Transcription
Uh, yeah, I see a couple of friends over there. But, uh, it's, um, it's wonderful to be with you all, and it's wonderful to be able to moderate this faculty, uh, panel. We have an all-star, uh, lineup of Harvard and Kennedy School professors with us to discuss the state, uh, of global politics and global, uh, economics. I'm going to run very quickly through their bio so that we can get straight to the point. I'm pretty sure you know them very well, uh, but just in case.
So, uh, Gita Gopina, who's here to my left, is the Gregory Anania coffee professor of economics at Harvard. Uh, her research focuses on international finance and macroeconomics, and she's a leading voice on dollar dominance, exchange rate trade, investment, international financial crisis, monetary policy, and debt. Uh, from January of 2022 to August of 25, she served as the first deputy managing director of the IMF. And before that, uh, she was the IMF chief economist, uh, for two or three years.
Um, to our left is Ricardo Houseman, who's the Rafi Haridi Professor of Practice of International Political Economy here at the Kennedy School. He's also the founder and the director of Harvard's growth lab. Uh, of course, Ricardo is a leading voice on economic complexity, growth diagnostics, and international finance. And from 1994 to 2000, he served, uh, as the chief economist of the Interamerican Development Bank. And before that, he served as minister of planning of Venezuela and as a member of the board of the central bank of Venezuela.
And Danny Rodic is the Ford Foundation professor, uh, of international political economy, uh, at the Kennedy School. He's an expert, of course, on globalization, employment, uh, economic growth, and his research focuses on economic development and political economy. He currently serves as the co-director of the school's, uh, reimagining the economy program and as co-director of the economics for inclusive prosperity, uh, network. Uh, he also served, uh, as president of the international economic association, uh, where he helped found the women in leadership in economics, uh, initiative.
So, uh, we were just discussing before the panel how we could, uh, run this. Uh, there are two big blocks that we want to discuss. One has to do with the structure of the economy and the international system and its implications, and then we'll move, uh, into tech innovation, AI, and its implications. Uh, so two big, uh, large topics that are occupying headlines constantly. Maybe after the discussion, we can extract some order and structure to this bombardment of change and news and alarming news that we're constantly being exposed to, and then we'll have time for a Q&A.
Um, so let me just, uh, let me just, uh, kick this off, uh, with a question for you, uh, Gita, about this. So, uh, we have, uh, the dean, I think, described this moment as extraordinarily, as an extraordinary moment, uh, in American and global politics. Uh, we seem to be living through the restructuring of the international economy. This seems to be true in trade, in investment, currency, energy markets. I mean, across the board. If you were to describe the economic phase that we're in and we're living, how, how would you help us sort of navigate, uh, the, the current moment?
>> Yeah. No, firstly, thank you. Thanks for having me on this panel. Firstly, the way I see the world, we have distinctively moved away from the world order that we had of the last, you know, 50 or 60 years of ever closer integration. So, I don't think this is a matter of a particular personality or one who's going to be the next president of the United States. We, regardless of what and who comes, we are in a different world order.
So then the question is, what does this look like compared to anything that we've seen maybe in the past? And, you know, there were two striking periods. One was during the inter war period when we had the breakdown of global trade. Then everybody put up protectionist barriers, and of course, wars followed, the second world war followed. Uh, and then we had that period from the 1950s until 1990, when you had the cold war, where you had the, the western block, which is the US, US block, I guess, and you had the Russia block, and you looked at decoupling between those two blocks. And then the question is, where are we now?
So we are kind of a bit of a hybrid mix of both of those. We have absolutely a plain protectionist streak, which is also coming from the preference of the US president for protectionism. It, I mean, regardless of where we were in the world, I think he likes tariffs. So that's where we are. And so that we have tariff, we have tariffs in, uh, back to levels we haven't seen in, you know, in decades. But unlike that period, we have not had this tit fortat tariff war that ensued in 20, in the 1920s. So what was, what's different now relative to back then was that then every single country started putting reciprocal tariffs on each other, and then there was this complete, you know, rush to the bottom.
This time, China is going certainly toe-to-toe with the US in terms of putting tariffs, but the rest of the world is not. And I think that is a reflection of the fact that people understand that that just takes us all into a very bad place, that there is, there are gains from globalization. There are benefits of globalization that we need to keep. And so there is this, you know, coalition of the willing. I think Europe and Canada and, uh, you know, a few countries there are the ones who are trying to say, can we hold, can we keep what we have right now? So right now, I think it's helpful in a kind of a positive way to think that about still over 70% of global trade, with 75% of global trade, takes place under WTO rules. So it's still, it's not hopeless.
The sec, in comparison to the cold war, again, we now it's not clear what the blocks are, and you can see it not just in the narrative and in the, but in terms of the data and the analysis, in terms of firms making decisions about where to invest and who to trade with, it's, you know, there is no, clearly there is no US block in terms of US Europe, that's, that is changing dramatically. Uh, and then, of course, there is China there. So it is a different place, which what this means is there's no simple analogy to the past one can look at. We are at a, we are in different times. And I think we each one of us has to follow these trends fairly carefully and see how each of the different elements of the world order affect the work that we do.
>> Can I, can I ask you a quick follow-up, and then I'll bring Ricardo and, and Danny in as well? So, I mean, if, if you were to project this forward, say 10 years, uh, how do you think trade patterns and investment patterns are going to look? So, is there a world where the US can be excluded from most of these things, or it's a, it's a real anomaly because of its tariff levels and barriers to access the US economy, and others, and the rest continue to trade? Because from a corporate point of view, I mean, this is a very different world. So, we, we've lived for over 30, 40 years in a hyper globalizing international environment. The logic for private and public actors was really interdependence, right? So, we were all signing FDA, building infrastructure. Our companies were expanding globally. And what you're telling us is, well, now, now this is a little bit different, right? Because you have a geopolitical and political risk lens that you need to apply to your decisions. Is, is that lens going to be exclusively applied to the US space, and the US will remain, I don't know what it is, 14, 15% of global trade, a little bit on its own category, and then the rest of us will fall under WTO rules? All of this will continue. I mean, can you help us project this forward a little?
