---

Brad Setser’s writings on the international economy are must reading for policy makers, academics, and students of economic policy making. His iconoclastic views on China, trade and foreign currency have been influential with the Obama and Biden administrations. During the Obama administration, Setser was deputy assistant secretary for international economic analysis when the Treasury was pressing China to revalue its currency and shift away from a reliance on exports. In the Biden administration, he was a counselor to U.S. Trade Representative Katherine Tai, who has worked on resolving disputes with allies begun during the Trump administration, so the U.S. could try to create what Biden called a “united front” to challenge China. He is now a senior fellow at the Council on Foreign Relations where he publishes the blog ‘Follow the Money.’ This interview is part of Rules of Engagement, a series by Bob Davis, who covered the U.S.-China relationship at The Wall Street Journal starting in the 1990s. In these interviews, Davis asks current and former U.S. officials and policymakers what went right, what went wrong and what comes next.

Q: When did you start working on China as a research project?

A: It was roughly 2003 or 2004. I was working as a research assistant for Nouriel Roubini [Roubini is a prominent international economist credited with forecasting in 2006 that the U.S. was headed toward a financial crisis.] Nouriel and I went to China for a conference in 2005. China was a really enormous story unfolding over those years. Chinese exports were booming. Chinese reserve growth was exceptional. Japan had stopped intervening [in foreign exchange markets] so China was by far the biggest source of foreign exchange reserve growth, and the biggest intervener in any market. There was a very real debate about whether China’s peg to the dollar was at an appropriate level given all that was happening.

Back in 2008, Treasury Secretary Hank Paulson famously had to assure Vice Premier Wang Qishan that China’s money was safe in the U.S. during the financial crisis. The U.S. concern seemed to be that the Chinese would pull their money out, although if they had pulled their money out, wouldn’t they have taken huge losses?

The concern was twofold. From 2005 to 2008, China was buying far more agency bonds — those issued by government-sponsored entities such as Fannie Mae and Freddie Mac — than they were buying Treasury bonds.

Paulson had a very particular concern that China would shift from agencies to Treasuries, and that that would create a lot of stress in the agency market. So, it was not that China was necessarily dumping all of its U.S. assets. It was that China would move from agencies to safer assets, and in the process add to financial stress. It was a very legitimate concern. [In 2008, the federal government took over the Federal National Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage Corp. (Freddie Mac) after both had become insolvent when the market for their bonds collapsed during the financial crisis.] 

In fact, China did shift its portfolio over time from agency bonds to Treasuries. That shift was made possible, without greater disruption, because the Fed had its first quantitative easing program, QE 1, which was basically buying agencies. [With quantitative easing, the Federal Reserve purchased massive amounts of financial assets to try to stimulate the economy.] The Fed stepped in to maintain orderly markets. But at the time, QE was not an accepted policy tool and there wasn’t the same level of comfort that if China sold so much that it moved the market the Fed would step in.

When I started at the National Economic Council in the late spring of 2009 [during the Obama administration], the concern then was that China was going to keep its currency pegged to the dollar and return to a large build-up of reserves. When the Obama administration came in, it had at least the rhetoric that it wanted a new kind of recovery — one built around investment and exports. It didn’t want a new housing-driven bubble. To have a recovery built on exports, demand from the global market has to flow to you not to China. [China repegged its currency during the crisis.]

So the idea was that the Chinese yuan was undervalued and needed to be revalued? [A stronger yuan would make Chinese exports more expensive and less of a competitive threat to American exporters.]

At the time it was unambiguously true that the Chinese yuan was still undervalued.

How much of China’s amazing growth in the first decade of the 2000s was the result of it having an undervalued currency for several years?

A significant part of China’s boom phase is tied to an undervalued currency. There are estimates that suggest that by 2007 and 2008 the yuan was undervalued by 30 percent. That is the period when China was getting several percentage points of growth out of net exports — and normally net exports contribute nothing to growth.

That was a period when Chinese export growth was like 30 percent year-over-year consistently and global trade was growing at 10 or 15 percent. So China’s trade was growing at two to three times the global rate. That was because the yuan was flat against the dollar until 2006 and the dollar was dropping in value [against other currencies.] In early 2007 and late 2006 , the yuan was weaker against a basket of its currencies than it had been when China joined the WTO. During that period, China’s boom really did come from exports.

The global financial crisis was a shock to that growth model for China. China pivoted to a model where most of its growth came from domestic investment and loose credit/big infrastructure projects, and the build-up of China’s housing stock. For the 10 years after the global financial crisis, exports were a much more modest [contributor to Chinese growth].

