do calculus with bespoke derivatives that are well-suited to your problem on spaces that are a good deal more interesting and complicated than the simple spaces like the XY plane, the XYZ space, and more intricate than simple functions because these things are more derived from the graphs of functions than maps from a set A to a set B. So in all three of these cases, we're trying to say that Samuelson was correct, but you can actually use visual aids for economists to realize you guys are in the best situation you could possibly be in. Everything that you do is gauge theoretic. You just don't know it yet. And you're angry, and you're petulant, and you're controlling, and you're really, really nasty. And it, it's important to recognize that Arrow and Samuelson were none of the things that you see on Econ Twitter They were gracious. They were intellectual. Already with Becker, you see a very strong personality who's ideologically driven rather than intellectually driven. And, you know, my feeling about this is we need to go back to a more collegial, genteel, high trust, individualistic version of academics. And if you want, I would be happy to bring this to any, any group or department that knows the history. Bob, am I right that I've sent you a great deal of material indicating that these are very real problems that have been recognized in the field for-

01:43:59
Bob Murphy: Yes, definitely. Yeah. So folks, yeah, Eric had sent me a bunch of, you know, quotations and excerpts from various things to establish the history of thought on this. So yes.

01:44:10
Eric Weinstein: I'm just going to say it once. I would love to help, but if you want to fight me, and if you decide that this is stupid, or that this is charlatanism and obscurantism or anything like that, I'm happy to take you on ten thousand to one, and I'm going to win, and you're going to lose because you just don't understand your own subject. You don't understand the importance of the problem. You don't understand the history. And it's embarrassing. And quite frankly, I would really prefer that we did it the right way. But if you want to do it the wrong way, just let me know.

01:44:38
Bob Murphy: That's a good way. Let me just before we wrap up here. So thanks. I think those graphics are going to really help some of the viewers make a click. You sort of as an offhand remarks there said something about communism. So let me say some things, Eric, and then, you know, you obviously chime in if you think I'm getting it wrong. But at least the way, you know, I was taught this stuff is that, yeah, if you go back and read, ironically, so my dissertation, Eric, was on Bombaverk, who was like second generation Austrian, one of the guys who popularized the subjective marginal revolution. And some of his examples and things, he does have a notion of cardinal utility. It's kind of ironic. You wouldn't have thought that. And then it sort of gets expunged. And I think in Hicks, his value in capital, I think it's thirty-seven, he shows how we can do modern price theory without relying on this notion of cardinal utility. And he kind of like I think he says Pareto developed it, you know, taking the XYZ and then just the surface of the utility function, you drop it down and that's, you know, the indifference curves. Okay. And so then and part of the reason is that, yeah, there's this philosophical issues about Benthamite utils and whatever. But also, as you're saying, Eric, there was this notion that if we actually are walking around saying people have cardinal utils, well, then you that opens it up to say, hey, we took some money from a rich guy and gave, you know, spurred on a hundred poor guys, they would probably raise aggregate utility in society. And so that's one reason you wouldn't want to go down that path.

01:46:06
Eric Weinstein: Yeah.

01:46:06
Bob Murphy: And that's what you're getting at.

01:46:08
Eric Weinstein: Very much so. I mean, it's not that that argument is in a sense wrong. It's that the argument is incomplete. And my claim is that what Becker and Stigler and, you know, we should actually close it out on Dale Jorgensen and what happened to this work at Harvard.

01:46:24
Bob Murphy: Yeah, yeah, by all means.

01:46:26
Eric Weinstein: What happened was that you had a bunch of priests who pretended they were doing economics when they were doing policy. And I am not unsympathetic to the idea that we need to push markets and freedom worldwide. I'm very sympathetic to that, even as a guy on the left. I am as hyper-individualistic as you could meet anyone, anywhere. The problem is that cheating to do it is a sign of intellectual weakness. And that's what we have. We have a culture in economics in which the top people in the field are decorated so that they can lie much the way people are in public health. Instead of announcing that Tony Fauci is in part the bioweapons czar because we signed a bioweapons convention, the Geneva Convention on bioweapons in the 1970s, and then we fell into some weird situation where we're not allowed to do offensive bioweapons research, but we are allowed to do defensive weapons research, and then we do some of it in China, and then suddenly millions of people are dead and we can't admit to what we're actually up to and doing with Ralph Barrett's lab and the EcoHealth Alliance and all the stuff that we all know, then you have to call everybody who notices this insane. And frankly, that's disgusting. It should not happen within the academy. Could you bring up the wild versus the mild slideshow of Robert Gordon?

01:47:53
Bob Murphy: Yep.

