The global trading system has been broken for decades. A well-functioning trading regime would permit neither the large, persistent trade imbalances that characterize the current global trading system nor the perverse flow of capital from developing economies to advanced economies. The system needs new rules that encourage a return to the benefits of free trade and comparative advantage.
Until this happens, trade imbalances will persist. This matters especially to the United States because of the role it plays in anchoring global imbalances. Countries that run large, persistent trade surpluses must acquire foreign assets to balance these surpluses. American assets are particularly attractive for this purpose, and the United States allows nearly unfettered access to these assets. As a result, surplus countries prefer to acquire assets in the United States in exchange for their surpluses, which also means that the United States must run the corresponding trade deficits.
This has important implications for U.S. manufacturing, unemployment, and debt. It means that the U.S. share of global manufacturing must decline while that of surplus countries must rise. Because surplus countries are those that subsidize their manufacturing at the expense of domestic consumption, American manufactures are forced indirectly to subsidize U.S. consumption. This is why, during the past five decades, manufacturing has consistently migrated from deficit countries (mainly the United States) to surplus countries (mainly China). Until global rebalances are resolved, this will continue.
It also means that for all the talk of reshoring and friendshoring, the U.S. trade deficits cannot decline as long as surplus economies can continue to acquire assets in the United States with the proceeds of their surpluses. The United States, in other words, has no choice but to run deficits to balance the surpluses of the rest of the world.
What’s more, while many mainstream economists assume that foreign inflows lower U.S. interest rates and finance U.S. investment, as occurred in the nineteenth century, this hasn’t been the case for decades. Foreign inflows instead force adjustments in the U.S. economy that result in lower U.S. savings, mainly through some combination of higher unemployment, higher household debt, investment bubbles, and a higher fiscal deficit.
To rebalance its economy toward manufacturing while reining in debt and generating higher-paying employment, the United States must either transform the global trading regime or unilaterally opt out of its current role. Not only would this benefit the U.S. economy, but it would also benefit the global economy by eliminating the persistent downward pressure on global demand created by the surplus countries.
This won’t be easy, however. Any meaningful resolution of global trade imbalances will be strongly opposed by surplus countries and would result in a diminished global role for the U.S. dollar.
How does international trade affect the U.S. manufacturing sector?
Last month, Yao Yang, former dean at the National School of Development at Peking University, said on his blog that “America’s industrial base has already been hollowed out. How can it possibly compete [with China]? The United States has obviously made a strategic mistake.”
He’s right, but perhaps not for the reasons he thinks. While manufacturing comprises roughly 16 percent of global GDP, according to the World Bank, the manufacturing share of China’s GDP is 28 percent, among the highest in the world, whereas for the United States it is 11 percent, among the lowest for any major economy.1 The opposite is true for consumption. While consumption accounts for 75 percent of global GDP, it accounts for 80 percent of the United States’ GDP and only 53 percent of China’s GDP.
To put it another way, while China comprises less than 18 percent of global GDP, it accounts for over 31 percent of global manufacturing and less than 13 percent of global consumption. The United States, which accounts for 24 percent of global GDP, accounts for less than 17 percent of global manufacturing and nearly 27 percent of global consumption.
While the differences in the two countries’ manufacturing and consumption shares of GDP may seem unrelated, it turns out that they are different expressions of the same imbalance. China and the United States are extreme representatives of a common pattern in the global economy. Manufacturing typically represents a disproportionately large share and consumption a low share of the GDP of non-commodity economies with large, persistent surpluses. The reverse is true for advanced economies that run large, persistent deficits.
This clearly isn’t a coincidence, but in which direction does the causality run? Do countries have larger manufacturing sectors because they are surplus countries, or do they run surpluses because they have larger manufacturing sectors? For many years, economists have argued that it is the latter. Surplus economies, they claim, have a comparative advantage in manufacturing that leads them to produce tradable goods more efficiently, and this is why they export more than they import. Deficit countries like the United States, on the other hand, have a comparative disadvantage in manufacturing.2
But this misunderstands altogether the meaning of comparative advantage. As I explain below, surplus economies run surpluses mainly because of industrial policies that implicitly or explicitly force households to subsidize the manufacturing sector. Their competitive advantage in manufacturing comes not from comparative advantage but rather from transfers that distort comparative advantage and reduce domestic demand.
What is the relationship between competitive manufacturing and weak domestic demand?
In these persistent surplus economies, weak domestic demand is simply the flip side of policies that result in manufacturing competitiveness. The manufacturing sector is subsidized directly or indirectly by households, which leaves them more competitive and leaves households less able to purchase a substantial share of what they produce.
But in order to balance these surpluses, the opposite transfers must occur in the deficit countries. Just as consumers are forced to subsidize producers in the surplus countries through various explicit and implicit transfers, producers are effectively forced to subsidize consumers in the deficit countries.
There are many forms these transfers can take, but the easiest one to understand is through currency values. An undervalued currency, typical of surplus countries, affects trade imbalances by raising the cost of imports and increasing the profits of exporters. It results, in other words, in an implicit transfer from importers to exporters. Because households are all net importers, and because net exporters are mostly manufacturers, these implicit transfers subsidize the manufacturing sector at the expense of households. This makes the manufacturing sector in that country more competitive while reducing the capacity of households to consume.
