China’s Trade Surplus, Part III

How should policy respond?

Live view of Chinese exports

This is the third entry in a series on China’s immense trade surplus. In a saner world — a world in which the United States weren’t ruled by a madman — this surplus and its disruptive effects would be the biggest issue confronting the global economy. Unfortunately, we must cope with Donald Trump. Thus I paused the third entry in this series by a week to survey the results of a year of Trumponomics.

But regardless of how crazy U.S. policy gets, the challenge of China’s enormous and growing trade surplus isn’t going away. As I wrote in Part I of this series, China has an enormous economy and is relying on its export prowess to make up for its domestic failures. And the political economy that has set China on this path isn’t likely to change any time soon. As I pointed out in Part II, there is a theoretical case for simply buying the cheap stuff China supplies. But in practice China’s massive surplus creates serious problems for the rest of the world, and therefore requires a serious policy response.

“The rest of the world” for this discussion mostly means the United States and Europe, who are the key global economic players outside China. While other nations also have choices to make, only the US and the EU are big enough to influence China’s behavior.

I realize that offering serious policy advice to the current U.S. government is like preaching to baboons: You won’t get heard over the hooting, and even if they did hear, they wouldn’t understand. But these baboons people won’t be in charge forever — at least I hope not. And European policymakers, while often divided and ineffectual, are still compos mentis.

Beyond the paywall, then, I’ll address the following issues:

1. Reprise: Why China’s trade surplus is a problem

2. The case for industrial policy

3. Tariffs versus subsidies

I will discuss which industries should be supported and how — but not today, because this post is already long and I don’t want to overburden everyone.

Reprise: Why China’s trade surplus is a problem

China’s trade surplus, which reached approximately $1.2 trillion in 2025, is, as Donald Trump would say, LIKENOTHING ANYONE HAS SEEN BEFORE. What’s crucial and unprecedented is not the dollar figure; it’s how big that surplus is relative to the world economy. In particular, China’s trade surplus in manufactured goods — exports of manufactures minus imports — is a larger share of world GDP than any country has ever captured in the past:

Source: Weilandt, Bauer and Setser

I don’t want to engage in crude mercantilism and portray China’s trade surplus as an unmitigated bad for the rest of the world. China isn’t “stealing” from other countries by selling more than it buys. If anything, China is subsidizing the rest of us by selling us goods cheaply. Also, because the balance of payments always balances, the counterpart of China selling more goods than it buys is China purchasing more overseas assets than it sells. In practice China buys an enormous quantity of U.S. Treasury bills and other safe assets that pay low interest rates — and helps keep those rates low.

As I’ve written before, China is like a store that offers merchandise at highly discounted prices and also offers buy-now-pay-later plans with low financing charges. In effect, China is subsidizing the rest of us.

That said, China’s soaring trade surplus creates three major problems in importing countries.

First, China’s surging exports are economically and socially disruptive. The “China shock” caused by the rapid growth of Chinese exports between the late 1990s and around 2010 eliminated well over a million jobs in the United States, with job losses concentrated in a relatively limited number of communities. In terms of overall US employment, these job losses were offset by job gains in industries not in the path of the Chinese export surge, such as in healthcare. But most of the workers and communities displaced by the China Shock were not able to take advantage of these new opportunities. So while overall U.S. employment and economic growth do not appear to have suffered from Chinese competition, significant numbers of workers and their communities did.

And now we are experiencing another China shock, which is larger than the first.

The second reason that China’s enormous trade surplus is harmful to its trading partners rests on an old argument against completely free trade — the importance of retaining domestic capacity in industries crucial to national security. This concern has much more force now than it did, say, 30 years ago. As Abraham Newman and Henry Farrell have argued, we are living in an age of “weaponized interdependence,” in which governments that have control over economic “chokepoints” — crucial nodes in the world production system — can and do use that control to throttle, or threaten to throttle, geopolitical rivals.

After winning the Cold War, the United States controlled most of these chokepoints. And until Trump II, the U.S. exercised restraint, mostly respecting international agreements and the rule of law. As a result, the targets of weaponized interdependence tended to be rogue states like Iran.