>> So, I think what's clear is we've absolutely been moving away, and this is not just US specific, but this is true across all countries. We're moving from a model where we were all about efficiency and buying from the cheapest source to a world where we are going to build redundancies, and for reasons like national security, for supply chain resilience, it will happen. So, every country is going to now try and do the best they can on in defense production so that they're not entirely reliant on one or two economies. Rare earths are going to be, you know, refined in many more places, semiconductors. So, the, we, we are certainly going to be devoting more and more resources towards also, I think, a bit of a return of industrialization to ensure that there is enough, there's excess, that we have resilience, and some of that is fine. I mean, I think I agree with, maybe we, you know, the, the model went too much to one extreme of efficiency that Germany became 100% exposed for its energy to one country and being held hostage for it. So, it absolutely makes sense for some correction. But, of course, we're moving the other direction now.
I don't think it will stop at just the US and tariffs if we continue in the way we are, because what's happened right now is the trade between, uh, China and the US has dropped quite considerably, and some of it is rerouting, but not all of it is rerouting, and you have a flood of goods that are going into other parts of the world, including Asian economies, Europe, Africa, everywhere. Many of these countries are just, they are not going to sit back and let this happen on a, in a, in an indiscriminate way. So, I think if there is not the coalition of the, these countries coming together and figuring out how they're going to not export as much and dump as much in each other's country, there I expect to see tariffs will go up on, you know, on Chinese imports from other parts of the world. So, we're not settling yet into, we're not settled yet into an equilibrium. We would see more of this come up.
I don't see a world where the US is an inarchy and everybody else is trading with each other. That ain't going to happen. The US doesn't want that itself. What they want is smaller imbalances. They want, they'd be happy to trade. They just want less imbalances. Now, of course, we can debate about how to go about getting that, uh, but that's not the idea, is not to isolate, uh, the US, and I don't think that will happen. And, you know, just leave it at, there's all this talk about what's happening with the dollar and dollar dominance and so on. I think for all practical purposes, at this point, we are in a world where there is, I suspect, dollar dominance will continue, uh, unless there are even bigger political institutional shifts happening in the US, in which case, I would say, frankly, dollar dominance is the least of our problems.
I'll, I'll get to that in a second, and I'll ask you, I'll, I'll try to weave in, uh, also the, the Iran question and others that, that have, I think, a currency implication as well. But so, Ricardo, you work, uh, with governments around the world. You work on the growth models. You work on challenges that they face. So, you must have, you, you must be looking at this evolution in international economics and the international system and its impact on governments around the world. So, what, how do you think about this when you're thinking of growth, competitiveness? I mean, this new economy that we're entering, which is much noisier, less organized, less structured than the one, less integrated than we've had for a long time.
>> Well, a great, great question. Thank you. And I think I just came back for last week from Canada, and obviously, I think that Mark Carney more or less defined a little bit the moment we're living in in his speech in Davos, which is probably one of the most important political speeches in a very long time. And, you know, the southwest of Ontario is being devastated by the fact that, you know, that now they face 25% tariffs going into the US for the car industry, and firms are making those decisions. So, you have a major upheaval in, in Canada that, that needs to respond to, to something.
So, I, I have asked myself the question, is this kind of like a new equilibrium, that it's the new world, and, and consequently, the US is playing its, its best hand, and, and the world needs to respond with its best hand, and that will lead to some new equilibrium? Or is the US just making a, a strategic mistake, and, and, and this is a strategic mistake? Eventually, mistakes are often corrected. No, we don't know exactly how, but in part, they are corrected because the world responds to your behavior, and then you don't like where you are. So, it may take a while. It may take a while, but let me put it to you that the world we were living in was a world in which that, that emerged from a competition between two, two blocks, right? The Soviet block and, and, and the US-led block. And the Soviets had a universalizable ideology. I don't know that the word universalizable exists, but it, you'll, you understand what it means. Anybody can belong, right?
>> It exists in Spanish.
>> As, as many other things. So, um, so they said, you know, workers of the world unite, everybody can be communist, etc., right? And the US kind of was forced to develop a universalizable ideology, and which was, um, you know, freedom, democracy, rule of law, open markets, and anybody can belong, right? And that implied that the US would impose some disciplines on itself to sustain the, the coalition. And one of those disciplines was, you know, I'm the biggest economy in the world, say the US thinks. If I were to negotiate bilaterally with every country, I could extract a pound of flesh based on how much more powerful I am in every relationship. But I'm not going to do that because if I do that, then they don't, they won't have too many incentives to play ball. And as a consequence, I'm going to be end up in a world where the world is communist or something, right? So, so I needed to create a world where there was, you know, it was, there were incentives to play, everybody played the game on this side.
So, come Trump, he says, who are these idiots that, you know, agreed to abide by some rules and most favored nation? What the hell is that? No, I'm going to negotiate country my country and extract every pound of flesh, and I'm going to tell the Europeans, you impose zero tariffs. I impose 15% tariffs, and if you don't like it, see what you're going to do about Ukraine. Right? So, so that's kind of like the new equilibrium. Right? So, the US uses its power to, to the fullest, tries to extract upon the flesh in every relationship, as, as Carney said, weaponizes any interdependence. And, you know, and then, the world must suffer what it must suffer, as, as Tuscidi said, right?
So, I think this is going to lead the world to do exactly what go Gita said, which is to, to try to find new forms of alliances between middle powers to, to try to protect other things and inevitably to get closer to China because now you have to derisk from the US. Because if you thought, well, you know, the US main conflict is with China, then, you know, it wants to create a coalition to kind of limit China. But if its first attack is against Canada, and the second one is against Denmark and Greenland, and the third one is against Panama, and now he's, Trump posted Venezuela is the 51st state, right? So, so then, you know, everybody realizes that the US is a major systemic risk, and you have to derisk, and what are you going to do about that? So, so I think that, um, the dynamics will lead the US to be in a more uncomfortable situation. It will regret some of its current things, and it may self-correct eventually.
