News Ticker

Internal and external balance

Michael Pettis, 9 June 2015

I was recently asked by an Australian economics journal to write a review of a book I had already read,The Leaderless Economy, by Peter Temin and David Vines (published in 2013). Because the book is a great place from which to start a discussion on the links within the global economy, I decided to base this essay on the book. I had already read Peter Temin’s Lessons from the Great Depression (1991), The Roman Market Economy (2012), and Prometheus Shackled (2013), and I know his work fairly well.

There is substantial overlap between the way Temin and Vines view the global economy and the way I do. I am a regular reader of Diane Coyle’s blogThe Enlightened Economist, in which she reviews books, usually on economics-related topics but also on anything else that might catch her attention. Besides having said very nice things about my bookThe Great Rebalancing, and including it in her shortlist of top books for 2013 (I am bragging a little) she has also reviewed The Leaderless Economy, and has noted the similarities in the way both books approach the balance of payments.

She is right. Both books analyze the global economy in pretty much the same way, as an economic system in which any country’s domestic economy is inextricably linked to other economies through the balance of payments mechanisms. The ability to place events within their global context is consequently crucially important in understanding any country’s economic performance, but actually doing so tends more to be the exception than the rule.

To take one example, I remember in the early 1990s I read a book about the Panic of 1873, in which during September and October of that year gold reserves in New York plummeted, banks folded, and the stock market crashed, with the New York Stock Exchange actually closing, on September 20, for ten days. It was in many ways an excellent book about the events leading up to the crisis and the various factors that “caused” it, but except for a brief mention of the impact of the crisis on the British banking system, you could have easily assumed, as I did then, that the 1873 crisis was purely an American affair, and that events in the rest of the world barely mattered.

It was only a few years later, while I was reading Charles Kindleberger’s A Financial History of Western Europe that I learned that the 1873 crisis actually “began” with a stock market crash in Vienna in May, four months before the New York markets fell, which spread to Germany, England and other countries, and the subsequent depression was perhaps the first “global” panic and depression in history. It was in Kindleberger’s book that I also first learned about the impact of the Franco-Prussian War of 1870-71 and the subsequent reparations payments on global financial markets (which I discuss extensively in a February blog entry) and in unleashing the final stage of a global liquidity bubble that ended with the various panics of 1873.

The Panic of 1873, in other words, was clearly not only, over even mostly, an American affair. It was global, and its causes could not have been limited to American causes, as the book seemed to imply. On the contrary, I would argue that it is impossible fully to understand the crisis, whether the focus is on the evolution of the crisis in the US, the UK, continental Europe, or Latin America, without understanding the full global context.

The Leaderless Economy makes the same point I try to make in The Great Rebalancing: the economic analysis of any country is largely useless if it ignores, or treats as a minor issue, links with the external sector – i.e. other countries – and this is even more true today than in the past. Even something as important, and as seemingly “domestic”, as the US savings rate (which for most people is assumed largely to reflect cultural preferences towards thrift among American households) is not determined primarily by American households but rather by its links with savings distortions abroad.

This might seem a profoundly counterintuitive statement, but in fact you only need to understand two or three accounting identities to be able to work logically through the explanation. I showed why one country’s savings rate is as likely to be determined by domestic distortions as it is by distortions abroad in a blog entry on whether a savings glut must cause global savings to rise. In another blog entry, I explained why a low savings rate in countries like Spain was a necessary outcome of policies in Germany that effectively restricted wage growth. And finally I showed how – because of the role of the dollar in reserve accumulation and, more generally, the unrestricted ability of foreigners to buy US assets – the US savings rate is determined largely as a residual needed to balance net capital flows, in a blog entry on the so-called exorbitant privilege, and in a follow-up blog entry.

Global balancing mechanisms

In their book Peter Temin and David Vines don’t discuss the US savings rate but they do a lot of the same thing by changing the focus slightly. In my book I argue that because global savings and global investment must balance, if any country’s savings exceeds its investment by some amount, investment in the rest of the world must exceed savings by exactly the same amount. Temin and Vines express the same relationship between conditions in one country with conditions abroad in a way that can be summarized in two sentences. First, the domestic imbalance in any economy – i.e. high levels of unemployment or demand, large gaps between savings and investment, and so on – must be consistent with and equal to that country’s external imbalance at all times. Second, that country’s external imbalance must be consistent with and equal to the external imbalance of the rest of the world at all times

These two ways of seeing countries within their global contexts are basically identical. Any country’s capital account of course is simply the gap between its domestic savings and its domestic investment, and because the capital account must balance the current account (the two always add to zero), to say that the gap between savings and investment in any country must be equal to an opposite gap between savings and investment abroad is simply to restate the far more intuitively familiar claim that every current account surplus in the world must be matched by a current account deficit. Until we begin trading with extra-terrestrials the total must add up to zero.

