Taiwan’s government agencies are battered by 5 million cyberattacks every day. China is holding invasion drills at a replica of Taiwan’s presidential palace in Inner Mongolia. Last week, the PLA openly rehearsed an encirclement of Taiwan in so-called “punishment drills.”
What happened to deterrence in the Taiwan Strait? Can the status quo be saved?
To discuss strategies for avoiding WWIII, ChinaTalk interviewed Jared McKinney of the Air War College and Peter Harris of Colorado State University, who recently co-authored a monograph entitled, “Deterrence Gap: Avoiding War in the Taiwan Strait.”
Co-hosting today is ChinaTalk editor Nicholas Welch.
We discuss…
Why deterrence is decaying;
The difference between constraints and restraints in a deterrence strategy;
Whether symbolic solidarity with Taiwan does more harm than good;
The values and costs of strategic ambiguity;
Indicators of a successful deterrence architecture, as well as falsifiable and objective measures to test hypotheses in political science;
… and more!
Collapsing Constraints and Incentives to Invade
Jordan Schneider: Can you start by explaining the history of Taiwan deterrence, and why you think it’s faltering now?
Jared McKinney: The mood in Taiwan, with a few false invasion warnings aside, seems to be one of the status quo. Sometimes, Taiwanese people will say that Taiwan has been under threat from the PRC since the late 1940s. They think, “If we’ve had scares for the past seventy years, why should we be unusually afraid today?”
They understand that the PRC is a threat, but it’s been a threat for a long time.
Part of this is true and part of it is not true. Contrary to many people’s interpretations of the past, Taiwan was in a pretty secure position from the late 1950s to the early 2000s. In fact, from the mid-1950s to about 2004, a PRC invasion of Taiwan could not have happened. It would have been sunk in the strait. There would have been the proverbial million-man swim, and Taiwan would remain de facto independent.
This secure position was enabled by a combination of Taiwanese and American military power. People tend to talk a lot about American military power, and of course that’s a real factor — but they tend to minimize Taiwanese military power during this period, which is a big mistake.
In terms of American military power — after the Korean War, we became de facto engaged in the defense of Taiwan from a communist invasion, and then de jure or formally engaged in the defense of Taiwan from a Chinese invasion with the 1954 Mutual Defense Treaty onward. As a result of the de facto and then de jure commitment, we interposed US naval power regularly in the Taiwan Strait in a dedicated unit called the Taiwan Strait Patrol. We station tens of thousands of US soldiers in Taiwan.
Later on in the Cold War, we actually put tactical nuclear weapons there — a fact that is now declassified. We had a formal commitment. We backed up the commitment. And then there were two crises in which China signaled some aggressive intent in 1954 and 1958. Both times, we threatened nuclear war in response to PRC saber-rattling. Nuclear war was not something China was interested in, especially since China didn’t have nuclear weapons until 1964. Those were some pretty robust deterrents, even if there were some scares in this period.
Usually, when people talk about preventing war, they discuss deterrence, but we think that it’s actually more useful to think about deterrents.
Taiwan’s deterrents played a significant role during this period — for example, Taiwan’s air power.
Guess who got the first air-to-air kills in the history of air combat using precision-guided weapons? It wasn’t the United States. It was the Republic of China Air Force in 1958, which fired Sidewinder missiles to shoot down PLA fighters in the strait. They achieved kill ratios of about 30:4.
The PLA wasn’t competitive because they were firing guns and trying to do dogfights. They were being shot down by the seeking missiles that the US had transferred or sold to Taiwan.
The Republic of China Air Force at the time was awesome. Maybe it could have even stopped an invading Chinese force on its own — but it was never on its own because we had multiple interlocking deterrents at the time.
Even though people like Mao were borderline crazy, deterrence was basically assured.
Nicholas Welch: One thing I liked about your monograph was the bifurcation of deterrence into constraints versus restraints. Can you talk about what that adds to our understanding of deterrence?
Jared McKinney: In very simple terms, a constraint is something prohibiting you from doing something or limiting your ability to do it — while a restraint is something internal to your own mind, prohibiting or limiting you from doing something.
