 Good evening, everyone. I'm very pleased and honored to welcome Danny Rodrik to deliver the Jean-Marie Lecture on the real-world economics of globalization. I think it's fair to say, as a matter of introduction, that Danny has shaped the way we think about the interaction between trade and development and the development economics more generally. You're probably one of the very few scholars who can address these multiple dimensions of globalization across all dimensions, trade, finance, and, importantly, the political economy underpinnings of globalization, which I'm sure you're going to address tonight. And these are all matters which really underpin shape impact a lot of what we're doing here at the ECB. So I very much look forward to listening to you. And we'll have time to interact with the audience at the end of your lecture. So thank you very much. The floor is yours. Thank you very much, Benoit, for the invitation and for asking me to give this lecture. Thank you also, Luc Levin, for the invitation. Giving this lecture associated with Jean Monnet's name is a true honor. I've always thought of the European Union as the premier act of institutional engineering in the post-Second World War period of the world economy. And no one worked harder to make that happen than Jean Monnet. And in part, stimulated by this invitation, I read his biography recently, autobiography. And it's really quite striking the amount of energy he devoted, shuttling among the various European capitals and trying to get all those political ducks to line up, to advance the process of European integration. He always maintained his hope and this optimism that, in fact, Europe would end up integrating politically and not just economically. And I think it's a hope that seems to have received a few negative shocks recently, but I personally hope that hope will be fulfilled at some point. So thanks again for this opportunity. I'm going to be talking about, I gave the title, Real World Economics of Globalization. And that may sound a little bit like I'm going to be suggesting that there is the economics of the seminar room or the conference room, like what we've been having here. And then there is economics of the real world, and those two are quite different things. Actually, I'm going to suggest something very different that the way that globalization has worked out is pretty close to what some of the first principles of economics would have told us. And perhaps we should not have been surprised by the political backlash as much as we have. I think one of the key teachings of conventional trade theory is that stark redistribution or sharp amounts of redistribution is really the flip side of the gains from trade, particularly during the advanced stages of globalization when barriers tend to become small. I think thinking about many of these issues of compensation would have made us fairly skeptical about the possibilities of compensation, both for economic and political economy reasons. Looking closer at the kind of globalization we're having would understand that it created a second layer of issues having to do with institutional arbitrage and the clash of values that that generated. So it went beyond simply income redistribution. And so all these considerations then, really, which I'm going to be developing and going over, then suggest the final question that I'm going to ask if all of this should have been obvious, why didn't we pay more attention to, in fact, what our teachings are predicted? And so I want to end the talk with a few comments about where I think economists have gone wrong in the hope that maybe there are some lessons here for the future. But I think it might be useful to start with a little bit of a historical detour, going all the way back to the end of the 19th century. Because it's a way of making a couple of connections. One is to underscore that, in fact, populism is not something that is new. In fact, the very first self-consciously populist political movement in history developed in the late 19th century in the United States. It was the People's Party. William Jennings Bryan, who ran for the presidency twice for the Democratic Party, was very much a symbol of the populists and spoke for the People's Party. And many of the complaints of the populists and the populist movement was based, essentially, among the arose out of the difficulties and the complaints that farmers in the southern and western parts of the United States had. Because they were highly indebted during a period of price deflation. And they felt very high real interest rate burdens. And they blamed the northeastern capitalist and financial establishment and, of course, the gold standard for their problems. Because the gold standard was a conveyor through which the price deflation was transmitted into farm prices and the decline in corn and wheat and other commodity prices. And so we can see, from that very first episode of the rise of populism, that there was a connection between advanced stages of globalization. Here, the gold standard and the kind of populist movement that arose in response. This bit of a quote that I'm sure you all are familiar with. And it's probably the most famous piece of political oratory in US history, this famous statement, you shall not crucify mankind upon a cross of gold. So it was this early cry from the heart against what today we would call policies of austerity. And sort of making the tight connection between a hyper-globalized system of international finance and the rules that that imposed on national economic outcomes and the distributional consequences that it had on large segments of society. But let me come back to the present and particular sort of the main theme of my talk, which is sort of what does basic economics say about opening up and globalization? And I would say that there are three key