The Story of Europe's Decline Isn't So Simple | Odd Lots

Yeah, I mean, there are some aspects here. So especially in countries like Spain and Italy, there are a lot of informal hours that are not captured by the headline figures. There are adjusted series for that. But my own crunching the numbers says it doesn't make that much of a difference for the bulk of the European economy, the richer parts of the European economy. So that's Central Europe, the Low Countries, and Nordic countries, that they do work much, much less.

Thanks for having me.

I mean, you know, if you look at how Europe has done in the last, you know, last few decades of the booming tech sector elsewhere, despite never having been the major innovator itself, maybe very early before the 90s, Europe may have contributed some of the foundational research that of course American companies are very good at extending and commercializing. It seems that consumers have pocketed the welfare nonetheless by lower prices for like tech goods and services, and that's despite the fact that US tech companies are, you know, pretty decently entrenched, you know, moats in Europe, legally entrenched ones. But I think with AI, it's going to be much the same. That's my impression, at least. I think it's probably true that Europe will not be able to create a frontier model any time soon. I think it's a good case that it shouldn't try to do so, and that it should instead focus on this issue of sovereignty, of not being dependent on either the US or China for its tech stack, which currently is very much focused on whatever actually suits the business needs of most European corporates. My impression talking with people who are in this field, who regulate AI at the European level, in their opinion, you know, near frontier models like Mistral are totally fine for the vast majority of business needs of most European firms. And the interesting fact, of course, that the fact that Mistral is launching models with a Chinese chassis, so with basically the weights of a Chinese model, because Mistral doesn't have the compute to actually train their own weights, raises all sorts of other dependency issues, I suppose. But so far I'm pretty sanguine on Europe when it comes to the AI race.

I mean, in a way that was the last sort of, it was one strand of a set of arguments that was pretty much, that were quite similar. They're all about, you know, in relative terms, Europe is falling behind and it's losing a competitiveness race versus the US. Jamie Dimon has been making these remarks frequently. He's saying he came to Europe last year and in a couple events this year as well, said you're losing because, you know, European GDP was 90% of the US ten, fifteen years ago. Now it's 70%. This is not good. Now, there are a couple of reasons to assume that it doesn't really matter if that were true, because as I said, you have to, it's a very producer-centric view of competitiveness and also of welfare, because the debate is kind of relating competitiveness to future welfare, saying if you're going to fall behind on competitiveness, and we define competitiveness by, you know, output or output per hour, in other words, labor productivity, your living standards are going to go down in material terms relative to the US. And you're not going to be able to maintain your social model. That's I think the main part of the Draghi report, my understanding. That this is, I think there's a different case for why Europe may not be doing well economically in its own terms. But even if you accept that this sort of relative decline is relevant, I think there's a problem with the data. So the Draghi report was published in November 2024, I believe, was commissioned by the European Commission. It's obviously authored by the hero of the euro crisis, former ECB president Draghi, who's also a distinguished economist in his own right. It argues a lot of things that I think I agree with, that many other people would agree with, that Europe has to, you know, remove internal barriers to trade and the movement of capital and credits, it has to, you know, try to get capital investment up, but quite, it's very, very low compared to the US in particular, that's falling behind in the tech sector in particular. So Draghi, this is obviously written well into the AI boom, Draghi is emphasizing the tech sector as the main driver of the American productivity advantage. It's just a question of how big that advantage really is. And I think that's the centerpiece of the report is this notion that the productivity ratio, so GDP per hour in PPP terms, has gone from 90-something percent to 70 or 85%, depending how you measure it, since 1995, when the rate of the tech boom in the US begins, and that implies a downward trend, a very sharply downward trend, and that it will continue if Europe doesn't do X, Y, and Z. I think that's where the problem lies, because I don't think this would be enough evidence to actually suggest that. True, really. Even if you accept that Europe might have not been growing as much as the US in recent years.