The opposite happens in the deficit countries. An undervalued currency for one country is the obverse of an overvalued currency for its trade partner, and this overvaluation also represents an implicit transfer, in this case from net exporters (manufacturers) to net importers (households as consumers). Just as manufacturers are subsidized by consumers in the former, so are consumers subsidized by manufacturers in the latter, making their manufacturing sectors less competitive globally.
It is not surprising, then, that global manufacturing naturally migrates from deficit countries to surplus countries, while global consumption migrates in the opposite direction. This has nothing to do with comparative advantage. Manufacturers in both economies are simply responding to the direction of subsidies.
Although I use undervalued and overvalued currencies as an easy illustration of how these transfers between producers and consumers affect trade, they are not the only, nor even the most important, of such transfers. Repressed interest rates, for example, have often been far more important, along with overinvestment in infrastructure, wage repression, and several other implicit or explicit transfers that subsidize manufacturers at the expense of households. (See appendix 1 for a list of such transfers and how they subsidize manufacturing at the expense of households.)
Why don’t these transfers represent a kind of comparative advantage in manufacturing?
Comparative advantage doesn’t mean that some countries export and others import. It means that each country exports those goods and services in which it has a comparative production advantage to pay for imports of those goods and services in which it has a comparative disadvantage. Under a system of comparative advantage, in other words, the purpose of trade is to maximize welfare by maximizing the value of imports, which in turn means maximizing the value of exports used to purchase those imports. In a trading regime governed by comparative advantage, trade must be broadly balanced as countries exchange goods and services produced in their respective areas of comparative advantage for goods and services in which they are at a comparative disadvantage.
This is very different from a world of “competitive advantage,” in which a country’s subsidized manufacturing leaves domestic consumers unable to absorb domestic production. In that world, the purpose of exports is not to maximize the value of imports but rather to externalize the consequences of suppressed domestic demand.
Should the U.S. economy and other advanced economies shift from manufacturing to nonmanufacturing sectors?
Some economists have argued that advanced economies like that of the United States should de-emphasize manufacturing in favor of other sectors, including services and technology. They argue that the decline in the U.S. share of global manufacturing (relative to its share of global GDP) is not a problem but simply part of a natural evolution.
This argument might make more sense if that decline occurred in all advanced economies, but it seems to occur mainly in advanced economies that run persistent deficits, while the opposite occurs in non-commodity advanced economies that run persistent surpluses. As mentioned above, manufacturing represents 16 percent of global GDP, but only 11 percent of U.S. GDP, 9 percent of Canadian GDP, 8 percent of UK GDP, and 5 percent of Australian GDP. These are all advanced economies that have historically run trade deficits.
On the other hand, it represents 18 percent of German and Swiss GDP, 20 percent of Japanese GDP, 21 percent of Singapore’s GDP, 26 percent of South Korean GDP, and 34 percent of Taiwan’s GDP. Not coincidentally, these are all advanced economies that have historically run persistent trade surpluses.
But whether or not the manufacturing sector matters to an advanced economy, the important point is that under the current system, the United States doesn’t choose the sectors in which it specializes. As long as it continues to allow unfettered foreign access to U.S. assets, as I explain below, it must lose its global share in any sector that trade surplus countries choose to subsidize.
One way to think about this is that the global trade regime subjects the U.S. economy to a form of “industrial policy.” This industrial policy, however, is chosen not by American policymakers; instead, it is the obverse of the industrial and trade policies of the United States’ most aggressive trade partners. The United States, in other words, must follow whatever industrial policy accommodates the needs of its trade partners. That’s the reason—and not some natural evolution of the economy—its share of global manufacturing has declined so sharply.
Doesn’t an expansion of global trade always benefit the countries that participate in international trade and, more generally, the global economy?
No. There are conditions under which an expansion of global trade leads to an expansion of global production, mainly because under those conditions trade allows resources to be shifted to their most productive uses and businesses to exploit economies of scale. If the increase in production is distributed efficiently between consumption and savings, more exports automatically lead to more imports, and all parties are better off.
But there are also conditions in which an expansion of global trade suppresses global consumption through what in the 1930s used to be called “beggar thy neighbor” policies. When this happens, either global production must also contract and global unemployment rise or total demand must be maintained by rising debt.
The point is that under some conditions, an expansion of global trade can expand both production and consumption in the global economy, even if the benefits and costs are unevenly distributed. Under other conditions, an expansion of global trade can result in a contraction in global demand, which must be balanced in the deficit countries either by higher unemployment or higher debt.
How do persistent trade surpluses matter to the global economy?
Surpluses and deficits are neither good nor bad in themselves. What matters is the way in which they are created domestically and absorbed abroad. Because trade imbalances represent the excess of savings over investment in the surplus country and the excess of investment over savings in the deficit country, there are only three ways in which a trade surplus in one country can be balanced in the deficit country. It can be balanced with 1) an increase in investment; 2) an increase in unemployment, which reduces savings; or 3) an increase in debt (usually household debt or a fiscal deficit), which also reduces savings. I explain later why this must be the case.
If the trade surplus is balanced by an increase in investment in the deficit country, total sustainable demand in the global economy is unchanged in the near term and boosted in the long term by the increase in investment.
If it is balanced by an increase in unemployment, the global economy is worse off because of lower demand in the surplus country, although it is the deficit country that pays the cost of this lower demand through higher unemployment.
If it is balanced by an increase in the debt of the deficit country, global demand is maintained, but only at the cost of rising debt and so is not sustainable in the long term. In this case, it is again the deficit country that must pay for the cost—in the form of higher debt—for lower demand in the surplus country.
