Mark Cancian Explains the Battlefield Dynamics of a Taiwan Invasion

FISC: Could you start by giving us a brief summary of these simulations? What were your parameters, and what were the major findings that you think people should be aware of? 

​

Cancian: CSIS conducted a project that developed a wargame simulating a US-China conflict over Taiwan. There were three major elements to it. First, it was completely unclassified so that we could discuss assumptions and results very specifically. Second, we used a lot of hard data and did a lot of operations research to come up with combat results that were objective instead of relying on subjective judgments. Third, we ran the game twenty-five times in order to get a strong analytic foundation for the insights that came out of it and the eventual recommendations. Our overall finding was that the United States and its partners could sustain a democratic and autonomous Taiwan, but it would come at very high cost. In addition, the Taiwanese economy would be devastated. The Chinese would also lose heavily, sufficiently that it might endanger the stability of the Chinese communist regime. 
 

02   China's Mainland Unscathed?  

​

FISC: Speaking of the consequences for the Chinese regime, I was struck by the fact that in all of your simulations, the damage to mainland China was either nonexistent or strictly limited to some facilities that were being used to carry out the invasion. Is this correct, and can you talk a little bit about why even in a high-intensity conflict with the United States, with the exception of combatants and military hardware offshore, China appears likely to get through largely unscathed?

​

Cancian: Let me make a couple of points, but yes, your basic understanding is correct. The first is that one of our assumptions was that the United States would make strikes on the Chinese homeland, but we recognized that that was controversial, and we ran many excursions where the US military was not authorized to strike the Chinese mainland. The second is that the economic effects of war as a result of trade disruption would likely be quite severe. We didn’t measure it in this project, but we have a little follow-up project where we will be looking at economic effects.

 

Third, we note that the high level of casualties, which the Chinese haven’t experienced for many decades, might destabilize the Chinese Communist Party’s grip on power. We didn’t take it beyond that, but we did raise the question. We found that this is an all-or-nothing operation for China. Some people have suggested that the Chinese might launch an invasion, and if that fails, pull back, rebuild, and launch another one. However, the demands of an invasion are so great, and the losses are so great, that it would take many, many years to rebuild even if the Chinese Communist Party retained power. 

 

The final point is that even when US forces were authorized to strike the mainland, they focused on military targets, particularly ports and airfields. There were some teams that thought about strategic bombing campaigns that would attack Chinese infrastructure and political institutions. However, they quickly figured out that the demands of the conflict around the island, in other words the immediacy of the invasion, prevented them from making strikes on the mainland that would have long-term effects– effects that might be measured in months, for example. They needed to attack the Chinese fleet today because it was landing troops. So air attack caused relatively little damage to Chinese economic infrastructure.

​

03   Changes to Assumptions or Outcomes After 2026?

 

FISC: You ran the simulation for 2026, which obviously is a very close timeframe. If we were to push this out to 2027 or a decade or more into the future, what are the assumptions that you expect to change, and how do you think that might have an impact on the outcomes? The United States government has said many times that it thinks that Xi Jinping has ordered the Chinese military to be prepared for a conflict over Taiwan in 2027, which of course is not the same thing as making a decision to invade, and there are also some observers who think this might happen farther in the future.​

​

Cancian: We picked 2026 for two reasons. One was that it was inside the “Davidson window,” or the window of vulnerability that Admiral Davidson identified as occurring when the Chinese military buildup was continuing but before US capabilities would come online. 2026 was also the end of the Pentagon’s planning period, so we had a very good idea about what capabilities the United States might have. 

 

We thought about going further into the future, but it gets very speculative when you do that. It would be essentially a different project. You would have to look at different futures for both the United States and China. For the United States, for example, will we continue a military buildup, or will the defense budget plateau, and along with it US military capabilities? All of these new weapons that are coming online– would the US actually be able to build them and field them? On the Chinese side, what does their build-up look like? Is it going to continue at the level it’s been going at for two decades, or will it taper off as the demands of other elements of the economy make themselves felt? So it’s very difficult to make the prediction. We can say that any conflict, whether it’s in 2026 or 2030 or 2035, is going to be extremely violent, with high casualties in a short period of time. 