things that we typically stress. We're actually sort of we stress in decreasing order of significance. Of course, the first one of these things is that there are gains from trade, that reducing barriers at the border generally enlarges the economic pie. Now, there are all kinds of caveats here. If you're not in perfect competition, there are externalities and so forth. But leaving those caveats aside is a strong presumption of gains from trade from the basic competitive model. But secondly, that in fact in virtually every model of gains from trade or comparative advantage that we work with, with the possible exception of single factor models like the Ricardian model, that we're not going to get everyone to actually win. And they're going to have different parts of society that's going to be worse off. I'll come back to this in a second to emphasize the generality of this result. And third, that in the standard competitive benchmark model of the gains from trade, that in fact the amount of redistribution that one gets relative to the net gains or relative to the gains from trade become larger and larger in the advanced stages of globalization or as the barriers that we're going after to remove become smaller and smaller. So I think that last point in particular I think has some relevance to why advanced stages of globalization become politically more contentious because often it turns out that their first order impact is more likely to be redistributive than net gain creating and that of course is going to have some implications for possibility of compensation which I'll come back to in a letter later. But let me just say a little bit about these stark distributional effects because I think it's not always understood how general those distributional effects are. Of course the key paper here is the famous Stolper Samuelson theorem from 1944 which proved that not only that there were relative losses in gains but that in fact there were absolute losses and gains that in this particular case in a two by two model two factors of production, two goods and factors of production perfectly mobile between the two sectors that owners of one of the two factors of production would be made necessarily worse off with the opening of trade. So it's not that just one group let's say unskilled workers become relatively worse off or that they're not getting most of the gains. They get some of the gains that in fact they're absolutely worse off that they're in real terms that they've lost up. Now the point about this theorem is that in fact even though this one Stolper and Samuelson proved it in a very specific kind of a context where you have only two goods, two factors and perfect mobility of factors it is actually one that generalizes very, very easily to in fact a much broader set of circumstances and this generalization states and really the generalization relative to the sort of the competitive benchmark. The only thing that it really requires is that the importing country continue to produce the imported good. In other words that you don't have complete specialization that as long as that happens there's always going to be at least one factor of production that is rendered worse off by the liberalization of trade. So the Stolper-Samuelson theorem generalizes quite to a very wide range of circumstances as long as there is some overlapping of domestic production or very, very close substitutes with imported consequences, imported commodities and this is really a consequence of what trade theory is called the magnification effect that is that the factor price consequences are magnified relative to relative prices. And the question of sort of what kind of distributive margins will these distributive effects operate is very much a function of how we divide up the economy whether we think of different factors or different regions, different communities increasingly in models with heterogeneous firms of course different types of firms. So these gains and losses will be also distributed unevenly amongst them but this specifying those margins is obviously very specifying the right margins is obviously quite important and empirical work. I think one reason that early work by trade economists missed out some of these large distributional effects was that it focused on relatively large aggregates of skilled versus unskilled and that the reason that the more recent work has found bigger effects on the distribution is that they actually look at another margin that seems to matter more empirically which is that focusing across communities and it turns out that labor even in the United States is not that mobile across spatially. So some point number one here is just this distributional effects are real sort of the very models that we use to teach our students and inform the rest of the world about the wonders of comparative advantage and the gains from trade almost necessarily imply very stark distributional effects that some people will be worse off. Furthermore, the magnitude of these distributional effects tend to rise relative to the gains from trade as we advance in globalization or as the barriers that we're removing become smaller and smaller and that really is a consequence of straightforward public finance economics which is that when you think about trade tariffs on trade or trade restriction essentially as a tax is a tax on trade and we know that the cost of taxes rise with the or the efficiency costs rise with the square of the tax rate. So as the tax rate or the tariff that we're removing becomes smaller and smaller, the net gains from trade that we're generating become disproportionately smaller whereas the distributive effects that we generate are approximately linear in the change in relative prices. So in fact, the effects on redistribution remain linear whereas the gains become sort of disproportionately smaller. So