Remaking Globalization for an Era of Trade Wars

Dominik A. Leusder
Your work on trade has recently gained attention in US policy circles. Along with your coauthor Matthew Klein, you locate the origins of the current trade war in the growth models of large export-oriented economies such as Germany, Taiwan, Japan, and particularly China. Large and persistent imbalances in global trade and finance originate from “institutional distortions” in these countries. These take the form of weak welfare regimes and suppressed wage growth, which shift income to high-saving entities, such as firms and rich households, at the expense of workers. In other words: high inequality leads to a glut of savings in one sector and lower consumption in another.
In countries such as China, this is the result of concerted industrial policy: the “excess” of savings can be channeled into investment, and goods can be exported competitively. And China certainly is the most spectacular example of economic development driven by investment and net exports. But in a globalized economy, these domestic distortions are transmitted through the trade- and capital accounts, since globally savings and investment need to balance out the “excess” of both production and savings.
The United States, with its large capital market and consumer base, plays a singular role not just as a trading partner but as a destination of first choice for these “excess savings.” How do those capital inflows distort the US economy and how do they inform the current trade policy that we’re seeing?
Michael Pettis
Exactly. The paradox is that if instead you repress domestic wages and use your earnings to subsidize, say, manufacturing, you will grow more quickly. The problem is that your production will grow faster than your demand. And because in a closed system you can’t sustainably produce more than you demand, you end up running a trade surplus. At some point you will have to cut back on production and allow unemployment to go up. But in an open system like the globalized economy, you can run a trade surplus.
Robinson argued that it was a bad thing — but that’s not necessarily true. You can export your excess savings to developing countries who can use it to increase domestic investment. Because they have high investment needs and typically insufficient domestic savings, the reduction in your domestic consumption will be matched by an increase in investment elsewhere. And the world is still fine. Demand continues to grow, and businesses have to respond to that growing demand by expanding production and expanding productivity.
The problem arises from exporting the savings imbalance to advanced economies that aren’t suffering from saving constraints. In other words, if you export your excess savings to England or Canada or the United States — which together account for two-thirds to three-quarters of all of the export of excess savings — your domestic imbalance is being absorbed by countries that don’t have saving constraints. In that case, investment doesn’t go up.
If savings go up in, say, Germany, then that is relative to German investment. If, say, Spain was a developing country whose investment was constrained by lack of Spanish savings, then, driven by German capital flows, Spanish investment could go up, and the world would be in a better place, right? More investment in a country that needed it and growth as a result. But if Spain isn’t savings-constrained, investment won’t go up. And because everything has to balance, something must happen to allow higher German savings to be matched by lower savings in Spain. So something else has to adjust, and there are various ways they can adjust.
Robinson argues that Spanish savings adjust through higher unemployment in Spain. When you increase your manufacturing growth by repressing wages, then an increase in exports earnings isn’t recycled back to workers in the form of higher income, and so consumption doesn’t go up and imports don’t go up. So Germany runs a persistent trade surplus with Spain. [In an ideal trading system which didn’t allow for “beggar thy neighbor”–style export growth, trade imbalances are more likely to equilibrate if a rise in German export earnings translated into a rise in domestic consumption to the point that demand for Spanish goods and services rises.]
Robinson was writing during the gold standard, a time in which credit growth was constrained. But of course, we live in a different world where you can expand credit. So Spain has an alternative to a rise in unemployment: a rise in debt by way of the fiscal deficit. This would raise domestic demand in Spain. So now we don’t have a rise in unemployment, but we have a rise in Spanish debt. In addition, the increase in demand caused by the increase in debt gets redirected away from the tradable goods sector to the service sector, causing inflation and a rise in asset prices.
This is the basic problem: if Germany runs an export surplus because of distortions in the way domestic income is distributed and exports the excess savings to Spain, and if Spain does not invest those excess savings because there’s no saving constraint on investment, then Spain must respond either with a rise in unemployment or with a rise in debt.
