While Donald Trump has diverted his attention to other matters, here are some questions and answers to bear in mind, when he is back to talking about winning a trade war.

Q. Is there a “China problem” to be dealt with?

A. Yes. This was the theme of a piece I did just before the 2016 election: “China’s Great Leap Backward.” Its argument was that through the decades since the beginning of China’s post-Mao reform movement, in the late 1970s under Deng Xiaoping, two assumptions had guided U.S. policy toward China—but that China’s recent behavior was calling both of them into question.

One assumption was that, over time, China would not necessarily become more “Americanized”  or even “democratic”—neither of which anyone who’d ever lived in China could reasonably think. Instead the assumption was that China would become steadily less of an outlier to the international system, more aligned with international standards in realms ranging from financial and banking standards to environmental protection. Its people would become steadily more exposed to educational systems, pop culture, tourism, everyday liberties in the rest of the world.

As China prospered, it would not necessarily become a liberal democracy—even though all the other rich countries in the world, from Europe to North America to Northeast Asia, happened to function that way. But it would become more rather than less integrated into the world order for which those liberal democracies had set most of the rules. And this would apply in areas from economics to strategic and military interactions.

The other assumption was that, despite the certainty of economic frictions and the larger dislocations that came with China’s emergence as a great power, the businesses relationship between the U.S. and China could be more complementary, and win-win, than it was an inevitable Darwinian fight to the death. And this, in turn, was because the structures of the U.S. and Chinese economies were so different, offering the potential for at least as many constructive relationships as mutually destructive ones.

The contrast with Japan was instructive. Japan had already become a world industrial power by the 1930s. Its rebuilding effort after World War II was, like Germany’s, mainly a matter of re-establishing a world-class business, technological, and educational establishment that had helped power the fascist war effort, was destroyed, and then was redirected into non-military competition.

By the 1980s, Japan had large, technically sophisticated companies with their own research centers and internationally recognized brands. They competed head-to-head with their U.S. counterparts: Toyota versus GM, Toshiba versus IBM, Hitachi versus Intel and Texas Instruments, and on down a long list. In the years since then, despite its political and demographic challenges and supposed economic collapse, Japan has grown steadily richer—its GDP per capita is up 35 percent in the past 20 years, versus about 25 percent for the United States— and it has retained or increased its market share in most of the advanced-technology areas its policy focuses on. Japan’s population was less than half as large as that of the United States (and a tenth as large as China’s), and it was nothing like an all-fronts rival or competitor. But for its biggest firms and their American counterparts, it was a genuine head-to-head, win-or-lose type of showdown in those days.

China had never gone through the industrial and corporate revolution that Japan pulled off in the late 1800s. As it began its interaction with the United States 40 years ago, it was a mainly peasant country striving to get into the age of low-end, low-wage manufacturing. Over the next two decades, that is mainly what it did. (This is a story that I told in my book China Airborne.)

China is vastly richer now than it was a generation ago. Its population now includes hundreds of millions of industrial and service-sector workers, instead of nearly a billion peasants, as it did before. But it still is far short of Japan and South Korea, Western Europe, or North America in “commanding heights” corporations with international brand-name power and technology leadership. The gap is closing—but still there is no Chinese brand name on the world market comparable to Mercedes or Apple, to Samsung or GE, to Oxford or Harvard, to BBC or the New York Times. The Chinese aerospace ministry is trying its best, but it is many years behind either Airbus or Boeing.

Exactly how, when, and whether China closes that gap, and what rules it works by in trying to do so, is (or should be) the heart of the trade friction now. Over the past half dozen years, complaints from Western companies and governments have mounted about the tools the Chinese government is using toward this end: selective regulations, “buy Chinese” policies, industrial espionage carried out either through high-tech hacking means or through contracts that obliged Western companies to “share” technology if they want to enter the Chinese market. This has coincided with Xi Jinping’s tightening grip on China’s political system and parts of civic life in the country.

Are these real problems? Yes. Which leads us to …

Q. Are tariffs a good way to deal with this?

A. No. Or at least not in the way they’re being discussed right now.

Q. Why not? Aren’t you just being anti-Trump?

A. Here’s the problem. What the U.S. is doing now won’t work.

A rhetorically escalating threat-and-counter-threat between the U.S. and China may be emotionally satisfying to those who believe, as Trump has said constantly over the decades, that China is “cheating” and “robbing” the United States.

But if the goal is to get the Chinese government to change its behavior, there is very little reason to think that public threats, lectures, and showdowns will do the trick.