Trade surpluses, in other words, are positive for the global economy when they lead directly to an increase in productive foreign investment. They are negative for the global economy when they lead either to a rise in foreign unemployment or to higher foreign debt.
What does a well-functioning global trading regime look like?
In a well-functioning global trading regime, trade is broadly balanced and the purpose of exports is to maximize imports. What’s more, to the extent that trade imbalances exist, they do so mainly in the form of small, stable trade deficits in rapidly growing developing countries whose investment needs cannot be fully satisfied by domestic savings and so who must import savings from the capital-rich advanced economies. Other forms of trade imbalance are possible, but they tend to be short-lived: in an economy without major government-enforced distortions, both surpluses and deficits force fiscal and monetary changes that are automatically self-correcting.
The current global trading regime is very different. Not only has it been characterized in the past several decades by huge, persistent trade imbalances that don’t seem to self-correct, but, even more perversely, excess savings are generated in both advanced and developing economies and are mostly directed to a handful of advanced, capital-rich economies—with the United States, the United Kingdom, Canada, and Australia typically absorbing 60–80 percent of the total.3
What is the relationship between high savings and high trade surpluses?
A world of deep, persistent trade imbalances is usually a world of repressed demand.4 There is a widespread misperception that countries with high savings rates are countries whose households—perhaps for cultural reasons—value hard work and thrift. In fact, high national savings rates have to do not with cultural factors but mainly with the way income (GDP) is distributed among various sectors of the economy.
The distribution of income matters because different sectors consume different shares of their income. Ordinary households tend to consume a very large share of their income, while businesses consume none of their income and governments and the rich consume a very low share. That means that the total share of GDP that is consumed depends in large part on how GDP is distributed among these four groups.
In countries in which workers and ordinary households retain a relatively low share of what they produce, with businesses, the government and/or the rich retaining a high share, the consumption share of GDP must necessarily be low. In that case, savings, which is the part of GDP that isn’t consumed, must be high.
In other countries with the reverse distribution of income, in which workers and ordinary households retain a relatively high share of GDP, the consumption share of GDP will be high and the savings share low.
The share of GDP retained by households in the United States, for example, is roughly one-third higher than the share retained by households in China. (American workers get paid a higher share of what they produce than Chinese workers.) In that case, even if American and Chinese households had exactly the same attitudes and cultural dispositions toward consumption and thrift, the U.S. economy overall would still have a higher consumption share and a lower savings share of GDP than China.5
It is not a coincidence that most high-savings economies that run large, persistent trade surpluses are economies in which workers and ordinary household receive a disproportionately low share of what they produce. This low share is a consequence of direct and indirect transfers from households that subsidize the production side of the economy.
Are high savings rates good or bad for an economy?
It depends. In a developing economy with higher investment needs than can be met by domestic savings, industrial policies that force up savings can result in higher investment than otherwise—if properly channeled into the domestic financial system. Higher investment can in turn lead to faster growth, with households ultimately benefiting in the form of faster-rising wages.
This was the case in China in the 1990s and the beginning of the 2000s, when its extraordinary high savings rates fueled equally high investment in infrastructure, urban property, and manufacturing capacity urgently needed at the time. The high investment led to some of the highest GDP growth rates in history.
But in more mature economies, or in economies in which many years of high investment have satisfied most infrastructure, property and production needs, the high savings rate has a different impact. Production generates jobs and income, but the purpose of production is directly or indirectly to produce goods and services that can be consumed. If workers and their families earn too little (relative to their levels of productivity) to be able to afford the goods and services they produce—in other words, if the savings rate is too high—there is no point in producing them.
What happens to an economy when households cannot afford what they produce?
When households earn too little to be able to buy the goods and services they produce, or when domestic saving exceeds the amount of productive investment needed for the economy (which is the same thing), it just means that the total amount of goods and services produced exceeds the amount that can be absorbed in the form of consumption or investment. In that case, an economy has six options.
The best option is usually to raise household income directly or indirectly so that consumption and production can get back into alignment. When that happens, the economy saves just what is needed for productive investment and can consume the rest. There is no “excess” production or savings.
Redistributing income to households, whether directly or in the form of social transfers, is often politically very difficult, however, and requires that the subsidies that generated rapid growth in production be reversed. This would undermine the competitiveness of industries reliant on these subsidies. The irony is that the more growth depends on these subsidies, the more important these subsidized industries are to the economy and so the more painful it is to reverse them. Japan has been trying unsuccessfully to do this since 1986 and China since 2007.6
A second option is to boost domestic investment in property, infrastructure, and/or manufacturing capacity even if it is not needed. This investment allows the economy to continue to absorb high savings and is what the Soviet Union and Brazil did in the late 1960s and 1970s, what Japan did in the 1980s, and what China has been doing in the past ten to fifteen years. The problem is that because this activity is nonproductive, it creates rising amounts of nonrecognized losses and a surging debt burden—and is ultimately unsustainable.7
A third option is to reduce savings by cutting back on excess production and firing workers. In this case, while consumption will decline, production will decline even faster until the two are back in balance. This can be an extremely painful form of adjustment and is how the United States adjusted in the early 1930s and Brazil in the mid-1980s. It involves adjusting in the form of negative growth and higher unemployment.
Unsurprisingly, most governments try to avoid the second and third options if they can, which usually means resorting to increases in debt. In that case, the fourth option involves reducing interest rates and expanding the money supply in a bid to get households to raise their consumption by borrowing. This allows total consumption to rise back to levels needed to support existing production, but instead of being funded by household income, it will be funded by household debt. Because this creates rising strains on the household sector, it is ultimately unsustainable.