Today, under Trump II, America is looking more and more like a rogue state itself. However, I hold out hope that that won’t last. And, in any case, it’s not the topic of today’s post. The point instead is that massive Chinese trade surpluses threaten to give the Chinese government control of multiple chokepoints. And China is an authoritarian state that can’t be expected to refrain from weaponizing the rest of the world’s dependence on its exports.

For example, China recently used its dominance of rare earth production and especially rare earth processing to put the screws on the United States, Japan and the EU amid trade disputes. In a recent dispute between the Dutch government and China over unauthorized technology transfer, China threatened to disrupt the entire EU automotive industry by withholding critical supplies of semiconductor chips. So it’s entirely reasonable to harbor concerns over allowing China to acquire leverage over chokepoints in the global economy.

Third, there is the risk that China will lock in a long-term advantage in the industries of the future. To illustrate this concern, consider a case of international competition that doesn’t appear related to China: U.S. leadership over Europe in information technology. Last December I showed that it’s not America as a whole that dominates this sector. Rather, that dominance overwhelmingly arises from high-tech clusters in the Bay Area and Seattle, where an early lead has created a self-reinforcing “ecology” of skilled workers and specialized suppliers that makes it very hard for Europe to break in.

China already appears headed for a similar lock on a number of industries, such as solar panels and electric cars, and other industries will follow if massive trade surpluses continue. In that December post I argued that Europe hasn’t suffered from U.S. dominance in information technology because competition within that sector diffuses the benefits and the economic rents earned to consumers. But it seems particularly risky to assume that this will be the case if and when China dominates multiple industries of the future. For example, can we be sure that China won’t embed monitoring capabilities in its information technology – a risk that has been associated with Huawei?

In short, there are very good reasons for other nations to refuse to passively accept China’s huge trade surpluses. However, But this response shouldn’t take the form of crude protectionism. In the next section I will discuss policy options to counter China’s trade surpluses.

The case for industrial policy

There’s a widespread perception that economists have nothing to say about globalization other than “Yay free trade!” In reality, while economists do often act as cheerleaders for free trade, they have also devoted considerable attention to the conditions under which policy should deviate from pure free trade. Equally important, when those conditions are met, economics has a lot to say about how policy should deviate from free trade.

The case for free trade is the same as the general case for letting markets work. If importing a good is cheaper than producing it domestically, then limiting imports and/or subsidizing domestic production is usually a misallocation of resources. That is, those resources could have been used more productively making goods for the domestic market or for export, instead of competing with lower-priced imports. Consequently, government policy shouldn’t try to overrule the market by supporting domestic production against imports unless an industry satisfies a key condition: that maintaining domestic production generates sufficient benefits that aren’t taken into account by markets. If that condition is met, the government should engage in an industrial policy to support that industry.

For example, suppose that the US could cheaply import all its vaccine supplies from China. But in deciding whether to allow the American vaccine production industry to wither, a policymaker would have to take into account the possibility that imports of Chinese vaccines could be cut off and put the country’s health at risk. Hence it would be extremely unwise to allow cheap imported vaccines to undercut and decimate domestic production.

Technologically advanced industries in which there are extensive interdependencies between companies, and in which advances are made as a result of the cumulation of skill and knowledge, also satisfy this condition. Silicon Valley is the most famous example of this kind of industry, in which technology spillovers between companies lead to a self-reinforcing clustering of companies in a given location. The self-reinforcing nature of advanced technology means that it obeys a “winner-take-all” dynamic: once such a cluster takes off, it can be virtually impossible for another set of companies to catch up and compete with it. Thus, the European digital industry was never able to successfully compete with Silicon Valley because it started too late and the EU never engaged in policies to shelter it from American competition.

In the past, particularly when I was a student, trade economists were skeptical of interventionist trade policies based on concerns about national security or disruption of local labor markets. There were good reasons for their skepticism. Patriotism is the last refuge of scoundrels, and there’s a long history of politicians insisting that whatever they want to do is essential for national security. Meanwhile, arguments that we should limit disruption from imports because they disrupt labor markets can turn into arguments for blocking all economic change, which would mean blocking all progress.

But national security concerns, which seemed overblow in the 1980s and 1990s, are unfortunately alarmingly relevant now. There’s a full-scale war raging in Ukraine, while Chinese military action against Taiwan seems all too plausible. I’ve already mentioned the possibility that China, an authoritarian regime that plays a crucial role in many supply chains, might weaponize that position. And given the state of American politics, Europeans have reason to be nervous about their technological and military dependence on the United States.