if you calculate this simple ratio of what I once called the political cost benefit ratio of opening up which is assume all redistribution is politically costly, so you put it in an enumerator and you put in the denominator the net gains from trade. So the political cost benefit ratio is the absolute the sum of absolute redistribution that is done relative to the net gains from trade. This ratio becomes larger and larger as we end up chasing smaller and smaller barriers. How much? Well, I'll just show you a couple of very simple kind of back of the envelope type of calculation just to see how outsized these effects are before I turn to some of the empirical work that you might already know. One is in the simplest case, if you look at the effect of a tariff in a very partial equilibrium and there it's very easy to relate this ratio of redistribution to efficiency to three parameters only. These parameters are the share of imports in GDP, the import demand elasticity and the size of the trade barrier itself. It turns out that the political cost benefit ratio is just going to equal to the inverse of the product of these three parameters. If you look at the table there as to what happens as we reduce the tariff that we are eliminating going from 40% tariff to about a 10% tariff, this political cost benefit ratio really becomes very, very large by the time the tariffs are at the level of 10% which by the way is already higher than what manufacturing tariffs are presently. This political cost benefit ratio is already at the level of 20 and what 20 means is $20 of income is being reshuffled across different groups per $1 of net gain that is being generated. In general equilibrium you get sort of very similar kinds of results so here if you look at, this is a very simple, just a simple two sector model but if you just simply calculate the proportional change in the wages of low skill workers relative to the overall really income increase in the economy that rises from a ratio of about five to a ratio of almost 80 when you go from 40% tariffs to about 3% tariffs. All of this is just follows straight forwardly from the very standard assumptions of the baseline competitive model, the one that we teach comparative advantage with. Now there has been now as I'm sure you know sort of a number of empirical papers that have looked at retrospectively on some of the major trade shocks and economy like the United States has experienced. We have some evidence on NAFTA. NAFTA was a very, very small potato for the U.S. economy as a whole because Mexico is relatively small and U.S. tariffs were already pretty small relative to Mexico to start with. So if you were looking for the effects of NAFTA on the U.S. economy as a whole, if you ask what are the net gains from trade or if all the efficiency gains, basically people have had very hard time uncovering or finding out a very significant effect. So the most recent and the most sophisticated work on this finds that the net gains from trade that is created for the United States from NAFTA for the economy as a whole was of the order of 0.04 percent, 0.04 percent. So even less than half of one percentage points of GDP. About a recent study by Hakopyan and McLaren finds that the relative wage effects in fact were very large. In fact, a couple of orders of magnitude larger that wage growth in the most affected industries was reduced by about 17 percentage points in the most affected communities by about 10 percentage of points. So these are the communities or the industries that were the most exposed to opening up to trade to Mexico where it had the initial tariffs were the highest relative to start. Of course, the much better work, much probably much better known work of author and co-authors on the China shock and they found that large employment effects that were not only significant, but also they find actually quite sustained over a decade or more so that these employment effects lasted and had significant effects well into a decade later. If we want to understand, of course, sort of the distributional effects of globalization, we need to go beyond the conventional comparative advantage models of trade and goods. But of course, increasingly globalization also became a kind of an asymmetric opening up of the world economy where capital, both physical and financial capital became increasingly mobile while labor was for the most part was heavily, heavily controlled. And there are sort of immediate implications of this asymmetry and there are a number of empirical papers that have started to look at these. But again, simply on the basis of first principles of economics, if you look at sort of what would be the implications of this kind of asymmetric mobility, you'd conclude that there would be significant distributional impacts in terms of those industries were bargaining or rents were important in terms of the distribution of the enterprise surplus between capital and labor, between employers and the workers. There would be significant impact on the incidence of shocks as sort of, with capital being able to sort of mobile across borders. And of course, there would be significant impact on who bore, who was bearing the tax burden and the tax burden would shift from capital to labor. And of course, anyone who's familiar with what has happened to the corporate income taxes around the world, the explicit manner in which tax competition has been used as an excuse for reducing corporate taxes. Most recently, of course, in the United States is aware and of course the work, important work of Gabriel Zuckmann and his various co-authors shows the relative importance of this. There's a, I just wanna show you a couple of slides from a recent interesting work at the IMF that looks sort of in some way summarizes the kind of theoretical or conceptual points that I've been making with respect