The fifth option is similar to the third option except that it is governments that borrow rather than households. To make up for weaker consumption, the government will expand domestic demand by running fiscal deficits.
Finally, the sixth option is to externalize the imbalance. Rather than rebalance domestically by increasing wages, allowing unemployment to rise, or increasing household or fiscal debt, the economy can run trade surpluses that effectively allow it to export excess savings—or export its weak domestic demand, which is the same thing—to its trade partners.
All of these are simply ways for a country to resolve excess savings. The first five ways involve raising domestic investment or reducing domestic savings, either by redistributing income, raising unemployment, raising household debt, or raising the fiscal deficit. The sixth way involves transferring the excess savings on to trading partners by exporting it.
How does a surplus country externalize the cost of excess savings?
It does so by running a trade surplus.8 When excess savings are exported into another economy, the economy on the receiving end must adjust to those inflows by running trade deficits. Just as savings exceeded investment in the (surplus) country that is exporting its excess savings, so must investment exceed savings in the (deficit) country that is receiving foreign savings.
This is just another way of saying that when savings flow into a country, either its domestic investment must rise or its domestic savings must decline. One way or another, the recipient economy must adjust to the net inflow of foreign savings, and it must adjust in the same way that the country in which the excess savings were created would have to adjust if it didn’t have the option of externalizing the excess savings. In a closed system, savings cannot exceed investment.
Don’t foreign capital inflows lead to an increase in investment?
Not necessarily. Foreign capital won’t cause investment to rise if it flows into advanced economies in which business investment isn’t constrained by scarce capital. In fact, it may actually cause business investment to decline if business investment is constrained mainly by demand, as is the case in most advanced economies. In that case, to the extent that the accompanying trade deficit reduces demand for local producers, local producers may actually reduce investment.
This seems clearly to be the case for foreign inflows into the United States. It has been a much-discussed (and criticized) characteristic of the U.S. economy that American businesses sit on huge piles of cash and cash equivalents ($6.9 trillion, or 12 percent of total assets, according to one recent study), even after engaging in record levels of acquisitions and share buybacks. In that case, making even more (foreign) savings available for American businesses is unlikely to increase investment, especially if the increase in foreign inflows is the obverse of an increase in the U.S. trade deficit.
But the gap between investment and savings must still reflect net foreign inflows. While it may at first seem very counterintuitive that inflows can cause savings to decline, especially in a large economy like that of the United States, in fact as long as the United States doesn’t control its capital account, by definition it cannot control the gap between U.S. savings and U.S. investment, in which case if foreign capital inflows don’t cause domestic investment to rise, they must cause domestic savings to decline.
Given its high level of income inequality, shouldn’t the United States have a high savings rate and, with that, a trade surplus?
If the United States were a normal country, its high levels of income inequality would result in a high savings rate (because the rich save more of their income than the poor), and this would exceed domestic investment constrained by the corresponding weak domestic demand. Under these conditions, the United States would most likely run a trade surplus. This was the case in the late nineteenth century and early twentieth century.
The United States, however, isn’t a normal country when it comes to its role in the global balance of payments. It is the world’s top destination for excess savings globally because, in exchange for their trade surpluses, foreigners would rather acquire American assets, such as factories, farmland, apartments, stocks, and bonds, than those of any other economy. And because the United States has given up control of its capital account, it has no choice but to be a persistent net importer of foreign capital.
It is important to understand that this is not because the United States needs foreign capital but rather because foreigners need a safe place to absorb their excess savings. As long as foreigners prefer to acquire U.S. assets in exchange for their surpluses and are able to force the United States into a net capital account surplus, the United States must run a current account deficit. As I explain below, it does so mainly because rather than fund additional investment, foreign capital inflows force down the U.S. savings rate.
How can foreign inflows force down the U.S. savings rate?
The most obvious way is through a rise in U.S. unemployment. The mechanics are straightforward. As the U.S. economy sees its imports rise relative to its exports, part of the demand formerly directed to domestic factories is now diverted to foreign factories, with no compensating increase in foreign demand for U.S.-produced goods and services, and so factories respond by laying off workers.
Because unemployed workers still have to consume, production will decline faster than consumption, which automatically reduces the savings rate. Employed workers tend to have positive savings rates, while unemployed workers have negative savings rates. In that case, the United States adjusts through a rise in unemployment.
But to the extent that Washington intervenes to prevent unemployment from rising, it usually does so in one of two ways. First, Washington can expand the fiscal deficit and use increased spending to boost demand in the economy. This expansion of debt automatically balances the increase in foreign savings flowing into the economy by lowering domestic savings. (Debt is negative savings.)
Second, the Federal Reserve can lower interest rates and expand the money supply to encourage banks to expand household debt, which households in turn use to expand consumption. This allows consumption to rise back to levels needed to support existing production, but instead of being funded by household income, it will be funded by household debt.9 This process is often exacerbated by real estate or stock market bubbles set off by the combination of lower U.S. interest rates and foreign purchases of U.S. assets (see appendix 2). In that case, temporarily rising asset prices can further encourage higher household debt and reduced household savings.10
The point is that net foreign inflows must either cause investment to rise or savings to decline, and the ways in which they cause savings to decline usually involve some combination of higher unemployment, higher household debt, or higher fiscal deficits.11 Foreign capital inflows that drive down a country’s savings rate, in other words, are always bad for the economy.
