Should the United States Run a Trade Surplus?

Originally posted at China Financial Markets

Although standard trade theory predicts that highly advanced economies with sophisticated financial sectors, like the United States, should generally run trade surpluses, the country has run persistent, and often large, trade deficits for five decades. This can only be a consequence of significant global economic distortions.

There is usually a substantial difference between what I write in my newsletter and what I cover in this blog, with the newsletter generally more technical and focused on Chinese and global financial markets. In this post, however, I wanted to excerpt an older newsletter on why the United States runs a trade deficit when it should normally run a surplus. This is a companion piece to my February 7, 2019, blog post, “Why U.S. Debt Must Continue to Rise.”

This essay arose originally from a conversation I had several months ago with Barron’s Matt Klein during his visit to Beijing. Klein is one of the few analysts who understand the ways that changes in trade dynamics in recent decades have made much of the current debate on trade at least partly obsolete. In a discussion with some of my Peking University students, he pointed out that, under standard trade theory, the United States would normally be expected to run trade surpluses, and yet the country instead has run large deficits for nearly five decades. This should be surprising, he argued, and at the very least it strongly suggests the existence of distortions in the global trading system.

He is right, of course, and it occurred to me that working through the reasons might be an interesting way of understanding trade imbalances and their sources. In several recent blog posts, I have addressed this approach to trade from different angles (for example here, here, and here), and there will be overlap between this post with earlier ones. I apologize to my regular readers for this repetition, but the reason for discussing the topic from many different angles is because mainstream economists seem generally to misunderstand current trade dynamics.

For example, as I discussed in a May 2017 post, it is almost an article of faith among economists that the U.S. fiscal deficit contributes substantially to the U.S. trade deficit,1 and perhaps even causes it. But this claim is only true under certain conditions, which unfortunately most economists rarely bother to specify. If they did, they would probably see that these conditions no longer hold; and once we understand that the United States has little control over its domestic savings rate, we will see that the U.S. fiscal deficit is not a cause of the U.S. trade deficit as much as it is a consequence.

This flies so strongly in the face of conventional thinking among mainstream economists that it has to be repeated many times before they will consider it. So with apologies to regular readers for so much repetition, in this post, I want to approach the topic by arguing that the fact that the United States has run large trade deficits for several decades is surprising enough to require some explanation.


Why did Klein think that the United States would run trade surpluses in an open global system in which trade and capital flows are driven mainly by fundamentals? The reason is because investment should normally flow from advanced economies with high levels of capital, technology, and managerial know-how to less developed economies that need these resources, and in fact this has been the case for much of modern history.

Advanced economies—that is to say, mature, capital-abundant, and slow-growing economies—should have many decades of investment in high-quality capital stock behind them, so their current investment needs are relatively low. What is more, with their high income levels and sophisticated financial systems, their savings should be relatively high. For these reasons, savings should normally be pulled from these advanced economies into faster-growing developing countries, where capital is relatively scarce, investment more profitable, and institutional and technological resources lacking.

As the largest and most advanced economy in the world, and with by far the most sophisticated financial markets, the United States would normally be a net exporter of capital and technology to less developed economies: it should run on average a capital account deficit and its obverse, a current account surplus. This is just what the UK did in the late nineteenth century, perhaps the closest analog to the United States today. What is more, this tendency to run surpluses should be further exacerbated by the high level of income inequality from which the United States currently suffers—the highest since the late 1920s, when the United States, not coincidentally, ran the largest trade surpluses in history.

Why? Because income inequality at its simplest can be thought of as a distribution of income from low savers to high savers. It causes ordinary and poorer households, who are the high-consuming sectors of the economy, to have a disproportionately low share of total income relative to the rich, who on average consume a lower share of their income. Income inequality, in other words, forces up what economists call the country’s ex-ante savings (the savings that households plan to set aside at the beginning of a period of time) and, as part of the same process, reduces the consumption share.


