Oren Cass 与 Noah Smith 辩论关税的真实影响

The question is whether free trade with China advances free markets or whether it distorts and wrecks our free market. We have treated free trade as the natural extension of free markets. If you attempt to support free trade with a non-market economy, you are not actually advancing free markets in any significant way at all. You're actually dramatically hindering them.

People focus obsessively on the trade deficits and surpluses, on the imbalances, on the net amount. But basically, the more important effect is the gross—the total amount of exporting that we're able to do. In that sense, trade between us and Europe is positive-sum. Without trade being balanced, trade is positive-sum.

We're really excited to have you on because we've been talking about some topics over the past few months that you've been thinking and writing a lot about, and that you think have in fact informed, inspired, and encouraged some of what's actually happened. Before we get into some of the specifics, I wanted to give some of our audience who may not be familiar, especially on the a16z side, an opportunity to introduce your platform, so to speak, or the main topics and ideas that you're affiliated with.

We'll link to the book and some of the articles about a newer conservatism, but maybe you can help define it in broad strokes, and we'll get to some of the specifics.

Sure. I'll try to keep it short. Let me know if you want me to say more about anything. I founded American Compass back in 2020 with the mission to restore an economic consensus that emphasizes the importance of family, community, and industry to the nation's liberty and prosperity.

I suppose from one perspective you'd say, “Well, that seems self-evident and unnecessary.” But from another, I'd say it represents a fairly dramatic departure from the way economics has been conducted and economic policy has been conducted, especially on the right of center, but probably including the center-left as well. Therefore, it has really given shape to our economic trajectory over the past generation.

Our view is that I think we have seen excessive faith in markets and excessive assumption that whatever produces efficiency, whatever maximizes corporate profits, will also be best for everybody. That faith has failed in 2 respects. First of all, that is in fact not best for everybody. You can get very unequal and distorted outcomes.

Secondly, even if you had the economic system operating the way you might ideally wish it would, that's not actually going to take care of a lot of the things that matter most to people. The pairing of family, community, and industry may seem a little bit odd. Those aren't 3 things that always get paired together, but we see them as 3 things that are incredibly important to flourishing—both human flourishing at the micro level and national flourishing at the macro level.

Markets simply don't guarantee those things. Markets can uphold those things, but there's nothing in economics that says that markets will. The core of our argument is that if those are things that anybody, and especially conservatives, care about, then they have to be willing to think about an approach to economics and markets that goes beyond just trusting that whatever comes out of the market is going to be good. That then leads toward all sorts of interesting policy conversations that I think have been absent for about a generation now.

I would like to know how restoring manufacturing—because I've been calling for restoring manufacturing for a long time for various economic purposes, higher growth, national security, and things like that—would strengthen family and community in America. That's a nexus that I really haven't seen much evidence on.

I know, observing countries like Germany and South Korea, where the manufacturing share of industry has remained remarkably high, that they have extremely low fertility rates, low rates of family formation, skyrocketing divorce rates, and, in Korea's case, an extreme suicide problem. So I'm thinking: If we could restore manufacturing, which I would like to do, how would that restore the family?

I guess those are questions at very different levels of specificity. Let me say a little bit more about the broader question first, and then, if we want to dive into the manufacturing piece, certainly we can.

To your point, Erik, about this being Econ 102, I find myself back in Econ 102. It seems to me that a lot of the gap is between the oversimplified models that we teach in Econ 101 or Econ 102 and the set of factors that are actually relevant in a modern economy.

To give one example, if you are learning comparative advantage in Econ 101 and the wonders of free trade, you are almost certainly talking about an economy with 2 markets and 2 goods. They are almost invariably agricultural goods or something relying on natural resources. You're going to see that if you change how much each country produces and trade some of one for some of the other, then both can be better off.

That's great. I totally agree. No issue with that. The problem in the real economy is 2-fold. One is that, of course, most of what's being traded is not natural resources, where maybe you're a place that grows a lot of avocados or maybe you're a place that has a lot of fish. It's manufactured goods, and advantages in manufactured goods are a lot less likely to emerge from some sort of natural endowment of a country and much more likely to emerge from the explicit policy choices that the country makes.

A good example is Taiwan. There is no special silicon on Taiwan's beaches that makes it the best place to fabricate advanced semiconductors. If you are expecting to just see everybody have a natural advantage in something and that's what we trade, you'll be very disappointed. You'll find that the countries that strategically attempt to develop an advantage in something get to excel in that, and some things are more economically valuable, have more spillovers, and have more security implications than others.

The second problem, of course, is that you don't actually have to trade goods for goods at all. You can have the situation we have in the American economy today, represented by a trillion-plus-dollar trade deficit, in which you are exchanging goods for assets. If you imagine going back to your happy little example—I guess in my Econ 101 class it was fish and sweaters—and imagine that one country catches the fish and makes the sweaters while the other country just issues Treasury debt to buy fish and sweaters, it's not nearly as obvious to me that that is actually a welfare-enhancing exchange, certainly over the long run.

The one other thing I'd add that's even more conceptual, and that I think we've really failed to grapple with until recently, is this: To your point about being free-market enthusiasts, I, too, am a free-market enthusiast. The question is, does free trade with China advance free markets, or does it distort and wreck our free market?

We have treated free trade as the natural extension of free markets. If you are for free markets, you are for free trade. But I think what we've learned is that if you attempt to support free trade with a non-market economy, you are not actually advancing free markets in any significant way at all. You're actually dramatically hindering them.

That's one example of an area where I think there's a big problem. We do a lot of work also on industrial policy, financial markets and investment flows, labor policy, and competition policy. All of it ultimately follows the same template of asking: If we look at the world as it actually is, and at what economics actually can and cannot promise, where are the places where policy is probably going to have to play a role if we want to get the kinds of outcomes we would actually say are good outcomes?

Just to follow up on that point, your proposed policies—reduced immigration, trade protectionism via tariffs, policies to increase domestic industry and manufacturing, and policies to increase fertility—are they actually similar to China's? Are you kind of turning us into China, or kind of mimicking or mirroring its strategy?

