Tension between China and Taiwan could lead to two basic types of military conflict. China could seek to reunify Taiwan with the mainland by brute force invasion, or it could seek to use coercion to achieve the more limited aim of modifying Taiwanese behavior.
Should the United States seek to prevent a Chinese invasion of Taiwan, the island’s geography and technology conspire in ways that highly favor such an attempt—while these same two variables also make it much more difficult for the United States to prevent China from causing significant coercive pain to Taiwan. If the United States wishes to counter the Chinese capability to employ coercion against Taiwan, then it needs to make significant changes to current US military doctrine. If made, the resulting anti-coercion would cross a significant non-nuclear escalation threshold: attacking targets in an adversary’s homeland, using conventional means—in some cases on a large scale—in a conflict with a nuclear-armed opponent.
This is worrisome because many observers argue that the United States should abandon its traditional policy of ambiguity regarding its commitments to Taiwan and fully commit to come to the island’s defense in the event of any attack by China; perhaps the highest-profile example is Richard Hass and David Sacks’ essay in Foreign Affairs, aptly titled “America’s Support for Taiwan Must Be Unambiguous” (Hass and Sacks 2020.)
Given the potential for any US counter-coercion campaign to escalate, the United States should be wary of committing itself in advance of a conflict to defend Taiwan from all forms of Chinese coercion. If the United States does decide to end its policy of ambiguity regarding its commitment to Taiwan, it should do so only partially by committing in advance only to protecting Taiwan from invasion, while continuing to remain ambiguous regarding how it would respond to a Chinese coercion campaign that stopped short of invasion. Among the reasons to preserve such ambiguity is to preserve the incentive it gives Taiwan to work harder to obtain the best means of protecting itself from Chinese coercion.
Invasion or coercion
There are a number of reasons for China to be wary of launching an invasion of Taiwan, but the key military reason is that it cannot safeguard a properly sized, seaborne invasion force and the follow-on shipping necessary to support it during multiple transits across the 100-plus mile-wide Taiwan Straits. Protecting such an invasion force would require complete command of the sea surface of the Straits by China, but China cannot now and most likely will not be able in the future to come close to that level of command. This is because it has little or no capability to prevent extremely quiet American nuclear attack submarines (SSNs) from operating in the Straits, nor can it defend against long range, stealthy, anti-ship missiles (ASMs) launched from well east of Taiwan by long range strategic bombers operating from US bases in the Second Island Chain—most commonly defined as the chain of islands that make up the Marianas (most notably Guam), the western Caroline Islands, western New Guinea, and Japan’s Bonin Islands. In contrast, the First Island Chain—which is almost half the distance to the Asian mainland—is generally considered to be composed of the Kuril Islands, the Japanese Archipelago, the Ryukyu Islands, Taiwan, and the northwestern portion of the Philippines, finishing towards Borneo. (See Figure 1.)
Matters are quite different when it comes to Chinese capabilities to cause pain in an effort to coerce Taiwan. China’s People’s Liberation Army (PLA) Air Force strike aircraft, PLA Rocket Force conventional tactical ballistic missiles, and PLA Navy submarines can all be used for coercive purposes against Taiwan. Unfortunately, the United States today can only protect Taiwan from Chinese submarines, and even that capability is somewhat vitiated by the fact that the two major ports which serve Taiwanese ocean shipping are vulnerable to Chinese land-based air and missile attack. Previously, I have argued that the United States can protect Taiwanese shipping from Chinese submarines—as well as the shipping of other members of the First Island Chain—for two basic reasons: The United States has passive acoustic barriers deployed in the exits from the East China Sea and the northern third of the South China Sea that will prevent the covert entrance of Chinese submarines into the Philippine Sea and the southern two-thirds of the South China Sea in a crisis, causing Chinese vessels great attrition passing back and forth through those straits in a conflict. Second, only China’s shipping needs to pass through the northern third of the South China Sea; the shipping of all the members of the First Island Chain can avoid the northern third of the South China Sea by using the Philippine Sea (Cote 2019).