Most studies on the impact of trade deficits on the U.S. economy are unable to show this, however, because they assume the only possible negative impact is a rise in unemployment. When they don’t find this rise, they conclude that the deficit has no negative impact. But if the impact on unemployment is mitigated by a surge in household debt or the fiscal deficit, as is usually the case in the United States, the adverse impact of the deficit on the economy won’t be seen in the unemployment numbers. It will be seen in the rise in debt and an erosion in the manufacturing share of U.S. GDP.
What are the options for deficit countries? Can they protect themselves from absorbing foreign savings and running trade deficits?
There are basically three ways deficit countries can respond to beggar-thy-neighbor policies in surplus economies. They can aggressively subsidize their own manufacturing sectors and try to pass the cost on to trading partners (in other words, retaliate with their own beggar-thy-neighbor policies); they can opt out of the existing global trade regime, either unliterally or with other like-minded countries; or they can accept rising debt or unemployment and a further erosion of the role of manufacturing in their economies. The first option would lead to a global overproduction (or underconsumption) crisis like that of the 1930s, the second option would disrupt the global trade regime, and the third option would likely be politically unacceptable.
Probably the best response for the United States and the world in the medium and long term would be for the United States to intervene to reverse the beggar-thy-neighbor polices of the surplus nations—the second option. In that case, the United States either unilaterally or with other like-minded countries would implement trade policies to prevent surplus countries from externalizing the costs of their industrial policies (in other words, from running persistent trade surpluses). This could be done either with restrictions on the ability of surplus countries to dump goods into the U.S. economy or with restrictions on the ability of surplus countries to dump excess savings into the U.S. financial system. The most obvious of the former policies are tariffs, and the most obvious of the latter are taxes on capital inflows.
How do tariffs affect trade imbalances?
While many assume that the main trade effect of tariffs is to reduce the import of tariffed goods, in fact it is a more complex process, especially for the United States because of its special role in the global economy. As long as the United States acts as the great absorber of savings imbalances in the global economy, it has no choice but to run trade deficits. This is why the role of “absorber of last resort of global excess savings” is simply the obverse of the role of “consumer of last resort”—both mean the same thing.
But this has an important implication for trade policy. As long as the United States allows foreigners unfettered access to U.S. assets, U.S. tariffs can only reduce the U.S. current account deficit to the extent that they force a change in the excess of savings over investment in the rest of the world. This requires raising tariffs high enough that they force a contraction in foreign production in the short term, followed eventually by an expansion on foreign demand.
Tariffs don’t always do this. For example, when the United States imposed tariffs on Chinese imports in 2018, this did nothing to reduce the overall U.S. trade deficit because Chinese savings still exceeded Chinese investment by as much as ever, and China still exported the bulk of its excess savings to the United States.
In that case, while the U.S. bilateral deficit with China might have declined, neither the overall U.S. deficit nor the overall Chinese surplus declined. As long as China’s internal balance was unchanged, and as long as it resolved this internal imbalance by exporting capital to the United States, bilateral tariffs could only shift the bilateral trade imbalances of either country. Without changing China’s savings-investment imbalance, U.S. tariffs on Chinese imports could not change the overall U.S. trade deficit (or the overall Chinese trade surplus).
What are the costs associated with import tariffs on rebalancing the U.S. and global economies?
Tariffs work for a country such as the United States only to the extent that they can force the rest of the world to reduce supply relative to demand. Unfortunately, a uniform tariff on all imports doesn’t discriminate between countries that run surpluses and countries that don’t. Because a tariff only works to the extent that it forces a reduction of excess savings for the rest of the world in the aggregate, it basically ensures that economies that are more aggressively mercantilist will bear a smaller share of the costs than countries that are less mercantilist.
This problem can be resolved by making countries exempt from tariffs if they are part of a trade agreement among a group of countries that implement similar “anti-mercantilist” tariffs. This suggests that if the United States is to impose large, across-the-board import tariffs, it should be part of a larger reorganization of trade in which new trade agreements are created among countries that commit to maintaining broadly balanced trade. For countries whose trade accounts are broadly in balance, there is no reason to implement tariffs.
Is there a meaningful alternative to import tariffs?
Analysts often forget that in the current global environment, as long as foreign investors prefer to balance trade surpluses with U.S. assets and as long as the United States allows them unfettered access to U.S. assets, the United States has no choice but to run a capital account surplus (in other words, to receive net foreign capital inflows). That means it also has no choice but to run a current account deficit.
To put it in a slightly more technical way, if the United States doesn’t control its capital account, it cannot control the gap between U.S. investment and U.S. savings, which in turn means it can control neither its trade account nor its overall savings rates. If the rest of the world wants to implement industrial policies that suppress domestic demand and force up their savings relative to their investment, and if they are freely able to export those excess savings to the United States by buying U.S. assets, the U.S. trade account and the U.S. savings rates must adjust to those inflows.
This suggests that a more direct and focused alternative to tariffs is capital controls. If the United States were to tax capital inflows, or otherwise restrict capital inflows from surplus economies, then there would be no need for the U.S. trade account and its domestic savings rates to adjust to net foreign inflows. There would also be no need for import tariffs.
How would capital restrictions affect the global use of the U.S. dollar?