I don't think so. I guess I would say that just because China is doing something doesn't make it bad. But, of course, virtually everything you just described was also part of the American tradition until very recently.

America had very high tariffs and built its domestic industry behind a wall of protectionism. America aggressively restricted immigration. America had quite aggressive industrial policy, and so on and so forth.

Conversely, China—I believe that just last week they announced they might be doing some sort of pro-fertility policy—but historically, China has been the poster child for extremely unwise anti-fertility policy. So I wouldn't characterize anything we're doing as saying we should become more like China.

I would say that, to the extent that we are going to operate in a global economy, we need to have policy that accounts for the existence of China, and that is unfortunately something that we have declined to do.

Oren, I have a question. Do you think that Trump’s tariffs overall will increase U.S. manufacturing output and employment?

In the long run, I think it definitely will.

How long is long?

Well, let me answer the question. I think if the long run is anywhere upwards of a few years, it will. The question is how long it takes to actually change investment decisions and then how long it takes investments to come online. But I would expect to see the results of some of the investments we’re already seeing being made come online in the next few years.

A good proxy is how long it took Japanese automakers to set up in the U.S. after their imports into the U.S. were badly constrained. From there, a lot of it comes down to what we do on policy. Can we actually create a credible, long-term, consistent commitment to a tariff policy that makes people believe they should invest, and can we do the other complementary things—on the supply side, what I would call the industrial policy to actually support investment and workforce development? I think there are other things that we need as well.

So, to be clear, since Trump announced his “Liberation Day” tariffs, we’ve seen deterioration in manufacturing PMIs, purchase orders, investment plans—every indicator of manufacturing you’ve had is slumping now, notably beginning directly after Trump’s announcement of those big tariffs.

We have a pretty good idea, when we do surveys of people who run manufacturing companies, what they say is exactly what basic economic theory would say: They’re having a very difficult time sourcing intermediate inputs. There are all kinds of things that we get from supply chains that they can’t source. As a result, they’re cutting back on building factories, investing in and ordering more manufactured goods, et cetera.

So, if I’m hearing you correctly, the idea is that this is temporary, and that once we reestablish entirely within-country supply chains, this will reverse and manufacturers will start investing a bunch. We’ll have a manufacturing boom, and the manufacturing bust that we’re seeing now, apparently as a result of the tariffs, is simply temporary pain for long-term gain. We just have to stay the course. Is that an accurate summary of what you’re telling me?

Well, I guess I’d prefer to answer questions that you ask instead of trying to give a yes-or-no response to your statement of a case.

I would like you to evaluate the fact that all the manufacturing indicators in the economy are flashing red directly after the announcement of Trump’s tariffs.

Well, first of all, I would dispute your characterization that they’re all flashing red. I think, in general, the distance between the overheated rhetoric we have gotten from economists and the actual performance of the economy has been pretty dramatic on almost every front. There have certainly been a number of manufacturers who have said this has made life harder. There have been others who have said, actually, this is inducing us to make substantial additional investments.

We’ve already seen TSMC, as 1 example, in fact say that they are slowing investment in Japan to invest more quickly in the U.S. We’ve seen a very good story on drugmakers in The Wall Street Journal, I think it was probably last week, saying that, yes, they actually thought the effect of the pharmaceutical tariffs under consideration would lead to significant reshoring of drug manufacturing.

So, I guess if I look at both the economy’s actual performance over the last 6 months and the more anecdotal evidence, I would say it’s absolutely too soon to tell one way or the other. But what gives me a significant amount of optimism is that I do think you’re seeing companies, generally speaking, respond to the incentives in front of them by thinking about how they would adjust their businesses accordingly.

The intermediate-goods piece is definitely an important one that creates short-term challenges. But at the end of the day, if the goal is not to make America the assembly hub of the world, but to actually reindustrialize, you have to create incentives to bring back those intermediate supplies. You can’t just say we want cheap intermediate supplies so that we can, I guess, assemble for re-export.

I don’t know. I guess you’ve said all the manufacturing signs are flashing red. That has not been my perception of the economic coverage over the last month or 2.

Oren, what data sources are you looking at when you’re evaluating the state of America’s manufacturing in the last few months?

Well, there’s manufacturing employment. I guess one number I’m very interested to look at, that we won’t have for a little bit longer, is Q2 capital investment, which I believe we’ll get later in August. There are various manufacturing sentiment indicators. There is actual industrial output, capacity utilization, these kinds of measures.

We do an awful lot to measure the performance of the manufacturing sector. I think certainly we saw, in the immediate aftermath of the tariff announcements, a lot of downturns and a lot of concern. But I’m not sure how much of that has actually been borne out at this point, several months later.

You know what PMIs are, right?

Yeah.

The Institute for Supply Management, which is a consortium, surveys manufacturing companies. They call them up and ask them, “Are you investing? How much are you investing? Do you have more orders?” They do these surveys, and it’s pretty rigorous data.

Here, let me read from a Bloomberg article from the 1st of this month. It said, “U.S. factory activity contracted in June for a 4th consecutive month as orders and employment shrank at a faster pace, extending the malaise in manufacturing.” Bookings contracted by the most in 3 months and have been shrinking for the past 5 months. The employment index, which measures manufacturing employment, has fallen significantly.

The purchasing managers’ index, the PMI, is also below 50, which signals contraction in the industry. Bloomberg reports on this, The Wall Street Journal reports on this—everyone reports on this—as a reliable first indicator of how manufacturing is doing. It’s not perfect, but over the long term it’s pretty correlated with government statistics on how manufacturing is doing.

So I’m wondering: Where are the alternative statistics? Where are the numbers I’m not seeing that show good news in the manufacturing sector since Trump’s tariffs? Where is it?

Well, again, I just gave you a long list of other economic indicators we have that all seem to be holding up quite well.

They’re not. Wait, no, they’re not. Manufacturing employment is not doing well. We’re not seeing a rise in manufacturing employment. We’re not seeing a rise in manufacturing industrial production. We’re not—

Right. No, those are just exactly the things I said I didn’t think you would expect to see over the initial months of a policy like this, because it literally takes several years to do capital investment and actually build things out.

