← The Human Action Podcast
The Human Action Podcast
A Harvard Economist Tests Austrian Capital Theory
Loading summary
Transcript59 lines
- [00:03]
A
This is the Human Action Podcast where we debunk the economic, political and even cultural myths of the days.
- [00:09]
B
Here's your host, Dr. Bob Murphy. Well, Paul, welcome to the Human Action Podcast.
- [00:16]
A
Thanks, Bob. It's a pleasure to be here.
- [00:19]
B
So let's take care of this at the beginning on Twitter X. A while ago at this point I saw somebody announcing that, oh, these two Harvard economists have a paper that involves Bombavik. And I went and looked at it in the abstract. Right away it jumped out. I was talking about the average period of production. And we'll come back to this folks in a minute and say, well, that's, you know, a controversial element in Bomberkian thought. And I made some snarky remark about like, oh, geez, even when the mainstream is interested in us, they focus on the one thing that modern Austrians no longer, you know, dwell on. And so I want to formally apologize that I was just being snarky. And so thanks for coming here. And yeah, I mean, this is amazing that we've got two Harvard economists and it had this like, the draft I'm looking is like 103 pages of a paper about Austrian capital theory embedded in, you know, international trade. And it's under review at the aer, which is arguably like top journal in the world. So this is a very special event. So of course, folks, Paul was nice enough to come here and talk about it. Before we get into the paper though, or the, the draft, Paul, can I just ask you just a little bit about your background? Because I'm just curious. So you grew up in Spain and then went to mit. Can you explain, like over in Spain, what's the. What, like, what is M known for? Like, is it, does it have, like, what's the reputation? Is it considered ideological? Because coming from the right wing Austrian perspective, you know, we have our views of mit, but I'm just curious in general, for a young person going into economics in Spain, how they view mit, right?
- [01:49]
A
So, I mean, I'm not too sure how things are now. I mean, it's. We're talking, sadly, we're talking about 20, more than 25 years ago. So obviously things change. The short answer to your question is that back in 99, 98, 99, when I was deciding MIT was simply not known, okay? So, you know, I don't think that most people pay as much attention to academic institutions in the US as people here in the US would. Now obviously I was at a very good university in Barcelona and the folks that had taught me knew about it and had views about it. But there was no sense in which ideology mattered too much in that decision. I mean, there were views that Chicago, you know, it's sort of. I think it was more about the freshwater macro versus New Keynesian macro, less so about more like right or left type of stuff. But, but the short answer is that I had a hard time convincing my parents that this was a good decision, not because of ideology, but simply they could.
- [03:09]
B
They didn't object to the Cambridge capital controversy exactly.
- [03:12]
A
They could not understand why, having been admitted to Harvard, for instance.
- [03:17]
B
Right.
- [03:18]
A
I was deciding to go to a place with an acronym that they had never heard about. So that's the short answer.
- [03:26]
B
Okay, well then if you don't mind me asking, why did you. So you, you could have gone to Harvard to get your degree, instead you chose mit?
- [03:33]
A
Yeah, what you're saying. Absolutely. So, so the short answer is. Well, this, I don't know how short, but I. Harvard admitted me, I think in part because I had a Fulbright scholarship. So I don't think we operate this way anymore here at Harvard. But back in the day, if you had some financial aid from back home, that, that gave you a little bit of an advantage. And I was elated. Right. But then MIT admitted me and I started talking to some people back in my university that were telling me, you know, that's actually the best PhD program technically, blah, blah, blah, that's where you want to go at mit. I don't know why exactly. They saw something in. I mean, I felt like they did really see something in me and they offered me, without even me asking for it, a full ride, basically. And they said, look, we want you to come here. We think that when you finish your PhD, you will want to stay in the US and be an academic in the US and the Fulbright scholarship has some ties and we don't want you to take it because we want you to be able to choose. And it was four years down the line and that was very persuasive. I went to Harvard and I told them, look, that's what they're telling me there. I'd love to come to Harvard maybe, but you know, that's. Obviously, there's no strings attached after I finish at MIT and Harvard, obviously, because probably my ranking was not all that high. I had been admitted precisely because of the funding. They basically said, well, you know, you have a Fulbright. If you want to come here, use a Fulbright. So that's one. I mean, there were many other things, right? I mean, an MIT is really a fantastic PhD program. And I'm not sure what I would have decided under equal financial packages, but you know, it obviously if you can make a decision without tying your future, that, that sort of. And that's what I did. And, and, and it was clearly the right decision exposed in terms of I had a blast there, the great program and God only knows what would have happened if I had gone elsewhere, of course.
- [05:42]
B
Right, okay, great. So then my next sort of warm up question here before we dive into the paper is how did you get interested in, you know, Bimbawk and I guess capital theory more generally, but like specifically in the Austrian Bomberkian tradition because I imagine at MIT they weren't courting you saying yes, we definitely want someone to come here because there's a hole in our department that no one's working on. Bamboo.
