Earlier this month, the Economist published a widely commented-upon article that warned about what it saw as a worrying change in American trade policy. For decades, according to the authors of the article, U.S. trade policy worked to “to tear down the subsidies hurting American exporters and gumming up global trade.” But now, “rather than trying to get other countries to cut subsidies, the Biden administration’s unabashed focus is on building a subsidy architecture of its own, complete with the kinds of local-content rules that American officials once railed against.”

“The economic thinking that underpins much of this logic is dubious,” the article goes on to claim, but the changing U.S. strategy is driven by a political momentum that makes even such dubious logic hard to reverse. What is worse, the article argues, it is shifting a global regime that once valued open trade and economic efficiency toward one that will be less efficient and more unfair:

For America’s allies, from Europe to Asia, it is a startling shift. A country that they had counted on as the stalwart of an open-trading world is instead taking a big step towards protectionism. They, in turn, must decide whether to fight money with money, boosting their subsidies to counter America’s. If the result is a global subsidy race, the downsides could include a fractured international trading system, higher costs for consumers, more hurdles to innovation and new threats to political co-operation.

What kind of trade imbalances would we see in a well-functioning trading system?

While there is much in the article that is intelligent and useful, the Economist is wrong to imply that we once lived in a global regime that valued open trade and economic efficiency. What is more, its approach to trade-related policies—in the United States and elsewhere—seems to rest on a bureaucratic rather than a functional view of what constitutes a well-managed global trade regime.

This is the wrong approach. The global trade and capital system of the past four to five decades is among the most unbalanced, distorted, and protectionist in history. I say this not because of the number of deviations from WTO-proscribed behavior but simply because of global trade imbalances. Trade imbalances in the past fifty years have been among the largest ever recorded. This matters, because in an efficient, well-managed, and open trading system, large, persistent trade imbalances are rare and occur in only a very limited number of circumstances.

We would expect, for example, that rapidly growing developing countries with very high domestic investment needs might run moderate trade deficits during their rapid-growth phases. This is because domestic savings are often insufficient to fund the investment needs typical of rapidly growing developing economies, in which case it makes sense for them to run moderate trade deficits (moderate only, because large capital inflows have proven historically to be highly disruptive) as a way of importing savings from abroad.

For them to import savings from abroad, there must of course be countries that export savings. One such set of countries would consist of mature, capital-intensive economies. In these countries, returns on investment are likely to be lower than those in rapidly growing developing countries, and so a low to moderate flow of capital from the former to the latter would improve overall global productivity.

Another set of countries that might naturally export savings includes countries with very small populations and very large amounts of exportable commodities (e.g., Arab OPEC nations). These countries might run persistent surpluses mainly because their populations are too small to absorb their export earnings when commodity prices are especially high. However, because their savings surpluses are likely to be invested not in developing economies but rather in less risky advanced economies, from where they can be recycled, these reduce the net surpluses required by advanced economies to fund the investment needs of developing economies.

With these few exceptions, we would not otherwise expect to see large, persistent trade surpluses or deficits in a well-functioning global trading system.1 If global imbalances are limited to the above, the suppressed consumption of the savings exporters (I explain below why savings exporters suppress consumption) would be matched by an expansion in investment in the savings importers, with net benefits for the global economy.

Are current trade imbalances “efficient”?

But this not at all what global imbalances look like today, or indeed what they have looked like in the past five decades. Instead global trade and savings imbalances have been distributed in ways that make clear just how inefficient and distorted the global trading system has become.

One obvious such distortion is that while many advanced economies do indeed run persistent surpluses, as we would expect, these surpluses are extremely unevenly divided, and in many cases are far larger than can be economically justified. Germany, for example, has run trade surpluses since 2005 that typically range between 5 and 7 percent of its GDP. France, on the other hand, has run trade deficits over that period of up to 2 percent of its GDP.

A second and even more “surprising” distortion is that some of the largest trade surpluses in the world are generated by rapidly growing developing economies in Asia, precisely the countries that should be importing foreign savings through trade deficits. The largest among these of course is China—whose trade surplus is roughly 4 percent of GDP, but was as high as 9 percent of GDP in 2007—but they include several others, including relatively poor countries like Vietnam.