You could go through lots of game-theory reasoning to get to this point—for instance, that while the Chinese economy is in principle more exposed to the pain of a trade war than the U.S. is (because China is poorer, and more dependent on exports), in reality its government faces no mid-term elections, or presidential re-election runs, or public elections at all. Also, it can more easily offset an economic downtown in the short run just by ramping up the infrastructure machine— building another new high-speed rail line someplace in China, building ports and railways through the rest of the world via its “One Belt One Road” program, or in other ways turning on the job spigot for long enough to last out a trade showdown.

And, as Paul Krugman has spelled out in The New York Times, even though manufacturers would adjust in the long run to a shift in international prices (from tariffs and other factors), right at the moment, factories to produce much of what’s made in China don’t exist in the United States. Companies could decide to build them—but that’s a matter of years of planning and construction, and billions of dollars, reliant on the idea that Trump-era tariffs would last.

I have no idea whether the strains and contradictions affecting the Chinese system of government will prove unsustainable within the next few years (as Minxin Pei has argued forcefully) or whether it will veer and manage its way through the next set of crises as it has so many before. But when it comes to withstanding short-term domestic pressures of a trade slowdown—a drop in growth rate in China, versus pressures on farmers, aircraft manufacturers, and ordinary shoppers in America—I’ll bet on the Chinese any time.

Also, if yelling at and threatening a Chinese government were a shrewd way of changing its behavior, presumably it would have worked at some previous time in history. I’m not aware of such examples.

Q. So what does work?

A. Ah, you would ask.

There’s a reason certain problems are considered “hard” problems. Holding down health-care costs. Reducing gun violence, in a country that already has hundreds of millions of firearms and where people have radically different views about gun “rights.” Reducing the influence of misleading information and phony news.

Establishing the right long-term relationship with China is a genuinely hard problem. That was the point of the article I wrote late in 2016: Whoever led the U.S. government, with whatever set of policies on other issues, would have to deal with a China that was too big to ignore, too strong to insult, too valuable a potential partner (on issues ranging from climate change to North Korea) and too threatening a potential adversary to approach with anything other than a well thought-out plan.

The core of such a plan, I said, was trying to shape reality in ways that encouraged and discouraged various forms of Chinese behavior. The much-loathed Trans-Pacific Partnership, or TPP, was one of those possible shaping tools. The TPP was the rare policy that Donald Trump, Hillary Clinton, and Bernie Sanders all opposed, but its idea was to ally the United States with enough other Pacific Rim trading nations to create a rule-based economic reality too great and influential for China to ignore. Long-term industrial strategies, like the one the Clinton administration applied (successfully) in the early 1990s to revive the U.S. info-tech industry relative to Japan, also can have effect—and can include precisely thought-through tariffs as part of their power. Different times and different details require ever-shifting strategies, but they have nothing in common with tweeted-out demands and threats.

It’s been 18 months since I wrote this passage, but I’d say more or less the same thing now:

Lectures and public scolding of China have no record of ever changing its government’s behavior; if anything, they make it worse.

What may work, however, is a strategy one former Western-country ambassador to China described as “shaping reality in a way that makes it unattractive for China to maintain its present course.” The clearest recent example involves the Chinese military’s hacking of U.S. corporate secrets.

A year ago [in 2015], when Xi Jinping visited Washington (just after Pope Francis, who drew more press and crowds), President Obama is widely believed to have informed him that the United States had had enough on this front. Government-on-government spying and hacking? Sure, that’s normal. But governmental spying on foreign companies, to help their domestic rivals, was different.

And if it didn’t stop, the U. S. government would find ways to make life more difficult for Chinese companies. Through use of America’s own formidable tools for cybermeddling? Through impediments to investments? Through shifts in visa policies for influential Chinese families and officials? Obama could leave the means to Xi’s imagination. It wasn’t specific, it wasn’t directly threatening, and it wasn’t public, but Obama’s talk was apparently effective. By most accounts, Chinese military hacking of U.S. corporations has decreased.

The hacking decreased then. (Last fall Wired ran a fascinating account of what changed, and what didn’t.) Public shouting and desk-pounding played absolutely no role in a quiet, coordinated, decisive plan to shape the reality within which the Chinese government made its choices.

If Chinese writers were putting this principle into a traditional chengyu, the four-character phrases that are popular vehicles for homilies and wisdom, they’d presumably work with the four characters meaning voice small, force great. Or, in turn-of-the-20th century American parlance, “speak softly, and carry a big stick.”

The tariff policy so far has been voice big, power small.  Yell loudly, armed with a twig.