- [06:05]
A
No, this came much later. It came much earlier and much later and earlier in the sense that I would imagine that relative to the median academic economist in US universities, I'm perhaps a bit more versed in the history of economic thought. When I did my undergrad in economics in Spain, as you may know, there's much more specialization there. So I took like dozens, literally like more than 40 courses on economics and less on other things. And one of those courses was on the history of economic thought, which was taught by a very entertaining individual who had been in politics and you know, he was, he was obsessed with Schumpeter and the Austrian school. He was one of the introducers of Schumpeter's ideas about, into, into Spain in the 50s, 60s Etc. And so he did a very entertaining course in the history of economic thought. And we talked about, you know, Manger and schumpeter and Hayek, Mrs. Et cetera. Right. So I, I had a familiarity of where the school at least up to the 1950s, 60s, where it fit in the history of economic thought. And I knew how important it had been for like neoclassical economics for instance. And I've, I've regularly kept sort of reading History of Economic Thought. Now that's like way back right now. If then you have like 20 years of my career that are focused on very different things. I mean I've been working on models of global production, multinational firms, contract theory applied to general equilibrium models, very little capital theory and certainly not a lot of history of economic thought. But then 20, 19, I think, so not that long ago I was in Vienna and I go to Vienna regularly because my wife is from Bratislava, which you may know is maybe 50, like 40 miles from Vienna. So I spent chunks of the summer there and I co organized with a good friend of mine there a conference every year where we bring up a bunch, a bunch of international trade economists and have a couple of days of seminars. And I was always amused how folks go there. I don't know if you've been to Vienna, but it's sort of for. For young folks that maybe don't like fully appreciate the history of the city. They're shocked just how monumental the city is, right? It's like, wait, Austria is a small country. Why does it have this absolutely stunning capital? Now you explain a bit about the empire and this and that, and then you're like, well, but you know, we're having this at the University of Vienna. You know, this is the cradle of like modern economics. And so we were having sort of these discussions and over dinner we were having the discussion of just the importance of the Austrian School for the Development of Economics. And a local professor there recommended a book which is cited in the footnote of this paper that you mentioned, which is by. I wrote it because I always forget names. Janek Vassermans. I don't know if you've read this book. It's the Marginal Revolutionaries. How Austrian Economics Economists Fought the War of Ideas. Okay, So I read that in 2000 during COVID actually, and I found it, you know, some of the stuff I knew, but some of the stuff I hadn't fully appreciated. It was a fascinating read. And that sort of started reading the chapters on Bon Va work and like remembering what he was working on. And it just sort of made a lot of sense, this notion of thinking hard about the temporal dimension of production, thinking about the role of sort of time and sort of not only as a sort of a cost of time in the traditional way, but sometimes something that is necessary, that is productivity enhancing. And having thought a lot about sequential production processes and how they become global, I felt there was something, some concepts there that seemed potentially important that I hadn't seen in the type of formal work that I. That I typically I'm involved with. So that sort of led me to think, you know, you know, I do this with this, but I many other topics I see what's happening in the world. And as an academic that paid to think, well, we think, right? So, so what's happening? How do I think about it? What is the framework in my toolkit that I'm going to grab to understand this? Most of the time I have a framework. We can debate whether it's the right one or not. But I have a way to understand this to make sense of this, to criticize it or to praise it, to be worried about it or to be excited about it. But in this particular ideas, I didn't know about a formal model that would capture that and that led me down that path.
- [11:06]
B
Okay, great. Yeah. And to answer your question, I saw you cite that book and sort of given the history in the paper as to what led us down this path. And I haven't read that particular book, so I'll have to check that out. That's interesting. I have been to Vienna. I was there one time. I gave some kind of a presentation in actually the. I don't know if they call it central bank, but like a bank office in Vienna. It was kind of funny because I was basically giving a presentation saying why the central bank shouldn't exist or something. But anyway, so it was interesting. All right, well, let me, I think maybe Paul, for the benefit of the listeners here, maybe I'll just read the abstract of your paper and then we can start going through. Obviously we're not going to get into like equation 16 says such and such, but maybe just to give them the, the big, the big picture. So the title is an Austrian and Austrians in quotation marks Model of International Specialization. So it's Paul. And then you've got a co author, Adrian Coleza, also from Harvard. And here's the abstract. We develop a general equilibrium model of international trade in which the temporal structure of production is a key determinant of comparative advantage. Building on Bambovic's theory of capital, the model formalizes the idea that production processes with longer average periods of production or apps entail higher financing costs due to the time lag between input payments and revenue realization. We embed this insight into a multi sector Ricardian framework with endogenous interest rates under autarky. Countries with more patient consumers or more developed financial markets exhibit lower equilibrium interest rates and higher wage rates. With international trade. These countries typically gain a comparative advantage in sectors with longer apps, though the model can also generate multiple equilibria and unconventional specialization patterns. We extend the framework to include trade costs, global value change and international capital market integration. And then finally, empirically, we present evidence showing that countries with more developed financial systems export disproportionately more and more in sectors with longer apps, even after controlling for standard neoclassical and institutional determinants of comparative advantage.