This is pretty much what happened. After the German labor reforms in the early 2000s, Germany became a surplus country. That money poured into Spain, and Spanish household credit expanded very rapidly, and Spain went into a fiscal deficit from a fiscal surplus. Once Spanish debt could no longer rise after the 2008 crisis, Spain still had to adjust. But then it adjusted in the form of a rise in unemployment.
The same holds true for South Korea or Japan: if they implement policies that lead to persistent surpluses, specifically surpluses driven by the repression of domestic wages, and if they then invest the proceeds from those surpluses in the United States, Canada, and the UK, then what we should see is that those countries run persistent deficits backed either by rising unemployment or, much more likely, by rising household or fiscal debt. This is exactly what happened.
Michael Pettis
Yes. The rules of the eurozone made it hard. Before the euro, Spain could have adjusted by devaluing its currency or by changing domestic interest rates. So there were a lot of things Spain could have done, but with the existence of the euro, they couldn’t. And the important lesson there is not that the euro is bad. The important lesson there is that if your economy is open and you create a distortion in your internal account, it will create an external imbalance. And if mine too is an open economy, your external imbalance becomes my external imbalance, which means that my internal economy must shift in a way to accommodate that imbalance.
For example, if my country engages in industrial policy in order to increase the manufacturing share of my economy, then whether you like it or not, my industrial policy becomes your industrial policy in reverse. You must reduce your manufacturing share of your economy and shift from tradable goods to non-tradable goods, whether that’s your policy or not.
This argument was originated by both Joan Robinson and John Maynard Keynes. But more recently, Dani Rodrik has made a very similar point. He notes that in a hyper-globalized world, one in which frictional trading costs and frictional capital costs are very low, every country has to choose between more control over the domestic economy or more global integration. So if you and I both agree to choose global integration over economic sovereignty, we can have some kind of equilibrium. But if you choose global integration and I choose economic sovereignty, then I control both my internal imbalances and your internal imbalances.
Michael Pettis
I think it’s very likely, in fact. It’s what we saw in Japan in the 1980s and ’90s, in Brazil in the 1960s and ’70s, and in the Soviet Union. Namely, if domestic consumption is much higher than we think it is, there’s a puzzle. And there are only three ways you can resolve that puzzle: investment is much lower than we think it is; GDP is much higher than we think it is; or the country doesn’t actually run a trade surplus but a deficit. If you believe any of those things, then you can argue that consumption is much higher as a share of GDP.
But the problem is this: what do we mean by GDP? If Chinese investment were productive, then their measure of GDP would be correct. But as the Financial Time’s Martin Wolf asked about a month or two ago: How is it possible for a country to invest the equivalent of 43 percent of GDP and only grow by 5 percent?
Let’s look at Malaysia. At the peak of its growth, when it was growing much faster than China is growing today, investment was 33 percent of GDP at its peak. And that makes sense if you look at very rapidly growing countries with high productive investment — typically, investment is about 30 to 34 percent of GDP on average. It’s about 25 to 26 percent of GDP on average globally.
The only way you can resolve that very high investment with relatively low growth in China is by assuming that much of the investment isn’t productive. In practice this means that investment is unprofitable. One hundred renminbi of savings are invested but they create something only worth, say, eighty renminbi.
There are many reasons to assume that’s been happening. It is impossible for debt to rise faster than GDP, because the increase in debt should be matched or exceeded by economic growth. That was the case in China before 2007, where you had a rapid rise in debt but the debt-to-GDP ratios were pretty stable. It’s only after that you saw an acceleration in debt and a deceleration in GDP growth.