But I want to add another thing, which is the world today is seeing the boom in AI, but the world 10 years down the road is going to have a much bigger AI. At least if you believe stock market valuations in, in a oped I wrote with our former colleague here, Andreas Velasco, and we calculate that by 2035, to make sense of these valuations, the international sales of the 11 largest AI companies, starting with Nvidia and ending with OpenAI, Anthropic, Broadcom, and so, um, sales abroad will have to go up by $2.5 trillion a year in, say, 2035, 2036. $2.5 trillion is more than 100% of US exports of goods, is more than twice the US current account deficit. So, it means that the world is going to be buying. I mean, stock market valuations being correct. Dita has argued that they are, it's a bubble, but I'm just saying if that's the world, then the, the world, the US is going to want the world to buy a lot of AI. So, and how is the world going to pay for that? Well, it will have to export things to the US, right? So, I don't know that this love affair between MAGA and the tech bros can last because the tech bros will want the US to be able to have, to have countries earn enough to buy all the AI services they're going to want to buy. So, my optimism comes from the fact that I think the US is making a strategic mistake.
>> Well, thank you. Uh, let me, let me, that's a very convincing way of putting that. The, let me bring in Danny to the conversation because on top of these, I think, overarching questions to US policy on US centrality to global politics and diplomacy and the function of the US as a constructor of the international system and underwriter of the international system. Questions about the economic sustainability of the US growth model if it amputates itself, it sort of cuts itself off from international trade. I have an even broader conundrum that I know you've been working on, which is all of these things happen in a moment of extraordinary abundance. So that there is a paradox between the convulsions that we're seeing in our politics and in the design of our domestic and international policies, uh, and rates of growth in the aggregate. I mean, we've never been more prosperous than we are today in this country, in many European settings, at the aggregate level, in GDP terms. We've never had more access to information. We've never lived longer lives in the aggregate. So, so, so there you, if you had to say, and, and you've written extensively about this, sort of the discontents of globalization, challenges, you know, the tension between globalization and sovereignty, but why, why do you think we are living through this extraordinary moment of revision, uh, particularly led by a country that has been arguably the greatest beneficiary of the system that it built and, and, and felt itself constrained by and maybe no, no longer?
Well, I think the, I mean, we, the kind of globalization, um, we experienced after, um, the 1990s, um, you know, that, that I call hyper globalization to distinguish it from the previous model of the sort of more the Brettton Woods model of globalization, um, you know, created very, um, uneven, uh, benefits. And I think in the, you know, you ask the United States, why is there so much discontent? Um, if you look at the distribution of the benefits, um, you see, um, obviously, sort of, you know, the, the coastal elites and financial sector, multinationals, large firms, professional groups, uh, the tech sector, um, have been, um, uh, huge beneficiaries. Uh, but a lot of the, um, uh, parts of the country that, um, had to, uh, experience a sequence of, of trade shocks, whether sort of the NAFTA the consequence of NAFTA or the China trade shocks, and there sort of very visible and empirically, you know, um, uh, measurable effects of, of joblessness and, um, continued erosion of the middle class. Um, and so there was a kind of this, this economic divide that opened up in this period between those who had the, the skills and networks, the assets, the professional connections to benefit from a much more interconnected world, um, and those that did not, and either felt behind literally or felt like they were, um, feeling be, being left behind.
But I think politically that was layered over by also the sense that the centrist, the mainstream political elites were leaving, uh, these segments of workers or these parts of the country, uh, behind and weren't really responding. There was a sense that you had to, you know, you know, grin and bear it. You know, sort of globalization was here. There was nothing we could do about it. Um, you just had to adjust, and if you lost your job, you, you don't, if you lost your job, you either you had to get, you know, trained or move somewhere else, but don't ask us to reconsider the rules, even though, of course, the rules, all the trade agreements and so forth you mentioned, they were also man-made, they were made by these same governments. So, this sense of losing control, um, which I think resonated a lot when the far-right and the authoritarian populace made it, um, their rallying cry, I think, was, was fundamentally rooted in, in a real sense of political disenfranchisement that a lot of the, um, electorate felt.
And I think sort of we had two transition points. Exciting. One was, um, the, uh, global financial crisis in, in 2008, uh, where, um, the, the, the sense of, of, you know, you know, you know, those the elites who sort of, you know, came out relatively well versus the, um, those who, who lost their homes, and, and that I think had lots of political resonance, um, and then ultimately, um, 2016, when you have simultaneously sort of the political turning point, I think, with, uh, with Trump's election the first time around, and, and, and the Brexit referendum. But 2008, I think, is important because economically that's sort of the first time you see trade beginning to stagnate, falling behind global GDP. Um, uh, China's share of trade and GDP had reached its peak just before that, and then it continued to decline. It's actually declined by, by an enormous 15 percentage point since then. Um, and, and crossber financial flows also reached their peaks just before the, um, and so economically, you see that it's really the transition comes in 2008, and then the political effects, uh, play themselves out longer.