to the effects of globalization, both on inequality and on the gains from trade. So this, with respect to inequality, looks at the effect of inequality which is the vertical axis against the measure of how globalized a nation is. And the key finding here is that the more globalized you are, the more adverse the impact of globalization on your domestic income distribution. This one looks at the effects of the growth benefits of globalization. How much growth do you get from globalization? And it shows nicely sort of this diminished return, diminishing returns to globalization that the returns to globalization are greatest when you're relatively little globalized but they tend to converge towards zero when you're very highly globalized. Now, when we talk about these distributional effects, one reason that I think economists don't pay much attention to these distributional consequences and I'll come back to this is because in the back of our mind we have this notion that if the pie is expanded, if there are net gains from trade, that those gains could be redistributed, that there could be compensation. And then this possibility of compensation allows us to focus on the efficiency consequences. But we know that compensation is likely to be distortionary on its own account that lump sum transfers are not going to be possible. And therefore in practice compensation is going to involve economic distortions at several margins. There's only one paper that I know that has actually looked seriously at this in terms of what would you suppose that you were actually compensating individuals through the tax system and that these taxes were affecting labor supply or other irrelevant distortionary margins. How much would it eat up from the net gains from trade to compensate the losers? And this paper by Polantros and co-authors finds that significant amount in this case about 20% of the gains would be eaten up. I would say that this is a kind of understatement because of the way that actually trade barriers are treated in this paper because they are simply of the iceberg variety. So the revenues that are created by tariffs or quantitative restrictions, those themselves are not taking into account and putting them back in would create larger ones. If you just simply look at, if you take, go back to the kind of simulations I presented about how in our simple baseline or benchmark models how much redistribution we get per efficiency dollars, per efficiency gains, and you put in very sort of moderate values for the excess burden of taxation, such as 10 cents on the dollar, then in fact it's very easy to get to the point where you're eating up more than the gains from trade just to compensate the losers. So that simply from the perspective of feasibility, when you get to the advanced stages of globalization, it becomes very, very difficult to actually undertake that compensation. And there's of course a further problem with respect to compensation, which is that typically compensation turns out not to be time consistent. This is very clear in the case of the United States where the compensation takes the form of trade adjustment assistance, and every time the United States signs on a trade agreement, workers are promised compensation through a particular or an expansion of a trade adjustment assistance. Now of course the problem with this is a typical pull to the economy timing consistency problem that once the trade agreement is signed and therefore the status quo is almost irrevocably altered. Of course, we know with Trump that you can get out of trade agreements, but it is not costless. And therefore what that means is that you've actually irrevocably also changed the nature of the political game and it's going to be very difficult for that, the incentive to carry out that promise is not going to be carried out. This is by the way exactly what has happened with the way that TAA or trade adjustment assistance has worked out in the United States that Congress repeatedly has not provided the kind of funding that would make TAA an adequate compensation mechanism. And every study that I've seen shows that the TAA as a result has been really a very ineffective kind of compensation mechanism. That doesn't mean that compensation never works. I think in fact it works in Europe and it has traditionally worked in Europe, but it's worked when it is part of a kind of a political settlement or a constitutive political bargain and therefore is embedded in the social policies of the country so that in Europe of course as you know there isn't separate adjustment assistance policies for the losers and the losses are taken care of through the welfare state arrangement and because they are embedded in the welfare state arrangement, then to some extent of course you overcome this time inconsistency problem of compensation. But of course these welfare state arrangements in turn have become weakened in recent years because of in no small part because of the difficulty of taxing capital and employers in a world with financial and capital mobility where the tax burden has shifted increasingly onto labor and it's become much more difficult to sustain these generous compensation mechanisms and in turn there's work of course that shows that the weakening of these safety nets in part helps explain some of the rise of nativists or extreme right parties. This wasn't just sort of a kind of a textbook globalization. This was a globalization that also went significantly in the direction of what one might call a kind of a skewed institutional arbitrage. That if we think about this traditional model of trade liberalization under the GATT system being a model of what my colleague Robert Lawrence calls shallow integration that is that you basically are going after barriers at the border, the trade barriers, import tariffs and non-tariff barriers. Increasingly