The U.S. dollar is the most widely used global currency mainly because of the depth, liquidity, and flexibility of U.S. financial markets along with the country’s relatively strong protection of foreign investment. This is another way of saying that it is the U.S. role as absorber of last resort of global excess savings (“consumer of last resort”) that accounts for the overwhelming domination of the U.S. dollar in global trade and capital flows.12
Restricting net capital inflows would reduce the global use of the dollar and move the world to one in which no currency plays the role that the U.S. dollar currently plays. But while this would benefit American farmers, workers, the middle class, and domestic producers, it would hurt three very powerful U.S. constituencies.
The first is Wall Street, whose global dominance is underpinned by the global dominance of the U.S. dollar in trade and capital flows. The second is the foreign affairs establishment, who can use the dominance of the U.S. dollar to impose sanctions on countries that oppose U.S. geopolitical interests. And the third consists of large corporations that benefit from the easy transfer of investment out of the United States. This means that Washington must balance the long-term interests of the U.S. economy overall against the shorter-term interests of three very powerful constituencies.
Doesn’t the United States benefit from buying subsidized foreign goods?
Mainstream economists often argue that if a foreign country wants to subsidize its exports to the United States, American consumers should be happy to buy them because there is no cost, and a clear benefit (the lower price), to the subsidized good. The standard argument is that if someone offers you something at an artificially low price, you would be foolish not to take it.
This consumerist fallacy, common enough in American economics, assumes that if something applies at the individual consumer level, it must also apply systemically at the level of the economy. It shows that economists are often befuddled on trade and balance-of-payments issues.
The cost of an imported product to consumers isn’t the only thing that matters to an economy. The cost to producers also matters. What is cheap to households as consumers is only cheap if the cost of the implied subsidies is absorbed by the exporting country. If it is absorbed by the importing country, it is not “cheap” to households because households are not just consumers—they are also producers.
This argument would make sense only if the export revenues generated by the exporting country were converted into an equal amount of imports. In that case, the United States would be able to buy more imports for the same level of exports. It would indeed be better off.
But this is not the case when the revenues generated by the country subsidizing exports are not converted into an equal amount of imports but are instead converted into claims on U.S. assets. In that case, what temporarily benefits Americans as consumers is more than fully paid for by Americans as producers in the form of the export of claims on assets.
The recent electric vehicle (EV) controversy is an excellent example of this confusion. One side argues that China’s heavy subsidization of EV production is good for the world because it will result in a faster transition from gas-guzzling vehicles to electric vehicles, and it will make it cheaper for Americans to buy EVs. The other side argues that Chinese subsidies for the EV industry makes it impossible for foreigner producers to survive.
Both sides are right, but the problem is not that China may or may not dominate specific industries, like EVs. That is a separate issue from that of the overall economic impact of trade imbalances. The problem is that when a country dominates certain sectors, it matters whether or not its surging exports in that sector are matched by surging imports. There’s a difference between subsidizing supply in some sectors of the economy at the expense of supply in other sectors and subsidizing supply in one or more sectors of the economy at the expense of demand.13
The former shifts supply from one sector to another in the same country without affecting that country’s total demand, while the latter suppresses total demand and forces the cost of weak demand onto its trade partners. If a country just exports, it satisfies demand that had been supplied abroad, so that the increase in its exports doesn’t represent an increase in global supply but rather the replacement of supply in the surplus economy for supply in the deficit economy.
To put it differently, if China subsidizes EV exports, American consumers of EVs do indeed benefit from cheaper prices. But whether Chinese producers or American producers pay for the cost of these subsidies depends on whether higher Chinese exports result in higher Chinese imports or higher Chinese surpluses. In the former case, one set of Chinese producers pays for the subsidies delivered to another set of Chinese producers. In the latter case, it is U.S. producers that pay for the subsidies delivered to Chinese producers.
The problem with trade isn’t trade itself—it is unbalanced trade, in which the supply created by a country’s exports isn’t matched by demand created by its imports. Unbalanced trade and competitive advantage don’t expand global production. It is balanced trade and comparative advantage that expand global production.
Can trade intervention be used to make trade freer?
There is no meaningful difference between trade-oriented policies and most forms of industrial policy. Any economic, monetary, or fiscal policy that affects the balance between a country’s domestic savings and its domestic investment must necessarily affect that country’s trade balance, and through its trade balance, it must necessarily affect the balance between the domestic savings and domestic investment of its trade partners. In a closed global economy, where savings must equal investment, any policy that forces up the savings rate in one sector must be balanced by either higher investment or lower savings elsewhere.
That’s where trade intervention can lead to freer trade. Trade surpluses that are caused by beggar-thy-neighbor industrial policies—designed to improve international competitiveness by suppressing domestic demand—can only exist to the extent that they are matched by trade deficits in other countries. In that case, if the deficit countries implement interventionist trade or capital polices directed at reducing their deficits, these will automatically force surplus countries to reverse their own beggar-thy-neighbor policies. This in turn will force an adjustment in the global trading regime such that global trade is once again based on comparative advantage and contributes to expanding global production, not to suppressing global demand.
The point is that there are a wide range of policies that can cause global trade distortions, and while some of these policies can target trade, many of them don’t do so explicitly. That’s why in the interests of a well-functioning global trade environment it is better to target overall trade imbalances, as John Maynard Keynes proposed at the Bretton Woods Conference in 1944, than it to target specific trade violations.
While the World Trade Organization and other existing trade regulatory entities have focused on the latter, they have left us with a world of massive, persistent trade imbalances and perverse capital flows. These conditions are prima facie evidence that existing trade regulatory entities have failed to manage global trade appropriately. That’s why the United States, and most of the rest of the world, would be better off with a radical reorganization of the global trading system.