So, I think it's, it's, um, it's a combination of, of the economics and, and the sense of, of political, um, loss of control, um, that, um, and, and, and I agree with, with what, you know, Gita and Ricardo have said, that I don't think there is going back. Uh, I think we're moving to a different era, but also, I think, that we're in a moment of transition. We don't know exactly what that new system is going to look like. If you remember, sort of, when, you know, Bretttonwoods be, you know, started to fall apart in the early 1970s, it wasn't really until the 1990s that we had the emergence of a new, uh, global economic order with the WTO on the one hand, and the, um, and, and the promotion of financial globalization on the other hand. Um, so we are in a kind of system now where I, I, you know, it's not entirely clear. Um, and, and I, I sometimes say, sort of, the future scenarios is the good, the bad, and the ugly. Um, and, and, and the ugly is the one that, that Gita was talking about in terms of the inter war period, uh, when you have a significant collapse of international trade. Um, uh, I'm sorry, that's the sort of really the bad equilibrium. Um, that's the bad outcome. Um, I don't think we're going to go there because I, I, I don't think that, you know, that the cost of that are too high, and for exactly the reasons that I think Gita discussed, I don't think countries are going to be willing to disengage themselves from the world economy to that extent. Um, the, the good equilibrium is when we sort of return to a kind of a healthier balance between the prerogatives of the so of, of national autonomy and sovereignty and the requirements of, of, of an international economy. Something like a, a kind of a Brettton Woods regime redux, uh, where there's a, there's more room for n for national economic management and for building the kind of resilience and national industrial policies and so forth. And the ugly scenario is when, where I think, um, uh, geopolitics really takes over everything, that, that trade becomes, you know, the majority of trade begins to be viewed as, as zero sum. Uh, everything is about, you know, who's winning and who's losing. Um, trade increasingly becomes weaponized. Um, and, uh, and, and that's, you know, would not be complete decoupling, but it would be like, uh, you know, sort of, u, more and more of, of the kind of the world that we're in. So, I don't know where which one I, I, you know, I don't think we're moving to, um, uh, the, the bad outcome. You know, whether it's going to be the good one or the ugly one, is it's hard to say, but we are in a moment of transition.
But it's, it's interesting because on the driver, I mean, this analysis of yours, it's a, it's a political economy, sort of, economic issue as well, hollowing of the middle class leading to basically the hollowing of the center of the political spectrum, polarization, and then this has broad implications for domestic policies and politics and for international affairs because this politics is breaking the international system as we have known it. And I have a couple questions to weave in the tech piece here before we open to the Q&A. Uh, one is, I mean, I have an overarching question, which is, now that these policies are being implemented, do they help or do they worsen, uh, the drivers of the crisis in the first place? Because I would assume tariffs, because they're inflationary, they hit low-income households more severely. Uh, many of the measures that are being implemented actually are very regressive. I mean, some of the t fiscal policies, tax measures, and others. But above all, I really wonder what the effect of AI and tech will be in these trends if we're worried about effects on income distribution, uh, about the decline of labor income within our economies, which is one of the underlying drivers of this. I mean, in advanced economies, one of the, in fact, pretty persistent trends that we've seen over the last few decades is a decline in the labor income share. So, we're moving to these capital income economies quite clearly, and, and wages are the big distributive mechanism that we've had. And if you look at the effects of AI, the little rigorous, I mean, work that we have done with the latest models that have been published, that has been published, seems to suggest that AI will exacerbate this issue. Uh, you're going to see huge concentrations of productivity in companies with very small number of employees. So, so this idea of the erosion of the wage, uh, channel and that then fueling further fracturing. So, when, when you look at particularly AI, its deployment and its impact on the economy, what, what, what do you think are, I mean, I don't know, just to bring you back into the convers, are the main, um, components of that when you look at?
>> I mean, firstly, uh, thinking about just now in terms of what AI is doing, AI is very responsible for offsetting all the negative effects of tariffs and tremendous uncertainty around that's happening in the world that usually, usually would slow global growth. The fact that global growth is basically what, in fact, in fact, slightly higher than what it was projected to be, is because of AI, right? And so right now, because of the large amounts of investment that's taking place in AI, whether that pays off or not, we don't know, but it is real money being spent on real things. Investment in AI is a big factor. If you look at where trade is holding up, it's truly in AI trade. So anybody who's, any country that's part of the global supply chain for AI, you know, Taiwan, Korea, you know, in the chips trade, that's your exports are booming. And so, and though Korea is being hammered by what's happening in the Middle East, they, their economy looks great in terms of the numbers, overall numbers, because of, uh, the fact that they're part of the AI chain. The stock market gains that we're seeing in the US and many other countries too, is generate, generating wealth effects that's holding up consumption. So right now, AI is really the good news in that sense for the world economy, and we certainly haven't seen unemployment go up yet, any of that happen.
Now then there's a question of where do we go, where do we head? And I think this is important because, uh, I do worry that the whole China shock paper, which is a very nice paper, but I, but the, but it got picked up and made incredibly political and made it seem like that was the reason trade is the reason why manufacturing shrunk in the US, and there is absolutely no evidence of that, that's not the case if you look at any analysis of that China shock period and if you look at what's been the drivers of decline in employment in manufacturing. It, it's the China thing is a tiny, tiny blip. Uh, but what has played an important role is also, is automation. Automation has been big, and so AI, which is now automation of just not just manufacturing but automation of services, pres, you know, presents that challenge in terms of job creation. But again, I, I'm a little careful about whether this, I'm going to conclude that this will necessarily end with fewer people working and that this is going to be necessarily that negative because we have had cycles of technology where we worried about it.
I think I love the classic example of radiology. Radiology, but, you, know, this was 15 years ago, this everybody was saying nobody should do radiology because we already have machine learning. After all, radiologists read, you know, images. We can have machines do it. But right now, it is there are way more people, way more radiologists right now. They're earning higher incomes. And what basically happened was machine learning made it very cheap, reduce the price of radiology service. So the demand for it went up, and you had more, more increase. Right? So I, I think, you know, it's the, the, the idea that this will necessarily lead to job destructions is not clear. I what I do worry about though is that there are periods of transition where things can go wrong. And I think one of the things that we know from previous waves of automation is that close to, I think, like 90% of automation related job losses happen in the first year of recessions. And we saw that right after the great financial crisis when we had the previous wave of automation. And, you know, when, when, when times are good and, and companies are making money, they don't really want to be out there firing workers. Some people, some companies do, a lot of companies don't. They're like, anyway, we're very profitable. Even if the worker is not really adding that much value, we, we're making enough profits, and we, but then you hit a recession and you're in cost cutting mode, and that's when you say, okay, well, this is when we lay off workers, and you have jobless recovery. So we saw that again after the great financial crisis, we had the US economy bounce back very quickly in terms of overall GDP terms, but employment took much longer to come back up, and so you have these jobless recoveries. So, and it's very hard to forecast where we will be, you know, with AGI and what, where, what it's going to look like, or we, we have a bubble and therefore things are going to crash. I just don't know. But I think it is real that we could have costly transitions where in the next recession, which will come, and we're trying very hard to make it happen, it seems that, that we will, that we, we could see many more job losses that unfort, that unfortunately don't come back.