Should the Obama administration have done more to push China to revalue its currency?

Unambiguously, the Bush administration under Paulson should have done more. That was the period of time when there was a surge in imports coming into the U.S. from China, and there was a special safeguard provision in the WTO accession agreement to introduce tariffs in the face of a surge in imports. I think it was a mistake not to use that safeguard.

But immediately after the global financial crisis, U.S. imports were coming down, so it was hard to argue then that you needed a safeguard against Chinese imports. When conditions and imports started to pick up, I personally believe we would have been better off had we used that safeguards provision — and tied that to an argument around China’s currency policy.

There’s a separate question about whether you would have called China a currency manipulator. The 1988 statute defined the crime of manipulation ambiguously, with lots of wiggle room, but it didn’t actually specify any remedy.

The idea would have been to make a determination that China’s actions in the foreign exchange market were impeding adjustments in the global balance of payments. That could have been called manipulation. Then, as a result, we would initiate a set of special safeguard provisions dealing with surges in imports that we believed were attributable to the undervalued currency.

During these ‘Rules of Engagement’ interviews, the issue of whether the U.S. should have used the safeguard provision in China’s WTO accession agreement has come up quite a bit. [Under the import surge provision in the WTO agreement, which was in place until 2013, if industries could show that Chinese imports were disrupting markets and threatening industries, the U.S. could impose exceptionally high tariffs to block the imports.] Former U.S. Trade Representative, Charlene Barshefsky, who negotiated the safeguard, definitely thinks they should have been used and faults the Bush administration for not doing so. Bob Zoellick, USTR during the Bush administration. thought that would have been protectionist. 

I have a couple of reactions. Regarding the argument about seeming protectionist, at the time, there was still a sense that it was a great achievement for the U.S. to get China into the WTO system. And you didn’t want to welcome China into the WTO system with a barrage of cases [alleging WTO violations]. There was also the self-image of the policymakers that they were free traders, and they didn’t want to take protectionist actions.

It’s difficult to find WTO cases where there is a specific Chinese action that violates WTO commitments and is also significant enough that if you won the case it would matter [for the broader economic relationship].

For a special safeguard, however, all that you needed was evidence of a surge and some impact on the U.S. That’s a lower standard. If the U.S. government tried, it could have found cases where surges lead to a material impact. If you want to send a signal, you have to take a few of those cases or you have to self-initiate cases [meaning the U.S. would file the surge cases, not wait for companies to come forward].

Back in 2009, Secretary of State Hillary Clinton was in Australia. She said to Kevin Rudd, who was prime minister then and a China expert, ‘how do you deal toughly with your banker?’ meaning China. What kind of vulnerability did the U.S. have?

Secretary Clinton’s statement was a reflection of the Hank Paulson concern about how much China could disrupt the U.S. market by selling agencies. I don’t think China has actually ever developed a true capacity to translate its holdings of Treasuries [into power over the U.S].

China very clearly realized around the time Secretary Clinton made that statement, that it wasn’t getting any leverage over the U.S. by being the United States’ banker in the sense of buying U.S. government bonds. It wanted to become the banker for other countries. That’s when China started to build out the Belt-and-Road Initiative. 

China certainly thought it got leverage when it was acting as a banker providing loans [to developing nations] to build infrastructure. The U.S. ceded way too much space by not being willing to be a development banker anymore. The World Bank stopped growing. It stopped financing a lot of infrastructure. It just wasn’t competitive with the China Export-Import Bank and the China Development Bank (CDB). CDB is as big as the World Bank. Ex-Im is as big as the World Bank. So China has created two World Banks in the past 10 years while the World Bank barely increased its balance sheet.

Now China is unfortunately an overextended banker that is trying to find ways to get his money back from a set of economies that aren’t in a position to pay China back. So, it’s unclear if it worked. 

…China has so many different vehicles for supporting the advancement of the interests of the Chinese state, that it was naive to think that it would only be going through the CIC. 

When you were at the Treasury, you were also looking at sovereign wealth funds [including China Investment Corp., which invests China’s foreign exchange reserves in various projects]. This was then a new phenomenon. The issue was, would they operate in a commercial fashion, or would they be an instrument of state power?