I think one thing that really frustrates me in a lot of these discussions is that I think back to the argument from the other side, right? When the argument was, “Okay, we’re going to aggressively expand free trade. We’re going to embrace China.” And, oh, we got a lot of initial negative results from that. But that’s okay. Anybody who focused on those doesn’t understand the long run and the way that everybody’s going to actually end up better off someday.

But then when we go in the other direction and say, “Hey, actually, reshoring is going to be a process. You have to shift incentives. There’s going to be disruption. Here’s the better thing we might build toward,” you just get stuck on a podcast with somebody reading you the results of last month’s phone-survey data.

So I do think we have to evaluate a little bit more: What were the things that we did or did not expect to happen over the initial months, versus what are the things we did or did not expect to happen over several years? And to say that most of the economists’ predictions of everything that was going to go wrong initially have essentially been wrong—I mean, that—

What about the data I just read you? I mean, that was what economists predicted: Tariffs would interfere with supply chains, trade, and intermediate goods, and this would lead to a contraction in manufacturing. That’s exactly what we’re seeing now.

Yeah.

I don't understand how that prediction has gone wrong.

Yeah, I guess I meant more broadly: everything from the direction of the dollar to the performance of the stock market to the level of retaliation we should expect. There was a sort of consensus model of, “Here's what's going to happen if you impose tariffs, and why it's extreme; it's all folly,” and that world has not played out. So I do think it's important for those who were sure they knew exactly what tariffs were going to do because they took Econ 101 to actually step back and recognize that the world does seem a little bit more complicated than that.

I agree that the world is complicated, and I agree that there's a lot of uncertainty over the impact that tariffs will have. I can also tell you that the predictions I know economists got most wrong are about the direction of the dollar. That is one prediction that went the opposite way.

However, I do think that if we're talking about important effects of tariffs, the effect on manufacturing has got to be an important one. The whole idea of Trump's tariffs—maybe not Biden's tariffs, but the whole idea of Trump's tariffs—is that this is going to lead to a renaissance in manufacturing. I understand that, yes, in 5 years you could see something different. If that's the argument, if that's what we're placing our hopes on, then let's be explicit about that. Let's say we expect to see short-term pain for long-term gain.

But I will say that economists were right about that.

Pain in the manufacturing sector—and that's exactly what all the data are showing us now. Yeah, I've been extremely explicit throughout, in discussing tariffs, that there's short-term pain associated with them along various axes. Frankly, I'm surprised that we haven't seen more pass-through to consumer prices than we've seen. I would certainly expect to see an impact there as well.

The long-term gain here—I mean, a few minutes ago, I said I would expect to see this over the long term. You said, “How long?” I think I said 3 to 5 years. That's a reasonable range in which to expect to actually see a manufacturing sector that is performing differently. The good news is that there are other dimensions on which we should see things sooner. Capital investment, construction, and so forth are going to be leading indicators of that.

I'm not sure who out there was saying, “If you impose tariffs, manufacturing employment will magically materialize in factories that haven't been built yet.” Maybe someone was saying that, but it seems a little bit unfair to take that as the pro-tariff argument when it's obviously not the actual basis on which the argument is made.

But I'm saying that economists were right about this—the problems in the manufacturing sector that we're now seeing. It was a correct prediction by economists.

That sentiment would turn down in the initial months.

I'm not sure that anyone gave a month timeline for it, but the idea that this would be generally bad for manufacturing and would cause a contraction in the sector—that's what we're seeing.

And we know the mechanism by which this happens, which is disruption of intermediate-goods trade.

And, um, go ahead. I think Oren is saying, or acknowledging, that there would be some short-term pain for long-term gain, as you described it. I think he's also acknowledging—feel free to edit this—that, yes, it does make sense that economists predicted some sort of decrease in employment because the factories haven't been built yet, et cetera. But you're also saying, hey, economists didn't get right what would happen in the broader market and with the dollar, to which Noah also sort of acknowledged, especially the part around the dollar. Feel free to edit any of that characterization.

I have a broader question for both of you: What would cause you to change your mind about the long run? What would need to happen or occur such that you have a different view of the long-term effects of these tariffs?

Well, I think, as I was saying, the intermediate question that I'm most interested in is what happens to capital investment. At the end of the day, the sequence of events that you're trying to influence with tariffs is to induce relatively more capital investment domestically, to expand domestic production capacity.

I think, frankly, in a lot of ways, that was already at elevated levels because of other things that Noah and I agree on, like the CHIPS Act, which I've often highlighted as a dollar-for-dollar extremely effective way to start boosting investment. If we start to see those kinds of investments now in other sectors of the economy, and if we see that at a sustained, elevated level over an extended period of time, that would be the kind of sign, over the next year to 3 years, that this is actually doing what we want it to do.

The benefit that we would get, if it actually works out, is that we would see, in the longer run, a significant expansion of manufacturing capacity. We would see that in output, in manufacturing as a share of GDP, and in employment. In a lot of ways, what I'm most concerned about is that, if it's working, we need to see it in the productivity data, because manufacturing productivity has actually been falling over the past decade, which is a huge problem.

If we are actually seeing investment in a renewed and revitalized manufacturing sector, that absolutely is going to be—and we better expect will be—much more automated than the kind of manufacturing we've had in the past. That should show up in the productivity data as well.

I have another question here, which is that you've often talked about what you see as the shortcomings of economics as a discipline for predicting the effects of tariffs and economic policy in general. When you're predicting the effects of tariffs, what do you rely on? How do you make those predictions? Do you have some theories, other countries' examples, historical examples, some other discipline like sociology, or perhaps just some stuff you thought of on your own? Where are your predictions coming from?

Well, I guess I'm happy to answer that in detail. I figured, Noah, that you were going to touch on the prior question also, but I'll just add—

That's the prior question. I'm sorry. It was basically: Noah, what would—then let's circle back to Noah's question after this. But, yeah, no, let's hear your take on what would cause you to be excited about the long run for tariffs. What would need to change, or what would you need to see, for you to be like, “Oh, actually, it's having a long-run positive effect”?