Even if it got fully involved in defending Taiwan, the United States has little or no capability to prevent large-scale attacks against Taiwan by Chinese strike aircraft and conventional tactical ballistic missiles. There are, however, measures that the United States could take to counter such Chinese missiles and aircraft on behalf of Taiwan. All of them have two things in common: They would rely on existing or already planned technology but need a new doctrine for exploiting it; and they would be highly escalatory, involving medium to large-scale conventional strikes against Chinese targets ashore. Consequently, the United States should be cautious about committing itself in advance to coming to Taiwan’s aid if China pursues a military campaign of pure coercion—such as a blockade, air and missile bombardment, or anti-satellite warfare. The reasons for this are many: because the future innovations needed to counter such a campaign may not occur; because the escalation that would likely result from their being employed could provoke Chinese nuclear escalation; and because the United States should do nothing to reduce Taiwan’s incentives to invest in already existing technologies that would significantly increase its own counter-coercion capabilities—such as advanced, anti-tactical ballistic missile defenses.
That said, while the United State cannot prevent Taiwan from being attacked, the United States should still be able to defend Taiwan militarily from an outright, full-blown invasion by Beijing. This is due to the capabilities of US Navy submarines to successfully operate against Chinese amphibious shipping in the Taiwan Straits despite Chinese anti-submarine warfare forces; the ability of US Air Force bombers armed with long range, stealthy, anti-ship missiles to operate against the same shipping from US-controlled bases in the Marianas Island Chain in the face of Chinese aircraft and missile attacks; and the unique synergies that arise when US submarines and bombers are used as a team rather than independently. Let us discuss each of these items in turn.
US Navy submarines
Sonar, both active and passive, is the key sensor for anti-submarine weaponry, whether it be for surveillance or for tactical engagements. The enemy of all sonar propagation is shallow water—and because the Asian continental shelf extends far out to sea, any attempt to invade Taiwan across the Taiwan Straits would require Chinese amphibious shipping and the Chinese Navy ships escorting them to operate in very shallow water, mostly less than 600 feet in depth. Beyond that continental shelf, water depths fall sharply away to depths measured in tens of thousands of feet, which is the case both for the Philippine Sea and the southern part of the South China Sea.
In shallow water, acoustic energy tends to reflect repeatedly off the bottom and the surface— whereas in deep water, acoustic energy can find its way into deep sound channels where it is refracted between warmer water near the surface and denser water near the bottom.
Acoustic energy loses much more of its strength when it is reflected because the reflector absorbs energy from the sound wave, whereas with refraction the sound wave is bent without losing energy. In the Cold War a faint Soviet submarine tonal that might propagate for many hundreds of miles in the deep sound channel of the Mid-Atlantic might propagate for only five to 10 miles in the much shallower Barents Sea. At the same time, active sonars encounter a different problem in shallow water: Their powerful signals cause reverberation, or the generation of multiple echoes from the bottom and the surface, which cannot be distinguished from a target echo.
All this is to say that neither China nor the United States will be able to prevent their opponent from operating its submarines in the Taiwan Straits. The key issue is what those submarines can do once deployed in the Taiwan Straits in the event of a Chinese invasion attempt. Chinese submarines cannot be used as direct anti-submarine escorts for amphibious shipping because that would seriously expose them to fratricide at the hands of Chinese Navy surface ships also escorting the invasion fleet, and they cannot be used to form barriers to prevent American nuclear-powered attack submarines from entering the Taiwan Straits because China lacks the ocean surveillance assets needed to enable such a barrier (Cote 2019). As a result, Chinese submarines will not play a major role in preventing American submarines from operating against Chinese invasion shipping.
The primary Chinese anti-submarine weapons in the Straits will be the Chinese Navy surface ships and patrol aircraft, and their capabilities will depend directly on the degree to which American nuclear-powered attack submarines compromise their inherent stealth by aggressively engaging Chinese amphibious shipping and their Chinese Navy escorts with traditional approach and attack tactics that require the subs to close with their targets using organic, hull-mounted sensors and fire on them with torpedoes after a periscope identification. Such attacks are highly “indiscreet” and they create high profile “flaming data”—to use two examples of the undersea warfare vernacular. They therefore significantly reduce the area that must be searched to find the submarine, which inevitably would produce some level of attrition against American submarines operating in such fashion. Even if such attrition was relatively low by historical standards, it would be problematic given the relatively small number of American submarines that would likely be available in many scenarios.