These two distortions are only possible because of a third obvious distortion: in the past few decades, typically only 20–30 percent of excess global savings has been directed toward fast-growing developing economies. Most of the excess savings has instead gone to a particular group of advanced economies, with the United States absorbing nearly half of the balance, and much of the rest directed toward the other Anglophone economies (primarily the United Kingdom, but also Canada and Australia).2

Why has this been the case? It turns out that the main qualification Anglophone economies have for running trade deficits is their extremely deep, flexible, and well-governed financial markets. Governments, businesses, and the wealthy elites in the surplus economies must accumulate assets abroad as the counterpart to their trade surpluses and, perhaps not surprisingly, are primarily interested in accumulating these surpluses in the form of safe and liquid assets in low-risk economies.3 The United States and the United Kingdom are by far the most attractive such economies, followed mainly by other Anglophone economies. Because they attract the largest net capital-account surpluses, they must do so by running the largest trade deficits.

This is clearly not the way the global trading system is supposed to work. Excess savings are created through the explicit or implicit suppression of consumption, and so they create downward pressure on global demand. These excess savings can only benefit the global economy if they flow into regions where productive investment is constrained by scarce savings and expensive capital.

If instead they flow into advanced economies in which productive investment is constrained by weak demand rather than by scarce capital, they may actually reduce investment because of the downward pressure they exert on demand. In that case, the suppression of consumption needed to generate excess savings in one country is not matched by an increase in investment, and so global demand is reduced.4

Why do countries today run persistent trade surpluses?

By definition a country’s trade surplus represents the excess of its domestic savings over its domestic investment.5 Countries run trade surpluses because they save a larger share of their production than they can productively invest at home, which is another way of saying that countries run trade surpluses because they consume too low a share of what they produce (savings is simply the obverse of consumption).

We often hear that countries that run large persistent trade surpluses do so because their workers are harder-working and thriftier than those of their trading partners, but this is nonsense. If they were indeed harder-working and thriftier, their reward for faster export growth would be an equally rapid increase in their standards of living, and this would necessarily involve faster growth in imports. In trade surplus countries, however, the improvement in living standards has systematically lagged behind the growth in the economy, so that in these countries the benefits of export growth are not higher domestic living standards but rather higher profits for businesses and more income channeled from workers to the rich.

And that is precisely why they run trade surpluses. Their trade surpluses just mean that workers receive a disproportionately low share of what they produce in the form of wages, unemployment compensation, and other direct and indirect transfers. This is ultimately why these countries can only consume or reinvest a portion of whatever its workers and businesses produce, and must export the balance.

There is nothing natural, in other words, about the behavior and culture of countries that run large, persistent surpluses, nor do they have the comparative advantage of low wages, as is often claimed.6 They run surpluses because a wide range of domestic policies and institutions effectively channel income away from workers and households and toward businesses, government, and the rich.

This can occur even in rich, highly productive countries, like Germany, the Netherlands and Japan, suggesting that it is not low wages that drive surpluses but rather low wages relative to productivity. In a well-managed trading system, differences in wages would broadly reflect differences in productivity, and so individual countries would not be able to treat global demand as a zero-sum game from which they can extract disproportionately large shares. The system should instead allow each country or region to specialize in producing goods for export that would allow it to acquire through imports more goods than it otherwise would have been able to produce domestically.

Beggar-thy-neighbor policies

There is unfortunately a lot of confusion about what kinds of policies and institutions distort trade. Some of these may be of the kind that the WTO and the Economist explicitly define as protectionist, but in fact whether or not they are so defined is irrelevant to the actual functioning of the global trading system. This is the key point that so many analysts miss. Policies that create trade surpluses by reducing the share of income retained by workers and households make the manufacturing sector more competitive only by allowing them to grab a greater share of global demand.

They are, in other words, classic beggar-thy-neighbor policies, and they can take many forms, whether or not these are formally proscribed by trade bodies such as the WTO. They include the following:

  • Tariffs and import quotas. Because households are net importers, and exporters are producers, by raising the cost of imports and increasing the profits of domestic producers, tariffs and quotas transfer income from households to businesses.
  • Currency depreciation. This works like a tariff on all imports and a subsidy for all exports. Like tariffs, currency depreciation effectively transfers income from households to manufacturers.
  • Financial repression. In a financially repressed system, the allocation of credit is typically administered by financial and economic authorities and interest rates are artificially suppressed, benefitting borrowers at the expense of lenders. This can result in implicit income transfers from households (as net lenders to the banking system) to businesses as net borrowers. This was a major cause of trade surpluses in Japan in the 1980s and China in the 2000s.
  • Low pensions and a weak social safety net. These reduce indirect transfers to households in favor of businesses, who are usually expected to pay directly or indirectly for the social safety net.
  • Environmental degradation. This lowers costs for businesses while raising healthcare costs for households.
  • Weak workers’ rights. Policies that penalize workers and labor unions to the benefit of businesses can prevent wages from rising in line with worker productivity. In that case the distribution of the benefits of economic growth is distorted so that businesses capture a larger share of productivity gains and workers a smaller share. These policies include restrictions on the mobility of workers, most famously China’s hukou system, which can reduce the bargaining position of workers relative to employers.
  • “Surplus” capture. Policies that capture agricultural surpluses or the benefits of commodity production redirect income away from households and toward businesses and investment.
  • Excessive spending on logistics and transportation infrastructure. This may seem at first counterintuitive, but countries that overspend on infrastructure—most famously Japan in the 1980s and China in the past decade—effectively reduce business costs while forcing households (usually indirectly) to subsidize these costs.