A boom in manufacturing investment, for sure. If we see that—if we see the U.S. manufacturing sector actually get revitalized in a major way—that'll be exciting. So far, we're seeing the opposite. If that turns around at some point, that'll be great.

Imagine a world in which it does turn around. What needs to happen for that to happen?

Oh, you mean what do I think would actually cause that to happen? I think one thing would be industrial policy. I also think that, in terms of tariffs, we should have essentially entirely free trade with allies. I think that would be a wise thing to do, because Japanese and European manufacturing don't actually threaten America. In fact, we gain much more from pooling our markets so that we can attain scale similar to what China enjoys.

China is pretty much the only competitive threat we face, and all our allies are facing it at the same time. We're fighting with our allies instead of pooling our resources to stand up to China. But that's a bit of a sidetrack for me. I wrote about all these things pretty extensively during the Biden years, and I thought this is what we ought to be doing.

I think industrial policy is a big missing piece. There are also various capacity-building policies, like infrastructure and education. When you train a manufacturing workforce, one really reliable thing we see is that when we have really good vocational schools, you get better factories. Just train people to work in factories. Tim Bick has done a lot of work on that.

I think it's pretty clear what to do to affect a manufacturing revival right now. One of the big indicators that I look at is factory construction, and factory construction has to be adjusted for the price of building new factories, because we don't want to just make a chart of inflation.

When you look at factory construction—real factory construction—you saw it basically flat at a flat level for decades and decades, at a fairly low level. Then you saw this enormous boom during the Biden years, with all the factories concentrated in the areas where Biden's industrial policy was incentivizing factories: chips, batteries, and other things like that. It was just this enormous boom.

Now, since Trump got into office, you've seen that boom go into reverse. You're seeing the pace of new factory construction falling since “Liberation Day.” That's a bad sign to me, but I accept that in 5 years we may see something else.

I'm just genuinely curious: when I mentioned that I'm very interested to see in August what the Q2 factory-construction data looks like, what data is already out there on that?

It's called total construction spending on manufacturing in the United States. You can look at that on FRED, or you can look at it anywhere, and it shows the figures in dollars. You can also look at it as a percentage of GDP if you want, but I prefer to look at it in dollars. Then you have the Producer Price Index for New Industrial Building Construction, which is the appropriate price index for factory construction. You can see that under Biden this increased a huge amount, and then under Trump it has decreased a bit so far. We'll see if it keeps decreasing, but to me, that's a bad sign so far.

Okay. I think I would just echo some of what Noah had just said about some of these other factors that I think are really important as well. In terms of what it takes on tariff policy, if you really want these tariffs to work, stability and certainty are incredibly important. This is a criticism I've had of the Trump tariffs from the beginning: if what matters is not what people think the tariffs are going to be this week versus next week, what matters is whether people actually believe these are what the tariffs are going to be 3 years from now, when the investments would start to pay off.

That's where I think both stability in the tariffs—ideally, needing to legislate some of this instead of having it subject to legal challenges and a lot of changes in executive orders—and certainty are really important. Then, as Noah said, I think the industrial-policy side and the workforce side are incredibly important as well. If you ask me, do I think tariffs are an important positive step toward reshoring manufacturing, I would say yes, absolutely. If the question is how far we are toward the policy package that's going to get us maximum bang for our buck, I would say there's certainly a lot of work still to do.

And that maybe connects to Noah's other question about, well, how do you evaluate these things? In my mind, I think the right way to evaluate them is—I think a lot of the frameworks that economics provides us, in terms of thinking about incentives, particularly thinking at the margin and so forth, are exactly the right way to assess economic policy. I think where economics as a discipline has really gone off the rails is in overreliance on a set of assumptions that don't necessarily hold, and then overreliance on a set of models that require those assumptions.

This is where we've seen, frankly, a lot of very bad forecasts come out of—for instance, the prediction that free trade with China will benefit American workers and that we'll be making and selling more to them than ever before, and so on and so forth—when that sort of assumed a certain policy environment in China that did not exist. In a sense, the right perspective is to look at things from the perspective of a business owner, to look at it from the perspective of capital, and ask what the best opportunities are to earn a lot of money. This goes all the way back to Adam Smith and The Wealth of Nations and the invisible hand, where his argument about the invisible hand is not that this somehow automatically works by magic. His argument is that if you actually have incentives aligned—in fact, the first one he lists is if you have people preferring to invest domestically over investing overseas—then their pursuit of private profit will also advance the public interest.

For me, the question is: what are the incentives for people pursuing private profit? If we have policies that make it the incentive to offshore, to try to bring in cheaper labor, or to use financial markets to extract additional capital out of firms, then those are the things that people will do. At the margin, I think we have to ask whether the policies we're pursuing are pushing incentives in better directions or pushing them in worse directions.

So let's go back to this idea of where you get these ideas—the idea that tariffs will do this and that to help manufacturing. Where does this idea come from? Where did you get this? Did you just sit there and think, common-sense-wise, it sounds legit? Or are you using an analogy to say America's early growth in the 1800s, when we had high tariffs—we grew fast then and we had high tariffs then, so high tariffs will increase growth now? Where do you get this idea that tariffs will help manufacturing and the incomes and prosperity of the average American?

Well, I guess there are a few ways to answer that question. I think, one, as you noted, there have certainly been times in the past when we have used a much more robust tariff agenda. Conversely, our decision to drop tariffs and forgo them has produced a lot of the things that I think we've been a lot less happy with. If you look at somebody like Paul Samuelson's work, I often point out that in his seminal textbook, Economics, he goes through the arguments for and against tariffs and basically, at the end of the day, acknowledges that any country could pursue a beggar-thy-neighbor strategy. I'm sure there's a whole field of optimal-tariff theory, but the core case for the United States to embrace free trade—and Fred Bergsten makes the same point in the 1970s—isn't actually an economic case; it's a foreign-policy case.