This isn’t just true about income inequality. Any condition or policy that causes a transfer of income from one sector of an economy to another can affect the economy’s savings and consumption shares. Consider the table below, which divides an economy into six sectors, and describes each sector in terms of what share of its income is saved or consumed (all income is, by definition, either saved or consumed). As income is shifted from one sector to another, the differing tendencies of the two sectors to save or consume their incomes will change the overall savings rate of the economy.

Rich householdsConsume a small share of their income and save a large share
Ordinary households (older)Consume a large share of their income and save a small share
Ordinary households (younger)Probably consume a larger share of their income than older people and save a smaller share
BusinessesDo not consume, but save, all of their income, which is either invested or distributed to their owners
Local governmentsConsume a small share on behalf of local citizens and save a large share
Central governmentProbably consume a smaller share on behalf of citizens and save a larger share

The economy could be further subdivided into additional sectors with different saving propensities, but the above division should be enough to make the point clear: a country’s saving mainly reflects the way in which income is distributed. Notice the implication. Savings in a country usually don’t rise because the citizens of that country decide suddenly to become thriftier, nor do savings decline because citizens suddenly become more profligate. Savings rise and fall mainly as income is shifted among groups and sectors with different saving tendencies.

An example of how this occurs would be helpful in explaining the overall process. Among recent examples was the 2003–2005 Hartz reforms in Germany, after which German wage growth slowed sharply relative to GDP growth while business profits exploded. This is the equivalent of a transfer of income from workers and ordinary households, who consume a large share of their income and save a low share, to businesses, who effectively save all of it. As this transfer occurred, German savings soared. The press responded by expressing admiration for the ways in which German culture worships thrift, but the rise in German savings actually had almost nothing to do with a cultural inclination toward thrift.

It had to do with the shift in income from workers to businesses. In any country, whenever the business-profit share of GDP rises at the expense of the household share, the ex-ante national savings rate will automatically rise. It is this transfer that ultimately powered the huge subsequent increase in the German trade surplus. Contrary to popular opinion, in other words, Germany’s trade surplus does not reflect the fact that German workers are hard-working and thrifty (though they probably are, as indeed most workers everywhere are). Nor does the German trade surplus indicate that Americans, as EU budget-commissioner Günther Oettinger foolishly explained, love German cars more than Germans love American cars. It is mainly the result of a reform that allowed German businesses to profit at the expense of German workers.

From the table above, it is easy to see why income inequality also forces up ex-ante savings, and does so in the same way. It effectively represents a transfer of income from a low-savings sector to a high-savings sector—that is from ordinary households to the rich.


These kinds of wealth transfers within an economy are among the major determinants of a country’s balance of payments—its current and capital accounts. This is not just because of how such transfers affect ex-ante savings but also because of how they might affect investment (which, along with the capital account, determines how such transfers affect ex-post savings—the actual amount of savings).

A country that saves more than it invests must export that balance. This is shown in the following accounting identities:

  • GDP = Consumption + Savings, and also
  • GDP = Consumption + Investment + Capital Account Surplus, therefore
  • Consumption + Savings = Consumption + Investment + Capital Account Surplus, and because
  • Capital Account Surplus = Current Account Deficit, therefore
  • Current Account Deficit = Investment - Savings, or, which is the same thing,
  • Current Account Surplus = Savings - Investment

This is why the effect of income transfers is more complicated than we might first assume. Most economists assume that the lower consumption created by income inequality is matched by higher investment, or they think it is at least partially matched by higher investment; this is because they assume that higher investment is the result of an increased availability of savings at lower costs.2

But the impact of additional savings on investment varies greatly from economy to economy. In an economy whose substantial investment needs have been unmet because of scarce capital at high interest rates—mainly developing economies—the increase in savings is likely to be met dollar-for-dollar with an increase in investment or, put differently, the reduction in consumption is likely to be met dollar-for-dollar with an increase in investment. In such cases, income inequality would have little or no net impact on the current or capital accounts.