So I think you can definitely make, I mean, there's a fundamental difficulty with comparing economies. If you think that's important to organize your policy efforts around, you might want to have, you want to compare yourself to your peer economies, and you can do that statically. So at every given year you can compare yourself, or as a ratio. So what is the US's share of global GDP versus Europe's share of global GDP? But the specific statements that people make, including Jamie Dimon, everyone else, Draghi, you, we're also quite online. So you know that this is a constant theme online as well. Is there's a trend for the level trend over time. And the trouble is that because economies are very different, because they measure economic activity quite differently, we can't actually say anything about levels over time. So, so just to get to the actual source of the beef, which is the PPP adjustment of the GDP figures. You can't obviously just look at nominal GDP and exchange rate at market exchange rates like Dimon does, for instance, because all you're measuring is how the dollar has fluctuated over time, right? It's nothing to do with real output or anything. So you need to control for these exchange rate fluctuations, you control for things like the differences in working hours between the US and Europe, which are institutional, they're political choices about how you want to utilize labor inputs, how you want to trade productivity gains for leisure time. The US doesn't have any statutory limitations on working hours, and Europe has consistently decided to work less, but has remained employed at the same rates basically. So this is totally a choice to try to work less. And if you adjust working hours, and if you adjust for the relative price level, which is what purchasing power parity should do. So in other words, what does a French euro, how many French euros do you need to buy a dollar worth of consumption basically. These are constructed by the World Bank, and you have sort of a, now you have a comparison that allows you to compare price levels across time, in theory at least. The problem is with Draghi's measure, which is a frighteningly common measure, but I think this is well documented to be the wrong measure, but it does so by fixing prices at a given year. So basically it's measured in the report in 2010 dollars. So if you express this in international dollars, which is a synthetic currency that has the same purchasing power as the US dollar, and it's benchmarked to a year in which people have gone around and collected prices in different countries and created these ratios that are consistent with each other. But if you fix prices to a certain year, you have to use something to extrapolate between these different benchmarks. And the Draghi measure uses the real growth measures that the national accounts produce. So in the US and the European countries, what happens as a result is that you get this upward trend. So that data shows us a decline over time that you don't see in PPP measures that don't use the national account real growth rate, but they use relative inflation ratios. So how is the inflation ratio and inflation rates? What have they compared to the US inflation rates in Europe? And the problem is that we calculate these real rates differently. So in the US, what the BEA does more aggressively than the European offices, for economic analysis, is that they adjust prices, particularly in the tech sector. And this is why this is very relevant for Draghi's particular argument, much more aggressively for improvements in quality. So if the computers, I'm currently on a MacBook for instance, sell at the same or similar nominal price, even though the specs are much better, that's then measured as a decline in the quality-adjusted price. So the same amount of nominal spending on tech services and goods shows up as more real growth in the US than in the EU, because applying that deflator more aggressively means the inflation in that picture is lower. So real growth is higher at that point. But we don't know exactly how much that is, about 2.2 to 2.3% annually according to some studies. But over time if you compound that every year, and that creates this wedge between the EU and the US, and that is, at least in theory, to a large part due to typically these method differences rather than actual underlying productivity growth. And that's I think the main issue. And then from there on, I think the other issue is, well, should we then look at the economy in purchasing power terms at all, or should we look at it purely in real gross terms? And that's a whole different discussion altogether. And that's I think where the discussion has been going, between Paul Krugman and some European economists who are sort of trying to argue between each other over whether we should reform it, become a bit more like the US, which would result in, you know, loosening labor regulations, cutting welfare spending, and of course, making it easier for tech startups and so on to spend and invest in everything we can.