You can also look at the policies of a lot of other countries around the world today that have obviously had a lot of success promoting manufacturing and industry. They certainly tend to do it in a more protectionist way. I guess it seems to me there are a lot of signals all pointing in the same direction. I might ask you the converse: what would be the reason to believe that free trade—certainly with a country like China—was going to be good for our manufacturing sector?

Well, I don't necessarily know people who did predict that, because I was a kid, and when I look back at economics papers on the topic, I don't see many economic papers saying that free trade with China would be good for the manufacturing sector specifically.

But is the implication that tariffs on China would be better for the manufacturing sector?

What I'm saying is that I don't know. When you're asking me to justify the supposed predictions of people, I don't know who made those predictions, and I don't know economic models that predict that. For example, if you take the dumbest, simplest Econ 101 model of trade in existence, which is David Ricardo's comparative-advantage model, that would predict that trade with China would be bad for our manufacturing sector. The idea is that if China has a comparative advantage in manufacturing, and we have a comparative advantage in services—or maybe agriculture, I don't know—we'll trade the things we're good at for the things they're good at. It's not a very smart model. It's pretty basic. It describes some things that happen, but not a lot of things, right?

And then if you use a slightly more advanced model called the Heckscher–Ohlin model, which predicts that capital-intensive countries like America will do more capital-intensive things, while labor-intensive countries like China in the 1990s and 2000s will do labor-intensive things, that predicts that we'll lose lots of manufacturing jobs to China. So those basic models would predict that the U.S. manufacturing sector—or at least U.S. manufacturing employment—would do poorly in the face of Chinese competition, due to free trade with China, but that America overall would become enriched because of other effects in the service industry and other industries.

When you're talking about these predictions that free trade with China would be beneficial for the manufacturing sector, I have to say I don't know where they are. I don't know who made them, and I don't know why they would have made them.

Yeah. I guess I'm making the point more as a counterfactual. We could either take the view that tariffs are probably helpful in protecting a manufacturing sector, or that they're counterproductive. You were asking me why I would believe that these things help, and I'm saying, well, I think there are a lot of sources that would point to them potentially helping. What I'm less clear on is what the argument from the other side is.

What's the argument that, actually, if you have tariffs, we should get rid of them to help the manufacturing sector? Presumably, one or the other has to be true: either tariffs are good for the manufacturing sector or tariffs are bad for the manufacturing sector. I'm suggesting they're good for the manufacturing sector. I took you to be saying you think they're bad for the manufacturing sector, but I'm not sure that's where it breaks down for me, right? Because, at least with China, as you were just saying, I don't know who is saying they're bad for the manufacturing sector.

Whether tariffs on China could have been used to help the US manufacturing sector in the face of Chinese competition in the 1990s or especially the 2000s, when the China shock happened, I don't know of any analysis of that. There are definitely reasons why tariffs would have hurt the manufacturing sector, and there are reasons why they could have helped. For example, if you had used tariffs as a kind of bargaining tool to force China to stop undervaluing the yuan, which it did to a significant degree in the 2000s, that might have worked. In fact, I think we should have done that.

But notice that that's using tariffs not as a policy in and of itself whose direct effects encourage manufacturing, but as a sort of bludgeon—a tool, a threat, a tool of economic warfare—to force China to stop doing something that harms our manufacturing sector. The best outcome would have been us using the threat of tariffs to force China not to undervalue the yuan, but then not actually following through with the threat because the threat worked. So I think when you look at Paul Samuelson's work on strategic trade, he talks about this: basically using the threat of something that would harm both countries, but harm them more than us.

Using the threat of that, but not the actual thing, because the threat works.

Right. And so I think that using tariffs as a threat and using tariffs as an actual policy—a protectionist policy to grow our own manufacturing industries behind protectionist walls, which is something lots of people talk about, and which Alexander Hamilton certainly did talk about—those are 2 very different things. So I think, if you're asking me what I personally would have done, going back and doing the 2000s over again, I would have said threatening China into not undervaluing its currency would have been a good idea for our manufacturing industry and for us in general. But that doesn't mean that tariffs are a policy that, if actually implemented, helps manufacturing.

But surely you can't use tariffs as a credible threat unless you're willing to actually impose them. I mean—

That's right. The idea is they hurt us, but they hurt them more, and so that's why they are a credible threat in the strategic trade idea.

Right. So if we—

We're hurt, but it hurts ourselves. That hurts us.

But if we take China today, which clearly is not playing by any of the rules of free trade and shows no openness or willingness to do so, how then are we supposed to react to that in your framework?

Well, the idea of game theory is that actors are rational, and that if you have a credible threat, the threat doesn't need to be used in equilibrium because if you make the threat, the other person will back down because they're rational actors who give a best response to a best response. However, if you have irrational actors and you make a threat to harm them by double the amount they harm us, but to harm both people, and then you end up carrying through on the threat, the fact is everybody gets hurt.

Even if they get hurt more, in a war, that's what you want. In a war, you take casualties in order to inflict greater casualties upon the enemy. If we're looking at economics as a war where hurting China is more important than helping ourselves get rich or have higher living standards, more prosperity, whatever, then fine. That's certainly an argument you can make. But if we're talking about helping our manufacturing sector and helping our people's living standards in general, hurting them double the amount we hurt us is not a particularly attractive proposition, as I see it.

Americans don't think of trade with China as a war where hurting them is more important than helping us. They just want stuff; they want prosperity. And I think so far, the short-term effects of tariffs have been bad. Again, if we're going to say, "Stay the course 5 years, 10 years, eventually things will get better," okay. But that's what the communists said, and it didn't end up that way.

My understanding of your view was that you were more sympathetic to, and maybe even in support of, some targeted tariffs toward China, but that tariffs toward our allies, too, just didn't make any sense. Is that right?

Exactly. That's right. The reason is very simple to explain: it's the reason of scale. Now, you won't get this in Econ 101 or even 102, but you will get this from Paul Krugman's Nobel Prize-winning work, where he talks about the importance of scale for manufacturing industries and for industry in general. Manufacturing is the most important industry that this works for: the more stuff you can sell, the more units you can make, the lower your costs go.