- [13:11]
A
All right, so there was me realize that I need to make it shorter.
- [13:14]
B
Oh, was. Yeah, there's probably some sort of period of period round. Yeah, it was a roundabout exposition or Summary of your paper. Okay, so I was going to be kind of snarky and say, like, oh, so it's like Bombavik meets Krugman. Let me explain what I mean by that and then you can tell me like whether like. Yeah, that's kind of right or if like. No, that's totally wrong. But just for the. Because I think some of the listeners don't even know in terms of like, modern.
- [13:38]
A
Sure.
- [13:38]
B
Trade. Like, for example, Paul Krugman wins a Nobel Prize for his work on international trade. If you read some of his stuff where he's kind of like giving a big picture summary of what has changed. I think part of it is something like this where, for example, the case of the US we can understand why, if you're looking at Alaska versus Florida, Florida is going to send them oranges and Alaska is going to send them oil. That like just makes sense. And that's, you know. Why is that? Well, it's obvious because of natural endowments. But more generally, you could have two identical regions with the workers are exactly identical, whatever, at time zero. And if there are economies of scale in the production technology, just through happenstance, if one country goes down one path and another goes down a different path, you could see why 20 years later they each have comparative advantages and different things and they specialize in export to each other, not because the terrain or the resource endowments were different, but just because, you know, path dependence and one thing specialized in one versus the other. So, you know, partly the models, you know, model things like that and you know, combine different types of literatures. So is what you're doing just a particular example of that latter approach where you're saying one of the things that could make one country, you know, specialize in certain things is historically they've invested in longer production processes?
- [14:57]
A
Yes and no. Okay, so everything you said is absolutely correct. It's a very nice description of, of some of Paul Krugman's work. I would, I would say yes, in the sense of we have the,
- [15:17]
B
what
- [15:17]
A
we call neoclassical theories of specialization, Ricardian models, the Hexralin models, and then the Kruppman, what I'd like to call the Helpman Krugman revolution, so called Nutri theory was motivated by facts that we didn't see in the data that we, that were visible in the data that were not rationalizable with neoclassical models. And you gave a very good example of scale economies or sort of history matters and things like that. So in that vein, yes, what we're doing here is Saying some of the traditional determinants of comparative advantage, you know, a notion of capital abundance, of skill abundance, coupled with differences and say skill intensity across sectors are not sufficient in the sense that there's going to be residual variation in trade patterns that is explained by additional factors. Right. And Krugman might be just accidents of nature or country size or something like that. Where I would say that, that I'm less, we're less close to Krugman is in the sense that our model is actually other than this bombardic feature. It's absolutely neoclassical in the sense of having perfect competition, constant returns to scale. So there's no like every single assumption that Krugman built into models to kind of generate this new predictions, they're not there. Our model is actually in a lot of the papers in trying to argue that despite being a model in which we're going to have labor and capital, okay, as in the Hexralin model, which is one of the benchmark neoclassical models, the fact that we adopt this sort of Austrian notion of capital that is less tight to a homogeneous stock of physical capital, whatever that may be. And that would bring us to the famous capital controversies. If you think about capital as associated with working capital with credit, which is necessary to kind of generate this sort of longer production line length, then you get predictions even in an otherwise neoclassical model that are quite distinct. Okay, so that is what I would sort of draw the distinction that yes, it's a twist on neoclassical theory, but it's within neoclassical theory. And then there's more subtle issues, which is maybe there's a connection in Krugman in that we also have multiple equilibria. You refer to multiple equity. There are multiple equilibria that have to do with the nature of capital in those frameworks, which is very much an Austrian nature of capital, which emphasizes the dangers of over aggregate like aggregating capital into single things that may be convenient in some settings, but mainly to rule out situations that may well arises in equilibrium, which is what we illustrate in the paper.
- [18:08]
B
Okay, great. Let me ask you this then. Is your baseline result. Are you starting with identical regions and just showing if. Except perhaps maybe for prep consumer preferences for the timing of consumption and saying to people that are more patient, this is how that's going to unfold? Or is that not your approach? Like, do you. You get.
- [18:29]
A
You nailed it. That's, that's. I mean we do it more generally, but sort of the baseline case would be this one where it's not like the two countries are Completely identical. This, there is a, you know, which in the Krugman world, you can generate trade in completely identical countries. We need some difference. We need some different. What's, what's different here is that it's going to be coming from, as you said. Exactly. Patients. Doesn't need to be only patients with our model allows for interest rate differences that are coming from variation in the efficiency of the financial system. In some countries, funds are transferred from savers to investors in a more efficient way. Right. There's less loss along the way. That would also generate interest rate differences. But you need a little bit of a difference in that sense. Okay. And that's going to generate differences in availability of financial capital across countries and that generates the interest rate differences. And then the interest rate differences are going to affect the decisions of agents of just how long to kind of let production processes mature that are going to generate differences in production structure.