globalization or trade agreements and financial globalization became about rewriting the rules behind the border and so that was sort of a more of a deep integration kind of a model and when you think about sort of what that kind of institutional arbitrage does is often creates clashes in values about sort of who's going to have property rights over who writes the regulations, the value of institutional diversity, the value of democratic voice in setting your own regulations and so forth and that those are the kinds of things that are not properly or adequately or legitimately addressed through compensation or financial transfers because they're much more sort of entail clash in values. And I think here it's something that maybe is a little bit further away from the first principles of economics but I think it's only a kind of a slight extension of our horizon that when we think about the kind of fairness issues that arise in the context of institutional arbitrage is that there's a very big difference between certain types of trade that takes place across a national border versus a trade or an exchange that takes place in a domestic market economy because in a domestic setting there is a level playing field in terms of two competitors A and B operate under the same ground rules under the same set of legal and institutional setting but when one of those firms B interacts with a firm in C across the border in some other setting there's necessarily both a kind of gains from trade which is the one that I've been talking about the conventional gains from trade but there's a second effect that comes from kind of institutional arbitrage to some extent that trade or outsourcing is a channel through which some of the institutional arrangements in the other country in the home country of the exporter is a seeps back into the importing country and you can see that in terms of sort of I've already mentioned the corporate tax competition but there are implicitly all of these channels in terms of how you deal with sort of financial arbitrage or regulatory arbitrage in terms of arbitrage in domestic norms about how you deal with workers what's an acceptable redistribution sort of employee employment norms, labor standards and so forth and sort of all these I think raise legitimate concerns about whose standards should apply whether it is the home country standards or the host countries and of course this is an issue that the European Union is very familiar with and occasionally the European Union has in fact as in the case of the posted workers agreement very recently has actually decided to settle that it would be the host country rules that would apply even if you would actually undercut some of the conventional gains from trade as a result in other words Polish workers coming into France would have temporarily would have to operate under French rules that reduces the gains from trade it's sort of like saying goods produced in China or in Cambodia or in Myanmar have to be produced with labor standards that apply in Europe or in the United States so it does reduce the gains from trade but it's a way of dealing with this kind of institutional arbitrage. There is evidence that in fact people see these different kinds of trade differently and they have different views on how government should respond so from an economist perspective whether a worker gets displaced because of some demand shock or some technology shock or some trade shock shouldn't matter if a worker is displaced you deal with it if you think you should provide compensation or social insurance you do it but you don't really ask why did that worker get displaced in the first place? In the real world I think people care about this recently Rafael de Tela and I ran a survey where we asked respondents to respond to a variety of different kinds of shocks we gave them a newspaper story about a factory closing and workers being at risk of losing their jobs and each one of these treatments gave a different story about what the source of the shock was and those shocks are shown at the horizontal axis one of them is simply technological change automation another one is a kind of a demand shock where there are changes in consumer tastes a third one was a case of bad management the plant might close because the managers had made mistakes or mistakes and the last two were trade shocks one was a trade with a developed country and the last one was a trade with a developing country and what's striking is in fact how different, how diverse were the responses in terms of whether there should be any compensation at all whether the government should help those who were losing their jobs and whether government should respond by providing import protection two things, one is that what's striking is that when the firm, the treatment is a kind of a trade shock not only were respondents not interested in providing financial compensation in fact their willingness to provide financial compensation relative to the control case actually dropped and all of these results are statistically significant by the way so when you're dealing with trade shocks somehow the conventional moral intuition of the ordinary sort of respondent is that you don't deal with these with trade shock with compensation, how do you deal with it in fact you deal with it very much through protection and the second point of course is that there's a big difference, quite significant difference between how they respond to trade with a developed country as opposed to trade with a developing country that the trade with a developing country calls forth for a much more protectionist response relative to trade with a developing country. So now let me, after having suggested that in some, in a real sense, many of these sort of, the political consequences can be read off from some of the first principles or