The United States, of course, is not a developing country. Today, American investors can easily access capital at among the lowest rates in history, and yet few seem interested in raising money to invest in the economy. Investment in the United States doesn’t seem to be constrained by scarce savings at all. Still, most economists would assume that the increased availability of capital at lower costs would cause at least some increase in domestic investment, although not as much as the increase in savings (or, to relate this again to consumption, the reduction in consumption is likely to be met with a smaller increase in investment).

In such cases, when income inequality in the United States increases the gap between savings and investment—that is to say, if it causes savings to rise more than investment rises—the capital account deficit and the trade (or current account) surplus must automatically rise. As an advanced economy, the United States should normally run a trade surplus, in other words, and this surplus should be expanded by its high levels of income inequality.


And yet the country runs a trade deficit. Before explaining why, I want to digress again to discuss in this and the next section what I think is another important but misunderstood consequence of income inequality in advanced economies like the United States. What determines the full impact of rising income inequality on the trade balance is not just how it affects savings but also how it affects investment.

The standard argument is that forcing up ex-ante savings is positive for investment because, even in economies in which savings are already abundant and where interest rates are low, it lowers the cost of funding, however marginally. If businesses can borrow at a lower rate than before, the argument goes, there is always some productive investment opportunity that only becomes profitable at this new, lower borrowing cost, and so this must (the thinking goes) lead to more investment. More investment, of course, should lead to more growth over the long run.

This is the basic argument behind supply-side economics, and it is the implicit justification for President Donald Trump’s 2017 tax cuts. Most economists agree that investment levels in the United States are low and that the country would grow faster over the long term if businesses could be encouraged to invest more. Given that one of the most efficient ways to boost investment is presumably to make more capital available to businesses at lower costs, proponents of this viewpoint claim that tax cuts for the rich will eventually benefit the rest of the country as the additional wealth generated by higher investment trickles down.

Can supply-side policies that result in greater income inequality nonetheless leave a country better off? It turns out that the answer, again, depends on the relative availability of savings in the economy. In an environment of capital scarcity, typically the case for developing economies, policies that force up the domestic savings rate can result in a substantial (even one-for-one) increase in domestic investment for every unit reduction in consumption. In such cases, total spending is unchanged (lower consumption is matched by higher investment); the economy grows as quickly as ever in the short run while also getting wealthier in the long run.

This isn’t necessarily the case, however, in an environment of capital abundance, a condition that applies to most advanced economies today. In those cases, most economists would agree that every unit reduction in consumption is likely to be matched by a smaller increase in investment, meaning that, in the short run, total demand declines.

This means that while supply-side policies can reduce growth in the United States in the short term because they cause a drop in total demand (as lower consumption is only partially matched by higher investment), as long as at least part of the reduction in consumption is matched by an increase in productive investment, it is still possible to argue that the country is better off in the long run because investment increases productive capacity. In such scenarios, the rich benefit immediately from tax cuts for the rich, while the rest of society benefits eventually. That is how wealth is supposed to trickle down.


But—and this is what may seem counterintuitive to most economists—it might be a mistake to assume that conditions that force up the ex-ante savings rate must always lead to some additional investment. There are conditions in which such conditions may actually lead to less investment; this is especially likely to be true today in most advanced economies.

All it requires, broadly speaking, is that all or most investment falls into one of two categories. The first category consists of projects whose value is not sensitive to marginal changes in demand, perhaps because they bring about very evident and significant increases in productivity, or because the economy suffers from significant underinvestment. The second category consists of projects whose value varies as a function of future expected changes in demand. I discuss this topic further in the “Where Might This Argument Be Wrong?” section of a previous blog post. I show in that post that in economies like the United States, in which the profitability of most investments is a function of changes in demand (the second category) rather than in the cost of borrowing (the first category), rising income inequality and higher ex-ante savings can actually result in less investment rather than more.

