Rethinking the opening with China. By Matthew Yglesias

Read time: 11 minutes

Rethinking the opening with China

Mistakes were made! W was the worst!!

I’m sort of over cold weather at this point and looking forward to some proper spring, and I’m not even someone who minds the cold all that much!

Some good news this week includes an ambitious rezoning in Burlington, VT; Elon Musk endorsing One Billion Americans, congestion pricing is (finally) happening in New York, and the Biden administration is going to re-open a nuclear plant in Michigan.

My post earlier this week on state-level climate targets wasn’t really about backup power for data centers — the point was objectors were using them as a pretext. But this comment from Eric on the actual subject of backup power as interesting. I love to learn from readers!

It's worth noting that modern data center don't actually need backup generators as often as one might think. The multinational companies that host them have now gotten so big that, if one particular data center loses power, they can simply route traffic to another data center somewhere else. The likelihood of dozens of data centers in different geographical areas, connected to different power grids, all losing power at the same time is essentially zero. And, for the few minutes it takes to shut down the servers cleanly to avoid data loss, they can just run everything off batteries (such batteries cannot be replaced by backup generators anyway, as large generators take several minutes to start up, compared to battery response time, which is almost instant).

On to the main event:

Lost Future: Was the US opening China up to trade a huge mistake? If you could go back in time to the 90s and the Clinton Administration would heed your wise council, would you tell them not to let China join the WTO, grant MFN status to them, etc.? There was a lot of optimism that increased trade & capitalism in general would sand the rough edges off of the CCP, obviously that looks very naive now. Would leaving them in poverty be a good or wise US policy? Would it even have worked? Maybe Europe would've simply traded with China sans America.

On one level, I think obviously yes, mistakes were made.

For starters, it’s pretty clear that policymakers underestimated how big of a deal China’s entry into the WTO was going to be. That doesn’t, on its own, mean that it was a mistake. But an important component of the analysis was wrong, and other things followed downstream from that. Beyond that, I think the early proponents of deeper engagement with China sincerely believed that commercial ties would lead to political liberalization in China and a more peaceful relationship with the United States. Or more to the point, they believed that if those optimistic forecasts didn’t come to fruition, it would probably be because the commercial ties themselves ended up not being that deep. The analysis was, in retrospect, riddled with errors, and clearly if everyone had understood the situation better, they would have made different choices.

That said, taking the question literally and traveling back in time 25 years, I’m not sure that convincing Bill Clinton that Permanent Normal Trade Relations was a mistake would have accomplished very much.

The anti-PNTR argument was already being made at the time, mostly by protectionism-oriented labor union groups, and most Democrats in congress voted against Clinton on that issue. It’s easy to imagine a world where you bring him a vision from the future that convinces him to side with the unions, but then George W. Bush comes along and signs the PNTR bill in February 2001 instead of September 2000. That hardly alters the trajectory of world history.

To my way of thinking, the really big mistakes come not when Clinton signs the PNTR bill, but a year later when al-Qaeda knocks down the Twin Towers.

That set the United States off on more than a decade of Middle Eastern adventures that left the country much worse off and largely stopped us from updating on whether the Clinton-era assumptions about the geopolitical implications of trade with China were correct. But I would also note that while, under Donald Trump and Joe Biden, we’ve largely ditched those incorrect assumptions in favor of a more skeptical view of China, what we’ve actually done on the trade front is toss the economics textbook out the window and open the door to totally unprincipled protectionism. Trump raising taxes on Korean washing machines was not a smart strategic move against China. Biden has been more restrained than Trump, but similarly, when he set out to impose tariffs on metal used to make food cans, he hit Canada, Germany, and South Korea along with China.

Conversely, in his second term, Obama got really serious about the challenge of China and started pushing ambitious initiatives like the Trans-Pacific Partnership that were meant to address it. But the real shift in congress that happened between Clinton and Obama wasn’t a new view of China, it was a new view of trade, so suddenly people didn’t want to deepen commercial ties with Australia and Japan.

All of which is to say that rather than one mistake — we opened up trade with China and we shouldn’t have — I think there has been a rolling series of mistakes that continue to this day.

The China shock

Something that I think is underrated in this debate is that many analysts who looked at the China trade issue in the 1990s didn’t think that World Trade Organization accession would make a huge different to American manufacturing. The basic reality was that American tariffs on Chinese imports were generally low already. The legislation under consideration was to give them “most-favored nation” status or “Permanent Normal Trade Relations” (PNTR), which would smooth the way for Chinese entry into the WTO. But on the US side, this overwhelmingly meant ratifying the trade status quo as a stable diplomatic commitment. There was no lowering of gigantic American tariff walls, because no gigantic American tariff walls were in place.

In fact, because the United States played such a large role in shaping the WTO, membership largely amounted to China needing to align its policies in areas like intellectual property protection and pharmaceuticals more closely with American policy.

So it’s not that the Clinton administration didn’t realize Chinese import competition was a thing. But they didn’t think PNTR would have an especially large quantitative impact on it, especially not relative to its impact on American exports.

A lot of people cite David Autor’s research on the China shock as having changed their minds about US-China trade. But what really shifted my mind was work from Justin Pierce and Peter Schott on this question of how big a difference PNTR made. They show that the switch to PNTR did make a big difference, that the rise in Chinese exports wasn’t just due to improving Chinese policy and productivity. And they say that’s probably because the permanence really did matter for investment flows. Prior to PNTR, yes, companies would buy stuff that was made in China because it was cheap. But after PNTR, companies would deliberately invest in shifting their supply chains to China, and they didn’t just wait to see if a Chinese option became available. If you think about Apple’s Chinese sourcing, for example, they’re not just buying stuff off the shelf that Chinese companies come up with. They are constantly working to push the envelope on components and materials, and they are working with Chinese suppliers to do so. And so were tons of other companies across tons of other industries. The permanence was game-changing.