- [19:27]
B
Okay, so again, I'm taking the readers along for the. Right here, Paul. Okay, so I realize, yes, you cover a lot of stuff. It's 103 pages, a lot of appendices. So it's a, it's. You cover a lot of material. But, but yes, one way of thinking of it is just to get the listers and also like trained economists too, who are just curious, like, what are these people? What are these guys doing? I want to check this paper out. So yeah, one baseline result is again, two otherwise identical people, you know, groups at time zero, except that the consumers in one, you know, discount future utility less than the other ones do and so they're willing to save more. That can cause these things. But now am I right in thinking that like you could say, okay, but where's the bomb of Urkian capital theory come in? So the production technology in your framework, the people input labor hours and the, the, the gap, like the duration, like how long the labor hours are gestating in the production process before the finished good comes out. That's like one of the choice variables, right?
- [20:28]
A
And yes, I mean it's the, the bombaric. What do you call it? Verkin. That's a big, that's a big word aspect. Yeah, comes from the, the links be the link between the, the length of production. Okay. How many periods. I mean we have continuous time, but doesn't matter just how the interval, you know, how many seconds you spend in production and the productivity. And, and that's a little bit of a history dependence of the paper because in the initial version of the paper, which I started on my own and by the way, Adrian, my co author, is my student here and he joined the team at some point. He was making great contribution, so it made sense to bring them along. We started with this link, right? Then if you read the current version of the paper, a lot of the results really are more about the time phased nature of production, that some things, for whatever reason, they take longer than others. As long as there's interest rate differences across countries, obviously that's going to penalize longer production processes more than shorter ones if you're facing a higher interest rate. So you know, obviously the time phase nature of production is something that we associate with one Bavaric, but at some level goes back to Ricardo and Marx, they wrote about that quite a bit. I mean, I think about Bomber as sort of really taking this idea and running with it and arguing, well, that happens. But it's also important for understanding productivity. You want to understand how interest rates are determined. This productivity effects might be relevant for that. So you know, that's, that's in there. But, but a lot of it is more basic. Linking to the capital controversy, which is the time phase nature of production generates a lot of complications in thinking about what capital is and how do you, how you can aggregate it.
- [22:19]
B
Okay, so then my next question, just to clarify exactly how the Bombaverkian and I, I don't think I made that up. I think I've seen other people use that element. Makes your, your approach distinct. Let me put it this way. Couldn't a standard neoclassical growth theorist say. What, what do you mean? Like if you had a, like a solo model or something and then you posited that people had different intertemporal preferences and you, you know, had them separated, I could see how the one people that they used to have a higher savings rate so they accumulate more of the homogeneous capital stock and that increases.
- [22:57]
A
Yeah.
- [22:57]
B
You know, the level of consumption, you know. Anyway, could, is there some reason that. No. If capital is homogeneous, even though saving more makes production higher 20 periods out, that doesn't give rise to the specific things you're seeing.
- [23:13]
A
Right. Great question. So you sound a bit like my referee with my referee.
- [23:18]
B
Hopefully not referee too.
- [23:19]
A
Yeah, you're referee too. I could tell.
- [23:24]
B
Yeah.
- [23:24]
A
No, that's a good question. So the way I think it is as follows. If you take a solo model, which is a dynamic version of a model with a homogeneous K that maps to the Hexralin model, where also K is a homogeneous thing, and then you have predictions that say countries with lots of K relative to labor are going to produce and export capital intensive goods. And you test the theory by trying to measure this big K at the country level, which is hard, and then at the production level, at the industry level, you'd want to get a measure of capital spending, like spending on structures, on equipment, et cetera, relative to labor payments. Right. We're saying, yeah, maybe that might have some predictive power. But this story is different. It's like, don't try to measure physical capital. Maybe credit measures of credit over gdp, say, might be the, or measures of financial development, just how efficient financial systems are. This is the relevant country dimension. And at the industry or product level, what you want is not a measure of the capital intensity in a physical sense, but a measure of the, say, working capital intensity, or more directly a measure of just how long on average, this average period of production, how long does it take to kind of collect revenue relative to when you're spending on inputs. Right. And that variation, you could say, well, that's. Isn't that physical capital intensity? It need not be. Okay, that we show in the paper that this connection might break and in the data, actually, surprisingly, we can talk more about empirically how we measure the app. Our measure of the app turns out to be negatively correlated with standard measures of capital, of physical capital intensity. So that's why I see empirically it is a distinct source of comparative advantage. It's not about machines, it's not about buildings and things like that, enhancing productivity more in some sectors than others. It's really about the temporal dimension of production and the associated working capital needs and how it interacts with the financial system in different countries.