The point is that U.S. income inequality could increase the gap between savings and investment by even more than we might otherwise assume. Not only does it cause the savings of the rich to rise faster than investment, but it might actually cause investment to decline. This isn’t just theory. The increase in the savings share of German GDP after the Hartz reforms was matched by a reduction in the investment share, not the expected increase. And while it is too early to make the same claim about the Trump tax cuts of December 2017, which were supposed to boost investment by boosting savings, so far they seem to have done nothing of the sort.

Be that as it may, whether investment actually declines or merely grows more slowly than the decline in consumption, in a closed system like the global economy, savings and investment are by definition equal. This implies that in any country if a policy causes savings in one part of the economy to rise and investment to rise more slowly, or even decline, the domestic imbalance between savings and investment can be resolved in one (or some combination) of only two ways:

  1. The excess savings can be exported, in the form of capital account deficits along with the corresponding trade and current account surpluses.
  2. Something else must happen to cause savings in another part of the economy to drop, so in the aggregate there is no net increase in savings.

It is clear that the first of these two conditions does not apply to the United States. The country has no control over its ability to import or export savings. Its capital account is largely determined abroad.

Why? Because the United States has deep, completely open, and highly flexible capital markets and very strong governance, so the country ultimately absorbs a large part of the excess savings of the rest of the world—roughly 40–50 percent of the sum of foreign capital account deficits in recent years—the extent of which is only partially determined by domestic U.S. conditions or policies. As long as the United States runs a capital account surplus, and as long as this surplus is determined by foreign conditions and policies the country largely cannot control, the United States cannot be a net exporter of any excess savings accumulated through its high level of income inequality.

This means the second condition must apply. This is not a theoretical proposal but rather an accounting identity that cannot be broken: if the United States cannot export the excess of savings over investment, it cannot have these excess savings.

That being the case, something else must happen to cause savings in another part of the U.S. economy to drop by enough to absorb the sum of excess U.S. savings caused by income inequality and excess foreign savings imported into the United States. This has been the hardest part for even trade experts to understand, but there are many ways in which conditions that drive up savings in one part of the U.S. economy can drive them down elsewhere. I discuss some of these ways in the “What Drives Down Savings” section of my previous blog post.


What this all means is that, in one way or another, distortions (either in the U.S. economy or abroad), have transformed the United States from what should have been a reasonably persistent surplus economy into the world’s largest deficit economy. As I have explained before (including here, here, here, and here), ultimately the United States must respond to the distortions created by net capital inflows, and the consequent current account deficit, with either more unemployment or more debt. This explains the real relationship between the fiscal deficit and the current account deficit: if the United States is to avoid higher unemployment, either the U.S. government must run a fiscal deficit or U.S. authorities, including the Federal Reserve, must create conditions under which private American (mainly households) run deficits and raise debt levels.

The extent of the distortion can be enormous. Assuming that the United States should normally run a current account surplus of roughly 2–3 percent of GDP, in line with that of other rich, capital-exporting countries and substantially lower than the UK’s surplus at the end of the nineteenth century, the American current account deficit of roughly 3 percent of GDP implies a distortion equal to 5–6 percent of U.S. GDP, or a very high 1 to 1.5 percent of global GDP. This suggests that, while the United States currently absorbs 40–50 percent of the world’s current account deficits, it may be absorbing up to two-thirds of all the world’s excess savings.

So the answer to the original question of why the United States isn’t a trade surplus country is that it should be, but because of its deep, flexible, well-governed, and completely open capital markets, in a world of excess savings and insufficient demand, the United States absorbs a substantial share of excess savings from abroad. These savings, in turn, create distortions in the domestic economy, which in turn force down U.S. savings and cause the United States to run the largest trade deficits in the world.