I don’t think anyone can credibly argue that [Saudi Crown Prince] Mohammed bin Salman’s Public Investment Fund is operating at arm’s length from Mohammed bin Salman and the Saudi state. But over time, it has become clear that there are sovereign wealth funds, like Norway’s Norges Bank Investment Management fund, that operate as well-managed investment funds. They take positions, buy portfolio equities and buy into a few real estate funds, but they aren’t actively instruments of the state trying to buy strategically important assets.

There was certainly a concern when China set up the China Investment Corporation, that it was going to be the vehicle that supported the advancement of the interest of the Chinese state. What we’ve learned over time, is that China has so many different vehicles for supporting the advancement of the interests of the Chinese state, that it was naive to think that it would only be going through the CIC. 

In fact, CIC didn’t turn out to be the most significant of those vehicles. Other parts of the state financial system, including funds managed by state banks, definitely played bigger roles. Funds were also set up that were supported by China Development Bank and Export-Import Bank lending.

One area that you have taken the lead on is examining whether China is a golden market for U.S. exports. I read your research to suggest China isn’t the export market that people make it out to be for the U.S.

I’m also making the argument that China is not the export market that people in Europe make it out to be. So, it’s a broader argument.

All arguments need to be appropriately situated and caveated. China has become a very big import market for a host of commodities. China is an enormous export market for Australia for iron ore, and for Australian [natural] gas. It’s an enormous market for Brazilian iron ore and soybeans. And it is an enormous market for American agriculture, soybeans in particular.

It was also a growing market for American commercial aircraft from Boeing. That has now changed enormously because of the combination of the 737 Max’s problems, the pandemic, and the trade war. Boeing is not getting orders from China anymore.

But for other products, China simply has not been a good market for U.S. manufacturers. Part of that is because while China does buy a lot of imported semiconductors, over time, American companies became semiconductor design firms and subcontracted production to TSMC [Taiwan Semiconductor Manufacturing Co.] and other Asian fabs. So, the actual product sold in China was not made in the U.S. The design was made in the U.S. The intellectual property was probably notionally located in a tax haven—probably not taxed in the U.S. — and the products were manufactured by TSMC [and others] and sold to the Chinese for electronics assembly.

But for a whole host of other products, if you wanted to sell in China, you needed to produce in China. If you look at the details of China’s trade — and this was a surprising finding to me — China’s imports of manufactures actually peaked in 2004 as a share of its economy. In China, they split data on manufacturers into processing and non-processing. Processing is typically thought of as parts for re-export. Since 2004, the non-processing imports of manufacturers have come down as a share of China’s GDP.

This is stuff that is imported into China that stays there and is used in China. That particular measure is below where it was when China joined the WTO. So, China just has not empirically been that big a market for imported manufactured goods. 

Once a good establishes a foothold in China’s market, there’s a lot of pressure to produce it in China. Companies that produce in China get an advantage there. There are tariffs you can avoid. And you’re treated by buyers in China more favorably, including state buyers. If you go back and look at some of the things that people were talking about in 2005 – that China was going to become an enormous market for U.S. and European solar panels, for wind turbines — go on and on down the list. It turns out that China set up its own companies. Those companies got privileged access to Chinese power contracts because solar farms feed into the grid. Before you know it, China becomes a big export source of supply for the world and it’s not importing anything.

Across a number of industries that basic story replicates. The U.S. used to sell a lot of construction equipment to China. Now that is made in China, and Caterpillar makes some things in China for sale to the global market. Cummins used to sell a lot of big engines to China. Now it makes a lot of the parts for U.S. engines in China, and it uses its Chinese supply chain to supply the U.S. as well as China. Empirically, Chinese imports of manufactures for its own use are actually quite small [compared to the size of the Chinese economy] and they are shrinking. Chinese imports of manufactures have grown less rapidly than China’s economy, so as a share of GDP, they’re just coming down. That is not the conventional view, but it is a well-documented reality.

What are the policy implications? What should the U.S. do?

If China wants to wall off large parts of its domestic economy from foreign imported goods, at the end of the day China should export less. That would imply a stronger currency. That’s the most efficient way of getting to that outcome. 

The alternative view is the Trump administration’s view of systematically trying to break down all these structural barriers so that China becomes more like a normal country with an open market.

How do you get China to agree to have its currency float or grow stronger?

The complexity now is that the Chinese economy is a little weak, unlike some points in the past. It’s not clear what direction the yuan would move in right now if it were allowed to float. If you want China to have a stronger currency, you need China to do more to support household consumption.

The unique characteristics of the Chinese economy over time have been a low and falling share of manufactured imports. I have put it provocatively by saying China has been de-globalizing since 2004. 