- [25:38]
B
Okay, that is very interesting. And yeah, just so you know, Paul, and for the listeners too, right. I wanted to kind of work through like the theory and then a big chunk at the end. I want to give you time to explain empirically, like why do you think this approach is fruitful? But let me read something from your paper that prompted me to ask you that so you make sure you understand where I'm coming from. You got this line. These peculiarities arise because while financial capital is homogeneous and freely mobile across sectors, the physical capital goods it finances are heterogeneous within and across sectors due to variation in the temporal structure of production. And so that's why I was wondering, like, is it how much do your results depend on, like why your results differ from a standard model that really you are not just assuming that, oh yeah, you can summarize the capital stock at time T with, you know, big K subscript T as Opposed to. There's. Because that's a. Just so you know, Paul, that's a standard thing that Austrians, a lot of times when they're trying to explain to people, well, why is our. Why do people. Why should people care about the Austrian school almost as. Oh, because capital goods are heterogeneous. And some people, like, yeah, yeah, yeah, whatever. We, we kind of know that with some. And so anyway, that's partly why I'm wondering, like, how critical is that to what you're doing in this paper?
- [26:52]
A
It depends on what you want to take from the paper. Right. So when we talk about the empirics and what we look into the data, what we find in the data, it would be this idea that indeed, in the data, longer production processes tend to be carried out in countries with lower interest rates. I don't think those differences are crucial, okay. In the sense that we could have. This is a prediction that comes out very cleanly precisely when, even though measuring the aggregate K could be complicated, the natural way to aggregate them would deliver roughly the right answer. In the paper, we also show that in situations in which that aggregation becomes trickier, the model generates a series of more surprising results. For instance, we show that it could be the case that countries with lower interest rates could actually end up producing goods, that when you measure working capital intensity, that working capital intensity is not higher. Okay? So that there's some sort of, what we call ranking reversals that can arise in the situation. Perhaps more interestingly, whether a country,
- [28:28]
B
whether
- [28:28]
A
the more patient country, say, ends up with a lower interest rate or not, it's a bit less obvious. There's multiple equilibria. Because if you happen to be, and I'll try to explain this in plain words, suppose you're a country with patient consumers. Then you, you tend to think that in that country interest rates are going to tend to be lower. But suppose that leads to, according to the model, that country to specialize in lots of sectors that have long production processes. Well, that's going to entail a lot of demand for financial capital because we need to sustain these long production processes with limited capital mobility across countries. That's going to generate a lot of local demand for capital which might bid up the interest rate. And it's not inconsistent that then you'd end up with a higher interest rate, but then you no longer have comparative advantage in, in this long production processes. Okay, so then you can get into these loops where you end up with, with a patient country, actually with a higher interest rate, which, which is not Something you'd get in like a solo style looking neoclassical model. So, you know, what we haven't done in the paper yet is, okay, what's the evidence for this sort of new, like completely new predictions that are hard to generate in neoclassical models? We haven't done that. The predictions we have, I'd say are distinct from available frameworks because they tell us we should be measuring K aggregate K and the variation across sectors in working capital intensity differently than we had been doing in the Hexcherle model. But we are not testing the sort of funky results that we can get because of that thing.
- [30:19]
B
Okay, fair enough.
- [30:22]
A
Super quickly. I mean, you may know there's this sort of related literature on re switching effects. Sam Wilson had a very summing up paper, classical paper, showing funky stuff can happen when you think you can aggregate K and you can't. Right. So we know that's true. And actually Samuelson was one of the few people going on record saying Joanne Robinson was right. Similarly, I guess the Austrians were right and that it's typically not possible to aggregate K. And here's proof of that. Now, the literature, as you say, they moved on and people say, whatever. Why? Well, because maybe we think these are funky situations that are not typically arising and we understand it's all approximations, which is models are just an abstraction of reality. And yes, it's not exactly right, but maybe ballpark right. So I'm not sure I have a view on this, whether this sort of funky things that can arise here are first order importance or not. But we thought that this is interesting, we might as well lay it out, maybe people will pick it up and test it in the future.
- [31:31]
B
Okay, great. So you actually just segued right into one of the things I was going to ask you anyway, Paul. So let me bring the viewers along to make sure they're not getting lost. Okay, so partly why I was pleasantly surprised to see people, you know, nowadays working on stuff and referring the average period of production. Was that a standard history of thought treatment says this is just dead. Like Mark Blogg and his history of economic thought. I think he. This might not be an exact quote, but it's definitely close to it. When he's talking about Austrian capital theory, he says that this, this re switching controversy was the, the nail in the coffin of Bombavik's original conception. And so just for the people at home here, the big idea is, the idea is, okay, if you take labor and natural resource inputs and if you're willing to invest them in physical processes that take Longer, in general, you can always find a higher physical yield, right. That, you know, you want to get water out of the stream. If you just go with your bare hands and keep walking back and forth to your hut, you're not going to get many gallons per hour that way. But if you first take the time to go fashion a shovel and you dig a ditch and you do this and that, you know, it might take you a year to get that whole thing up and running. But once you do, the amount of gallons per hour of your expenditure, all things measured, is much higher. Right. So the longer you're willing to, you know, make the production process take the, the more yield you can get in physical terms for the amount of inputs you put in. Okay, and so then. And then Bavaria, you know, tries to link that to. Oh, so that's why a more capitalistic country that, you know, they save more, they invest more in longer processes, they have a higher standard of living and all that kind of makes sense. But the re switching controversy, what they discovered in the famous, you know, Cambridge capital controversies in the mid 20th century is that you can come up with very simple examples. And Samuelson came up with a simple one, and he was throwing in the towel, by the way. He was. He had been on the wrong side of the debate originally said, okay, yep, yep, These critics were right saying that you can imagine a production thing where at high interest rates, one production process is favored over another, and then an intermediate range of interest rates, it flips that a different one is, you know, more cost effective. But then if you push it as race down even more, it flips back to the other one. And so he's saying there's no unambiguous way to rank production processes in terms of being more or less capitalistic in physical terms. Hence this story to try to connect like, oh yeah, the capitalists save more, it lowers interest rates and that's good for the workers. And that kind of justifies ideologically the, the interest payments to the cat. Like, no, that's all crazy because you. We can just show that doesn't work in general. So I'll stop there, Paul. So. And then, you know, blocks and see. Nail in the coffin. Bombaver was wrong. Do something else. If you want to talk about capital and interest theory. And so you guys are bringing that up again. So what's going on?
