With China’s reticence to support oil embargoes against North Korea, public attention has coalesced around a pivotal question: Will the Trump administration initiate a trade war against China? In his speech at the Asia-Pacific Economic Cooperation forum in November, President Trump stated “The United States will no longer turn a blind eye to violations, cheating or economic aggression. Those days are over.”

The stakes are high. Regardless of the metric of interest, the US is closely tied to China with China surpassing Canada as the US’s largest trading partner in 2015. Year-to-date total bilateral trade also reached $342 billion by July 2017, a 9% increase over the same period in 2016. From an investment standpoint, the stock of Chinese foreign direct investment (FDI) in the United States reached as high as $110.1 billion in 2016, whereas the US has invested $142.6 billion according to the Rhodium Group. To borrow the Chinese idiom, the American and Chinese economies are as close as lips and teeth, and the high-stakes nature of the issue has led to considerable consternation over the future of bilateral relations.

As many analysts and pundits have come to realize, efforts to predict the direction of US policy under the current administration are often thwarted. President Trump’s views on the bilateral relationship have oscillated between campaign statements that China is “raping” the United States in trade and calling President Xi a “terrific guy” for whom he holds “great respect,” mere hours after the death of Chinese social justice activist Liu Xiaobo.

However, the flipside of this unpredictability is that the President demonstrates a willingness to retract or not act upon a number of ill-advised statements. For example, the US has yet to withdraw from the North American Free Trade Agreement, and the President has not yet delivered on his promise to show Pyongyang “fire and fury” on an unprecedented scale. In each of these cases and numerous others, he seems to have realized that reneging on international agreements or risking the ire of our allies does not serve US strategic or economic interests.

We must then look past the rhetoric and consider a fundamental question: Would a trade war serve the interests of the United States, or otherwise help to consolidate President Trump’s own political support? Though it is inherently difficult to predict the long-term consequences of a trade war, the US seems most likely to come out on top as China exhibits an asymmetric dependence on exports for economic growth. As Eurasia Group president Ian Bremmer asserts, China relies on the international economy for tens of millions of jobs and trade itself accounts for almost 40% of China’s GDP. In the US, the figure is less than 30%. Particularly in an environment where much of the CCP’s legitimacy rests heavily on its ability to drive economic growth, a trade war could spell disaster for Beijing.

Yet this does not imply that a trade war would be beneficial to the US. Rather, the interdependent nature of the economic relationship between the two countries seems to preclude this possibility. The extent of damage would depend largely on which measures are pursued, but experts largely agree that the US would not escape unscathed. A pre-election quantitative analysis conducted by the Peterson Institute of International Economics found that if President Trump implemented the full slate of protectionist measures proposed during his campaign, the US economy would fall into a recession: 4 million jobs would disappear and the effects would resonate throughout every sector of the economy. Though the report did not focus solely on China (the listed numbers include the effects of a trade war with Mexico), it underlines the inherent danger of any extreme measures.

Simply put, an all-out trade war would not serve President Trump’s own best interests. Any sweeping measures would result in economic stagnation and severely undermine his domestic political position, which is already under siege. Short-term political gains from protectionist policies would be far outweighed by the long-term economic damage. Such a development would pose a serious challenge for any president, but this is especially true for one who staked his public persona on the image of a successful businessman. Moreover, as Bremmer asserts, “China’s leaders can read an electoral map,” and will exert pressure on companies and voters in states most crucial to Trump’s re-election.

Still, many indicators point to widespread domestic support for some form of economic measures against Beijing. There is an increasing consensus in Washington that trade policy with China has been inadequate and that the US must pursue a more aggressive stance. Meanwhile, the business community, which has long served as a source of ballast helping to crest the most dangerous waves in bilateral relations, has exhibited “a clear trend of declining optimism” due to perceptions of unfair competition and an increasingly negative investment climate.

Thus if the political calculus decreases the likelihood of an all-out trade war, President Trump may instead choose a middle ground—what one may call “trade skirmishes.” This path would entail limited yet high-publicity measures that are hawkish in appearance but comparatively moderate in overall effect. These policies might include continued expansion of targeted tariffs against banks and entities which help North Korea to engage in illicit activities and launder money for its nuclear weapons program, or an expansion of restrictions on foreign investment in artificial intelligence and biotechnology through the Committee on Foreign Investment in the United States (CFIUS). Executed with appropriate caution, these measures could increase pressure against Beijing without isolating the United States.

There is also considerable support for retaliations against Beijing’s intellectual property (IP) violations. Analysis from both the National Bureau of Asian Research and the Office of the Director of National Intelligence assert that Chinese IP theft costs the US economy at least $400 billion per year. In order to combat these violations, the administration initiated an investigation under Section 301 of the Trade Act of 1974, a law which enables the United States to unilaterally raise import duties against trade partners who engage in unfair practices.

Yet President Trump would be well advised to avoid sanctions through this mechanism, as some experts view Section 301 as a unilateral “cudgel” which would undermine the legitimacy of the World Trade Organization (WTO). The EU, Japan, Germany, and Canada have all expressed concern over China’s IP violations, so rather than engaging in unilateral action, the administration should pursue recourse through the international system. Similarly, anti-dumping steel tariffs under Section 232 are too broad and risk harming US allies more than Beijing.

By eschewing unilateral action and pursuing a more moderate course, President Trump would be able to walk back his more extreme protectionist rhetoric and have an increased chance of effecting long-term change in Beijing’s behavior. In response, China would likely implement similar high publicity retaliatory measures, but President Xi would likely choose to maintain a balance preventing the outbreak of a full-blown trade war. Though Xi exceeded expectations in his ability to consolidate power at the 19th Party Congress in October, a trade war with the United States could deal a devastating blow to his own efforts to transform the Chinese economy, cripple growth, and potentially even jeopardize regime stability.

President Trump may well seek to appease his base by implementing highly publicized measures like bringing cases against China at the WTO for intellectual property rights violations or by reforming CFIUS. He should not, however, follow through on unilateral tariffs, risking an international trade war that will only serve to undermine his own interests.

Brian Bumpas is a second-year graduate student at Georgetown University’s School of Foreign Service. Contact him at bumpasb@gmail.com.