Supposedly, my car has the capability to drive 120 miles an hour. But there are restraints preventing me from driving that fast: I don’t want to hurt anyone, I don’t want to kill myself, I don’t want to go to jail.
In the context of deterrence, there is actually similar logic in which people sometimes choose not to do things even though they have the capability to do so. If you have a holistic deterrence equation, you need to look at not only imposing constraints, but also creating incentives for acting with restraint.
Nicholas Welch: When looking at the interlocking constraints and restraints, it’s very striking how in the twentieth century, the US and Taiwan relied almost exclusively on constraints like US naval dominance, nuclear deterrence, and Taiwanese air superiority.
Then, at the turn of the century, most of those constraints disappeared, and now we almost exclusively rely on restraints like China’s positive expectations about trade with Taiwan and the US, or Taiwan’s semiconductor manufacturing capabilities.
Can you talk a little bit about why the twentieth century was so constraint-heavy and what’s changed going into the twenty-first century?
Jared McKinney: Before the rapprochement with China in the 1970s, there weren’t a lot of positive incentives in the relationship. It was hard-nosed deterrence focused on constraints.
However, once we go through this process of rapprochement in the 1970s, the potential for positive incentives is created. This is true for first the United States and then for Taiwan itself, which, beginning in the 1990s, begins a relationship with the mainland that deepens and widens — starting with sending mail back and forth, and then widening out to include tour groups and trade. Then, a relationship of technological interdependence emerged in the 2000s, centered on Taiwan’s silicon industry.
Peter Harris: Leaders in Taiwan have also made policy mistakes. They should have invested more in the military in the 1990s and 2000s to maintain that external lock. It was a mistake to rely upon self-restraint on behalf of Chinese leaders.
Because you never know. The difficulty with deterrence is that it’s really hard to find observable implications that deterrence is working. You never really know what is deterring an adversary from pursuing an undesirable action.
It’s prudent to make sure that there are as many layers of deterrents as possible. Even if you think, or you hope, that economic interdependence or Chinese concern for their reputation or something is working to restrain them, it’s still prudent to maintain those external locks.
Jared McKinney: In the mid-1990s, Taiwan was spending 5%+ of its GDP on defense. Then that number fell to below 2% until recently. Then in the last two years, it’s gone a bit above 2%, maybe heading in the direction of 2.5%.
That’s the right direction, but there’s a lag time between investment and actually procuring military power.
Jordan Schneider: It’s certainly easier to spend money on a military when you’re a military dictatorship.
Jordan Schneider: Figuring out what drives people’s decisions is a fascinating and impossible question.
Can you explain the trends that you believe are the most worrisome regarding deterrence today?
Jared McKinney: A big problem with the current discussion on Taiwan is that we don't know what’s in Xi Jinping's head. We don’t know how exactly China interprets many trends. We don’t know if we’re reading information about the PRC correctly. It is a mistake, I think, to approach the topic with too much confidence.
We want to push people to analyze ideas in a way in which they can be falsified.
As part of our argument, we lay out at the end of the book how we could be proven wrong. We’re watching for that evidence as it emerges. We actually have nine indicators, and we would be willing to reverse our position if the majority of them prove false.
This doesn’t mean that we’re going to be able to establish scientific certainty. But I think that this approach moves us away from just fear-mongering.
The problem with fear-mongering is you can cry wolf only so many times before everyone stops believing you. We believe that the constraints that externally bound the PRC have mostly decayed.
The term “decay” shows that deterrence is a temporal thing. You can establish it for a while, but then an adversary can try to circumvent what you have threatened them with. Deterrence decay has actually happened in a sort of quantifiable way.
Taiwan’s Air Force was superior to China’s until 2004. Today, Taiwan still has more or less the same air force, though the F-16s have been updated a bit.
In contrast, the PLA has fifth-generation fighters, advanced air defense systems, advanced naval SAMs, KJ 500s, and flying airborne radars, all of which pose a lethal threat to fourth-generation fighters.
That constraint has decayed. It’s been circumvented. The same happened with American military power.