The other characteristic is an enormously high savings rate, an enormously low share of output that is accounted for by household consumption, and a fiscal system that is brutally punitive for low-income workers in the formal sector. There are tax collections from VAT, but almost no taxes on wealth. It’s really quite regressive. [For example], there are way more out-of-pocket costs for medical care than there should be. [Chinese patients] pay at the point of service, and then have to hope they’ll get reimbursement.

…during periods when China is intervening in the market, the U.S. should be a little bit more aggressive in calling out that intervention and pushing back.

In order for China’s currency to appreciate, and for China’s exports to be commensurate with its low level of manufactured imports, you need the household consumption side of the economy to fire up. I also think — and this always gets me into difficulty with my former colleagues at the Treasury — that during periods when China is intervening in the market, the U.S. should be a little bit more aggressive in calling out that intervention and pushing back. That includes being willing to look at the growth in the net foreign asset position of the state commercial banks, and the role China Ex-Im, and CDB have played as backdoor conduits for moving foreign exchange out of China that would otherwise be showing up in reserves.

The state commercial banks have a net foreign asset position of a trillion dollars. The policy banks don’t disclose enough, but from what they do disclose, they have the foreign asset position that is somewhere between half a trillion and a trillion dollars. Those are big sums. You have to think about the role that foreign asset accumulation has played in keeping the yuan weak during periods when the market wanted a stronger yuan.

Explain how this winds up weakening the yuan.

When the Chinese government intervenes to keep the yuan from appreciating, it doesn’t typically intervene to weaken the yuan. It lets the market drive the yuan down and then it resists periods when the market will push the yuan back up. So, it’s always technically resisting appreciation rather than pushing its currency down, but the effect of its action in the market to maintain its currency at an artificially low level. The key is that the government is buying dollars, holding them and keeping its own currency from rising 

For years and years and years, the U.S. has been pushing China to rely more on domestic consumption and it hasn’t gotten very far. How could the U.S. pressure or convince China to do what you’re talking about, particularly when the two sides have just started talking again about economic issues?

I am hawkish in one respect. It is helpful to just be clear that there are limits on the rest of the world’s willingness to absorb yet more Chinese goods. Now, it was complicated over the past few years, because of the supply chain disruptions [during the pandemic]. Shifting global demand towards goods, which was putting upward pressure on goods prices and inflation, meant Chinese supply was welcomed. But at some point, there should be a message that China has been able to grow on the back of exports and that can’t go on. Net exports have contributed over one percentage point to Chinese growth on average over the last three years. That’s a big, big contribution to growth.

I can think of two ways to do this. One is tariffs—really high tariffs. The other is an idea of [Peterson Institute for International Economics founder] Fred Bergsten, who called for the U.S. to intervene in such a way as to counteract Chinese currency intervention to drive up the value of the yuan.

Those are your tools. I have some sympathy for the Bergsten argument around counter- intervention. It is complicated by the fact that China’s financial account isn’t fully open. And if you do counter intervention, you’re putting money into China [by buying Chinese assets to drive up the yuan]. If there were a set of events around Taiwan, your investments in China would be an obvious hostage. So, there are certain risks from doing that. 

But if China is intervening in the market and building up its foreign exchange reserves, counter intervention should be an option. Make it more costly for China to keep intervening.

The difficulty with tariffs is demonstrated by how the U.S. and global economies adapted to the Trump tariffs. We now have non-trivial tariffs on the majority of trade with China.

A couple of things happened with tariffs. A lot of companies just paid them and sometimes passed the cost on [to consumers], and sometimes they did not. The other effect, which is more difficult to get around, is that it’s very easy to take a few screws out [of a product made in China], ship the screws and the other components separately to Vietnam and then do enough work in Vietnam to count as a material transformation, so that the product becomes characterized as a Vietnamese good [and avoids U.S. tariffs].

If you really want to keep Chinese content out by using tariffs, you can’t do that by tariffing China alone. Obviously extending those kinds of tariffs to all trade that has high-embedded Chinese content is an enormous change with lots of material side effects. I personally prefer the more targeted approach.

Targeted in what way?

Focus on pushing back hard on Chinese intervention, focus on the currency. In order to achieve broader adjustment, my view is you have to have a macroeconomic adjustment.

If you look at what has happened in the five years since the U.S. imposed tariffs on China, you find that China’s global goods exports have gone up by a trillion dollars. U.S. tariffs were utterly ineffective at changing the broad structure of China’s economy. They did not make China less export-oriented. It’s actually more export-oriented now than it was before. So, if the goal was to change the overall direction of China and make China into a different kind of economy, [the Trump tariffs] failed. 