01:47:53
Eric Weinstein: What really happened to this work, let me just say what I believe. This is all joint work, but it appeared in the thesis of a young female from the developing world. And in my opinion, this person's career was immediately ended by Dale Jorgensen and by the Harvard Economics Department. There's something the Harvard Economics Department has called the jobs market meeting, which is basically a conspiracy and restraint of trade, where the department decides who it will be pushing to get the plum jobs and the prize position. And because Dale Jorgensen was in the process of committing academic malpractice through the Boskin Commission, it was imperative that a work on inflation that happened to appear in the first revision of the CPI at this level since the Stigler Commission in the early 1960s, late 1950s. So somehow we had gone from around 1960 to the mid-1990s without a commission on inflation. Bob Packwood and Daniel Patrick Moynihan came up with a scheme, which was they figured out that since the 1970s, we had automatically indexed entitlement payments to inflation because of the inflation of the age. Previously, I think it had required an act of Congress. And then in the 1980s, we got around to indexing tax brackets. And so these political gentlemen, if you know the phrase in Washington that politics is the art of the possible, and if you ever want, you know, to commit mass murder, cut up the bodies, and distribute them so they won't be found, the art of the possible is the phrase. So many crimes are hidden in that phrase, I can't even begin to tell you. So what they did is they decided they didn't want to touch the third rail of politics, which is slashing benefits and raising taxes. So they get five economists to do it. They turned immediately to Michael Boskin at Stanford, but more importantly to Dale Jorgensen and Zvi Griliches at Harvard. And then we had one of the great crimson washing experiences, which for some reason Robert Gordon, who was one of the other five Boskin commissioners, chose to discuss in a presentation called The Wild versus The Mild, where the Boskin commissioners were wild, they were unsupervised, they had no bickering, there was no peer review, they made stuff, they pulled stuff, you know, out of their backsides. They made stuff up. And the most important slide in that presentation-

01:50:28
Bob Murphy: Can you see this right now, Eric?

01:50:30
Eric Weinstein: Yeah, I can.

01:50:31
Bob Murphy: Okay. Do, do you want me to go with this right here?

01:50:33
Eric Weinstein: Let's, yeah, let's go. Big difference. Boskin lack of review. So please tell me again about peer review, all my dear colleagues in economics. We were a creation of the Senate Finance Committee, and especially of the soon-to-be-disgraced Bob Packwood, and then the elegant Daniel Patrick Moynihan in between a backburner activity. We only had fee five meetings, two of which were with top BLS people. Next slide.

01:50:58
Bob Murphy: And, and this is from, again-

01:50:59
Eric Weinstein: One of the Boskin-

01:51:00
Bob Murphy: ... Robert Gor- Robert Gordon giving-- explaining to people this is what my experience was on this commission, just to make sure-

01:51:05
Eric Weinstein: Oh, yeah

01:51:06
Bob Murphy: ... people didn't get lost.

01:51:07
Eric Weinstein: How decentralized it was. We were very agentic, you see. The Boskin Commission was built on trust. Trust of the population? No. Trust of the profession? No. Trust of peer review? No. It was just five commissioners set up by Packwood and Moynihan. We had no arguments, and there was no bic- bickering. Now, this is one of the greatest slides in the history of economic theory. Dale said one point one percent implies one trillion dollars in Social Security savings over ten years. In other words, he had backed out a clean decade and a clean trillion dollars, and he said, "You know-"

01:51:42
Bob Murphy: And, and this is Dale Jorgensen, also at Harvard?

01:51:45
Eric Weinstein: Yeah.

01:51:45
Bob Murphy: Okay.

01:51:45
Eric Weinstein: This is the, this is the chairman of the economics department at Harvard. This is a university professor above even department level, which now Larry Summers and Eric Maskin, I believe, are the university professors. I think that Larry sort of has Dale's position. Somehow, somehow our separate efforts came up with the one point one percent bias norm. In other words, the target that was backed out from we need to save one trillion dollars in Social Security by cutting benefits over an even ten years, we get an unrounded number. And somehow we broke into two groups. But Zvi and I did not question the substitution part. And Mike, Dale, and Ellen didn't question the zero point six quality change part. But when you added the two pieces together, it miraculously came up to the number, the number, notice the definite article in front of one, that Dale Jorgensen said. This is, ladies and gentlemen, is why your field is not trusted by anyone. These are hitmen. These are not academics. They're not serious people. They're not grown-ups. They're not colleagues. They're committing academic malpractice. And the result of this is that what I consider to be one of the most important theses in the history of mathematical economics, effectively beginning a second marginal revolution, was buried so that this slide could govern. And I believe there's a quote of Greg Mankiw after the fact saying something to the effect of, "This was really about balancing budgets. It wasn't about analyzing the CPI." And if you look inside of the Boskin Commission, they give an example with chicken and beef, and they say that the economy is too dynamic for a fixed basket. But if you, if you say, "Okay, assume that the, the example is generated by Cobb-Dougl- Cobb-Douglas preferences," you will find that the exponents shift. In other words, ordinal preferences are in fact changing in the Boskin Committee example. They can't handle that. Then they appealed to a guy named Erwin Diewert or Diewert at UBC, claimed that he had a theory of superlative index numbers which allows you to avoid taste change and taking in tastes, neither of which are true, because it relies on something called homothetic aggregator functions. And as a result, the Bureau of Labor Statistics is lying, saying that it has moved to a COLA framework, which it hasn't. It doesn't take in taste data. Superlative index numbers don't solve the problem. There is no subject here. And what I'm looking for is you need to rehear the fact that your field was innovated at Harvard thirty years ago. And what we did instead is we allowed for the malpractice and exercise of power by Dale Jorgensen, Zvi Griliches, and their fellow Boskin commissioners, and we buried real economic theory as a result of it. And that is why you're going to find that this is a laughing stock on Econ Twitter. It has nothing to do with the quality of the thinking. It ha- doesn't have to do with the seriousness of the issue. Doesn't have to do with the fact that this has been well-recognized by all of the top people in the field. It has to do with the fact that what we now have is we have a coin-operated subject rather than an academic subject. I'm on the other side.

