Despite the heated rhetoric of the past few days, a trade war between the U.S. and China does not seem imminent. But it may be inevitable.

Almost immediately after the Trump administration announced its plans to impose tariffs on a broad array of Chinese imports, with an eye towards compelling the Chinese government to address intellectual-property theft and other alleged trade abuses, Chinese officials responded by threatening tariffs of their own, shrewdly training their fire on U.S. imports from constituencies crucial to Republican political fortunes. Many observers have thus concluded that we’re on the cusp of a devastating economic confrontation, which will badly damage U.S. interests.

Of course, nothing is set in stone. Just as President Trump dialed back his steel and aluminum tariffs, to unruffle the feathers of various allies, it is likely that something similar will happen this time around, particularly if the Chinese make concessions, as I expect they will. Beijing recognizes that they are more vulnerable to a disruption in trade flows than their U.S. counterparts, thus giving them a strong incentive to moderate their stance. In this instance, the U.S. and China will likely step back from the brink.

In the longer term, though, the interests of the two countries are set to collide, for the simple reason that the Chinese recognize that they find themselves in a profoundly unfavorable position. While Xi Jinping is quite willing to deploy cosmopolitan rhetoric—witness his many paeans to global free trade—it is always in service to Chinese national interests as understood by the leadership of the Chinese Communist Party, which maintains, correctly, that we live in a rivalrous world. Given America’s naval dominance, it would be foolish of Beijing to rule out the possibility that the U.S. might one day subject them to an economic blockade, or worse. This would be true even if the current occupant of the White House communicated in dulcet tones rather than in bellicose late-night tweets. To preserve their autonomy, the Chinese believe it necessary to do everything in their power to substitute for sophisticated imports, whether the world’s already-rich market democracies scream bloody murder about it or not.

Quite often, U.S. politicians portray the Chinese as unscrupulous economic predators, who have managed to game the rules of international trade to grow at America’s expense, usually in connivance with feckless U.S. elites. Trump in particular placed this narrative at the heart of his presidential campaign, as have countless others. And I don’t doubt that there is some truth to it. But it is worth looking at the world through a different lens: Consider the possibility that it is not the Chinese who are the predators; rather, it is the savvy multinational business enterprises that have long taken advantage of China’s relative poverty, and who for decades have leveraged its low-cost workforce and its cooperative government to extract tax and regulatory concessions from their home-country governments, and to keep domestic workers on their toes. There’s some truth to this story, too.

It helps to look to the origins of China’s economic opening. Ultimately, the wrenching decision to surrender centralized control of many aspects of Chinese society was rooted in the fear that as neighboring states grew wealthier, they’d gain in relative power. China would grow more vulnerable to foreign coercion, of the sort seen during the country’s “century of humiliation.” Throughout the 1960s and ‘70s, companies in Japan, South Korea, and Taiwan learned how to sell their wares in the U.S. and other faraway markets. By the time China opened its economy in the late 1970s, containerized shipping was a more mature technology and China was desperately poor. Moreover, the international trading system had evolved, thus giving Beijing less room to maneuver. The Chinese didn’t really have the option of simply mimicking what their neighbors did, at least not at first. Instead of building its own champion exporters, China cautiously opened itself up to foreign companies that wanted to offshore low-level production. And then mid-level production. And then, eventually, pretty much all production, from soup to nuts.

In Playing Our Game, political scientist Edward Steinfeld lucidly describes how offshoring was a two-way street. One of his central observations is that “in the networked world of global production, there inevitably arise lead firms and follower firms, rule makers and rule takers.” In short, multinationals were the rule makers and Chinese firms were, by and large, the rule takers. While multinational firms offshored elements of production to their Chinese counterparts, the Chinese in effect offshored governance of the production process to foreigners—a marked departure from the autarkic ethic of Maoism.

How did offshoring work in practice? Whereas a factory in the U.S. or Germany might have built products using a similar mix of labor and capital, factories in low-wage countries, China very much included, would have done things quite differently. Milton Ezrati offers a useful illustration in his book Thirty Tomorrows. Say a piece of production equipment has been shipped from Germany to China to be used in a Chinese factory. Once assembled in China, the equipment would be “de-engineered” in various ways to disable some of its higher-end automated functions, because an abundance of low-wage labor meant that you could do more things by hand, and to make it easier to use, because low-skill workers would generally be better suited to using a simpler process.