This should all be obvious, but if an imbalance in one country must be matched by an opposite and equal imbalance abroad, there are only two possible explanations. Either the imbalances in every country are set endogenously, and, by some miracle, at every point in time, they balance out perfectly, or an imbalance in one country can force imbalances onto other countries. The first explanation is obviously absurd, so the gap between the amount a country saves and the amount it invests is as likely to be determined by external conditions as by internal. This is why it is possible to argue that a country’s savings rate (or its unemployment rate, or its investment rate, etc.) is determined by policies abroad as much as by policies at home. This is what it means to live in a “globalized” world.

Most analysts tend automatically to look at the current account as the balancing mechanism, but I think this is perhaps why it is so often easy to get confused about the balancing process. By explaining the adjustment process in terms of the capital account, as I do, I think it becomes much easier to understand how the direction of net capital flows determines the balancing mechanism. It also helps analysts avoid falling into the trap of looking at bilateral trade balances as meaningful when in fact, and very counter-intuitively, they are almost totally useless when it comes to understanding how distortions in one country might cause distortions elsewhere. For example, last year Stephen Roach, one of the most knowledgeable commentators on the Chinese economy, fell into the same confusion when he wrote the following:

The US trade deficit is a multilateral imbalance with many countries – 102 in all – not a bilateral problem with China. It arises not from the alleged manipulation of the renminbi, but from the simple fact that America does not save. Lacking in domestic savings and wanting to grow, the US must import surplus savings from abroad, and run massive current-account deficits to attract the foreign capital. And that leads to America’s multilateral trade imbalance

If China runs a capital account deficit and the US a capital account surplus, and these are roughly equal to net purchases by the PBoC and other Chinese government entities of US government bonds and US assets, China will run a current account surplus exactly equal to its capital account deficit. The US will run a current account deficit exactly equal to its capital account surplus.

This may show up in the form of corresponding bilateral trade surpluses and deficits between the two countries, but it is not necessary, and in fact extremely unlikely, that it would. While we cannot say just from looking at the numbers whether low US savings relative to investment forced the Chinese into saving more than they invest, or whether high Chinese savings relative to investment forced the Americans into saving less than they invest, we can say that the savings rate in one of the countries was determined in large part by distortions in the other.

The main point is that if you want to understand the causal relationship between Chinese surpluses and US deficits, or between German surpluses before 2009 and peripheral European deficits, you have to look at the direction of net capital flows, and not at bilateral trade. I will return to this, but to get back to The Leaderless Economy, this book is underpinned by the same deep familiarity with financial history that Peter Temin brings to all his work, and perhaps this is why the book should leave everyone terribly frustrated on two counts. First, the authors show that while the 2007-08 global crisis and its aftermath were unquestionably large and complex affairs, there was nothing about the crisis that was unprecedented – we have experienced similar events before – and we understand far more about what might or might not have worked than subsequent policymaking might suggest.

The authors show for example that anyone who had a reasonable understanding of the current and capital account pressures of the 1920s whose inconsistencies doomed Germany should have been able to understand why downward pressure on German wage growth, in the several years before the global crisis was set off in 2007 by the US subprime crisis, would ultimately force huge internal imbalances onto Europe during that time. More importantly, this understanding should have been enough to convince European policymakers that unless Germany responded to the crisis by reflating domestic demand sufficiently to generate large current account deficits, it would be all but impossible to prevent a decade of anemic growth and extraordinarily high unemployment in peripheral Europe.

It’s not that Germany in the 1920s was anything like Germany in the 2000s, or even that Germany suffered in the 1920s from the kinds of pressures it forced onto peripheral Europe in the 2000s (although this is closer to the truth). The problem was that Germany in the 1920s was required to export large amounts of capital, because of extremely high reparations demands, while its ability to run current account surpluses was constrained by European post-war policies.