The United States isn’t completely impotent on this issue. U.S. policies and conditions can have some effect on the amount of foreign excess savings, and a somewhat greater effect on the extent to which these savings are exported to the United States, but these effects can be counterintuitive. For example, if Washington were to reduce the U.S. fiscal deficit, and if a lower fiscal deficit increased the attractiveness of the United States as an investment destination, net foreign inflows into the country might actually increase. This would mean that a lower fiscal deficit could paradoxically result in a higher current account deficit, directly contradicting the view of the many economists for whom it is an article of faith that lower fiscal deficits must result in lower current account deficits.

Under current conditions, however, the United States’ ability to control the amount of foreign excess savings that is invested in the country is very limited. As long as it has a completely open capital account, the U.S. current account deficit is likely to be a residual, mainly reflecting factors abroad. To the extent that the amount of excess savings in the rest of the world is determined partially or mainly by conditions and policies abroad, the United States cannot control or manage its current account deficit as long as it does not manage its capital account.


This, by the way, is not just a problem for the United States. It is also a problem for the UK and the other Anglo-Saxon economies with broadly similar and equally open capital markets, all of which have tended to run persistent deficits since the 1970s, contrary to basic trade theory. This was also a problem for countries, like Spain and other “peripheral” European countries in the years leading up to the 2008–2009 crisis, whose monetary conditions left them open to capital exports from large surplus countries in Europe, mainly Germany. Finally, it may also be a problem for developing countries during periods of major global liquidity expansion, especially to the extent that they cannot control the enormous liquidity inflows that typically afflict developing countries during these periods.

The United States has been running trade deficits for so long that we have forgotten how strange this is, and if we think about it at all we tend simply to dismiss the problem as a consequence of American profligacy. But even if profligacy were truly a problem, in a well-functioning world of global trade, if a country like the United States were to run deficits, after a fairly short time these deficits would force changes (mostly monetary but also structural) in the U.S. economy that would eliminate the deficits. Persistent U.S. trade deficits are unnatural, and there has to be a reason they exist.

What is more, these U.S. trade deficits force the United States into accepting either higher unemployment or a more rapid increase in debt. If the country wants to escape this condition, the United States must reduce its trade deficit with the world, but not by addressing the trade deficit directly through import tariffs or quotas. Instead, the United States must address foreign capital inflows directly, perhaps by taxing them.

For what it’s worth, in September 2019, Yale University Press will publish a book in which my co-author and I argue that these distortions should not be seen as representing conflicts between countries so much as conflicts between economic classes. Trade war, in other words, is really a disguised form of class war, but more on that in September.

Aside from this blog, I write a monthly newsletter that covers some of the same topics. Those who are interested in receiving the newsletter should write to me at chinfinpettis[at]yahoo[dot]com, stating affiliation.


1 Technically, throughout this post, I should refer to the current account rather than the trade account, but the latter is usually the main component of the former and absorbs most, if not all, of the variation. Although it is technically incorrect, I will generally treat the trade account as interchangeable with the current account.

2 I was told by a reader a year ago that some of his friends, academic economists, dismissed this approach to savings and investment on the grounds that it is based on the dreadfully unfashionable loanable funds approach to money creation, rather than endogenous money or MMT.

This only indicates how confused many economists are about the meaning and implications of MMT, and about money creation more generally. Savings, or rather deposits, according to MMT, are not exogenous to the banking system and distributed by banks in the form of loans, as is commonly thought (although no one actually thinks this), but rather they are endogenous because they are created by bank lending.

But this is a completely different meaning of savings. It is monetary savings (that is, bank deposits) and not real savings. In the savings-investment framework used here, savings are simply the amount of goods and services produced by the economy that are not consumed. Investment is necessarily constrained, by definition, by the amount of available savings. The savings-investment framework has nothing to do with loanable funds, endogenous money, or indeed any other theory of money creation. It is about the classification of real goods and services. In the case of an economy that produces three widgets, if it consumes two widgets, it cannot then invest two more widgets, whether or not money is created by banks or merely distributed by them.