They did introduce frictions on trade between the U.S. and China, and some of those frictions are helpful. Having tariffs on auto trade with China insulated the U.S. from China’s EV breakout. With a protected home market, China built up some really efficient EV companies and induced global manufacturers to move to China. They discovered an efficient Chinese electric vehicle supply chain and now China’s auto exports are off to the races.

But if you want to change China’s economy so that it is less of an export powerhouse and less of a draw on global demand for manufacturers, you have to do more than just have tariffs on China.

Do you think that the Trump tariffs should be eliminated or rejiggered?

If I had my choice, there is some rejiggering, and some adjustments that to me make sense. It’s a political judgment about whether you want to do so, but rejiggering rather than elimination, or the extension of all tariffs is the right approach.

Year after year, Treasury secretary after Treasury secretary has said that a prosperous and strong China is in the U.S.’s interests? Do you think it is anymore?

Yes. A China that grows on the back of household consumption, that absorbs inside China the productive capacity of the Chinese economy, and that faces pressure on its currency to appreciate because it has a strong domestic engine of growth is unambiguously in the United States’s economic interest.

It’s more complex than just saying any growth in China is in the United States interest. A China that grows on the back of exports, to me, is not in the United States’s economic interest. A China that can finally grow from an expanding domestic consumption base and from a growing middle class that can afford to buy all the EVs that China clearly can produce is better than a poor China that is producing EVs for the United States.

But if China doesn’t change, then what? You get into talk of decoupling or disengaging. 

There will be a set of products — and that a set of products will grow over time — where through some mix of trade policy and regulation and incentives, the U.S. simply won’t rely on China for supply. So, the scope of trade will be limited. It probably won’t be zero. There is a natural trade for clothes and for soybeans. That may pose some strategic issues for China in depending on the U.S. for food for its swine herd and chickens, but it doesn’t pose any real issues [for the U.S.].

But there will be a growing set of areas like telecommunications equipment, electric vehicles, wind turbines where the U.S. won’t be as reliant on China. If the U.S. is really concerned, eventually it will have to do really difficult and expensive things like build a serious solar photovoltaic industry that doesn’t rely on Chinese polysilicon or the underlying wafers that go into the cells.

And also building a capacity to assemble iPhones outside of China. I don’t think that would actually end up producing higher prices and iPhones because Apple is very effective at charging the maximum it can. But I do think it would reduce Apple’s market cap and its profit margin. Those are the kinds of things that would happen over time.

To put it in the current jargon terms: There’ll be more de-risking.

Is there any difference in your mind between de-risking and decoupling other than sounding less aggressive?

I don’t think we were ever doing decoupling, even with tariffs at 25 percent. Tariffs don’t decouple. Too many people were just paying the tariffs. Too many Chinese components were coming in through imports from Southeast Asia.

De-risking implies that in specific areas, you go well beyond just putting up a tariff. You’re much more serious about developing alternative sources of supply. 

The administration says it is coming close to announcing its review on outward investments. What do they need to watch out for?

A review of outward investment is a better remedy for the underlying problem identified in the Trump 301 investigation, which is technology transfer. Tariffs don’t stop technology transfer. They punish you for engaging in the practice.

…in an era where sadly the U.S. and China are wargaming against each other, and high-end computing power can be used in developing unmanned systems that potentially could change the military balance in key parts of Asia, [technology reviews] are necessary.

A review of outward investment would be an impediment to technology transfer in key sectors. The challenge is identifying what’s defined as an investment. Intel building a top-end semiconductor factory in China would clearly qualify. But what about a U.S. private equity fund taking a 25 percent participation in a Chinese facility to build higher-quality memory chips?

And what part [of the review] involves notification — where the U.S. government has to look at an investment and see if it’s okay? And what part of it is an absolute prohibition? But as a general principle, in an era where sadly the U.S. and China are wargaming against each other, and high-end computing power can be used in developing unmanned systems that potentially could change the military balance in key parts of Asia, [technology reviews] are necessary.

But how does the U.S. do a review without also including Europe, Japan, Korea. Say you’ve stopped Intel investing, but then Samsung invests in the same thing? 

It would be most effective if other countries have similar reviews that are broadly convergent. If you’re in a non-cooperative world, then you would look at the embedded U.S. content that goes into that facility and use export control [to block investments]. But it is always better if you can act in a coordinated way.

---