No country can run a capital account deficit unless it runs a current account surplus, and these enormous countervailing pressures forced Germany, among other things, into an unsustainable borrowing path. Similarly since 2008-09 peripheral Europe has been under pressure to run capital account deficits (it must repay substantial borrowings and fund flight capital, but it has trouble borrowing without implicit ECB guarantees). Its ability to run countervailing current account surpluses, however, is constrained within Europe by Berlin’s refusal to allow a reflation of domestic demand and it is constrained outside Europe by Germany’s enormous current account surplus, which prevents a currency adjustment. This forces peripheral Europe into both rising debt and high unemployment, and it is only because Europe as a whole has forced the problem of weak German demand onto the rest of the world that conditions in Europe are not even worse.

Are trade surpluses virtuous?

Before I explain the second source of great frustration that this book provides, I want to make a quick detour on the subject of current account, or trade, surpluses. Any suggestion that Germany’s trade surplus is too large and has become a source of global distortions elsewhere seems to infuriate a lot of people, for many of whom the idea that a trade surplus can be too large is the moral equivalent of castigating people for working too hard and being too prudent. A trade surplus, they believe, is the reward a country gets when its people work hard and save their money.

But would a nation of productive and hard-working people, let’s call them ants, normally run trade surpluses? The consensus seems to be that it would. Running a trade surplus means that a country sells more to foreigners than it buys from them, and there seems to be an implicit belief that exports are what a hard working country produces, and imports are the equivalent of its consumption, so that a trade surplus means that the country earns more than it spends, and the larger the surplus, the more likely the ants in that country are especially productive, thrifty, morally upright, and perhaps fond of sensible clothing.

Countries that run large and persistent trade deficits, on the other hand, buy more from foreigners than they sell, and because this seems to suggest they are living beyond their means, we usually think of these people, let’s call them grasshoppers, as lacking discipline and prudence. Grasshoppers excite our scorn, even if we sometimes envy their carefree ways and their bohemian morals.

If households that are careful to spend less than they earn, making sure to save part of their earnings, are seen as admirably thrifty, while households that are too quick to pull out their credit cards to buy things they cannot afford are seen as foolish, and will in the end implicitly require that their prudent neighbors bail them out, the same, we assume, should be true of countries. It is unfair that ants have to bail out grasshoppers, and so naturally we tend to abhor policies aimed at favoring grasshoppers over ants, and by extension we also abhor policies that seem to favor trade deficits over trade surpluses.

But countries are not at all like households and, more importantly, a country’s earnings are not equal to its exports. A country earns the total amount of goods and services that its households produce, of which exports are only a part – and usually a small part except in trading entrepots and in mercantilist countries. The richest and most productive countries are not the ones that export the most. They are the ones that produce the greatest amount of goods and services, and we usually, although not always accurately, measure this number as gross domestic product (GDP).

A country of hard working, productive ants, in other words, will not be rewarded with a high trade surplus, but rather with a high GDP, which should translate into a higher quality of life. They might export a great deal, but the reason they export is so that they exchange their exports for foreign goods that either they cannot make at home or that take more time and money to produce at home than to import from abroad. This should be obvious when we consider that while it is possible for every household to work harder, produce more, and so consume and save more, it is impossible for every country to work harder, export more, and import less.

A country of ants could run either trade surpluses or trade deficits, depending on conditions that have nothing to do with how hard working it is, but if it did run surpluses for many years, its currency would eventually rise, allowing ants to exchange their products for even more foreign products than before. This would cause the trade surplus to disappear as the ants consumed a larger amount of foreign goods that they could acquire for less work than before.

So if ants are not rewarded with trade surpluses, what explains the fact that some countries run persistent trade surpluses year after year while other countries run persistent deficits? The reasons, it turns out, has to do with the relationship between savings and investment. Countries that save more than they invest must export these excess savings to other countries, through the capital account, because the savings cannot be invested at home.

But something else must happen. The ants must either work to produce goods and services to be consumed, or they must build roads, factories, airports, and other kinds of investment that are funded by their savings – i.e. the share of their production that is not consumed. This statement is usually characterized as:

Consumption + savings = Consumption + investment

If the ants save more than they invest, then by simple arithmetic the total amount of goods and services they produce as they earn their incomes must be greater than the total amount of goods and services they produce in order to consume or to invest. Put differently, the total supply of goods and services they produce must be greater than their total demand for goods and services, and so they must export the excess by running a trade surplus.