All of these policies and conditions work in the same way. By explicitly or implicitly transferring income from households to manufacturers and producers, they increase the international “competitiveness” of domestically based businesses, not by encouraging them to invest in productivity-enhancing technology but rather by allowing them to restrain wage growth so as to subsidize production. The result is that manufacturers in the tradable goods sector grow more rapidly because these various subsidies make them more competitive in the international markets.

By preventing household income from keeping pace with production, these policies force down the consumption share of GDP or, which is the same, they force up the domestic savings share. Notice this has nothing to do with a culture of hard work or thrift, or even with comparative advantage. If the growth in wages and household income were allowed to keep pace with domestic productivity growth, domestic consumption would rise more quickly than it has, the savings rate would automatically decline, and increases in exports would be matched by equivalent increases in imports.

While economists may decide that certain forms of such transfers, e.g. tariffs, should be considered illegitimate, and other forms, e.g. overspending on infrastructure or mobility restrictions on workers, perfectly acceptable, in fact from a global trade point of view this distinction is meaningless. In either case policymakers expand domestic production by taking a larger share of foreign demand, and they do this by implicitly or explicitly forcing households to subsidize the international competitiveness of domestic manufacturers.

If, in preventing domestic businesses and households from converting exports into imports, these economies recycle their surpluses into developing countries that convert their imported savings into higher investment, they shift the locus of import-creation to developing countries, so that there is no net contraction in global demand. If they don’t, and recycle them instead to advanced economies in which investment isn’t constrained by scarce savings, the global economy is worse off even though the trade-surplus countries are better off, which means that their trade partners—the deficit countries—are likely to be much worse off.

In a well-functioning trading system this would not happen. If Mexico had a comparative advantage over the United States in the production of widgets, for example, it would be perfectly natural that the production of widgets should migrate from the United States to Mexico. But if Mexican workers were paid in line with their productivity (e.g., if they were one-fifth as productive as American workers and were paid one-fifth the wages), as widget production migrated to Mexico, Mexican workers would be able to raise their consumption, which would in turn directly or indirectly raise Mexican imports by enough to match the increase in exports to the United States.

As that happened, American workers who formerly worked in widget factories would lose their jobs, but they would find new jobs producing the goods that Mexico now imported. This is a point that too many trade economists do not seem to understand. There would be nothing wrong with moving production from the United States to Mexico as long as the consequent increase in Mexican exports to the United States were matched by an increase in overall Mexican imports.7 When that happens, the world benefits from an open trading environment. Both the United States and Mexico shift production in ways that exploit their respective comparative advantages, and so total production in both countries, along with workers’ productivity and wages, will rise.

The problem occurs only when an increase in Mexico’s exports to the United States is not matched by an equivalent increase in Mexico’s imports. In that case, Mexico will run a beggar-thy-neighbor surplus, and this surplus will simply reflect the fact that Mexican workers were paid too little to convert their higher production into higher consumption. I should note that this is not actually the case with Mexico because Mexico typically runs trade deficits—as it should, in order to import foreign savings to fund domestic investment needs—so that increases in its exports are likely to be matched with equivalent increases in its imports.

What is protection?

Large, persistent trade surpluses exist only because businesses in certain countries are able directly and indirectly to underpay domestic workers and households in order to become more competitive internationally and to grow more rapidly. The result, as John Maynard Keynes pointed out in the Bretton Woods negotiations in 1944, is a net reduction in global demand. The rest of the world must manage this reduction in global demand either with an increase in unemployment or, what has been far more likely, with an increase in household and fiscal debt to make up for the demand absorbed by the surplus countries.8

If the Economist is determined to fight protectionism and create a better global trading system, it should uncover and oppose any policy distortion that leads to large, persistent imbalances and, above all, it should recognize that protectionism is not caused by deficit countries. It is caused by countries whose growth policies are designed to generate large, permanent trade surpluses. In that light, it is ironic that in its very next issue the magazine quoted South Korea’s trade minister, Ahn Duk-geun, who shares its view, as saying that because of the “protectionist” industrial policies implemented by the United States, “Japan, Korea, China, every country will engage in this very difficult race to ignore global trading rules.”