The American submarine force is already taking several steps to reduce this tradeoff between lethality and survivability. A prime example of these efforts is a small, battery-powered unmanned air vehicle that could be launched from a submarine’s countermeasure dispenser tubes and be used to search 10-20 miles in advance of the submarine for surface targets (Blake 2020). Once found, surface targets could be identified and torpedoes or anti-ship missiles could be launched against them from well over the horizon—and therefore well over the horizon from the sensors of the target vessels that the submarine is shooting at. In this way, US submarines could “stand off” from the dense but very short-range anti-submarine assets escorting Chinese amphibious shipping without sacrificing the lethality of their weapons—particularly their heavyweight torpedoes, only one of which is needed to break most surface combatants or commercial cargo vessels in half.
US Air Force bombers
In addition to its submarines, US bombers also pose a relatively unanswerable threat against Chinese amphibious shipping, largely because of the unique capabilities of the United States’ new Long Range Anti-Ship Missile (LRASM), and the ability of US Air Force bombers to operate from the Second Island Chain—where the volume of potential attack by Chinese aircraft and tactical ballistic missiles is radically reduced compared to the vulnerability of air bases in the First Island Chain.
LRASM is an air-launched anti-ship missile, derived from the Air Force’s stealthy, long-range, air-launched, land attack cruise missile known as the JASSM-ER, which has a range of more than 300 miles. Its stealthiness makes the LRASM essentially immune to Chinese Navy and Air Force air defenses, and allows it to remain at a high altitude during its initial approach to its targets—unlike essentially all other cruise missiles, which drop to extreme low altitude flight well before coming within line-of-sight of their targets and their accompanying air defenses. The ability to remain at higher altitudes for longer enables the second capability that makes LRASM unique: It can use its multispectral sensor imaging suite to search for, locate, and identify all the potential targets in its field of view by literally “measuring” them and comparing the results against a data base of ship types. Thus LRASM, and perhaps more appropriately, a large salvo of LRASMs, can be programmed to choose the best targets in a target-rich environment without the need for high-quality offboard targeting support. In some ways, the LRASM (which entered development in 2007) is one of the first and perhaps most significant indications of when the US military truly began “pivoting” to Asia.
The LRASM achieved its initial operating capability in 2018, when the US Air Force incorporated the missile in its B-1 heavy bomber force; it will also be integrated with B-52s and Navy F-18s and P-8s. One B-1 can carry 24 LRASMs and one B-52 can carry 20. Assuming that more than 50 US bombers are equipped with LRASM for the long haul, we can see that such a force would pose a significant threat to Chinese ships operating essentially anywhere in the Western Pacific, as long as they can safely operate from bases in the Second Island Chain.
Today, the main base used by US bombers in the Second Island Chain is Anderson Air Force Base on the island of Guam in the Marianas Island Chain, approximately 2,000 miles away from the Taiwan Straits.
China currently has two means of launching conventional attacks against this air base; the Chinese rocket force’s intermediate-range ballistic missiles armed with conventional warheads, and bombers with conventional, long range, air-launched cruise missiles. The Chinese rocket force has about 200 intermediate-range ballistic missiles in total and its air force has 60 H-6 bombers, but both these forces also have two other missions: anti-ship attack against US Navy ships, and nuclear attack (Office of Secretary of Defense 2020). Conventional and nuclear land attack payloads for intermediate-range ballistic missiles are probably interchangeable, but anti-ship variants are probably dedicated to that mission. H-6s can in theory carry any one of the three different types of air-launched cruise missiles but it is unknown whether H-6 units train for all three missions or whether some are dedicated to one mission at the expense of the others.