A country’s trade surplus (more accurately its current account surplus, which includes things like royalty payments, tourism, returns on investment abroad, and other things, but we will pretend they are the same) is exactly equal to its total savings minus its total investment, which is equal to the amount of savings it exports. This is why we refer to the balance of payments. The total money that ants receive from their trade (current account) surplus is perfectly balanced by the capital account deficit, which is simply the amount of savings they send abroad.

Why save more than you invest?

Countries run trade surpluses, in other words, not because their citizens are hard working ants, but rather because they save more than they invest. Of course ants are supposed to save more than grasshoppers because they are prudently willing to make some sacrifice today in order obtain a better outcome tomorrow, but this means that they also must invest more than grasshoppers because the only way a sacrifice today results in a better outcome tomorrow is if we save part of today’s production and invest it in something that will increase our productivity tomorrow. The natural thriftiness of ants explains why they save more, but it cannot explain why countries save more than they invest, and so run trade surpluses, year after year. There are three reasons that explain most cases of large, persistent trade surpluses.

The first reason is a good one in that it results in higher growth and a better outcome for the world overall. The second two reasons, which are really variations of the same reason, result in lower growth for the world overall.

In the first case, there is a huge investment opportunity abroad, perhaps because a group of foreign ants have identified a great opportunity to increase productivity, and this opportunity persists year after year. The returns on that investment are so much higher than they are at home that ants at home are willing to save more than they naturally would to invest at home.

Because they save more than they invest, they export the excess savings abroad, where it earns an outsized return. Total growth in production for the world, consequently, is higher, and if that increase in production is shared between the ants at home and foreign ants, every one is better off. Notice however that the foreign ants will be importing savings because their investment needs exceed their savings, and so in spite of their hard work and rapidly rising productivity, they will run trade deficits for many years. In the 19th Century England and the United States played these two roles, with excess English savings pouring into the United States to fund growth in history’s most successful emerging market, and while the British ran persistent trade surpluses, and the US ran persistent trade deficits, both countries got richer.

The second reason a country might save more than it invests is because of high levels of income inequality. Rich people usually save more of their income than ordinary people, so that the more they keep of the total amount of goods and services produced by hard working ants, the higher that country’s savings rate and the lower its consumption rate. Because lower consumption discourages businesses from building new factories or otherwise expanding production, higher savings often come with lower investment, and so countries with highly unequal income distribution tend to run large trade surpluses.

The other reason has to do with the share that households receive of everything they produced. A country’s total production is divided between households, businesses, and the government. Although governments do make purchases that we can classify as consumption, almost all consumption comes from households. This matters, because in countries in which households are allowed to keep a very large share of what they produce, whether they are as thrifty as ants or as spendthrift as grasshoppers, their total consumption will be a high share of total GDP, and total savings a low share. If their high consumption encourages businesses to open new factories, low savings and high investment might even lead them to run temporary trade deficits, although only until the factories begin to produce.

But if households retain a low share of everything they produce, with governments and businesses getting the rest, then again, whether they are as thrifty as ants or as spendthrift as grasshoppers, their total consumption will be a low share of total GDP, and the country’s total savings, which is equal to GDP minus consumption, will be a high share. Once again if low consumption actually causes investment to drop, they will run large trade surpluses. Notice however, that the country has a high savings rate not because ants have saved a lot of their earnings. It has a high savings rate because businesses and governments are able to save the money that was not given to households.

This is what happened in Germany during the past decade. While Europe boomed, a number of factors allowed companies to reduce wage growth in Germany sharply, leaving German households with a declining share of what they produced and businesses with more than ever. As income inequality also rose during this time, German household consumption rates dropped and so total consumption rates dropped. Meanwhile, German companies decided to reduce their investment in Germany, perhaps because German households were not able to buy what they produced. Higher savings and lower investment caused Germany’s trade surplus to soar.

So as counterintuitive as it may at first seem, trade surpluses are not the reward for hard work or thrift. Trade surpluses can have many causes, too often just the low share of GDP that ordinary households are allowed to keep, but trade surpluses or deficits that persist for many years are always the result of significant distortions at home or abroad. Rising productivity and hard work can lead to strong export growth, but only net capital exports will lead to a trade surplus (or, more correctly, a current account surplus). If countries like Germany, whose surpluses have caused so much damage to Europe, were to arrange income more fairly, not only would Europe benefit, but Germany’s hard working and thrifty ants would finally get their rewards – a higher income, rather than a larger national trade surplus.