implications or relatively minor extensions of the kinds of things that we teach. Why in some sense were we not better citizens? We should have known better. So I think there's, in a way there's sort of two smaller questions in this, where did we go wrong? And secondly, maybe more controversially, why did economists sort of become cheerleaders for the wrong kind of globalization? Because you can imagine globalization that was more balanced, one that sort of really took compensation seriously and one that ensured that the rules did in fact ones that were broadly beneficial. I wanna say just two points in closing as to why sort of how I think about this. One is that I think there was a, what I would call a kind of a naive political economy model in our minds, in the minds of, this is maybe more true of international economists in general and trade economists in particular. And so this was, I say naive because in some sense it was never made explicit and making it explicit maybe would have made us a little bit more cautious. This was brought to mind, sort of brought home to me when I wrote this little monograph two decades ago that was called, has globalization gone too far? And I sent it around to a bunch of economists and one very liberal in the US sense and sort of economists sort of came back to me and said, well this all seems fine but don't you think you're going to be giving ammunition to the barbarians? So the story was that there were no barbarians on this other side who were the protectionists and sent by expressing some skepticism or doubts very much along the lines of what I've been talking to you about that this was going to be providing ammunition to the barbarians. So what is the model? What is the explicit political model? Political economy model under which that makes sense. I think the kind of implicit model that many of us or many economists carried in their mind was that you have a bunch of rent seekers, self-interested rent seekers, but that these groups are almost inherently on the protectionist or the anti-globalization side that these protectionists are the politically dominant force and therefore the resulting political economy equilibrium is biased towards to little globalization and therefore leaning on the globalist side sort of leaning against the wind so to speak is almost always desirable from the economic perspective. So this is the underlying political economy model. But I think in the real world there are really barbarians on both sides of the issue. I think the kind of the multinationals, the pharma groups or the international investors or the bankers who are pushing for certain kinds of rules had their own self-interest in mind as well and many of those were not necessarily broadly beneficial or indeed even net net net gain creating. So I think that was sort of having a more explicit and better based model of political economy of policy making in the international economy may have been better. The second thing is probably, for lack of a better term I would say a kind of a first best mindset. And I think this comes in two variants. The academic version of the first best mindset is that you can just assume the other complications and just focus on directly on the efficiency consequences. And it's sort of, so you can say you can present the gains from trade or compare the advantages in its own whole sort of glory, but not worry about sort of the possibility of compensation or the public finance complications that might arise as a result. Perhaps the more damaging policy maker version of this is actually much better informed, but I think in some sense worse in practice because the policy maker version of this is to always insist on a whole long list of complementary reforms, but never actually A, B clear that you have a closed set of reforms. In other words, whether you've actually specified all the reforms that are actually required as opposed to an open set where you keep adding onto that complementary list and the longer list becomes longer and longer. And secondly, whether you're actually checking that the proposed reforms are actually taking place. So I think we've seen this both in trade openness and in financial openness, and I still see this all over the place that the argument that of course, we never said it was just going to be trade reform, trade opening. We always said that you needed to do this alongside labor market reforms and governance reforms and maintaining competitive currencies and all the rest. And similarly, of course, the argument that was very actively made during the debates on the spread of Catholic unconvertibility and financial globalization in the 1990s, always the insistence on the importance of the complementary reforms that financial openness requires and the appropriate macroeconomic policies and the appropriate prudential regulatory frameworks and so forth. So that looking back, we can all say, we warned you all that this was going to be required all kinds of complementary reforms and the reason that it failed is not that you opened up to financial globalization and all you got was volatility and crises but that you didn't listen to the rest that you had to take all those other globalizations, all the other reforms too. Well, I'll end with the, you know, I think the statement that best characterizes what our attitude to this should be and this comes from Avinash Tikset, who said the real world is second best at best and that I think is the right sort of kind of attitude that we ought to have that we can't just assume that it's always going to be somebody else's job to take care of all the side implications of the world and I think had we paid a little bit more heat to this, maybe we might have ended up a little bit more on the right side and I think as a result maybe have kept a little bit more of the trust of the general public. Thank you.