Then you get to Autor’s research with David Dorn and Gordon Hanson.

Everyone always said that trade creates “winners and losers.” But the standard pattern of post-WWII trade relationships had been very small wins for the winner and very small numbers of losers. But China is really, really big. So instead of one or two industries shrinking in a way that allowed a few other industries to grow, a huge share of manufacturing industries underwent simultaneous decline. As a result, American manufacturing shifted from long-term stagnation since the mid-1960s into sudden precipitous decline.

They do confirm in the same paper that the flip side of the China Shock being large is that the benefits of trade with China were also large. If you happened to live in a metro area that was highly exposed to import competition from China, the consequences for you were quite bad. But if you lived in a rural area with a farming or natural resource economy, you got a lot of upside. Same if you lived in a high-tech, finance, entertainment, or energy hub. The shock was really bad for Ohio, but good for Boston and Dallas and San Francisco. It was just surprisingly big.

Things fall apart

The fact that “trade has winners and losers” was not new information. “Trade has big winners and big losers” means that trade is a bigger deal than you might think, but it’s not on its own a game-changer for how I think about the issue. In pure economic terms, the problem with the opening to China isn’t that the consequences were surprisingly large — Autor, Dorn, and Hanson confirm that the consequences were positive on average. The fact that the magnitude of the changes was bigger than expected doesn’t mean it was a mistake.

But then there’s the question of what do you do about it.

US-China trade in the Bush years created the opportunity for the federal government to increase the budget deficit significantly, without creating big problems in terms of inflation and interest rates. That could have been an opportunity to create a federal child allowance that boosted fertility while dramatically cutting child poverty, plus investing in a huge infrastructure program to re-employ blue collar workers and hopefully create some useful things of lasting value. Instead, we got $8 trillion in regressive tax cuts and a very expensive Global War on Terror.

The really striking thing about that GWOT spending, is we now see that it’s left the United States with very limited ability to manufacture ammunition or warships. It managed to be extravagant defense outlays without a meaningful buildup of military capabilities.

I sometimes feel a little cringe about my big idea about American politics being “George W Bush was a really bad president.” But he really was bad. Not because he’s a uniquely awful person, per se, but because the circumstances of 9/11 gave him some unusually large degrees of freedom in terms of how to use his authority, and he mostly1 used it badly.

Because this is my thing, I would also note that this is the period when the bad land use trends that started in the 1970s and 1980s really started to be macroeconomically significant. A lot of the metro areas that benefitted from the China shock were metro areas with tight zoning, so instead of experiencing rapid population growth and broad-based prosperity, they experienced a run-up in housing prices. So tech growth in the Bay Area wound up being great for tech workers and great for legacy landowners, but it didn’t generate the kind of soaring number of jobs in construction (and construction-related manufacturing) or rising wages for service workers that it should have. This piece of the puzzle was not Bush’s fault, but it didn’t help.

Chinese democracy

Beyond the economics of trade with China, the other thing that happened during the Bush years was the success of the “Golden Shield Project.” Nineties policymakers had hoped that China (and other countries) would need to choose between prosperity and authoritarianism, because prosperity required modern information technology and modern information technology meant embracing the open internet. As early as 2002-2003, China was showing that this was wrong as a technical matter, and there was nothing stopping them from creating a heavily censored but highly functional internet.

There’s no great shame in making an erroneous forecast, but the Clinton-era optimism about commerce and liberalization showed signs of trouble really early.

The Bush administration didn’t really update based on this new information, and the Obama administration didn’t either at first. But the issue here was actually quite profound. China was, obviously, growing very rapidly — enjoying the kind of “catch-up” growth that we’d seen previously in Asian countries like Japan, Korea, and Singapore. But unlike Japan and Korea, China was not adopting a congenial set of politics and values. And unlike Singapore, China isn’t tiny. Very rapid economic growth in a country of one billion people governed by a hostile ideological system is cause for alarm, and I think American policy is still under-alarmed by it.

A lot of people credit Trump with turning around the bipartisan consensus on this, but I think Trump’s policymaking was actually incredibly scattered and dangerous.

The president who did the best job of taking the China problem seriously was, for my money, second-term Obama. If you put together the whole picture of the nuclear deal with Iran, the diplomatic opening to Cuba, the (failed) reset with Russia, the (semi-successful) effort to stay out of the Syrian Civil War, and the Trans Pacific Partnership, you had an administration that was trying to set aside irrelevant old beefs and focus on the big new challenge of the day. In the case of Russia, the problem was that it takes two parties to quash a beef and Moscow fundamentally turned out not to be interested. But with Cuba and Iran, the problems were more on the American side — key domestic constituencies did not accept the idea that their specific hangups should be set aside in order to pursue a more strategic foreign policy. And the same thing happened on TPP. The critical domestic constituency for getting tough on China, trade-wise, just favored generalized protectionism. The idea of forming an anti-Chinese trade bloc had zero appeal to them.

And that’s really where we were under Trump and even more so with his proposals for a second term. The plan is blanket tariffs to partially offset the cost of a regressive tax hike, not any kind of strategic trade policy. Biden is much better than Trump in terms of thinking things through in a sane way, but still, I think, worse than Obama — a consequence of his fondness for everything bagels and mild skepticism of conventional economics.

So while I think most people today believe the opening to China was a mistake and that we’ve learned our lessons, my view is that we are basically still making the same mistake, just in new and different ways.