Appendix 1: How do indirect transfers affect trade?
While much trade conflict focuses on direct subsidies, it is the indirect subsidies that are much more problematic, although they mostly work the same way by transferring income from ordinary households to manufacturers and producers. An undervalued currency, for example, is a transfer from net importers to net exporters and because all households are net importers and all net exporters are manufacturers and producers, an undervalued currency effectively forces households to subsidize manufacturers. This increases the share of GDP retained by manufacturers while reducing the share retained by households, which results in less consumption and more savings.
Other subsidies work the same way. Ample credit at artificially low interest rates, for example, effectively transfers income from net lenders to net borrowers. In economies in which households are large net lenders (usually as depositors in the banking system) and borrowers are mostly manufacturers or investors in infrastructure and property, these low rates force the former to subsidize the credit costs of the latter, thus again reducing the GDP share of households and, with it, reducing the consumption share. Financial repression, as this is called, can force transfers from household savers to manufacturers that far exceed anything that might occur through undervalued currencies, as seen in Japan in the 1980s and China in the 2000s.
There are a wide variety of other such transfers. Low penalties for environmental degradation reduce costs for manufacturers but raise health and absentee costs for workers and households, forcing them to save more. Repressive labor laws obviously benefit employers at the expense of employees. Restrictions on migration, as in China’s hukou system, can penalize migrant workers while suppressing both labor costs for businesses and social benefit costs for local governments. In economies that overspend on transportation and logistical infrastructure, the country overall loses because the gross benefits of much of this spending is less than the overall costs. But because businesses and manufacturers nonetheless benefit from better infrastructure than could otherwise be justified, the net result is yet another indirect transfer from households to producers. This is an especially important form of transfer from households to manufacturers in countries such as China.
The point is that there are many kinds of transfers that have the same effect, but they are usually praised as effective forms of industrial policy rather than excoriated as forms of trade intervention, even though the consequence is the same. In a recent Financial Times article on Chinese policy on electric vehicles (EVs), Yanmei Xie notes that “China’s well-rehearsed industrial policy can be staggeringly wasteful but still produce stunning results.”
It is important to understand that precisely because much of this “industrial policy” spending is wasteful, this also makes it a powerful from of trade intervention that leads directly to competitive manufacturing, weak domestic demand, and persistent trade surpluses. That’s because while the Chinese economy may lose overall from government-subsidized spending, the losses nonetheless benefit the EV manufacturing sector directly.
As a result, this form of industrial policy represents yet another transfer of resources from non-manufacturing China to manufacturing China, in other words, a transfer from Chinese households to subsidize Chinese manufacturers in the same way that overspending on infrastructure, an undervalued currency, cheap capital, hukou restrictions, limited labor rights, and a weak social safety net do.
This has a triple result. First, the huge extent and variation of EV subsidies makes Chinese manufacturing extremely competitive in international markets. Second, the extent of direct and indirect transfers from households keep domestic demand too weak to support the economy or, more specifically, to absorb the production of EVs. And third, to externalize the cost of the subsidies and balance weak domestic demand with competitive production, China must run large trade surpluses.
While there is often ferocious debate about the extent of these subsidies, there is a very simple test for measuring them. Economies that heavily subsidize manufacturers at the expense of households should have larger manufacturing shares of GDP than their trade partners (as manufacturing migrates to where it is most heavily subsidized), they should have lower consumption shares of GDP than their trade partners (as consumers are forced to pay for the subsidies), and they should run large, persistent trade surpluses (as the repressed household income needed to pay for the manufacturing subsidies makes it impossible for domestic households to convert production into consumption).
When all three conditions hold, it is almost certainly the case that manufacturing has been subsidized at the expense of household demand.
Appendix 2: Does foreign capital reduce U.S. interest rates?
It is widely assumed that foreign capital inflows lower U.S. interest rates because it creates additional demand for a fixed supply of U.S. bonds, but this only shows how confused many are about the functioning of the balance of payments. If American businesses wanted to invest in expanding infrastructure or production facilities but were unable to do so because of scarce savings and the high cost of capital, foreign capital inflows would indeed help the problem by increasing the resources available to Americans (in the form of higher U.S. trade deficits) and so lowering the cost of capital for American businesses. This was the case in the United States during much of the nineteenth century.
But this is not the case if investment in the United States is constrained by weak demand rather than scarce capital. In that case, foreign inflows will affect the country’s internal balance in one of two ways. One way involves a rise in unemployment, as American producers lay off workers in response to the diversion of domestic demand to foreign producers in the form of a higher trade deficit. In that case, foreign savings simply replace domestic savings, and there is no net increase in demand for U.S. bonds.
The other way involves a rise in either household debt or in the fiscal deficit to counter pressure for an increase in unemployment. In that case, while the total amount of demand for U.S. debt will rise with the influx of foreign savings, the total supply of U.S. debt will rise by the same amount. There will thus be no predictable impact on interest rates.
This, by the way, should be obvious from the data. If foreign inflows really did lower interest rates, adjusted for levels of development, it should be the case that deficit countries would have lower interest rates than surplus countries, and the larger the deficit, the lower the interest rate. In fact, not only is this not the case, but it seems to be the reverse of what actually happens; it is surplus countries that tend to have lower interest rates than deficit countries. China, Japan, and the EU, for example, all have lower interest rates than the United States and the United Kingdom.