This means it is hard to calculate what the threat of a Chinese conventional attack is to targets like Guam’s Anderson Air Force Base in the Second Island Chain, but it is clear that the threat pales in comparison to that posed against targets in the First Island Chain, never mind targets in Taiwan. Steps are now being taken to improve the defense of Anderson air base and Guam as a whole against missile attack. If executed in full, they will go far toward dealing with the relatively limited missile threat described above, but that threat might grow in the future, albeit at great cost to both China’s rocket force and air force.
In addition to better protecting Anderson, the United States has an additional option to pursue in the Second Island Chain that would be much easier to execute than in the First Island Chain: to apply the concept of agile combat operations that the US Air Force is pursuing for its fighters in the First Island Chain to bombers and tankers operating from the Second Island Chain.
The doctrine of agile combat operations in the First Island Chain envisions each individual fighter squadron simultaneously operating from several austere bases using already existing commercial runways and bringing in the mobile repair, fueling, arming, flight control, and housing facilities needed to turn those runways into minimally functional fighter bases. This creates a sort of shell game for attacking aircraft and missiles.
An agile combat operation in the First Island Chain is challenging for two fundamental reasons (Priebe, Vick, Heim, and Smith 2019). First, it requires significant pre-positioning of equipment in the theater, which is expensive. Second, and more important, it requires the permission of one or two members of the First Island Chain to not only allow the pre-positioning of material on their territory, but it also requires guarantees from the host nation(s) that they will allow US Air Force access to those bases in a war over Taiwan with China.
In practice, the countries in question would be Japan or the Philippines—or perhaps both. It is more possible to imagine Japan providing such guarantees but Japan is geographically much closer to China than most of the Philippines, which consequently makes agile combat operations there more risky. The island of Luzon in the Philippines provides by far the best real estate for such a collaboration but, needless to say, it is highly unlikely that the Philippines will pre-commit anytime soon to allowing the US Air Force to operate from its territory in the event of a war occuring with China over Taiwan.
This dilemma does not hold true for the Second Island Chain, or at least in the Marianas archipelago, because the four Marianas’ main islands—Guam, Saipan, Tinian, and Rota—are all US territories. Because of this, the United States has the ability, if it chooses, to construct the infrastructure needed to disperse operations by US bombers, tankers, and airlift aircraft across a very wide area. Combined with the dispersal of some of the active and passive defenses already planned for Anderson Air Force Base and Guam, such a posture would provide a bastion against conventional missile attack from which US Air Force could safely operate long range aircraft in a conflict with China.
However, there are several potential flaws in the argument to all the above. First, although US submarines will likely not suffer significant losses operating against a Chinese naval invasion force in the Taiwan Straits, their effectiveness would be limited by the size of their torpedo and missile loads, which can only be replenished in port. Second, US bombers carrying LRASMs do not have the organic capability to target their missiles if they are to take full advantage of their 300-mile range. Though the LRASM is highly autonomous compared to other anti-ship missiles, it still needs a rough, initial targeting cue from another source to tell it where to fly to. The United States has a panoply of air and space surveillanceassets that could easily provide such cues, but many of those will not be available in the very high threat environment that would occur over the Taiwan Straits in a conflict.
But here is where the potential synergies of a submarine-bomber team come into play; specifically, tactical synergy and operational synergy.
The tactical synergy is that US Navy submarines can provide targeting cues to US Air Force bombers carrying LRASMs, or to any aircraft carrying LRASMs for that matter (Converse 2020). This addresses both the concern for magazine size and potential attrition for submarines, and the problem of offboard targeting for LRASMs.
The operational synergy is that the submarine-bomber team presents the Chinese navy with a dilemma: It can concentrate its amphibious shipping into escorted convoys to best protect it from attack by lone US submarines, thereby exposing them to large-scale attack by LRASM-equipped US bombers. Or, it can disperse its amphibious shipping into a stream of individual ships presenting a less lucrative target to LRASM-equipped bombers—but also presenting an ideal set of targets for US submarines no longer faced with concentrated, anti-submarine defenses and therefore able to pick and choose the most lucrative individual targets, maximizing the effectiveness of their finite supply of weapons.
One last point: The power of the US submarine/bomber synergies described above apply everywhere, not just in the Taiwan Straits, nor the Western Pacific, but anywhere US submarines and bombers operate—i.e., on a completely global basis.