- [34:23]
A
Yeah, no, I mean, two answers. So the short one is sort of discarding a theory because of sort of some unconventional results that can happen and particularly carefully chosen examples. I mean, I don't Think that's necessarily a conclusion that you can draw. And you could do the same with solo style, aggregate K theory and Hexralin models. If you have things called factor intensity reversals, we can absolutely have them with well behaved production functions that also generates all sorts of funky results. And we didn't discard the theory because of that. Okay, so that's answer number one. Answer number two, perhaps more constructively. So part of what we do in the paper is formally derive an app measure, which is not new, actually. It's somewhat distinct from the bombard one because of continuous discounting, like the compounding of interest rates. But this is something that actually you can trace it back at least to Hicks, his value and capital 1939. He has chapters on the average period of production and shows exactly how one should measure it with properly with discounting. And what we show, which we were very excited about, we thought we were the first ones to point this out. And then I found a paper, I'll tell you the story later, in the 1970s paper that makes it very clear that this notion of the average period of production, despite all these things that you mentioned with Samuelson and stuff, there's always a very tight connection between this definition of the app and how interest rates affect production costs. Okay, Technically, the. I don't know if you want technical stuff, but what we call the semi elasticity of the unit cost of production to the interest rate D log C over Dr. That is exactly equal to this definition of the app, okay. Which tells you that if you want to measure how sensitive costs are to interest rates, formally that is exactly what you want to be measuring. And it's a formal definition that is very much tied to bombard, except for this sort of better treatment of compounding, which, you know, can, you can trace it back to Hicks. Okay. And that's very important because in neoclassical trade theory, that derivative, I mean, how factor costs, in this case interest rates affect cost is what determines comparative advantage. If a country has comparative advantage in capital intensive goods is because changes in the rental rate of capital affect capital intensive goods more than labor intensive goods. That sensitivity is key. And that's a sense in which, yes, the app, we're aggregating all this complicated stuff into a number for a production process, right? It's an object, it's a real number. But that's actually the key thing to pin to, to figure out where that thing should be produced, okay, if it's in a high or low interest rate country. So that's a sense in which even Though funky stuff can happen. Okay. And I alluded to Samuelson in terms of multiple equilibria this and that. The link between interest rates and, and, and where high or low app sectors should be produced, that is, that is not subject to kind of funky counter examples. So that's a sense in which I'm. And by the way, I mean hicks. But I have it in front of me. I'm sure you're familiar with Peter Lewin. Lewin, yeah.
- [38:06]
B
And Nicholas Kaczianowski.
- [38:07]
A
Yeah, they have a very nice paper which I also discovered in this and which we cite, which. Which makes a very similar type of point. Like, actually, it's not that. Okay. I don't think they make this semi elasticity point that I'm mentioning, which is key in our paper. But, but, but they, but I, I didn't get a sense reading that paper. These are pretty well known. Austrian economists are gonna get a sense that the app was dead. Okay. That there was at least a sub community of the Austrian school that was still believing in, in this concept.
- [38:40]
B
Oh, yeah, I may have overspoke there. Right. So for folks. So Peter Lewin and Nicholas Kajanowski. And I have. I've had Nicholas on Live interview Peter as well. And. Right. That's one of the things they do. They do a lot on, like the intersection of financial economics and Austrian capital theory. And yeah, one of their things was to say if you giving away that you can correctly deal with the average period of production in a way that doesn't. Isn't subject to these reversals and it does what you want. Also, apologist is a fun anecdote. There's this Austrian economist, Roger Garrison, he recently passed and his original training had been in engineering. And so one of his complaints was he said that, you know, economists don't keep track of units. And that used to annoy him in the, you know, in the capital literature, that they didn't keep the units consistent. But he always thought that the capital re switching thing was like a big nothing burger. Because he said, yeah, that's just showing that a quadratic, you know, the quadratic formula has two solutions, like, big deal. You know, so he never thought that that was a big deal one way or the other. Okay. So I. Thank you for going. I'm glad I asked you that to go through that. So maybe can you. And you kind of already said it a little bit, but just for completeness, can you summarize to say, if someone's like, okay, what's the. What's the payoff? Why Do I want to wade through this paper empirically? What kind of things do you think that your approach can shed light on? Whereas again, the standard explanations from the neoclassical or even the new trade theory people don't quite fit the patterns that we see in some of the data out there.