The value of history

I said that for me there were two great frustrations that come from reading The Leaderless Economy. The first is that we have the tools to understand much of what went wrong before the 2007-08 crisis, and to understand why distortions in the current and capital accounts can lead to a build up of debt and a subsequent disruptive adjustment. We didn’t use these tools because we failed to work through the logic of the balance of payments, to the point where in some countries we actually applaud measures that reduce the share of production allotted to hard-working households on the grounds that their real reward is a higher national trade surplus.

The second frustration is closely related. By showing us how useful history is in understanding current events, the authors also reveal, indirectly, how little value there is in the vast literature of economic theorizing that emerges from the abstract and distorted world of academia, bereft as it is of historical context.

For most economists, history is usually little more than a data sequence whose values are plugged into mathematical models – whose implicit assumptions too often these same economists fail to understand. But as Temin and Vines show, history is much more usefully seen as the evolution of often complex institutions – financial, political, legal, cultural, and so on – through which economic behavior is mediated and which affect the ways in which recurring patterns of finance, commerce and trade unfold, and that without an understanding of history we lose so much complexity in our models that we often end up making very obvious mistakes.

Charles Calomiris, a professor at Columbia University whose work on banking systems and banking system risk is among the best in the world made much the same point at a presentation at the Federal Reserve Bank of Atlanta last month: “You have to think historically to think usefully. Traditional economic models don’t incorporate historical analysis, despite its usefulness in explaining factors behind crises or finding holes.”

The 2007-08 global crisis and its aftermath are so much easier to understand if the period is placed firmly within history, which the authors do, especially focusing on events in the 1920s and 1930s, two decades that have great resonance with the past fifteen years. One of their goals is to explain the insights that John Maynard Keynes provided on the workings of the global balance of payments as they work through the development of Keynes’ thinking in real time. Not only does this flesh out the logic behind Keynes’ evolving theories, but it also makes it easier to understand the events that inspired Keynes ultimately to overcome Ralph Hawtrey’s objections in a famous 1930 debate to his intuitive sense that excess savings and unemployment were related, although the debate itself is not discussed in the book.

Why does this matter? It is impossible to overstate the usefulness of the reasoning Keynes provided that links savings, investment, unemployment, and trade. John Hobson had already shown nearly half a century earlier how institutional distortions that force savings levels higher than the level of desired investment will affect an open economy, but Keynes created a formally logical way to think about it. By using this model we are able to understand today why in a world without a near infinite amount of credible demand for investment (the role the United States played in the 19th Century), high levels of income inequality, or downward pressure on the household share of GDP, must lead almost inevitably first to rising debt and later to rising unemployment.*

These savings imbalances are often at the heart of the great imbalances in global trade and capital flows, and in The Leaderless Economy the authors clarify immensely the transmission of savings imbalances from one country to another. Domestic imbalances in any economy must be consistent at all times with that country’s external imbalances or, put differently, any gap between a country’s total savings and its total investment must result in a current account imbalance.

So if households consume a declining share of what they produce, either they must increase investment commensurately, so that savings rise and GDP remains the same, or, if investment cannot rise fast enough (in fact in most cases it is likely to fall), unemployment must rise fast enough to keep total savings from exceeding total investment. Of course if they force the demand shortfall onto the rest of the world, they can keep unemployment down by exporting their excess savings and running a current account surplus, the counterpart of which is a current account deficit somewhere else. These are just accounting identities – the internal imbalance between a country’s savings and investment is at all times consistent with and equal to its external imbalance, and a country’s external imbalances is at all times consistent with and equal to the external imbalances of the rest of the world. To understand which country will run the corresponding external imbalance depends on the nature of the institutions through which economic behavior is mediated, and this is why a deep grounding in economic and financial history is so important.

* I explain the model formally in the “Appendix” to The Great Rebalancing: Trade, Conflict, and the Perilous Road Ahead for the World Economy (Princeton University Press, 2013)

Academics, journalists, and government and NGO officials who want to subscribe to my newsletter, which sometimes includes portions of this blog and usually (as in this case) does not, should write to me at, stating your affiliation, please.  Investors who want to buy a subscription should write to me, also at that address.

Reproduced with permission, originally published as

Leave a comment

Your email address will not be published.