Countering coercion, as opposed to invasion
Compared to the challenges of mounting an invasion of Taiwan if the United States intervenes on Taiwan’s behalf, China can cause a very large amount of pain to Taiwan in a coercion campaign that the United States cannot currently do much to contain. China has three primary potential instruments of coercion against Taiwan: a maritime blockade of Taiwan’s shipping by the Chinese navy; conventional strikes against the full range of Taiwanese targets on land by Chinese air force strike aircraft; and use of the Chinese rocket forces’ short-ranged conventional tactical ballistic missiles.
Space does not permit a full discussion of how the United States might counter a Chinese coercion campaign using these tools, but countering them would in some cases require significant innovation by the United States in how it uses its existing weaponry; the resulting innovations would in some cases involve crossing significant conventional escalation thresholds. I will briefly discuss one example of this tradeoff in the case of countering Chinese air force strike aircraft attacks against Taiwan.
Geography—or perhaps more accurately, political geography—prevents the United States from either placing its forward air bases in the First Island Chain or rapidly deploying to bases in that region the US Air Force’s large and growing fleet of fifth-generation tactical fighters. If the US Air Force were able to replicate the hardened, forward-base posture it enjoyed in Western Europe during the Cold War in the First Island Chain, then it would be able to defeat the Chinese air force in the air and gain command of the air over Taiwan, the Straits, and coastal China itself. The fact that it cannot do this at the outset of a conflict, and that it is the Chinese air force which will likely have command of the air—at least against non-stealthy targets—is the single largest benefit that China derives from geography in a coercion campaign against Taiwan.
The only other way to defeat the Chinese air force early in a conflict is to attack it at its bases, which in US Air Force terminology is called an “offensive counter-air” campaign. Such campaigns against land-based air forces were never considered very effective through the end of the Cold War for largely technical reasons, and they were not an item of speculation and analysis during the post-Cold War era given the absence of peer-to-peer competition or conflict. In the meantime, the technology to conduct relatively efficient offensive counter-air campaigns has been developed and in some cases already deployed. The two key technologies are large payload, all aspect stealth bombers, and cheap, standoff, glide weapons with GPS-independent, precision guidance.
The US Air Force is several years away from beginning to deploy at least 120 B-21s, a follow-on to the now three-decade old B-2 stealth bomber, of which only 21 were originally bought. The B-21’s capabilities—such as range and payload—remain classified but estimates are possible. The current building block for measuring US Air Force bomber payloads is the common strategic rotary launcher, of which the B-1 can carry three, the B-2 two, and the B-52 one (the B-52 can also carry weapons under its wings). One rotary launcher can carry 16 of the 500-pound Extended Range Joint Direct Arrack Munitions and 32 of the 250-pound Small Diameter Bombs. These devices are the main examples of the cheap, standoff, precision glide weapons referred to above. If launched from high altitude they can glide for 40 miles (and 10 to 15 miles if launched from low altitude), providing the capability to stand off from dense, terminal air defenses in either case. This means that if the B-21 is provided with two rotary launchers it can deliver from 32 to 64 precision glide weapons per sortie.
One can quickly see how the numbers of targets hit would rapidly escalate if dozens of B-21s were to fly multiple sorties over the course of several days or a week. If we assume that the Chinese air force has approximately 40 bases, each with 24 hardened aircraft shelters within unrefueled range of Taiwan, then that means that there are roughly 960 so-called “aimpoints” representing hardened shelters that could be covered relatively quickly (Heginbotham et al 2015). If these raids also included glide weapons with wide area sub-munition payloads, then Chines aircraft dispersed away from hardened shelters on ramp space within the airfield perimeter could also be attacked.
So far, so straightforward.