I, I want to bring Ricardo and, and Danny in, and then it's, it's open to a, to a Q&A for about half an hour. Uh, I, and I want to bring you in with an anecdote. So, about a week or a couple weeks ago, uh, we were lucky enough in Madrid to host the UN AI scientific panel. So, the UN has put together, it's a bit like the IPCC for climate, but they've done the same for AI. And one of the co-chairs, his name is Joshua Benjio. He's one of the godfathers of AI, touring prize winner, and we had a long exchange about how the, he sees the future of AI. Uh, so as you know, I mean, these tools, the frontier models, are now extraordinarily capable. So, they beat human performance across a whole range of functions. Pattern recognition is one. Uh, writing, I mean, verbal and, and written expression is another, uh, coding they're getting to human, uh, capacity. They're getting so good that it's now becoming hard to build benchmarks that can actually measure their progress. And when these are built, they beat them in weeks and things like that. And, uh, it's according to the most sophisticated computer scientists, there's actually no very clear end to their capability. There might be some end, uh, in the current models, but they're going to continue to evolve in this direction. So, they're going to be very, very capable. So, I think it's, it would be very interesting to get a sense from you of what the implications of this might be, uh, mostly for, for our economies, but also beyond, uh, our economies. I mean, we're sitting in an academic institution. One of the, I'll finish with this, and then over to you guys, but one of the things that they do very well is they do standardized testing extraordinarily well, right? So, already three years ago, they were performing on the 80th, 90th percentile of the, you know, the LSAT the GRE the GMAT, you know, these sorts of exams. So, I, I think they, these tools pose immense questions for the future of our economies and for the future of academic institutions, as well as pieces in the economic machinery, you know, but how do you think about this in your own day-to-day work and research?
>> Well, I think that people are either biologically optimists or biologically pessimists, and then, and then they process the information according to their biology. So, I am biologically optimist, so I'm probably not seeing too many things on the dark side, but my interactions with AI make me feel like Superman, right? My, my colleague Dan Levy says that the important question is not whether humans are better or not better than AI, the question if is if it's a human plus AI is better than a human, and, and that's the important inequality. And, and we, we have to try to build a world in which we are superhumans because we have this extra special power now that we didn't have before. And that should allow us to face certain challenges that we have, we have today, that we have not been able to solve, especially, for example, say, in developing countries. And I'll name just a few.
In most countries, we complain that we don't have adequate skills in teaching, right? And that, you know, in, in poor regions, poor neighborhoods, etc., we cannot get good teachers and so on. So, can we revolutionize the teaching process so that we increase educational quality dramatically? Secondly, we complain that we don't have enough doctors and so on, there weight times. Everybody is complaining about weight times and so on. So, can we relax those constraints and make more health available?
But I'm going to go to more economic things. Okay. So, I wrote a column, I think, last October or something, asking myself, I think, for example, if you were to analyze the financial markets, there's been tons of work on how to make banks better. Now, banks give loans, loans is debt. Bank look at the gearing ratios, the indebtedment ratios of companies, so they want to look at the debt equity ratios. So, we've done a lot in creating debt markets, but what about equity markets? The equity markets in most places don't exist. If, if they exist, it's only for the super large firms. It tend to be banks or telecoms or things like that. But for the most part, it's friends and family. Can we use the digitization of transactions plus AI to dramatically reduce the cost of equity and then democratize equity? Okay, so that's another, another thing that we should be working on, right?
I, I think that, that the question that, that we should be facing is, is, is let's look at human needs because in the end, what's a market? A market is a situation where you earn your living doing things for other people, and how much you earn depends a little bit on how much other people value what you're doing for them. So, this is an occasion to say, let's look at the needs of other people. How can we supply them in new ways with this superpower, new superpower that we have? And in that sense, I think the radiology example that, that Gita just said, for example, right now, we don't, we don't standardize having you a, a body scan of your whole body, say, every year, every couple of years, every three years to see if any cancer appears anywhere, right? But if things become cheap enough, that's going to be completely standardized, and that's going to be the future. And that's my, my colleague, my co-author, future co-author, and, and Luis Garano has a book coming out called messy jobs, and the idea is that human jobs have some combination of this thing that AI is good at and this thing that AI is not good at. So, what we need to figure out is what is the killer app for humans, and that's, that's what we should be focused on.
>> Thank you. Those are, those are two, I think, very optimistic, uh, answers. I have a friend of mine who's an entrepreneur that says, if you're a pessimist, you sound intelligent. If you're an optimist, you make money. So, I guess I, I guess that's why some of us are academics and make no money at all, and others do. But, uh, Danny, so what's your reading on all this?