​

04   Casualties for China

​

FISC: Can you talk a little bit more about those casualties? You spoke a little bit in the report about what you expect on both sides, but could you review for us the scale of the casualties that you’re expecting on the Chinese side, both in the scenarios that are more pessimistic from the Chinese point of view and those that are more optimistic?

​

Cancian: On the Chinese side, the losses ranged from 100 to 150 ships and anywhere from 20 to 300 aircraft. We didn’t make a specific estimate for personnel, but I think it may be close to 50,000, plus a large number of POWs on Taiwan. The POWs are an important point because in those scenarios where the invasion fails and the beachhead eventually collapses, there will be ten or twenty thousand Chinese who become prisoners. The sight of those prisoners marching down the streets of Taipei, for example, might provide such a visual and concrete view of the Chinese failure that it would destabilize the regime. 

 

The reason that there’s such a spread, particularly on the aircraft, is that a lot depends on the scenario. In the scenarios that are optimistic to the United States and Taiwan, the Chinese lose their ships very quickly, the aircraft not so much, because the conflict doesn’t last very long. As the scenarios get more pessimistic for the United States and its coalition partners, the casualties on both sides go up because the conflict goes longer.

 

It’s important to note that the losses described in the report should be viewed as a floor. The game runs for maybe 3 or 4 weeks of game time, not to the actual end of the conflict, and in most scenarios it would go on for another month or two until it was finally played out. There would also be a series of battles in the South China Sea as US forces in the Atlantic moved into the Pacific from that direction, so there would be additional casualties from there. 

​

05   Uncertain and Pivotal Variables

​

FISC: It was very interesting reading about your simulation design, and I appreciated that you hedged against uncertainty by assigning different values to some of these variables and then running different simulations with more optimistic or more pessimistic assumptions. You detailed this at some length in your report, but for those who haven’t read it, could you talk about some of the variables that you identified as both especially uncertain and particularly impactful in determining the outcome of a conflict?

​

Cancian: I’ll start with one variable, which we did not vary, and that is whether Taiwan would resist. The project assumed that the Taiwanese would resist strongly. If they don’t, of course, then all of the analysis is meaningless. 

 

After that, we had about 15-20 major assumptions, and for the ones that we thought were most impactful, we ran different scenarios to look at the effect of the alternative assumptions. To give you an example, Japan was a critical variable. In the base case, we assumed that the Japanese would allow the United States to use its bases in Japan, but the Japanese forces would not participate themselves unless the Chinese struck the Japanese homeland. Now, in 19 of the 25 cases, that did happen, but as an assumption, we tested that with many senior Japanese officials and military officers, and they were quite comfortable with it. They pointed out that this would be the path of least resistance for the Japanese– that is, if they did nothing when the war started, this is what would happen: the United States would use its bases, and they would not participate. They also pointed out that if they did not allow the United States to lose its bases, they would be tearing up the US-Japanese security arrangements that had safeguarded Japan for 70 years. They didn’t think that any government would be willing to do that. We did look at an alternative where the Japanese were entirely neutral, and that was extremely difficult for the United States. The United States lost, because there’s no way to get all of our fighter attack aircraft into the battle. All of those F-35s, F-15s, and F-16s are useless because we can’t get a base close enough for them to fly over Taiwan. 

 

We also ran some excursions where the United States was distracted by a conflict in Europe or the Middle East, so the flow of US reinforcements was lower. Additionally, one critical technological variable was the US missile called JASSM, which is a long-range precision attack missile. On the new version, which is called JASSM extended range, there’s some question about whether it can strike moving ships. In our base case, we assumed yes, but we did many excursions where we assumed no, and that turns out to be really important, because the United States has a lot of those missiles. Unfortunately, the United States has very few of the long-range precision missiles that are specifically designed to attack ships, called LRASM. US only has about 400 of them. By 2026, the US will have over 2000 of the JASSM-ERs, so if those can strike ships, it makes a huge difference.