Two potential problems loom over this argument. First, there is a doctrinal issue, and second there is the question about conventional escalation. In the interest of conserving space, I will not analyze these issues here but simply flag them so that the reader knows of their existence. From a doctrinal perspective, it is not clear whether the US Air Force would support or resist committing its B-21 force to a large, offensive, counter-air campaign against the Chinese air force. Certainly, it has historically sought to gain command of the air with its fighter force, and has generally sought to use its bombers and other strike aircraft against targets whose destruction would in theory produce independent, strategic or “war-winning” effects, such as the leadership targeting pursued in the First Gulf War (Haun 2021). Thus, even if it is feasible—which I argue it is—a large-scale offensive counter-air campaign may at least initially be rejected by US Air Force.
Second, the question of conventional escalation is fraught in the context of a war between China and the United States over Taiwan. To be specific, massive cyber attacks, attacks against satellites in space, and attacks against targets in the opponents’ territory can all be considered rungs in an escalation ladder that stops short of nuclear use, but which expands conflict well beyond attacks against opposing ships at sea or aircraft in the air.
It is possible, if not likely, that political leaders may refrain from crossing some or all of these thresholds because of fears that crossing them might lead to nuclear escalation. On the other hand, it may be that the shade thrown by nuclear weapons over all other types of conflict during the Cold War may be reduced in a Sino-American context, because both sides face disincentives to using nuclear weapons that are even higher than those faced by the United States and the Soviet Union during the Cold War. To put it simply, the United States’ stakes in a conflict over Taiwan are vastly lower than they would have been in a conflict over the fate of Western Europe, which essentially eliminates any scenario where the United States would use nuclear weapons first in order to protect Taiwan. By the same token, US nuclear superiority over China is and will remain so vast that even if China uses nuclear weapons first there is no scenario in which China could reduce the essentially infinite amount of damage the United States could cause it in retaliation.
The question of how much conventional escalation there would be in a conflict between the United States and China obviously cannot be settled here—but there is no guarantee that a conflict between the United States and China over Taiwan would be characterized by high levels of conventional escalation.
At the same time, there are arguments to be made regarding how the United States might counter all the means of coercion of Taiwan available to China, to include its mobile tactical ballistic missiles and its submarines. I do not make them here, but as with the offensive counter-air campaign described above, they would depend on some degree of doctrinal innovation in the use of existing technology, as well as a willingness to cross the escalation threshold of attacking the opponent’s homeland.
Therefore, ironically, it appears that if the escalation threshold barring attacks against the opponent’s homeland is crossed, the more conventional escalation the better for the United States, from a purely military point of view in a counter-coercion campaign. China has essentially no capability to project conventional power against US territory other than US territories in the Marianas. China’s main means of conventional escalation appears to lie more in the realms of cyber and anti-satellite warfare, which would have little direct effect on any US effort to defend Taiwan from invasion.
Consequently, there are many uncertainties regarding the United States’ ability to defend Taiwan from Chinese coercion. It is for this reason that I argue that the United States should not be in the business of making security guarantees that it might not be able keep in the event they are challenged. The United States cannot currently guarantee Taiwan protection against Chinese coercion. Whether the United States should work hard in the coming years to put itself in the position to do so is among the most important questions decision-makers on US military policy regarding Taiwan and China need to grapple with in the coming years.
 Refraction results from changes in the speed of sound in water that vary primarily with pressure and temperature. When the speed of sound is decreasing with depth, acoustic waves bend downward, and when it is increasing with depth, the waves bend upward. The speed of sound increases with both temperature and pressure, and when the two variables are working at cross purposes, i.e. the water temperature is dropping with depth and the pressure is increasing, temperature trumps pressure. A deep ocean basin usually has a surface layer of constant temperature and increasing pressure, a middle layer of decreasing temperature and increasing pressure, and a bottom layer of constant temperature and increasing pressure. Consequently, sound that gets into the middle layer can get trapped there and spread horizontally as its waves keep bending away from the surface and the bottom. This is what is known as the deep sound channel, and its axis is where the global underwater acoustic sensor network known as SOSUS is employed.
The Bulletin elevates expert voices above the noise. But as an independent, nonprofit media organization, our operations depend on the support of readers like you. Help us continue to deliver quality journalism that holds leaders accountable. Your support of our work at any level is important. In return, we promise our coverage will be understandable, influential, vigilant, solution-oriented, and fair-minded. Together we can make a difference.