>> Yeah, I mean, I have, I, I have, I have both a pessimist, an optimistic side on this, and a pessimistic side. On the optimism side, it's truly, uh, transformative. I mean, I, it's, and I don't buy the idea that, that, you know, it's going to, you know, take work away. I mean, everybody I know who's using, um, AI, including myself, is working a lot harder because there is so much more you can do with it. Um, and it just, it's, um, and, and, and there's a, there's again, on the optimistic side, there's one very important feature of AI, which is different from earlier waves of technological change, if you can, you know, so if you think about automation, what was automation about? Automation was directly about doing stuff that, um, people could do, you know, real humans. So, you know, we think about robots, robots came and they took away jobs of production workers in factories. Whereas the nature of AI is, is, is that it actually allows each one of us, including most importantly, those of us who are less skilled, less experienced, less educated, to have access to the knowledge and the practices and accumulated skill and wisdom of those who are much more experienced and educated. In fact, you know, the universe of that. So, if applied and, and used well, AI can be something that actually enhances, um, uh, the capabilities, the incomes, the productivity of those in the workforce that are actually sort of at the bottom in terms of the, uh, the wage distribution, earning distribution. So that, and we can see that, you know, sort of, uh, Ricardo gave some examples, but I think the potential for low-inccome countries is huge. I mean, the vast majority of workers in developing countries are basically self-employed. Uh, they're informals, and they do, you know, sort of, um, very precarious gig work, everything from, you know, care to, you know, stream market vendors to, um, you know, food and, uh, hospitality. And in each one of these areas, AI has a huge potential to just essentially enhance the capability of those sort of relatively, you know, less skilled, less educated, you know, just if you're able to provide entrepreneurs in low-inccome Africa with customized, um, basically coaches and, and, and, and, and business advice and consultancy services that can actually allow them to fulfill their full potential. Or if you basically provide, um, AI tools to gig workers or to, uh, street vendors so they can actually manage their supplies better, they can access much larger markets and do tasks that they weren't able to do before and earn much higher. So, it, the potential is, is, is actually very, very, um, um, the upside is, is, is very, um, uh, very large.
Now, where I get, um, where I get pessimistic is that I think currently we're moving in, in a direction where we've completely seeded, uh, the, um, the direction of AI and its regulation to these large, uh, um, tech companies. Um, I view that as being very, very dangerous. Um, and I do think we need, uh, significant regulation, uh, both for reasons of the kind of distributional implications that you mentioned, Manuel, uh, but also with regard to, um, about sort of risks and, uh, with, with regard to, um, you know, direction of, of, of deployment, um, and, and I'm really pessimistic that on our current trajectory, we're not going to take the steps that are required, uh, because of essentially defeatism. About, you know, whether it is the, we can't, you know, stop and consider what we're doing because otherwise China will win, so that there's a kind of geopolitical consideration. There's a kind of defeatism about, oh, you know, the technology is moving so rapidly, there's no way that the public sector or regulators can do anything about it, uh, for any of those reasons that we're not going to do something until something bad really happens. So, I mean, you know, once a disaster strikes, I don't know whether that's, you know, some AI system striking out half of the electrical grid of the country, or it is something with real loss to, to, you know, human life or, or something else. But I, I fear that it's just going to take a real shock, real crisis, a real moment of having to reckon, uh, with what AI is able to do. And at that point, I think there is going to be a, a kind of, you know, sort of an immediate backlash. It can go, you know, too far, but I think it will mobilize the forces, the political forces that will ask us, you know, let's stop asking how can society adjust to AI, but just reverse it and say how, how does, you know, how should AI adjust to society, um, which is really the question that we should be considering.
>> Yeah, well, the three very interesting answers, and they, they connect very clearly with the prior debate, but you don't seem to see AI as a massive accelerator of the prior trends of fracture and others. I mean, maybe in the intim as we grapple with the technology. I'm going to open it up to questions. I think the format is the, there are two microphones, uh, here if anybody has a, a question. If not, I have a couple of.
>> I've never been afraid to be the first one up here to speak. Uh, my name is Marie Chevier. Uh, MPP86, uh, PhD 91. Um, and my experience at the Kennedy School has really fed my leap from lower middle class to, um, elite status as a Harvard PhD. If I look around this room, I see everybody is, is now at, in terms of alumni, have made that leap. We are the elite, and it seems like very little, except for this panel, have really talked about the needs of the income gap, the wealth gap, and how our policies, in fact, have made that worse over time in this country and perhaps around the world. Um, and I, I, I think I probably have that, that pessimist gene, but sometimes I turn into a polyiana and think, "Oh, yeah, we can do this and we can do that." And sort of at the ground level, at the implementation level of what we can do as the elite to connect back with, um, poverty, uh, people who don't have a high school education like my parents, neither of them went to high school and realized that if they had had that opportunity, if they had had that education, and I worry a great deal about the effect of AI on education. I have taught in public, um, in public, uh, universities, not, um, elite universities, and people are no longer assigning books because students don't read them. Papers, not only in classes, but submissions to journals are increasing dramatically because of AI. And is that increasing the quality? No. So, I know that's a basket of things, but any of you can sort of jump in and, and, um, respond to that. Thank you.
>> I guess this
Is also a question about, you can't you? Oh, this, sorry, I dropped this. I guess this is also a question about the regulation and the governance of AI that you were saying. I mean, this transition period and others, and Danny was mentioning, so how does a properly regulated, integrated AI economy look like? You know, on the taxation front, on redistribution, on public services, you know, what are the measures? How is our governance of this going to look like in the coming years, and how is it going to lead to a society that is more equitable, more just?
>> Yeah, no, there was a lot there in that question. So, firstly, as a fact, I know we should never be satisfied where we are. We certainly have a lot of reasons to complain, but if you look at the long arc and where the world is right now, the drop in global poverty levels, the number of people who've been uplifted out of poverty, is just very striking. So that doesn't mean we should be pleased with where we are. I just want to make that point that there is a lot that's also happened that's been quite positive in terms of AI and education.
Actually, the point I wanted to make was, you know, you go around, and in previous waves of technology, the argument was, "Let's make sure that students are getting the skills they need to use the technology." I think this time around, I'm actually worried whether they will get any skills, that they will even not do the skills we have. Like, the skills we got through trial and error are basic cognitive thinking, which came from, you know, solving a problem, getting it wrong, trying, working on it three more times. And now each of them has like a friend who's way smarter than them, who can give them the answer instantly, and you have this fake sense of, "Oh yeah, of course I get it." But you didn't get it. You really didn't.
And so, I'm less worried about whether they know how to use AI, which they will know beautifully. It's very simple, actually. I think people make it sound very complicated. I use it. I use it. I think it's phenomenal. It's been great for my productivity. But I'm thankful that I had all these years when I actually had to learn it the analog way, which is really sit with pen and paper and understand it.