This cannot happen if you are investing productively. In a capitalist economy, nonproductive investment leads to insolvency. This is what economist János Kornai called a “hard budget constraint.” But in an economy with “soft budget constraints,” you don’t need to write your investment down. You can effectively pretend that investment is productive. In other words, you take one hundred renminbi of resources, you convert it into an investment project that only creates eighty renminbi of value, but you carry it on your books at one hundred renminbi. Now this changes your GDP calculation, because normally, when you take a loss, that’s a reduction in total GDP. And if you don’t take the loss, if you capitalize it, you’ll end up with a higher GDP than you would in a hard budget constrained economy. Hard budget constraints enforce discipline because, over time, nonproductive investment results in bankruptcy that effectively writes down the “fake GDP.”
If China’s GDP growth is much lower than the reported growth, then, it’s not because they’re lying. They’re calculating GDP the way the rest of us do. But this process can’t tell good investments from bad investments. Assume for a moment there are two Chinas that are identical in every way with only one difference. In the first China, you invest nonproductively but somehow you know you are and you immediately write down your investment. In the second Chin you don’t know. These identical Chinas will report very different growth numbers.
Now you might say, well, that’s a Chinese problem. Why should the rest of the world care? The world should care because of how it affects China’s trade account, especially after the property bubble collapsed. Property is one of the three big areas of investment in most economies. And GDP growth is equal to investment growth plus consumption growth plus the growth in the trade surplus or net exports.
If investment in property drops very rapidly — which it did in China — then either total investment must decline, or you must do something to prevent total investment from declining. But if you allow total investment to decline, then the GDP growth rate must decline unless you get explosive growth in your trade surplus or a much more rapid growth in consumption. So what did China choose? It chose to externalize the cost of the property sector collapse by reverting to net exports of manufactured goods. Dollar for dollar, the reduction in property investment was matched by an increase in manufacturing investment.
But China was already over invested in manufacturing and already produced more than it could domestically absorb. So those trading partners that don’t control their trade and capital account saw a contraction in manufacturing, because total manufacturing must equal whatever global demand is.
Michael Pettis
I have never said that. It is not true that American debt has nothing to do with domestic conditions.
The most important reason for why American debt rises has been articulated in a series of papers by economists Atif Mian, Ludwig Straub and Amir Sufi: income inequality and “the saving glut of the rich.” If income inequality goes up, consumption goes down. Because rich people consume a much smaller share of their income. So if you take $100 from a worker and you give it to a billionaire, the worker’s consumption will probably go up by $95 but the billionaire’s consumption will not go up at all. So income inequality raises, not the national savings rate, but the savings of the rich.
Now, how does that affect the United States? If it were a developing country like in the nineteenth century, having lots of rich people would be very good for investment, because there would be huge investment needs and rich people save. That’s what they do. That’s their function in the economy. But American businesses refrain from investing, not because they can’t access savings or because they can’t finance it. Interest rates were zero at one point in the United States, yet they still weren’t increasing investment. The reason they don’t increase investment is because building, say, an automobile plant in the United States makes no sense if it simply cannot compete with subsidized automobiles abroad.
So if the savings of the rich rose, but American investment did not, then savings in some other sector must have fallen. One way in which this happened is that Americans stopped buying American made cars and bought German made cars, and American car companies fired workers. And unemployed workers have a negative savings rate. So, the savings of the rich are matched by “dis-savings” among the ordinary people through a rise in unemployment.
If we don’t want unemployment to go up, we can either loosen monetary conditions, encourage a rise in household debt, or expand the fiscal deficit. Now you notice, this is exactly what Joan Robinson was talking about. She said, if you are not savings-constrained, and if foreign savings flows into your country, unemployment goes up. Or, in our current economy, the fiscal deficit or household debt. The research shows a rise in debt of ordinary Americans.
So I would argue that the American trade deficit adds to a problem that’s probably primarily driven by rising income inequality, in which the only way to keep the economy growing, to keep unemployment from going up, is by encouraging household debt and/or fiscal deficits.
It is not a coincidence that when American income inequality started to surge, so did American trade deficits, household debt, and fiscal debt.