And it depends on whether you think these differences between countries are actually that severe and why they actually occur. But you can of course. And we end up, by the way, with the notion that we don't have a good measure to compare trends over time. So whatever your judgment call is, the thing that's presented as a robust stylized fact in the Draghi report, put out by Dimon and other people, is simply not such a thing. It's not established by any sort of solid economic data. What we can do is we can use PPP in current prices to compare level snapshots. So we can say in the year 2024, what is GDP per hour, purchasing power adjusted, in the EU and in the US. That comparison I think puts the US slightly above the Western European aggregates. I also show that if you actually then exclude the top 10% of the income distribution, that advantage goes away entirely. So the US is slightly richer in aggregates, but not at least on this measure, but not quite at the level of Germany and some of the sort of lower Low Countries, Nordic countries. But it's very top heavy. And indeed the bottom half is actually significantly worse off than any single European country, but for Portugal. Now we can say, well, this is the most defensible output measure in my view, comparatively. But we can say, well, maybe people care about income, you know, money in the bank. And maybe politically we're so different that people don't care about leisure time that much. And people are fine with living in a more top-heavy society. How much money do individuals and households in the US actually have? Because I mean, consumption is a large part of what we could define as material well-being. Right. So I think it makes sense to look at consumption in particular, what's called actual individual consumption, which is, I think, a better measure than income because income is hard to compare as well. It's measured differently, but also actual individual consumption includes in-kind benefits and includes the benefits of publicly provided healthcare and so on. So it should control for these political differences that you were hinting at and the fact that in Europe there are sizable welfare states and, you know, larger protections and a greater amount of public goods provision. If you adjust that for working hours, the US is pretty unremarkable. And what you also have to do because of these political differences is you have to look at the way these consumption figures are distorted by some of the things that the US is completely unique in. I choose, I think, healthcare as an example and writing about this because it is quite unusual. It's the only advanced country that doesn't have a publicly funded system, despite it being very well established for many, many years by American economists that actuarial insurance of private insurance doesn't work for healthcare for various reasons, and you end up with very high costs and poor provisions. And indeed, the US spends much more on healthcare, both as a share of GDP than most other European countries. It's a function of having this private system that costs a great deal and still doesn't provide the same kind of coverage or quality. It's, I think, quite useful to look at empirically how healthcare spending actually compares to other OECD countries over time. And one model that I run myself on this is that the US spends around 50% more on healthcare per person than you would expect relative to what advanced countries actually spend on healthcare. What the consumption implies, in other words. So the level of consumption that you have for how rich you are, in other words, predicts a certain amount of spending. But the US residual, say the difference is 50% higher. And in total that results in around between 1.7 and about 1.9 trillion a year that's spent in excess of what you would expect of a country at that income level. And the difference is, I think this goes back to these political economy differences. The reason is prices, really, and the fact which is well established by people like, you know, Uwe Reinhardt, the great Princeton professor and health economist, that the US price is per unit of care just much higher. And utilization is very, very poor and outcomes are generally worse. If you look at average life expectancy, mortality rate for certain procedures. So this is part of this political economy which is more dominated by rent extraction and middlemen. So prices are high in the healthcare sector because you have hospitals, insurers, you know, these trade associations that can take a cut, administrative costs and all these things which make a large part of the income distribution of the top 10 to 20% very wealthy, but it creates enormous costs for them. Much of the rest of the population, especially the uninsured, which I think, I think 27 million by some accounting, and even those who are insured by their employers have large out-of-pocket costs. And the headline is, well, some people object to using purchasing power parity to actually adjust for these figures because, and indeed income figures, because they say, well, it's kind of hiding something that makes the, that is actually a sign of strength. In other words, they assume that's the reason why costs in the non-tradable sector, like healthcare, are very, very high is because, well, you have this very powerful productivity growth in the tech sector and elsewhere and very high incomes that people are sort of wage arbitrage themselves into the non-tradable sector. It's their high incomes that are inflating those prices in healthcare, legal services, education, whatever. That assumes that these prices are not a function of rent seeking, which unfortunately, I think the preponderance of the evidence shows they are.

Yeah, I mean, there are some aspects here. So especially in countries like Spain, Italy, that a lot of informal hours that are not captured by the headline figures, there are adjusted series for that. But my own crunching the numbers show it doesn't make that much of a difference for the bulk of the European economy, the richer parts of the European economy. So Central Europe, the Low Countries, and Nordic countries, that they do work much, much less. And if indeed it meant that the Europeans were employed at lower rates, that would be concerning, and it would probably point to the fact that, oh, this is a labor market failure, and people may not be able to work or they're not working because they're getting diminishing returns. But there's again, there's good empirical evidence, including by people who co-authored the long-run productivity figures that go into the Draghi reports that show, well, actually, it's sort of a substitution effect where productivity is causing hours to go down. And hours going down is actually increasing productivity. So it's not really, there's no evidence to think that's, first of all, it is real. People do work less. And there's no reason to think that that's somehow a sign of dysfunction underlying it. I think I would also just say it's kind of econ 101. So labor is a disutility, and it's precisely that because leisure time, the value of leisure time is greater than nil. If you assume otherwise, you're sort of contravening a very basic macro principle. I know, what I should say is that there's good reason to suggest that Western Europe is probably more wealthy than the US in the aggregates, even if it has probably declined relatively over the last six, seven years. And I do think there's robust evidence to show that, but of course internally there are huge differences. So I think we should care that Europe, in its own terms, has been doing incredibly poorly over the last ten, fifteen years. But that's due to policy choices, especially macro policy, which the US does comparatively very well. And of course, there are internal differences. We should care that the Italian South is extraordinarily poor. It has been declining in absolute terms, even when the entire economy has been stagnating for twenty years, unique in the G7 economies. But we shouldn't care, for instance, that Michigan in the US, we shouldn't care that Michigan has been outgrown by California for twenty years, even if Michigan has been doing well and it's been diversifying its economy. And this I think is the main takeaway when comparing the EU as an aggregate to the US, like we have our own problems, but the relative picture shouldn't matter that much. I don't think like there's a lot of work to be done, but it needs to actually mean becoming more like the US, which may not be politically amenable to most people here anyway.