So, the challenge, frankly, for education is how do we make sure that these kids actually learn anything coming out of college? And that's the... I'm less worried about whether, if you had all this knowledge, you would figure out how to use AI. I have no, I have no worries about that. I just don't know whether they will build the skills. And we're seeing it increasingly when you, you know, I just taught a class where I gave a written exam and then I also gave an oral exam because I wanted to, I wanted to see what they actually learned. And when you have a conversation, you realize that there was a lot of help from AI in the other stuff. Yeah.
>> On the issue of inequality, I think people... there is a framing of the debate that says, "There's inequality, so we have to tax the rich and have a universal basic income or something like that." And I would like to propose the idea that the inequality that we should most worry about is inequality in productivities. So, it's not that the economy is baking a pie of a certain size and the pie is being poorly shared. It's that different parts of the economy are baking pies of very different sizes.
And in many countries, people are trapped in very low productivity activities. And there's a... in the middle-income countries and low-income countries I mostly work in, the size, the places in the country, and the activities in the country that bake big pies is relatively small. So, the real question is, how can we incorporate more people into high productivity, right?
So, I make a distinction between what they call redistribution policies, which happen sort of ex-post, and say you take this inequality as given and then you try to compensate through transfers. I think that as palliative to some extent. And so, redistributive policies versus inclusive policies, which is about what are the strategies to include people in the networks that make people more productive. And I must say that people are not more productive; organizations are more productive. Your productivity is really not something that's inherent in the individual; it's mostly inherent in the organization that individual is plugged into.
So, I would put it to you that what we should focus, a little bit like Danny was saying, and there's a lot of work suggesting that the problem of low and middle-income countries is that they don't have enough professional managers to coordinate production at a higher scale. If you relax that constraint, it could have big, big effects in the availability of people to operate at much higher productivities.
>> So, I'm not trying to give rich people a pass in the sense that we're not going to tax them. An inclusive policy can be as expensive as a redistributive policy. So, I'm not discussing budgetary allocations, but the inclusive policy is growth-enhancing because it's raising people's productivity, right? And so, it marries growth with inequality reduction, while the redistributive policy puts us in this old tradeoff that I think we must find ways to relax.
>> Thanks. Do you want to?
>> Well, let me say, I think with regard to education, we face an interesting conundrum because, you know, we, of course, as university professors, we obviously have to think education is very important. And certainly, it's important, and it's going to remain important for citizenship and for, you know, broadening the mind. And increasing access to education is very important.
But I think, you know, the old, the traditional view that increasing access to education, college education in particular, is going to be the main vehicle through which we are going to enable a sort of increase in overall incomes and reduce inequality through spreading the productive opportunities that Ricardo was talking about, I think is becoming harder and harder to believe. And that's because of what's happening to the occupational structure of the labor force, both in the advanced countries and in developing countries.
If you look at the United States, you look at projections of the Bureau of Labor Statistics for what is the occupational structure of the US labor force 10 years from now, and you can go from the top, the largest occupation, and count the top 10 occupations. Of those top 10 occupations, the ones that are going to employ the largest number of people, only two require college education. The rest are things like care, food services, and other personal services, and a variety of other types of occupations where education is not required.
In developing countries, I estimate that about three quarters of the labor force are going to be in occupations that do not require college education. So, we have to figure out, for everything that, you know, education might be good for, if we are looking for policies for inclusion and where the benefits of growth and productivity are going to, you know, diffuse throughout society, we really need to figure out how we're going to increase the earning, the income-earning capabilities and productivity of these people who are in occupations where having a college degree is actually not going to necessarily enhance those capabilities or make them do better.
That's where I think, connecting with what I was saying earlier with new technologies and AI, that's really an opportunity because, you know, AI might give those people a leg up because, if deployed appropriately, they could be sort of income-augmenting, labor-augmenting forms of technology that might make many of these people more productive in those kinds of occupations.
>> Thank you. I have teeth in the back.
>> Good morning. My name is Ghanam Tiwari. I graduated 15 years ago from the MPA program here. With the core idea that inspires every one of us, "Ask not what your country can do for you," and worked in politics in India and education. We have a platform. We have five half a million students learning on that platform online every year.
Fifteen years back, it seemed like democracies and the forces in the world will power democracy. Fifteen years down the line, it seems like even the developed countries face cryptocurrencies where leaders can dently lie, stay unaccountable, and still be very popular. So, at one level, you see the weakening of the voter, and the weaker a voter is, the weakening is further and is institutionalized, and this across the world in many places.
Second, on the other level, we see a weakening of the worker in the world of AI, where AI can easily remove or make more and more worker redundant. Few companies have more and more power, both in stock market and in their ability to manage the world. So, in this scenario, the two foundations that power democracy, which are worker rights and citizen rights, are getting weakened.
So, how does one look at a positive view of the world where, on one end, you see democracies turning into cryptocurrencies, and on the other side, you see companies becoming more and more powerful vis-à-vis their workers?
>> It's an easy...
>> That's a very dy question.
>> No, I think, you know, that's a fascinating way of putting it. And I do think, and it also suggests the answer, the answer is that, you know, we've gone too far in the directions that have produced some of the backlash, and it has assisted and has produced, you know, dysfunctional politics and an economy where a lot of its participants feel it's not paying out for them.
Now, you know, I do take, again, I'm not too much of a pessimist. You know, if there are, I think, signs of some of the correction that might be... You might look, for example, at what President Biden was doing here in the US as trying to correct some of those things. I don't think you can say the Biden administration was corrupt. You know, I have a lot of disagreements with the specific policies that he pursued, but they were very serious about increasing worker power.
I think, you know, they put too much emphasis on re-industrialization and the manufacturing sector, whereas, as I was saying before, most of the workers had already shifted to services. So, you need a different kind of a strategy.
And with respect to democracy, interestingly, I think some of the more encouraging things are coming from the developing countries. You know, you have to take heart from the fact how Brazil, for example, bounced back from their version of Trump. You know, you have to take heart from the fact how South Africa bounced back from one of those corrupt autocracies that you were talking about. Those are very good signs that are happening in some of the middle-income countries. So, I do believe in the self-correcting power of democracy. And I do think that, you know, those are the kinds of changes that we need to fight for. And I think reminding us, as you did, that those are the things that we've been experiencing, is a place to start.