Michael Pettis
There’s nothing special about tariffs. For some reason, academic economists go absolutely crazy when you mention tariffs. But what tariffs do by raising the costs of imports is not just to act as a tax on consumption for some households and also as a subsidy for the production of others. Lots of things do that. If you depreciate your currency, then you’re doing exactly the same thing. If you are in a system like China or Japan, where the banking system does most financing and it’s oriented towards the production side of the economy, lowering interest rates is a much more efficient way of transferring income between households.
The question is what are you trying to do and what’s the most efficient way? What you’re trying to do in the United States and in England, presumably, is to incentivize production. And if you call it a subsidy, everybody panics because you are making it more profitable to produce and more expensive to consume — you’re taxing consumption. Now, many people say that’s terrible. Consumption is what poor people do. You’re hurting them. But consumption depends on production. The way for me to get you to consume more is not to lower the price of imports, but to get you to produce more. Then, whether import prices go up or down, you will consume more. The only way to separate consumption from production is temporarily with increases in debt. So that’s really what you are really doing.
Michael Pettis
I think the only way to implement tariffs is a blanket tariff, like a 20 percent tariff on all imports. When you start doing all of these bilateral tariffs and sectoral tariffs, all you’re doing is redirecting trade flows through other countries, given how supply chains work nowadays. It’s like saying the United States should depreciate the dollar against the renminbi, but not against pound sterling. That would make no sense. If you happen to own renminbi and wanted to buy dollars, you would simply buy pounds and then buy dollars. So capital inflows would continue, the same way trade flows would.
I would argue that the most efficient way to do this is to have a new global Customs Union along the lines that Keynes proposed at Bretton Woods, in which you penalize countries that run deep and persistent imbalances. Keynes’s argument was: I don’t want to tell you how to run your economy. You can be communist, capitalist, fascist, socialist, whatever you like, but whatever your domestic problems are, you cannot externalize them through the trade account. So we will not allow you to create persistent imbalances. That would be the best solution — a new form of globalization.
Many people say this can’t be done: at Bretton Woods everybody hated everybody, and the United States ruled the world. Back then only one country needed to be convinced, now it’s many. But I think that’s a little pessimistic. If you include the United States, Canada, England, and a few other developing countries with deficits, such as Mexico or Colombia, then you’ve got 80 percent of global deficits. If you create a Custom Union that says “if you trade with us, you cannot run surpluses,” that will force the whole world to adjust.
But if you can’t do that, then the next best way is unilaterally. This means that the United States should refuse to continue absorbing global excess savings. How do we do that? We put a tax on inflows, also known as a Tobin tax. [Named after its originator, Nobel Prize–winning economist James Tobin.] A Tobin tax is just a small transactional tax on capital flows. So, if you’re a “hot-money” speculator it can get brutally expensive. But if you’re going to build an automobile factory in Leeds, England, it’s going to take you twenty years to get your investment back. A small tax on inflows will have no impact on your investment. So you can impose a small Tobin tax that discourages short-term capital flows without discouraging productive foreign investment.
Michael Pettis
So there was a proposal by Senator Josh Hawley (R-MO) and Senator Tammy Baldwin (D-WI) a few years ago. Many people said it came from my work. This is not the case. But it was a good idea, and it had bipartisan support.
Regarding the effects on other places: remember that the Eurodollar market [the original terms for the offshore dollar market], which was centered in London, was created because of American taxes on inflows and outflows. So what that would really do is create an offshore dollar market, and, yes, that could be in London, or it could be anywhere else. It doesn’t really matter. The key point is that the United States would no longer play the role of absorber of global savings imbalances. I don’t want to be naive. It was the petrodollars from the OPEC countries, which they could not put into the United States due to these taxes, that created the Eurodollar market. The resulting imbalances did not affect the United States. Rather, it created huge trade deficits in Latin America, which then led to the global debt crisis of the 1980s.
So this proposal has its problems. But the point is we don’t want a world in which real economies must adjust to hot money flows. So anything that reduces them — and I’m saying this as an ex–Wall Street trader — is good for the global economy.