In a recent article for Foreign Affairs, I argued that we need a serious debate on whether the United States should maintain its commitment to Taiwan. This is a hard question, and reasonable people can disagree. Blake Herzinger’s response in Foreign Policy rejects my analysis. Unfortunately, his harsh critique does little to advance this necessary debate.

First, he fails to address the central pillar of the analysis. I argue that the U.S. commitment to defend Taiwan brings large risks; therefore, the United States should end its commitment to Taiwan based on the logic of retrenchment. Given the value that China places on unification with Taiwan and China’s increasing military capabilities, I believe there is a significant probability that over the next couple of decades China will use force against Taiwan, drawing the United States into a large war. This conventional war could escalate to nuclear war along numerous paths. If losing the conventional war, China might launch a limited nuclear attack to signal its resolve and coerce the United States to concede. Alternatively, if the United States were losing, it might escalate to nuclear attacks because it believed defeat would undermine its East Asian alliances. The dynamics of any outright war, involving large conventional attacks by both countries and the alerting of U.S. and Chinese nuclear forces, could create other pressures and incentives for intentional and accidental nuclear escalation.

I am far from alone in judging the probability of conventional war to be significant. In March of this year, Adm. Philip Davidson, who was then-commander of U.S. Indo-Pacific Command, stated in congressional testimony that “Taiwan is clearly one of their ambitions. … I think the threat is manifest during this decade, in fact in the next six years.” I am not this pessimistic; in the Foreign Affairs article, I wrote the probability was small. (I actually miswrote, saying incorrectly the risk was small.) But if there is, say, a 10 percent chance over the next two or three decades of a large conventional war between the United States and China, we are in a very dangerous situation. Herzinger says nothing about the probability of conventional or nuclear war. Thus, his response provides essentially no basis for judging whether the United States should retain its commitment to Taiwan.

Second, Herzinger incorrectly characterizes my recommendation as “inconsistent with the values that should be central to U.S. foreign policy” and concludes that “realism is not an excuse for callousness.” I believe, as I acknowledged, that retrenchment would be bad for the people of Taiwan and would therefore come at some cost to U.S. humanitarian and ideological values, which include supporting democracies. I also argued, however, that the commitment brings large risks to other U.S. values, most obviously the well-being of U.S. military personnel and U.S. civilians. A large conventional war could result in thousands or evens tens of thousands of American deaths; escalation to nuclear war could increase the toll by orders of magnitude.

Calling to end the U.S. commitment therefore is not inconsistent with U.S. values but instead reflects a policy tradeoff where not all U.S. values and interests can be achieved simultaneously. In international relations, as in many other walks of like, we sometimes must choose between bad situations—not between a good one and a bad one. Herzinger suggests otherwise by failing to address the risks of current U.S. policy.

 

He is also incorrect in attributing my conclusion to an allegiance to the realist school of international relations theory. The overwhelming majority of realists favor retaining the U.S. commitment to Taiwan. Moreover, although I have written extensively about what is termed defensive realism—which argues that the anarchic international system allows for major-power cooperation and peace under a range of conditions—my analysis hardly draws on it. Rather, my assessment reflects the degree of China’s determination to control Taiwan, China’s growing military capabilities against Taiwan, the United States’ much smaller interest in Taiwan, and the combined impact of these factors for deterrence. Deterrence theory, yes; realism, no. All of this said, defensive realism is optimistic that China can rise peacefully, except for the dangers generated by Taiwan.

Third, although Herzinger rejects many of my points about the United States’ ability to defend Japan, he says nothing about the relative risks of defending Taiwan and defending Japan. Even if Chinese control of Taiwan increases China’s military capability against Japan, the United States’ best option could be to accept this reduction in the U.S. ability to defend Japan. Unless U.S. capabilities were reduced quite dramatically—and likely not even in that case—the probability of China attacking Japan would almost certainly be far smaller than the current probability of it attacking Taiwan. To start, there is little doubt that China places much greater value on controlling Taiwan than Japan. Moreover, Japan would be much harder to conquer than Taiwan. Consequently, even if accompanied by some reduction in the United States’ ability to defend Japan, ending the U.S. commitment to Taiwan would be the United States’ best option.

Herzinger says I have simply been repackaging the argument I first made in 2011 without making it more persuasive. In fact, my current argument for retrenchment is significantly different from the argument I made in 2011—when I offered a broad-gauge analysis of why the United States should consider ending its commitment to Taiwan—and the argument in 2015, which made the case for a grand bargain in some detail. Although a grand bargain appeared possible if not likely at the time, China’s more assertive policies since then have made clear that China is unwilling to compromise on its claims in the South China Sea and also that it may be more willing to use force against Taiwan. Recent research on the value China places on controlling the South China Sea reinforces this conclusion. Retrenchment is quite different from a grand bargain, underpinned by its own logic. Analysis of complex international policy issues usually confronts a variety of uncertainties. With time, experience can provide information that narrows these uncertainties and shifts best estimates, and policy conclusions can, as a result, change. The evolution in my analysis—from supporting a grand bargain to calling for retrenchment—reflects newly available information.

The Economist’s May 1 cover story titled “The most dangerous place on Earth” is about Taiwan. This assessment is correct. How the United States should adapt its Taiwan policy to China’s rise deserves much analysis and serious debate. I hope we are only at the beginning of the process.