There used to be a consensus that if China invaded Taiwan, the US would just send in the aircraft carriers with F-18s.
Now, the PRC has circumvented the default American response. Starting in the 1990s, the PRC developed and eventually deployed an anti-access area denial capability that’s based on missiles which can kill aircraft carriers.
We’re in a position today where naval-borne air power is no longer relevant as a deterrent.
If constraints have mostly disappeared, then what about restraints?
The restraints are those elements that could incentivize the PRC to maintain the status quo. For example, Revolutionary chaos in China used to be a powerful constraint. During the Cultural Revolution, people talked to Mao — Nixon, Kissinger wanted to know whether China was planning to invade Taiwan. Mao essentially says, “There’s a bunch of counter-revolutionaries there. Why would we want to invade? We have enough counter-revolutionaries at home.”
In Deng Xiaoping’s 1984 “One Country, Two Systems” speech, he says, “It could take 100 years or it could take a thousand years, but peaceful unification is the way forward for Taiwan.”
A thousand-year timeline gives the Communist Party a lot of time to act with restraint — but that’s gone now. The PRC’s third historical resolution from 2021 changed that timeframe to 100 years, starting from 1949.
It literally says that the objective is to complete the complete reunification of the Chinese nation by the 100th anniversary of the founding of the PRC. There’s a cognitive shift from broad acceptance of a very long timeline for peaceful unification, towards a deadline that is only a few decades away.
These time pressures are concerning in that they indicate a certain inevitability about forceful unification that wasn’t present before.
Ambiguous Strategery, Symbolic Solidarity, and the End (?) of One China
Nicholas Welch: You identify that, since the turn of the century, there have been four restraints in particular which have been trending in the wrong direction. One of those indicators is the One China policy. Can you explain why One China is important for deterrence?
Peter Harris: In order for the PRC to opt against invasion, the status quo needs to be barely tolerable. There’s always a cost of restraint as well as a cost of war.
For a long time, the status quo across the Taiwan Strait has been more or less stable because of this One China discourse. Obviously, there’s a wide range of disagreement between the United States, Taiwan, and China over what the phrase “One China” means.
But this constellation of understandings about “One China” created a kind of middle ground, where leaders in all three polities could agree for a long time that the political status quo was tolerable.
From the PRC’s perspective, what that meant was there was no international move toward recognizing Taiwan as an independent and sovereign state, especially not in Washington, DC. Thus, to forego the option of invading Taiwan was not tantamount to de jure losing Taiwan.
This was based on a theoretical possibility that everyone was at least nominally open to the possibility of peaceful unification in the future. We go back and forth about how realistic that has truly been.
But we think that has changed.
One reason it’s changed is because of changing Taiwanese domestic politics, which the United States cannot control.
But more importantly for our monograph, there’s a perception in Beijing — not an unreasonable perception — that America’s One China discourse might be shifting. That gives cause for concern in Beijing. It creates an incentive for Chinese leaders to start thinking about a resolution to the problem. If the status quo is not stable, what are we moving toward?
They fear that the status quo is trending toward a future where an American leader brings Taiwan back under the American security umbrella. A “Two Chinas” policy or a “One China, One Taiwan” policy is not tolerable for Beijing.
Nicholas Welch: You mention “27 Firsts” that American leaders have done with regard to Taiwan.
For example, Trump picked up a call from Tsai Ing-wen before he took office. John Bolton met the secretary-general of Taiwan’s National Security Council. Taiwanese admiral Lee Hsi-min 李喜明 came to the White House in 2019.
You argue that these things obviously piss off the CCP and make it less inclined to believe the West’s commitment to a one-China principle.
What would you say to the argument that those “firsts” are beneficial? Is it possible that they strengthened the resolve of the Taiwanese people to resist invasion?
Jared McKinney: The resolve to fight for freedom is internally derived and not externally given.
America can’t breathe life into Taiwanese de facto independence. That’s something that Taiwan and its people are going to have to believe in.
Peter Harris: Even if we assumed for a moment that that would stiffen the resolve of Taiwanese people, it would come at the cost of convincing Chinese leaders that the status quo had already been overturned.