In addition to this blog, I write a monthly newsletter that focuses on global imbalances and the Chinese economy. Those who would like a subscription to the newsletter should write to me at chinfinpettis@yahoo.com, stating their affiliations. My Twitter handle is @michaelxpettis.
Notes
1 According to the World Bank, the only countries whose manufacturing levels exceed those of China are Ireland (38 percent), Liechtenstein (35 percent), and San Marino (33 percent), all of which are tiny economies whose predatory tax regimes sharply distort and overstate the reported value of their manufacturing sectors. Aside from these, Taiwan’s manufacturing share is 34 percent, the only economy more dependent on manufacturing than that of mainland China.
2 See, for example, Gordon H. Hanson’s otherwise very good “Washington’s New Trade Consensus,” published by Foreign Affairs in December 2023, in which he says, “[Former U.S. trade representative Robert Lighthizer] misattributes job loss to shady trade deals rather than to the simple truth that the United States has little comparative advantage in most areas of manufacturing. As the U.S. labor force has become more educated and U.S. technology companies, consulting firms, and other business service providers have established a commanding global presence in their industries, rising costs have priced American companies out of many manufacturing markets.”
3 The United States and the United Kingdom have run persistent deficits since the 1980s. Because Australia and Canada export large amounts of commodities, they will sometimes run surpluses when the prices of the commodities they export are especially high, but they generally run deficits. It turns out that what these economies have in common are deep, flexible, and transparent financial markets in which there is relatively little discrimination against foreign ownership of assets. It is mainly the quality of their financial markets that explains their high consumption shares, low manufacturing shares, and persistent deficits.
4 There can be exceptions, but they tend to have specific causes. In the period from the 1920s to the 1960s, for example, the United States ran large, persistent surpluses, while Europe (and later Japan) ran large, persistent deficits. But this was mainly because the latter, whose infrastructure and manufacturing capacity had been demolished by war, had substantial postwar reconstruction that needed to be financed, and the United States was the only economy that could finance it. Once reconstruction was largely completed, in the 1960s and the 1970s, the huge imbalances quickly disappeared. Clearly, that’s not the case today.
5 The relevant point here is not that wages in the United States are higher than they are in China, although they are indeed much higher, but so is American productivity. What matters is the relation between the two. Because the difference in wages exceeds the difference in productivity, American workers retain a larger share of what they produce than Chinese workers, and so they consume a larger share. In the Europe before 2008, on the other hand, wages and productivity were higher in Germany than in Spain, but because the difference in wages was lower than the difference in productivity, German workers received a lower share of what they produced than Spanish workers, and so they consumed a lower share. This is why the savings share of GDP is higher in China than in the United States and higher in Germany than in Spain. It has little to do with culture.
6 In 1986, Japan released the Maekawa Commission report, which argued that Japan should switch from an export-oriented economy to a domestic demand–led economy. In March 2007, then Chinese premier Wen Jiabao made his famous “unbalanced” speech to the Chinese National People’s Congress.
7 For more on this, see Michael Pettis, “China’s Debt Isn’t the Problem. It is a Symptom of the Problem,” Financial Times, December 20, 2023, https://www.ft.com/content/630f828c-ce4b-4f41-a867-9593bfaf0528.
8 There is no requirement, however, that bilateral trade imbalances must match bilateral capital imbalances. This means that the impact of a country such as China, Germany, or Japan on the U.S. trade deficit is not the same as the extent of its bilateral trade with the United States. Far more important is that extent of the bilateral capital flows. For example, if Japanese savings exceed Japanese investment by $100 and Japan invests the full amount of its excess savings in the United States, the United States must run a $100 trade deficit and Japan a $100 trade surplus, even if the two do not trade with each other at all.
9 Contrary to widely held assumptions, an explosion in consumer credit isn’t caused by a cultural propensity to consume. All it requires is a normal distribution of risky behavior among households and so can happen in any culture. That’s because if banks respond to monetary easing by relaxing lending standards, they will always be able to find optimistic households that want to borrow to boost consumption. What’s more, if foreign inflows cause the prices of domestic assets to rise, as is often the case, they can cause an increase in consumption through wealth effects that make it easier for banks to finance additional consumer debt.
10 This is not just a U.S. problem. Huge foreign inflows (mainly from Germany) into EU countries such as Spain in the early 2000s set off stock and real estate bubbles that accompanied soaring debt and, ultimately, a crisis in 2008-2009.
11 For more on this, see Michael Pettis, “Why US Debt Will Continue to Rise,” Financial Times, July 27, 2023, https://www.ft.com/content/61823af9-c6c1-4e3d-bcea-83b638c204a3.
12 See Michael Pettis, “A (Very Short) History of Global Reserve Currencies,” Financial Times, June 7, 2023, https://www.ft.com/content/c967ba48-f21b-4222-9f11-beb61ce710ae; and Michael Pettis, “An Exorbitant Burden: Why Keeping the Dollar as the World’s Reserve Currency Is a Massive Drag on the U.S. Economy,” Foreign Policy, September 7, 2011, https://foreignpolicy.com/2011/09/07/an-exorbitant-burden.
13 Notice that I separate the problems of trade from other issues. There may be very good political, security, or other reasons why the United States or Europe may want to protect specific industries, quite apart from their impact on trade imbalances. These reasons are legitimate ones, but they are outside the scope of this blog. My main point here is that subsidized imports may be paid for by either the exporting economy or the importing economy, and if it is the latter, households in the importing economy do not benefit from the lower prices.