So if you can make 1 million cars, you can make cars a lot more cheaply than if you only make 10,000 cars. The more you can make, the better you get at making stuff. Now, China has unprecedented internal scale because they're a country 4 times the size of America, right? So they can make tons of units of stuff. America can't match that with just our domestic market.

We can pay more per car, but in terms of the scaling effects, the scaling effects depend on how much you make—how many things, how many cars roll off your assembly line. And we are just such a small country compared to them. We're one-quarter their size. Us trying to match China is like Germany trying to match us. Germany can make really high-quality stuff, but they can't match us volume for volume. And so they'll never be able to get the scaling just from their internal market that we get. That's why Germany focuses on exports, exports, exports.

What we need to do is find some other countries where we can say, "Our manufacturers get scale from exporting to you, and your manufacturers get scale from exporting to us," and we both get scale. Then we have this mutually beneficial agreement to scale our manufacturing industries such that we're not playing only in this tiny domestic market, such that we have a large market too, so we can compete with the Chinese because we can match them for scale.

Now, the only way to do that, given our small size, is to get other countries to pool their markets with us so we can get scale, right? We need Europe and Japan and Korea, and maybe India, which is the biggest country of them all, to partner with us so that we can scale. But that requires some reciprocity, right? We need to have our manufacturers be able to scale by using their markets, but their manufacturers also need to be able to scale by using our markets.

And then, as Krugman showed, our manufacturers and their manufacturers make similar things but slightly different things—slightly different versions of the same things. We'll make Harleys; they'll make Kawasakis. Then competition happens by differentiation. That's fundamental to Krugman's model, and it's what we actually see in, for example, US-Japan trade. We make broadly similar stuff, but not exactly the same, and both can flourish.

That's drawing from advanced economic theory, not the basic stuff you get in Econ 101. That's what I would do to resist Chinese manufacturing. I would get scale by pooling our markets with our friends. And if you look at what, for example, Elon Musk has said, Elon Musk said we need a complete free-trade zone with Europe so that we can all scale.

Now, if there's 1 American who actually does know about manufacturing, it's Elon Musk, who's the only entrepreneur who's been able to successfully scale manufacturing in America, even in the face of international competition. It's been pretty credible because his supply chains are far more domestically sourced than most manufacturers. But when he says we ought to combine our markets with Europe's markets so that we can all scale together, that's directly out of Krugman's theory. It's directly out of lived experience in manufacturing. So that's what I would do, and then penalize China.

I should say I agree with almost all of that and would like to see that as the American trade strategy. I think the concern I would have is that those 3 countries that I think you rightly focused on—Germany, Japan, and Korea—relative to their GDP, our trade imbalance with them is as large as it is with China.

When you describe Germany's strategy as export, export, export, that is of course Japan's strategy and Korea's strategy, certainly vis-à-vis the US. Just talking about the size of the trade balance in goods between the US and Japan, Japan, too, imports a lot from China and others, but it seems to me a fundamental problem that the US is dealing with in ideally constructing that sort of trade model is that these key countries that we want to be our allies in scale are themselves pursuing this kind of export, export, export but don't import model.

That's why I think it was very interesting to see the EU, at the start of the agreement that they've been announcing with the US—you heard the EU leader say quite explicitly, "We acknowledge that we're going to have to do some rebalancing here." So I don't know. Noah, maybe you disagree that there is an imbalance there, but it seems to me that as we move toward that model that you're describing, if it is going to work for the US, there is going to have to be a significant shift in the behavior of these trading partners.

Right. I actually have an important response to this, which is something I think almost everyone in these trade policy discussions gets wrong: they focus on net instead of gross exports. Let me give you a concrete example. If we export $0 to Germany and Germany exports $0 to us, and then we go to exporting $10 billion to Germany and Germany goes to exporting $12 billion to us, we have opened a trade deficit with Germany. We now have a $2 billion trade deficit, but our exports have gone up by $10 billion.

This is a good outcome for us because it allows our exporters to scale because what matters is the gross amount, not the net amount. People focus obsessively on trade deficits and surpluses, on imbalances, on the net amount. But basically, if ours and Germany’s both grow, and Germany’s grows a little bit more, that’s actually a good outcome. Even though we have a trade deficit, from the perspective of scaling our manufacturing industries, increasing GDP, and gaining manufacturing power, deficits and surpluses matter less than the total market size available to our manufacturers.

Trade deficits are always a sore spot because people have an inherent sense of fairness, and they worry about this stuff. That’s fine; I understand that we need negotiations, and we can try to rectify those things. It’s complicated because there are also financial effects on trade deficits, so it gets hard. We can talk about how to fix trade deficits later.

But I think the more important effect is the gross amount—the total amount of exporting that we’re able to do. In that sense, trade between us and Europe is positive-sum without trade being balanced. I’m not talking about comparative advantage, your standard Econ 101 stuff. I’m talking about advanced Paul Krugman stuff. I’m talking about what Elon talks about. I’m talking about manufacturing and scale economies, and I’m saying the total amount we export to these allies is more important than the delta at the edge.

But it seems to me that that assumes that active trade itself is somehow expanding total demand for manufactured goods. What I mean is—

It is.

But not certainly not one-to-one. I mean, if we talk about cars as an example, and you ask in which case U.S. car manufacturers have a bigger market: the 0-0 case, or when the U.S. is exporting $10 billion of cars to Germany but Germany is exporting $12 billion of cars to the U.S.? Unless you’ve created an extra $2 billion of demand for cars, you’ve just reduced U.S. car sales. So, no—well, I mean, the point is that you do.

I’ll talk in numbers, but in dollars; let’s assume they’re all the same price of cars. If we export $10 billion more to Germany and Germany exports $12 billion to us, what matters for the total amount of cars our manufacturers can make is whether the induced reduction in domestic sales is larger than the induced increase in exports to Germany. If Germany exports $12 billion of cars to us and our domestic car sales go down by $6 billion, while our exports go up by $10 billion, our total car manufacturing has gone up by $4 billion, even though we’re running a trade deficit and Germany’s car exports to us have destroyed some domestic demand for American-made cars. But the amount they destroyed was not sufficient to balance out the amount our exports gained.