The irony is that South Korea is one of the countries that aggressively target trade surpluses as a fundamental part of their economic policymaking, to the extent that the recent decline in its trade surplus has set off a rather frantic search for remedies to reverse this decline. It is at best bizarre to assert that surplus countries that implement policies designed actively to maintain and increase their surpluses are not engaged in protectionist behavior, while deficit countries that implement policies designed to reduce their deficits are.

It is especially absurd to criticize the United States, the country that typically absorbs 40–50 percent of global trade surpluses, as “protectionist.” Instead, the United States should be criticized for allowing—strongly encouraged by its major financial institutions—the unlimited and uncontrolled entry of foreign savings into the U.S. financial system, which in turn allows global trade distortions to continue by forcing the United States to accommodate the trade surpluses of protectionist countries.9

But if countries like the United States were to implement policies deemed “protectionist” by the Economist, and these were able even partly to reverse the trade imbalances (a big “if”), they would actually improve the efficiency of global trade by reducing the imbalances. Protectionist policies work by altering the balance between a country’s savings and its investment, and to the extent that trade policies by deficit countries like the United States reduce their deficits, they do so by partially reversing the imbalances, which in turn forces the surplus countries either to improve the distribution of income to their domestic households or absorb internally the demand-contracting consequences of their beggar-thy-neighbor policies.

In an article this week in the Financial Times, Martin Wolf argued that while capitalism may be becoming somewhat less global, “an internationally open capitalism remains the foundation of future prosperity. It needs to be reformed. It must not be abandoned.”

I agree, and would add that the primary direction of reform should be to reduce the deep, persistent trade imbalances that undermine global trade. Among other things, this might involve reviving Keynes’s proposal at Bretton Woods to penalize economies that choose to run large, persistent trade surpluses. If U.S. policies force a reduction in global trade imbalances by reducing the U.S. deficit, even if they are deemed “protectionist” policies by trade bureaucrats, they are clearly not protectionist. In fact, by partially reversing distortions in the distribution of income in trade surplus countries, these policies will actually be trade-enhancing.


1 It is important to distinguish here between overall imbalances and bilateral imbalances. Because of the complexity of global supply chains and the differing specialization of different economies, we would not expect trade to balance at the bilateral level but only at the national level, with a country’s various bilateral trade surpluses and deficits broadly balancing out.

2 Benn Steil and Benjamin Della Rocca, Greenberg Center for Geoeconomic Studies, “Global Imbalances Tracker,” Council on Foreign Relations, accessed January 17, 2023, https://www.cfr.org/report/global-imbalances-tracker.

3 This is further explained in Michael Pettis, “Will the Chinese Renminbi Replace the US Dollar?”, Review of Keynesian Economics 10, no. 4 (October 2022), https://www.elgaronline.com/view/journals/roke/10/4/article-p499.xml.

4 See Matthew C. Klein and Michael Pettis, Trade Wars are Class Wars (New Haven: Yale University Press, 2020), https://yalebooks.yale.edu/book/9780300261448/trade-wars-are-class-wars/. Some economists, usually without realizing it, still operate in the pre-1970s world in which investment in the advanced economies is constrained by scarce savings, and so they argue that the United States runs a surplus on the capital account because of its low savings. In fact, they have the causality backwards. Foreigners invest the savings generated by their trade surpluses in the United States and, to a lesser extent, in the other Anglophone economies because of their deep, liquid, flexible, and well-governed financial and asset markets. By definition these net inflows must create a gap between investment and savings. As I explain in two earlier Carnegie papers, because investment in advanced economies like the United States hasn’t been constrained by scarce savings in decades, the result of these inflows must cause domestic U.S. savings to fall, either through higher unemployment or through an increase in debt.

5 Technically this is the definition of the current account surplus, but for the purpose of this essay the distinction is irrelevant.

6 Neither David Ricardo nor any of the major trade theorists ever argued that low wages are a comparative advantage, or that comparative advantage conveys trade surpluses. They argued instead that specializing in comparative advantage allowed countries to import more goods and commodities than they could domestically produce.

7 It doesn’t matter whether the resulting increase in Mexico’s imports is balanced by an increase in imports from the United States or from a third country. Because trade balances globally, if an increase in Mexico’s exports to the United States is matched, say, by an increase in Mexico’s imports from Brazil, U.S. producers would end up increasing their exports not to Mexico but rather to Brazil (unless Brazil engaged in beggar-thy-neighbor policies). What matters is the overall trade balance, in other words, not the bilateral trade balance.

8 See note 4.

9 The only way the United States can refuse to accommodate the trade surpluses of the rest of the world is to intervene in these capital inflows.