- [40:15]
A
Right. So, so there's a narrow, a narrow goal of the paper, like takeaway, and then a broader one. The narrow one is, yes, maybe we've missed that when thinking about capital and how it shapes specialization, international specialization. We might not just want to stick to this Heckscher alien physical capital idea. And we want to envision a notion where capital matters in a way that interacts with the duration of production. And empirically, part of the contribution, less of this paper because we just borrow this measure from a different paper I wrote with a different grad student, Italy Tubdanova, we suggest a way to measure this notion of duration, which is, I mean, if you think about it, it's hard, right? So if you go back to the Vixel timber examples of like the trees that are growing, you gave us somewhat different example, but same notion that if I wait longer for the tree to grow, I'm going to get more timber. I mean, you could imagine for particular sectors being very careful about how long they last, how long time improves their productivity. Think about whiskey or wines where we could get a sense of just how much productivity improves with time. Instead, we just. Take a different approach, which is partly motivated by theory, which is in the model, variation in the app should correspond to variation in things we can observe. Okay. And in particular what we show. And it sort of goes back to work that you may, you may have seen. Robert Dorfman, who taught at Harvard for many years, worked a little bit on this sort of Austrian models. And we kind of got the idea from there, which is use data, inventory data, okay? So if you are a firm that is producing things instantly, very, very quick production runs, you're going to produce and spit out of the factory like all the time, right? So there's, you know, there's going to be very little stuff sitting as inventory. When you are producing wine, say, and you have all these barrels that are, have different maturities and so on, you've got a lot of inventory in house, okay? So it's sort of natural, that inventory, and in particular what we call work in process inventory, not like raw materials or finished products, inventories, but this sort of work in process that's very much tied to this notion of there's a semi finished product that is at the factory. Okay. Or whatever plant. And it's still, there's still added value being added. So we have data for that. I mean, inventories data are widely available. And what we show is that the AP is very close in the data under some stationary conditions. It's very. It should be captured by the ratio of inventories to the cost of goods sold. Okay. And then those are two things that are always in financial and financial data you can find for any publicly traded company. That's very easy to do. And with Vitali, we just went and measured this. We measured this for all the firms in Compostat us various countries. We explored variation that was, you know, the extent to which firms in different sectors, like all the winemakers, are going to tend to have like, larger duration than all the milk producers. Okay. And indeed, we tend to see that there's a big sectoral component to this measures. There's a lot of variation. If you look at the top longer production processes and the shorter ones, it kind of makes sense. You can kind of sense that there's a temporal. You know, this industry is different in a lot of ways, but it does look like the temporal dimension of production matters. I think tobacco was one of the top five longest production processes. That had me a bit confused. And then at some point I gave a talk somewhere and people said, no, it makes a lot of sense. You need to like, and I'll put them in the sun or I don't know, I don't know anything about tobacco, but they were telling me, you know, that makes sense actually. It actually takes a while to do these things. So that's how we came to this measure. And then indeed it does look in the data that longer production processes, if you look at exports of countries at the country and sector level, countries with lower interest rates tend to export disproportionately of this, what we've measured being a longer production process. It's a statistical result. If you want to find counter examples, I'm sure you'll find them. It's sort of on average, this has predictive power and quite a bit of predictive power comparable to other standard determinants of comparative advantage like skill intensity, skill abundance, its interactions. That's in a nutshell, what we do.
- [45:14]
B
Okay, fantastic. So of course, folks, I'll. I'll put a link to Paul's paper, the version submitted. That'd be great. We just have a few minutes here left. Paul, I have to ask you. And this is like we're, we're chasing ratings here or whatever, but Given, you know, I was looking over your CV and how much in general you've done like on supply chains and things like even in the agricultural sector. Do you mind just sharing your thoughts? The. The Strait of Hormuz. And you know, there's different things like a lot of the analysts, not just professional economists, but just guys who write like newsletters for financial people and whatnot, they're just going over the raw facts about like the interruption, not just a crude exports, but fertilizer and things like that. And yet it seems like market prices are not nearly as worried about things as these analysts are. Do you have any comments on that?