Unlike the world-class Japanese companies that duked it out with American corporate dinosaurs in the 1970s and ‘80s, Chinese companies didn’t really compete with U.S. companies, or not until recently. They’ve been more like sidekicks that help U.S. companies earn higher profits by lowering their costs. By liberating U.S. companies willing to offshore their production from their near-total dependence on the U.S. workforce, Chinese workers helped supercharge the growth of U.S. multinationals, sending their value skyrocketing. Though the U.S. represents a shrinking share of global GDP, globalization has helped U.S. multinationals grow even more dominant than in decades past.

At the same time, however, the Chinese have sought to rise from their subordinate position, and to foster formidable multinationals of their own, with increasing success. The fruit of these efforts can be seen in the rise of China’s consumer internet giants, which have flourished behind the country’s so-called “Great Firewall,” an instrument of mass surveillance and repression that has also proven an effective tool of industrial policy. It has become routine for U.S. technology firms to cede the Chinese market outright to local firms, so cognizant are they of the unlevel playing field. Viewed from Beijing’s perspective, however, allowing the Alphabets and the Facebooks of the world to operate without constraint would be sheer madness: They’d expose their citizenry to external influences, which in turn would threaten the stability of the regime. Having closely studied the collapse of the Soviet bloc, and of various other one-party states around the world, the Chinese government has no intention of surrendering control. Thus far, expectations that rising affluence would lead to calls for liberalization, let alone competitive multiparty politics, have not been borne out, to the surprise and dismay of the U.S. foreign-policy community. Even if the Chinese government were to become more responsive to the mass public, it is not at all obvious that this would entail dismantling its distinctive brand of innovation mercantilism: Indeed, it might lead to its reinforcement, as ordinary Chinese citizens rally around the flag.

China’s desire to strengthen its indigenous enterprises is not limited to the consumer internet sector. The “Made in China 2025” initiative, for example, is all about establishing Chinese dominance in a number of emerging industrial technologies. The country’s “one belt, one road” (OBOR) effort can be understood in a similar light: as a means of shifting Chinese private investment from the U.S. to various Eurasian states that are more susceptible to Chinese influence, and that will evolve into receptive markets for Chinese goods and service. It is a very 19th-century vision, and there is no guarantee that it will succeed. But there is also no denying its logic. If China sees the U.S. as its most formidable rival, why exactly would it continue to rely on Boeing when it has the scale to manufacture its own sophisticated aircraft, which might one day be deployed against the U.S. and its allies?

I take no pleasure in the thought of a more confrontational relationship between the U.S. and China, and I certainly hope it can be avoided. The only way to do that, I suspect, would be for the Chinese government to drastically change its policies by promoting increased consumption among Chinese consumers, thus reducing its domestic savings and easing global imbalances, a prescription championed by the veteran China analyst Michael Pettis. Under the current dispensation, state-owned enterprises and the export sector are heavily subsidized at the expense of Chinese consumers. By cutting taxes on Chinese households, and strengthening the country’s threadbare safety net, the government would greatly improve living standards among ordinary Chinese. And indirectly, a consumption-led approach would increase China’s appetite for imported goods and services, thus addressing some of the grievances of its trading partners. At times, Beijing seems to be moving in this direction, albeit in fits and starts. Yet moving too quickly might cause fissures within the Chinese Communist Party, as the economic interests that benefit from the status quo remain inordinately powerful. Don’t expect an economic perestroika anytime soon.

It’s more likely that the symbiotic relationship between the U.S. and China, in which U.S. firms rely heavily on Chinese intermediate inputs and vice versa, will unravel in slow motion. Automation will, over time, offer U.S. multinationals an alternative to China-centric global production networks. And perhaps Trump’s America First agenda will give way to an Americas First agenda, in which U.S. dependence on Chinese manufacturing prowess is supplanted by deeper integration with neighboring economies, not unlike the way German industrial firms are enmeshed with suppliers in central and eastern Europe. All this may sound fanciful. But China only entered the World Trade Organization in 2001; our mutual entanglement is not quite old enough to vote. The forces pulling the U.S. and China apart are more powerful than those keeping them together.