 

Another assumption about national policies is that there would be no US troops stationed on Taiwan, either before the war or reinforcing Taiwan during a crisis. When we talked to people in the White House or in the State Department, they said they would absolutely not send forces in a crisis because that might precipitate the conflict they were trying to deter. We did run an excursion where the United States moved one of the marine littoral regiments onto Taiwan, and it was helpful, but the problem is that once the conflict begins, it’s very difficult to get any reinforcements or logistics onto Taiwan.

​

06   Surprising Findings 

​

FISC: Moving from assumptions to outcomes, there were a couple of things in the report that struck me as surprising. Aerial combat didn’t seem to be that important, for example, and nor did the distinction between fifth-generation and 4.5-generation fighter aircraft. It also seemed that ground-based Army and Marine units that have been recently designed for this kind of conflict have a limited effect. Could you explain some of these outcomes that might go against people’s expectations?

​

Cancian: One of the striking insights that came out was that 90% of US aircraft losses occurred on the ground due to Chinese missile attacks. The Chinese have an enormous inventory of missiles of all ranges, and periodically they would sweep the US and Japanese airbases and destroy literally hundreds of aircraft as they did that. The insight there was that the difference between a 4.5 and a 5th generation aircraft was much less important if you’re losing 90% of them on the ground. In terms of air combat, there is some air-to-air combat, but aviation is focused principally on attacking ships and ground forces. 

 

There were two things that people might find surprising in terms of naval operations. First, surface ships were very vulnerable when they were inside the Chinese defensive zone, so US and Japanese surface surface ships moved east to Guam to stay out of the range of Chinese missiles. The two US carriers often became casualties early on because they started the conflict inside this defensive bubble. Second, US submarines were hugely useful, and they got inside the strait and sank many ships. 

 

In terms of ground combat, this is not a theater where it’s easy to bring ground forces to bear.  The Marine Corps has built marine littoral regiments, and the army has built a similar kind of unit called the multi-domain task force. The idea behind both units is to operate inside this defensive zone with long-range strike missiles to attack the Chinese forces and particularly the naval forces. These can be very useful, but the problem is that as they get close in, where they can use these capabilities, it becomes very hard either to position them once the war starts or to resupply them if they’re in position. 

​

07   Resistance to a "Porcupine Strategy"

​

FISC: You talked in the report about the “porcupine strategy,” or the asymmetric measures that Taiwan needs to take to effectively prepare for a cross-Strait invasion. Could you tell me why, as you mentioned, there continues to be resistance to this approach from the Taiwanese military? It seems that US observers unanimously believe that it is the only way forward. 

​

Cancian: We talked a lot to Taiwanese national security officials and military officers, and there were two reasons why they’ve been resistant to moving towards this porcupine strategy. One is inertia. For seventy years since the KMT established the regime on Taiwan, they’ve had a balanced military with fighter aircraft, surface ships, and an army, and it’s hard to move away from that. The argument we make in the report is that that approach was appropriate for 50 years until the Chinese started to build maritime and air capabilities. Now the Taiwanese really can’t compete. 

 

The other argument they make is that these kinds of capabilities may not be useful in an invasion scenario, but they can be very useful in grey-zone and blockade kinds of situations. Having fighter aircraft that can go out and intercept Chinese aircraft that are moving towards the median line and having surface ships that can protect Taiwanese civilian vessels against Chinese harassment can be very useful. 

 

We’ve come to recognize that there needs to be a balance, but the Taiwanese still need to move that balance towards more survivable platforms. They built an amphibious ship, which doesn't help very much with blockades, and it’s still extremely vulnerable. They need to build more ground-based systems, anti-ship missiles, and anti-air missiles as part of this porcupine strategy. Those capabilities are much less vulnerable. 