The better approach would be to find ways to stiffen Taiwanese resolve to convince China of Taiwan’s commitment.
Jared McKinney: We need substance, not symbolism. The truth is that most of these interactions are almost purely symbolic.
The problem with symbolism is that it antagonizes the PRC. It pushes against the established framework, but does nothing to actually contribute to Taiwan's ability to defend itself.
Jordan Schneider: The two big variables here are Taiwan’s willingness to fight and America’s willingness to show up. Biden has said concretely that if China invades, then the US will put boots on the ground. The defense of that choice was to strengthen the belief that the US would show up, and bank on that as a deterrent.
What’s your take on the tradeoffs inherent in strategic ambiguity?
Jared McKinney: We argue that you get a better deterrent effect from ambiguity than you get from clarity.
If the PRC believes that America is very likely to intervene in an invasion of Taiwan, then the PRC would very likely will launch a first strike against forward-deployed American units, logistics, fifth-generation fighters, and destroyers. This first strike would do a lot of damage and create a pause in American military power which could potentially be exploited and snowball into a successful invasion.
If the PRC doesn’t know what we’re going to do, then they won’t launch a first strike. They’d load up 30,000 soldiers in amphibious vessels — 071s, 075s, ROROs — and send them out into the Taiwan Strait.
At that moment, if we chose to intervene, we could sink them all. If we did that, then in fact no invasion of Taiwan would be possible. But the PRC will get to that point of vulnerability only if they don't know what we’re going to do and thus decide not to launch a first strike.
Jordan Schneider: The counterargument is that the probability of those ships going into the water goes down if the PRC is certain that an invasion means WWIII against the USA.
Jared McKinney: Military commanders want certainty, especially in the PLA.
They’re not comfortable with tons of uncertainty. How do you report uncertainty to your commanding officers? Your answer, in reality, is that an invasion might be catastrophic or it might be okay. From a military commander’s point of view, that’s not very acceptable.
If you can generate uncertainty about the overall operational success, I tend to think that’s the more salient view variable. Even though there could be a larger war, a military commander is focused on that operational point of view.
Peter Harris: Even if we accepted the premise about strategic clarity bolstering deterrence, that would still be a mistake for the Taiwanese above all else, because what if we have a future president, like Trump or someone after him, who says that they wouldn’t come to defend Taiwan? Or, a war in Europe or the Middle East could cause China to start to doubt whether the United States can muster the capability.
Thus, to rely too much on that US-led deterrent would be a mistake. One of the lessons from Ukraine was that it’s much better to engage in a general, direct deterrence. For Taiwan, that means maintaining a standing array of autonomous deterrents that it can level against China.
I want to push back against the idea that only the threat of a cataclysmic world war is sufficient to deter the Chinese invasion of Taiwan. I just don’t agree with that. I think there are less serious threats that can be leveled against China that are serious enough to make an invasion irrational and to stop one from happening.
Those deterrents can come from Taiwan and regional partners. They don’t need to be US-led.
Richard Haas and David Sacks wrote this article in Foreign Affairs entitled “American Support for Taiwan Must Be Unambiguous.” It was very sober and kind of an arresting article, but there was a line in it where they said, yes, fighting China over Taiwan will be devastating for the United States even if we won. But they argued that we should threaten that anyway.
But I think it’s incumbent upon people who argue for that to then also explain how we are supposed to make that credible. We’re admitting that it would be devastating for us to fight a war against China, even if we win. That decreases the credibility of our threat to start that war, even in the best of times.
Because the usual commitment devices are not possible in the case of Taiwan — we cannot sign a treaty or permanently forward-deploy US forces on the island. If I were Taiwanese, I would probably prefer deterrence to be built upon a foundation that doesn’t depend on who is occupying the White House next year, in five years, or in ten years.
Subscribers get access to part 2 of this episode, where we discuss:
How Taiwan can optimize its deterrence posture;
Lessons from the dance of death between Iran and Israel;
Objective factors for measuring invasion risk, and whether the world should be scared about 2027;
How to analyze decision trees for fundamentally irrational decisions;
and more!