01:55:17
Bob Murphy: Okay, let me just make sure people are getting s-some of the takeaways from there. So when Pia Malaney is going through, she's working on her dissertation Dealing with, ah, yes, using gauge theory applied to economics, it, it as an offshoot among all the other things it can do, gives a new way to, to solve the index problem. And then right at this time, the Boston Commission's formed, and Dale Jorgenson, who's also at Harvard, so Eric Maskin is her supervisor, but Jorgenson's also there in the department. Oh, this is amazing. The stars have aligned right when we're coming up with this theoretical innovation to help economists really have a much better grasp on what's going on with index numbers right when the government needs a revision. And why is it relevant? Because among other things, Social Security benefits and then tax brackets are indexed to inflation at the, you know, consumer price inflation at this point. And so calculating what the CPI changes from year to year is that, that means a lot of money. And so you're saying that the Jorgenson knew ahead of time we need this Boston Commission to come up with the CPI has been overstating inflation.

01:56:24
Eric Weinstein: Real, real simple.

01:56:26
Bob Murphy: Yep.

01:56:27
Eric Weinstein: We cannot have two kids who happen to be in the bank the day we're robbing it for a trillion dollars saying, "Hey, methodology is up for grabs." You, you need-- There, there needs to be a national map of CPIs just the way we have pressure fronts and temperature and wind maps for, for weather. You would never accept a single scalar. Inflation is a field over agents, geography, and time, and it is a rich subject. The-- I mean, the good news for intellectual economics is that this is an open door to an orchard whose fruit has not been picked. The bad news for coin-operated economics is that there's a way in which the economics profession basically sanitizes the objectives, in particular geopolitical strategic ob-objectives of the United States, by trying to claim that we can get ought from is. That we, we ought to have high levels of immigration even though nobody wants it because we can determine that it is good for the country. There's a whole game that is played with Kaldor-Hicks preferences where you never actually do the redistribution, but you say that there's a Pareto improvement. Nobody cares about Pareto improvements. I don't want to be a dollar richer if you're gonna be a billion dollars richer. But there's this entire toolkit having to do with the Washington Consensus and free trade and immigration and all of these things that have been terrible for the United States, like the moving, the offshoring of all of our manufacturing. And you know, you and I are speaking right at a moment where Larry Summers has stepped back because he appears to have been under the influence of trying to curry romantic favor with daughters of people involved in the Road and Belt Initiative. And we're in a situation in which economists have to stand up for national interests. And economists like Brad DeLong, protege of Larry Summers, when pushing NAFTA, years later, he rips the mask off and says it was a social Darwin welfare function we were opera-optimizing. It makes you richer to the tune of the cube of your extant wealth. And you're thinking, "Oh, really?" This is that game that you play called esoteric exoteric, where you lie to the public, and you tell the truth in private, and you make up what the national interest is that you got is from ought. This is supposedly the game. I don't know if you know the story about Paul Krugman at the beginning of his career saying that there were places where you would want tariffs and protectionism because of infant industries and path dependence and increasing returns. And as the story goes, I don't know whether it's true, supposedly Jagdish Bhagwati called him up and said, from what I heard, "Wonderful career you have there. It would be a shame if anything happened to it." This is a dirty game, and it's not going to survive. It's not gonna survive the Jeff Epstein stuff that we saw playing out in the econ department at Harvard with Rosovsky and Summers. It's not gonna survive the revelations about what was behind the scenes at NAFTA and how we, we sold out our own people through the Immigration Act of nineteen ninety through an economist named Miles Boylan working in NSF. Basically, the economics profession is dirty, and it's got to go back to acade-academic economics, or it deserves to be broken up. And so what I'm advocating for is I think it's a marvelous substrate. There's a lot we could do, but it needs to be an honest and public-spirited field. It can't simply be about the interests of oligarchs, giant institutions, the CIA, the State Department, and things that are funded by USAID. So I just want to say, if you're looking for me on Econ Twitter, I will be welcoming you with open arms. But I can assure you this is quite serious business. And Bob, it would be fun at some point if, if there's a lot of anger about this to do a second show where we really dig into just how deep the gauge theoretic economics rabbit hole really goes. And it's been a great pleasure. Thanks.

02:00:43
Bob Murphy: Thanks so much. Folks, my guest has been Eric Weinstein, a, a jack of many trades, and here he was telling us all about gauge theories applied to economics and why it's elegant but also has a lot of practical implications. Eric, thanks so much for your time.

02:00:58
Eric Weinstein: Really appreciate it, Bob.

02:00:59
Bob Murphy: Thanks everybody for tuning in. We'll see you next time. [outro music]