 

To give the Taiwanese credit, they have moved some in that direction. They’re buying 400 harpoon missiles. One of our recommendations is that they buy a lot more and put them on mobile launchers because they can directly threaten a Chinese fleet. The bottom line is that they do need this shift towards a porcupine strategy even though there’s a rationale for retaining some of these other capabilities. 

​

08   Potential for a Protracted or Episodic Conflict

​

FISC: In your remarks and your report, you briefly mentioned the possibility of a protracted or episodic conflict. I was happy to see that, because a little over a year ago, I wrote an article where I argued that even if the United States were to “win” a Taiwan conflict, that might just be the beginning of the issue, not the end. You mentioned in the report that reconstitution would be long and difficult for China but also faster than American reconstitution. Is that something that we should bear in mind with regard to this potentially turning into a recurrent conflict?

​

Cancian: Sure, and I think there are a couple of points that come out of what you just said. The first is about conflict termination, which a number of game participants raised. We assume that the conflict will end when the military aspects end, but very often, conflicts don’t end at that point, and the combatants continue. That’s something that we couldn’t look at in this project, because we just didn’t have enough time, but I would love to do a follow-on project that looks at not just one month but what happens after 3 months, what happens after 6 months. At that point, both sides have lost so much equipment that they’re having to improvise. The nature of this improvisation will be very interesting, and quite different from what the United States has expected in future conflicts. 

 

There’s also the question about reconstitution. If there is an end to the conflict or some sort of ceasefire, both sides will start to rebuild, and the Chinese are in a better position to do that because their industrial base is much more robust than the US industrial base. One of the problems we’ve seen both in this project and some other work that CSIS has done is that the US military defense industrial base has been sized for efficient peacetime production and not for wartime requirements. It would take many years to replace aircraft and munitions that were lost in the conflict. For ships, it would take decades, and for something like an aircraft carrier, essentially, they can never be replaced. The industrial base is just too narrow to sustain a fleet and rebuild from losses. 

 

This means there is a risk that the Chinese would rebuild faster and be able to regain global influence at a faster rate than the United States could. The caution we would point out is that if the Chinese really were defeated as we envision, that would be a major blow to them, and they might become introspective as a result of the defeat. They might not be willing to rebuild and try again. 

​

09   Changes Since Early 2023?

​

FISC: In the 13 months since this report was released, have you seen any trend on the US side towards better procurement or better strategizing for a Taiwan conflict? Or does the outlook look largely the same?

​

Cancian: There have been a couple of things that are encouraging and a few things that have been discouraging. The encouraging thing is that the Pentagon has put funding in its 2024 budget and  future years plan to build up the defense industrial base, especially regarding munitions. So that has sunk in from the report and other reports and lessons learned from Ukraine. Another positive event is that the United States has signed agreements with the Philippine government to use bases. Those are focused on the SCS and Chinese assertiveness there, but the basing could be quite helpful in a Taiwan scenario. The third positive thing is that the Japanese have become more assertive, and they have expressed a desire to build offensive weapons and to increase the defense budget as they recognize the threats that are coming from China.

 

A discouraging element is that the US defense budget is plateaued at best, so it’s going to be difficult to bring on all of these new capabilities that the services have been developing and talking about. Additionally, the Air Force and Marine Corps and Navy have all intentionally reduced the size of their forces to save money so they can invest in these new capabilities. As a result, the forces will  be much smaller than projected when we put together the orders of battle two or more years ago. For a conflict that goes beyond a couple of weeks, that could be quite important. 

​

10   Closing Points

​

FISC: Are there any last findings that you want to bring people’s attention to?

​

Cancian: When we first started talking about the outcome of the wargames, some people said, what’s the big deal, we win? And we pointed out, yes, that is true, we can sustain an autonomous and democratic Taiwan, but it comes at such great cost that the United States should build capabilities to deter such a conflict and to win it more rapidly if it comes to conflict. In addition, there’s the risk of the US population becoming discouraged. We saw that after World War I. Even though the United States was on the winning side and its losses were relatively low compared to other combatants, there was a turn towards isolationism, which had disastrous consequences in the 1930s. We want to avoid that in the future.

​