Moreover, the China shock between the late 1990s and 2010 shows that reducing disruption from trade flows can be a valid policy goal. In principle we shouldn’t single out disruptions due to trade as opposed to disruptions due to technological change. But the fact is that the political economy of trade shocks is different from the political economy of technological shocks. Hectoring people in an effort to change that reality won’t work.

The bottom line is that free trade isn’t an inviolable principle. There are good reasons to intervene to in order to protect some industries from the effects of foreign, and especially Chinese, competition.

Tariffs vs Subsidies

If the goal is to aid an industry, tariffs are almost always the wrong policy tool. Subsidies are almost always a better choice.

The classic work in this field was Max Corden’s 1978 Trade Policy and Economic Welfare, which is still arguably the best one-stop manual for trade policy. I would summarize Corden’s message as being that trade policy should be surgical. What I mean is that if you think you’ve identified a reason markets are getting it wrong, allowing an industry to shrink when it shouldn’t, you should design policy to address the market failure, with as few side effects as possible.

This typically means that subsidies — possibly targeted subsidies that, say, promote employment — are almost always a better policy than tariffs. For while tariff protection may help a domestic industry facing import competition, it also raises prices for consumers. Furthermore, if the targeted imports are “intermediate goods” — inputs into the production process, like steel – tariffs will raise the cost of production for the final good. Hence Trump’s tariffs on steel imports have raised the cost of domestic production of autos.

That last issue — the effects of tariffs on production costs — is far more important today than it was in the past, because of the rise of global supply chains in which production of many goods is a multi-stage process, with the different stages often taking place in different countries. Approximately half of U.S. imports are intermediate goods, some of which — like steel and many products made from copper — now face high tariffs. The way these tariffs have raised production costs is one of the reasons manufacturing employment, which Trump’s tariffs were supposed to boost, has steadily declined since last spring.

Subsidies, which don’t raise prices, avoid this kind of collateral damage. If you say that subsidies to preserve an industry would cost too much, you’re really saying that the industry isn’t worth saving. Why? Because a tariff actually costs morethan a subsidy that achieves the same results, once you take the adverse effects on consumers and downstream industries into account.

According to standard trade policy analysis, the one circumstance under which tariffs are the right policy is when a nation has market power, so that exporting countries absorb a substantial part of the cost of a tariff by reducing their prices. Of course, Trump constantly asserts that foreigners are paying his tariffs. But data on import prices — the prices the U.S. pays for imported goods, not including tariffs — don’t show them falling significantly after tariffs went way up in April:

OK, I’d add a couple of additional circumstances under which tariffs or tariff-like policies might be justified. These are my personal views, not what the economics literature says. First, if you’re faced with a sudden import surge, tariffs are quicker to implement than subsidies, both because they’re administratively simpler and because in some cases U.S. law allows the executive to impose them without waiting for legislation. Yes, Trump has massively abused that privilege, but it was there for a reason.

Second, sometimes tariffs or tariff-like policies are necessary for political reasons. Biden’s Inflation Reduction Act mainly consisted of subsidies for green energy, which was extremely justified given both the threat of climate change and the fact that green technologies may well qualify as an industry of the future that will belong to China if we allow the Chinese to lock in their advantage. But these subsidies came with buy-American provisions, which free traders decried as a form of protectionism similar to tariffs — which they were. Other things being the same, the IRA would have been better without those provisions — but without those provisions it would never have passed Congress.

These caveats aside, the response to China’s trade surplus should mainly involve subsidies and other support for industries the West wants to keep, not tariffs.

The next question, clearly, is which industries should be supported. But answering that question would make this already long, hard post even longer. So for now let me lay out some criteria, then return to the topic in a future post.

A necessary condition for supporting an industry is that it be an example of at least one of the three basic reasons I gave that China’s burgeoning surplus is a problem. That is, it must be an industry in which a rapid decline would be strongly disruptive to workers; or an industry where retaining a strong domestic presence is crucial for national security; or an industry that may be key to the economic future.

Which industries appear to meet that criterion? To be continued …