Right?

The point is that when 2 countries start trading, the number of cars that people consume goes up. Absolutely. Does it go up by more than the number of cars represented by the imbalance, the trade deficit?

Cars go up.

Pardon?

Why will the number of cars go up?

Because cars are cheaper. Remember, we’re scaling. That allows us to make cars more cheaply. When cars are cheaper, people buy more. You get that second car, that third car for your family. Even people who don’t own a car could now have a car. You can also get more expensive cars, but that’s another story. But in terms of scaling, if we drive those costs down by increasing the size of the market, people buy a lot more cars overall.

If you look at vehicle ownership in America during the time that Detroit was facing this incredible competitive pressure, you actually saw Americans buying more cars, and you saw people in these other countries buying more cars.

But you also definitely saw a decline in demand for the American-made cars, right?

You saw a decline in domestic demand for the American-made cars.

No, total.

Right, but if you look at GM’s total vehicle production—

While you’re looking that up, let me make 2 related points. One is that this strikes me as a classic example of what goes wrong in our economics dialogues, where—

Oh, yeah, sales were going up and up. Total sales, even despite the competitive pressures from Japan and Germany, GM’s total global sales—

Yes. No, no—domestic production in American factories. That’s final assembly, by the way. You’ve also got to look at the total output. If you look at total manufacturing output in the American auto industry, it seems to me there are 2 problems here.

One is that we have the intuitive view that if you have a trade deficit, that’s probably bad for domestic producers. You have the seemingly thoughtful and nuanced economist view that, no, we have to focus on gross, not net, because we’re expanding the pie. Then you realize that that was relying on a bunch of assumptions about expansion in total demand and so forth, which aren’t necessarily true. In the aggregate, they are definitely not true.

The way that we know they’re not true is just by looking at the trajectory of U.S. industrial output, which has essentially flatlined since 2007.

Now, since 2007?

That’s right. But during the time of the China shock, it didn’t actually flatline. It actually rose.

Yeah, there was still a period of increase there. But if we’re talking—

That’s during the time when we lost the most jobs to China.

Oh, I agree. There were a number of different things going on. I’m just talking about the actual trajectory of the American manufacturing base. It’s not that demand for manufactured goods suddenly fell off. It is that virtually all of the increase in demand that historically a substantial share would have gone to domestic production has instead gone to imports.

And so when I look at that huge trade deficit, we see that it does in fact map directly to a decline in domestic output, and that’s the thing that people are, I think, rightly quite concerned about.

But if you look at when our trade deficit was big, our trade deficit was much bigger before 2008 and then shrank dramatically after 2008. That’s just a fact. Our trade deficit shrank enormously after 2008. However, that is when our manufacturing productivity flatlined, and that is when our industrial output flatlined—after 2008.

Now, when we’re looking at the 1990s and the 2000s, this era of hyperglobalization, we see industrial output rising and manufacturing productivity rising. But we see the biggest trade deficits we’ve ever had. So that’s a big fly in the ointment of your thesis. The timing there just doesn’t line up at all with the naive idea that trade deficits destroy U.S. manufacturing.

When I’m looking at this, I searched for and found an Economic Policy Institute report on the U.S. auto industry, and I see a chart reproduced from Goolsbee and Krueger (2015) showing seasonally adjusted light vehicle sales. I’m seeing an increase over the course of the 1980s and 1990s. Employment fell here, but output appears to have risen and risen as you got robots, machine tools, and things like that.

U.S. manufacturing grew and grew during the era of hyperglobalization and during the increase in competition from Germany and Japan in the 1980s, and even during the initial increase of competition from China and the newly industrialized countries in the 1990s and 2000s. Our manufacturing even grew a little bit during the China shock when our manufacturing employment fell. Our industrial production actually rose a bit, and that was during this era when trade deficits were much larger than they are now.

That timing doesn’t fit the naive story that we’re dealing with, that trade is a zero-sum game and that trade deficits destroy American manufacturing output. It just doesn’t line up. I think this is a case where actually thinking deeply about the economic ideas is a little more helpful in explaining some of these patterns than simply doing the common-sense thing of saying, “We’re selling $10 billion to them, they’re selling $12 billion to us, we’re losing, and therefore our manufacturing industry must be getting destroyed.” And that said, I do think trade deficits present a problem.

If they're too large, if they're for the wrong reasons, if we're just borrowing to consume, it's a short-termist consumption loan, basically, that we don't need. So I do think trade deficits are a problem, and it also violates people's innate sense of fairness and reciprocity. So I think that when we have trade with allies, we should work to make sure trade imbalances are minimized.

How would you do that?

How would I do that? Yeah. Just negotiations: “You know what? We'll open our markets to you if you help us rectify some of these imbalances.”

But what if our markets are already open and that, in fact, is contributing to the imbalance? I mean, what—

We can offer a lot of other incentives to these countries.

So, but we should offer—

Yeah.

I just want to understand the logic. Okay, so we agree there are imbalances. We agree that's a problem.

We could defend them militarily. Well, it seems to me we've been doing that, right? We could say, “All right, we'll step up our commitment to your defense if you help us do this,” or we'll—various other things, you know, our markets.

This fascinates me, right? Because it seems to me that we're—if you think about what the Trump administration's posture has been—

Are you giggling because you're imagining a world of free trade between the US and its allies, and you're very enticed by the possibility of the scale this would offer our manufacturers?

No, I love those things. I'm giggling because it is an interesting inversion of what we are seeing happening today, where the Trump administration's argument is, “Wait a minute. We have an extraordinarily open market. We have been defending these countries, and the result has been these imbalances.” If we are going to sit down and negotiate with them over that, then, exactly to your earlier point about game theory, we should use market access and our commitment to their defense as tools to force them into a position that we like better.