- [46:05]
A
I've comments. I'm not sure. It's, it's. It's. I'm perhaps a bit less comfortable talking about these things than about semi elasticities. Let me just say. I'll answer the question. Let me just say one thing that I, that, that I left hanging and I don't want to leave it, which is before in the previous question you were asking me about what I draw conclusions. Right. Or ways forward and I mentioned this one about. Here's a new measure of duration. The broader point I wanted to make, which I think should be exciting for folks in your school of thought, is that I do think that some of these ideas of the duration of production, the links between time and productivity, they have much broader applicability than just figuring out the implications for trade slopes. Um, and, and that's by the way, that is my sense. This is a paper that I've presented in many, many places, many audiences, some international trade audiences, but talking to macro economists, they find this stuff very interesting. Okay, so you might have the view that that sort of economists of this neoclassical tradition, they've sort of built on, on these ideas and they're immutable. No, I think people understand that this ideas. There was some power in some of these ideas, that it's complicated perhaps to work with models with heterogeneous capital. But you know, there's a lot of. I sense interest. And so it leads me to believe that maybe there might be broader applicabilities and that people are going to start thinking harder and harder about, you know, time, an explicit notion of time and how it shapes production, how it shapes interest rates, which is sort of at some level one way to kind of one of the key ideas in Bombard anyway that I wanted to leave that.
- [48:00]
B
Which is why guys like me should not be snarky on Twitter.
- [48:03]
A
When you see someone talking about Twitter, I mean that's. This is supposed to be you know,
- [48:08]
B
do you ask a bird not to fly?
- [48:10]
A
I've been snarky there, too. That's, that's the way that platform operates. So. And I didn't know. I, I didn't take it badly at all. Okay. Hormuz. It's a. You know, I think I've been surprised, as you have about us not having seen more of a direct impact of this right away. I mean, we just saw the inflation. I don't know if you've seen it. We just got inflation numbers today, 4.2%. Okay. So we are certainly seeing price action now. Why? Because, you know, we. It's well understood in this modeling of production networks that even though oil is a relatively small input into most production processes, there's not a. In the short run, there's not a lot of substitution. Okay. So you might end up with price effects that are large. Maybe they're going to take a while. There's other work that shows that, you know, all prices are going to take a while to kind of transmit to different parts of production. And you can have models of why that is in the sense of maybe the direct production cost is not very large, but it's affecting me when I drive around and then I'm paying so much I need higher salaries. That's going to lead me to ask for a salary increase that's going to, you know, wherever I produce, that's gonna. So there's all these rounds of adjustment that, that, that might lead to prices not just going up immediately, but slowly, but more worrisomely. And that's why I think a lot of expectations about rate increases at the Fed might, might come, which is that it, you know, this is not something that's going to disappear the day after the war is over.
- [50:10]
B
Okay.
- [50:11]
A
So that's very focused on oil and prices. But I know that's where a lot of the discussion is. I think in terms of more global commerce, I'm perhaps a little bit less surprised that that has been less, less of an effect because that's sort of, I mean, I think for oil, it's a particularly important bottleneck. For other commodities, I think it is less so. I've seen data, though, about, like, shipping prices going up a lot, too. So I'm not, I'm not saying, I'm not sort of trivializing that this might lead to a bigger shock that we've witnessed so far. But it's, but again, I mean, I would imagine with shipping costs, something very similar than, than with oil would have. It's a very It's a small part of cost, but it, it might take a few rounds of sort of updating wages and this and that for it to end up showing up. But yeah, I mean, and then there's, I mean if you do need to have an oil expert to, to speak about that, but obviously there's been like adjustments on reserves and things like that that have attenuated, have attenuated the response. And in the case of the U.S. i mean, the U.S. is a net exporter of oil. Right. So in the country where we are, I mean, economic theory might suggest that on, on aggregate that's not necessarily a bad thing in the sense that the price of something we export on net is going out. That's a term of terms of trade gain. Right. That obviously will not resonate with people that are paying close to 100 bucks to fill up their tanks these days.
- [51:57]
B
Right. But the people who own the oil fields, they'll be happy. Right.
- [52:02]
A
And I don't know, I. An interesting thing there is whether to what extent this oil companies that are probably their stock valuations through the roof, to what extent are these US owned or because as you know, I mean, there's a lot of foreign ownership of US stocks. So I don't know that windfall, I don't know how much we're keeping of that. But obviously the median individual in the US is probably not cheering the higher gas prices right now. But you know, from a macro perspective, it's less obvious that this should be something that is having a negative impact on standard aggregates. Right.
- [52:39]
B
Okay, great stuff. So thank you for that. So folks, my guest has been Paul Antras. Paul, thanks so much for your time here and explaining your insights on this fascinating paper.
- [52:48]
A
Thanks Rob. I really appreciate the opportunity. And if your listeners have any feedback, any complaints, any ideas for future work, you guys know where to reach me. Thanks again for your time.
- [53:01]
B
Okay, sure thing. And then so folks, like I said, well, we'll link to his paper also some stuff I've done on like the re switching debate. If you want to see that spelled out folks, we'll link to the Show Notes page. And thanks everybody for your attention. See you next time.
- [53:13]
A
Check back next week for a new episode of the Human Action Podcast. In the meantime, you can find more content like this on mises.org. Sam.