>> I want to jump in and make another point, which is another area where I think we are past peak. So, if we think... I think we're past peak globalization. I think another area where I think we're past peak is on immigration.
Because where we are right now is this environment where, I mean, we have populism in many, many parts of the world, either extreme right or extreme left. This is not the era of elites and, you know, "How do we help the world?" Nobody really cares to ask us that.
So, and I think a common feature that we have to recognize and grapple with is that a feature that is common across several countries where you've had these extreme levels of polarization is immigration. And that hasn't been solved about how do we end up, how do we, like, what's the optimal level? But it is the case, and this was a very important, this is a very important factor in the US, where it doesn't matter what's happening to the US economy and what's happening to US inflation. There are, you know, 35% at least of the population who believe that as long as the president gets immigrants out, that is good enough for me.
And so, I think we have to recognize, and, you know, at one level, there are, of course, tremendous benefits, and I'm an immigrant, from having immigrants and so on. But there are costs: economic, social, and political. And maybe this requires a clear... I think one of the things that as we are revisiting globalization and everything, one aspect of global integration has been people moving across, and certainly it has not been managed well, and we are in the place we are right now. It's... I just think that that is a common factor, and I'm not going to name countries. We can go country by country, and this topic is super important in many parts of the world.
>> Absolutely.
>> I'm Amy Porges, MPP 1980, and a question mainly for Professor Houseman. I'm curious about your reflections on the situation in Venezuela, sort of from the standpoint of economics, growth economics, geopolitics. What do you think are the biggest challenges that now face Venezuela? And what's the best path to reanimate Venezuela's economy and get it back on a path to durable economic growth? And is AI, globalization, whatever, making it, assuming that the resource curse is one of those challenges, is it getting easier to overcome the resource curse over time, or harder?
>> So, thank you for the question. You know, January 3rd started with jubilation in Venezuela because we had tried to get rid of our dictator in a thousand different ways. We marched, we protested, many of us led by Le Paulo Lopez, MP 96, who spent quite a few years in jail.
And so, we got rid of the dictator, and then the US government announced a strategy of stabilization, recovery, and political transition, they said, in that order. And I immediately wrote an op-ed saying that there's going to be no recovery without the reestablishment of rights because it is the absence of rights that destroyed the economy, and it's the absence of rights that is preventing people from investing, from going back, from betting on the future, etc.
So, the reestablishment of rights has to be put first. And, you know, a government that has the consent of the people is a fundamental element for even the big guys, in the big, big oil companies, to want to deploy billions of dollars in any setting.
So, my fear is that the US is currently dragging its feet on a return to democracy. That the US is asking Maria Korina Machado not to go back to Venezuela just yet and to wait. That, in the meantime, the Venezuelan kleptocracy that destroyed the country is making great alliances with the American kleptocracy and allocating oil contracts not in a competitive way, but in dark, you know, back rooms.
And so, I worry that we are not on a path to a return of rights and democracy, and that we'll probably not get back to democracy in Venezuela until Venezuelans go back to the street and impose a different reality. So, it will be on us. Democracy will come not from the great heart of Trump, but from the force of Venezuelan will.
>> Leopold. Do you want to?
>> Oh, thank you. So, my name is Leopold Lopez. I'm MP96, and I want to start by thanking... Thank you. I want to thank the Kennedy School because when I was in prison, it was very important, the support I got from here. And I also want to tell you that there are many people that have graduated from this school that have gone to prison, like my friend and brother Felix Mariaga, and there are many others.
So, I would like to point out the fact that the fight for democracy, and not just for economic growth, is something that we learn here, and that we went back to our countries and we struggle for that. I completely... I'll go to the Venezuela question, but just a comment on where we were 30 years ago when we were here. It was a moment of victory, and it was this idea that was the main idea: that market economy was going to drive to democratic reform, almost as trickle-down democracy. Well, 30 years after, it didn't happen.
So, I think it's very important to understand what Professor Houseman was saying. How much does democracy, how much the guarantee of rights that can only happen under the rule of law and democracy, is part of this new view of globalization? Yes, tariffs, yes, blocks, yes, spheres of influence, but how much does democracy and the promotion of individual rights and rule of law is important for this new wave of globalization?
In terms of Venezuela, I fully agree with Professor Houseman that we are much better off than what we were on January the 3rd. Much better off. Maduro is out. Maduro is facing justice, and that is justice for the Venezuelan people. But we are not where we want to be. We are not where we want to be because for us in Venezuela, democracy and freedom are not theory. They are not poetry. We don't drive our views of these issues from theory; we bring them through our own experience.
So, until we have democracy, until we have a flourishing nation that can provide the protection of individual rights, we will continue to struggle, and that's our challenge. That's the challenge of the democratic leadership, and I believe that that's the clear view and hope and commitment of the Venezuelan people. So, I am optimistic. I am optimistic that we will get there, but I know that it won't be just waiting and seeing that somebody else from the outside will bring us democracy. Democracy needs to be brought from the bottom up, from the people to the nation, and that's what we will continue to do.
So again, thank you to the Kennedy School. Thank you to my class, 1993-1996. It was great support that you gave me when I was in prison. Thank you.
>> Thank you. Thank you. Thank you, Leopoldo, for your words and for showing what the Kennedy School does and how much of a community this is, and how important it's been for many of us that have then gone back into very complicated circumstances, particularly folks like yourself.
So, we have to wrap it up. They're forcing me. I mean, despite being Spanish, I would have run this on a Mediterranean schedule. So, we would have gone for at least another 15-20 minutes, but, you know, this is a very Protestant environment, so I need to stick to that.
But anyway, it was wonderful to share, to be able to moderate this panel. Thanks to all of you for your time, and thank you to all of the alums for being here. Thank you very much.