Based on everything you'd said up until the last minute or 2 of the discussion, that would seem to me to be a position very consistent with your views. But then when we got to the point of actually saying, “Great, what should we do about Europe?” it was sort of like, “Well, what else can we bribe them with? Can we offer them even more market access? Can we promise to spend even more on their defense?” I'm just not sure why that's the approach we would take.

I don't know. If we're saying that we're going to use the threat of tariffs to establish a free-trade zone, that's—but because of game theory, because the idea is that actors are rational and we won't have to use the threats and hurt both us and our allies—if that's the idea, then I have to say, well, it's not really a credible threat. It's only a credible threat if we're nuts.

The idea that we would hurt ourselves and our allies and China—the idea that we would use tariffs as a threat against China, a country that we wouldn't mind seeing taken down a peg, seems rational to me because you could use that as a credible threat. If we follow through and hurt China more than we hurt ourselves, that's a credible threat. But hurting Japan and Europe more than we hurt ourselves—hurting both of us—actually just lets China win. So it doesn't seem like a credible threat to me.

First of all, it doesn't seem like a credible threat. So this doesn't seem like a good negotiating tactic to me, because if we go through with the tariffs, as we currently are going through with many of those tariffs—not all of them, but 15% or something on our allies—okay, if we go through with that, even as we surrender to China, but that's another story, it just means China wins.

So that's my first point in response to this idea. The second point is, if this is your argument, we're back in the realm of tariffs are bad, but they're a good threat because we're threatening to shoot ourselves in the foot in order to get this free-trade equilibrium that we like. The idea is tariffs are bad, free trade is good with our allies, but we need tariffs as this tool to threaten them—to get to free trade with allies through the back door.

That's very different from saying, “Behind these sheltering, protective walls, American manufacturing will regrow because economists don't know what the hell they're talking about.” Those are 2 important points: the fact that this is a noncredible threat, and the fact that tariffs as a threat are different from tariffs as a long-term, actual policy. And we appear to be doing the latter. So those are 2 important counterarguments, I think, to what you're saying here.

Yeah. I guess I would just say on the first point, first of all, it apparently, at least at the moment, is a credible threat. I mean, the administration is using it, and I think certainly the reporting on the EU negotiation indicates that they were quite frustrated by how credible the threat was and the extent to which it therefore forced them to make a lot of concessions.

I do think the sort of other way of looking at it—which is, well, the US has to, first and foremost, be thinking about everybody's well-being, so these other countries can adopt policies that really are beggar-thy-neighbor and we can't do anything about it—that was certainly the kind of post–World War II, “We must preside over a liberal world order” argument. I don't know how much sense it makes now, given the costs that the US is sustaining under the status quo. So I do think that the negotiations are not as clear-cut a game theory as you were describing.

And then on the second point, I would just say it's interesting, and perhaps quite satisfying, to the extent to which we are landing in a very similar place at the end of this. My view about certainly the so-called reciprocal tariffs and the way the administration has spoken about the reciprocal tariffs is that they, in fact, see them as negotiating leverage, and their goal was to strike deals with these countries, but to have these countries commit to balanced trade. I mean, that was quite explicit. I think Stephen Miran's remarks at the Hudson Institute are clearest.

We seem to agree that tariffs on China do make sense and that our relationship with China—

Not all, but tariffs on, I would say, strategic industries with China. I don't think we need tariffs on Chinese clothes, because if China makes clothes, that's not a national security threat to us.

Okay, but China is sort of a—

Toys.

I don't give a damn if China makes our toys. Let them make toys. I think that's fine. I think there remains a balance question there and a question of what core inputs still do and don't matter to us. But the point being, China is in a separate category.

I think maybe where we would probably end up disagreeing the most is on the question of a baseline tariff, and that's where my view has always been that a baseline tariff on the order of 10% is a skewing of the playing field in international trade in a way that helpfully reflects a preference for domestic production. If you could do it in a way that actually had long-term certainty and predictability, and if you could also then use it as a revenue source—which meant that, relatively speaking, you also need less revenue from elsewhere—it would be a very good element of what is otherwise a quite skewed global trading system.

And so that's how I look at it. I think certainly we don't agree on all the pieces of it, but it seems to me there are pieces of it that we do agree on. Also, the intuition of those who are pushing for this kind of approach, myself included—I think there's a lot to debate, but it is not the kind of baseless, unfounded, and incoherent way of thinking of things that it sometimes gets presented as.

Assuming tariff policy gets pretty much continued and that we have somewhere around the 15% to 20% tariffs that we're going for right now, suppose that continues and suppose our manufacturing sector continues not to flourish. Manufacturing investment continues to weaken and deteriorate. Manufacturing orders continue to weaken and deteriorate.

At what point—how many years will it take for you to say, “Oh, maybe I should have read an economics paper or 2. Maybe economics isn't all just hooey, and maybe the economists were onto something”? How many years must I wait for you to update your ideas there, assuming that all that continues?

Yeah. I think it's an unfortunate way to end the conversation, given that we both know that I've read lots of economics papers on these topics—

I was being figurative.

Yeah. Well, you also wrote a very long Substack that wasn't at all figurative, making the same accusation. So I don't think that's a constructive way to conduct the conversation.

On the substance of the question, I would say that I think we need a year or 2 to see whether or not you actually get the investment response. Secondly, and related to that, I think it's very important to see whether you get the useful accompanying policies. That is, if what people are saying is, “We really want to invest, but we can't because of workforce,” we do need to actually address the workforce stuff.

If a couple of years from now you are not seeing the investment response, and either we can't do the workforce stuff or, for political economy reasons, it is not feasible for us to do that stuff, then I would be the first to admit this is apparently not an effective strategy for pushing toward reshoring.

Got it. Well, that seems reasonable to me. Erik, take it away.

No, I was just going to ask if any of you wanted a last word, but I feel like this has been a constructive conversation. As someone who's not an expert in these topics, I feel like being able to hear the discussion and debate in real time and understand both positions better has been edifying. I really appreciate you guys coming on and having a discussion.

This was a lot of fun. Thank you for having me.