Restrain and Hedge: A New U.S. Nuclear Strategy for a Two-Peer World

What if fielding more nuclear weapons makes the United States less secure, not more? That question is now at the center of a growing debate as the United States confronts a nuclear landscape shaped by two major nuclear rivals.

China is rapidly expanding and modernizing its nuclear arsenal, while the New Strategic Arms Reduction Treaty (START), the last remaining nuclear arms control deal between the United States and Russia, has expired. In what appears as the beginning of a new, more dangerous nuclear age, some analysts believe the United States should increase the size of its deployed nuclear arsenal. Others believe that increasing the number of warheads on existing launchers is unnecessary and will prompt a new nuclear arms race. While deploying more nuclear warheads is tempting in the short term, it is strategically flawed. It would provide the United States a marginal deterrent benefit while further fueling a trilateral nuclear arms race for which Washington is ill-prepared.

Instead, the United States should exercise nuclear self-restraint. This means deemphasizing damage limitation within its nuclear strategy while pursuing a new nuclear arms control framework with both Russia and China. To hedge against the possibility that those two countries fail to exhibit similar restraint and are not amenable to striking a deal, the United States should improve its capacity to refurbish existing nuclear warheads, produce new ones, and manufacture nuclear platforms such as intercontinental ballistic missiles and strategic ballistic missile submarines. This approach offers the chance of putting a lid on an emerging trilateral nuclear arms race while also preparing the United States to participate in it should it prove unavoidable.

Deemphasizing Damage Limitation

The first element of a more restrained U.S. nuclear strategy is to deemphasize damage limitation as an operational objective. Taking this step would avoid intensifying a nuclear arms race for which the United States is ill-prepared and free it from needing to target every nuclear asset of China and Russia.

Deemphasizing damage limitation would not require the United States to radically adjust its targeting policy towards countervalue strikes against cities and civilian infrastructure. Counterforce strikes would already inflict immense punishment on an adversary. The threat of such punishment should be sufficient to deter Russia or China from launching a nuclear strike against the U.S. homeland. In addition, the United States could still include providing the threat of limited nuclear options in its nuclear strategy. These options inject uncertainty into the process of escalation that would make Russia and China wary of launching a nuclear strike against a U.S. ally or inflicting a massive non-nuclear strategic attack against the U.S. homeland or U.S. military forces.

To plan to limit damage to a meaningful extent, the United States should regard every adversary nuclear weapons platform as a target. At present, China is complicating U.S. damage limitation efforts by building hundreds of new nuclear platforms, such as 320 silos for the DF-31A and/or DF-41 intercontinental ballistic missile, along with 30 new silos for its DF-5 class of intercontinental ballistic missiles. The new silos are a key part of a current round of nuclear expansion in which the U.S. intelligence community assesses that China will field around 1,000 warheads by 2030. These missile silos could serve as a useful sponge for enemy warheads. Because weapon reliability is not guaranteed, the United States would want to devote two warheads to each target. China’s new nuclear silos could absorb up to 700 U.S. warheads.

Currently, the United States deploys around 1,600 warheads, meaning that China’s new intercontinental ballistic missile silos complicate U.S. efforts to limit the damage that both China’s and Russia’s nuclear forces can cause.

Advocates for U.S. warhead expansion, such as Keith Payne, Frank Miller, Vipin Narang, and Pranay Vaddi, have this targeting problem in mind when they raise the alarm about a two-peer nuclear environment. Expanding the U.S. nuclear arsenal to well over 2,000 or 2,500 warheads by putting stored but operational warheads onto existing launchers could allow the United States to account for China’s new intercontinental ballistic missile silos and still have enough weapons left over to threaten a damage-limiting strike against Russia’s nuclear forces.

The upside of arsenal expansion, however, is likely slim. While having more warheads would allow U.S. nuclear planners to plan to hit more targets, it does not allow them to hit every target. Some targets would certainly be mobile intercontinental ballistic missiles deployed in the field, which China and Russia both have. Finding, fixing, and tracking mobile missiles is difficult. Moreover, an adversary can take actions to make things even harder. China has developed decoy transporter-erector-launchers and can put covers on the back of these launchers to reduce their radar signature. China is also deploying a wide range of counterspace capabilities that could degrade the quality of space-based intelligence, surveillance, and reconnaissance for U.S. forces as a conventional war progresses. Finally, China is investing in more modern air defense systems that can identify and shoot down medium- and high-altitude long-endurance drones.

Against this small benefit of expansion should be weighed the strategic costs of deploying additional warheads. China (projected to have over 1,000 warheads) and Russia (around 1,700 strategic warheads) are unlikely to accept the United States having a much larger strategic nuclear arsenal than they do individually. Increasing the number of warheads on current nuclear platforms would therefore not be the end of nuclear competition. Instead, it would likely spur both Russia and China to build up their nuclear arsenals and further intensify a new arms race. If the United States only needed to engage in an arms race with Russia — whose economy is showing signs of strain due to the war in Ukraine — then perhaps the United States could succeed in a bilateral competition and therefore might welcome it. Yet, a nuclear arms race against a state as economically and technologically powerful as China is another matter. Thus, rather than increasing the number of warheads to pursue damage limitation, the United States should deemphasize damage limitation in its strategy.

Pursuing a New Strategic Arms Control Framework

In parallel with deemphasizing damage limitation, the United States should pursue a new arms control framework with both Russia and China to attempt to put a cap on an emerging trilateral arms race. The adjustment to U.S. nuclear strategy and the pursuit of a new arms control agreement would support each other. Russia has signaled it is willing to abide by the New START limits if the United States continues to do so. If Russia truly respects those limits, China’s nuclear expansion slows after the early 2030s (that the Pentagon dropped its assessment from previous reports that China will field 1,500 warheads by 2035 suggests this is possible), and the United States no longer sees a need to target every Russian and Chinese platform — then a foundation for a three-way deal could materialize.

A new strategic arms control accord could have two core features. First, at a minimum, it could limit the number of strategic nuclear weapons platforms and nuclear warheads each side deploys, like New START. That treaty capped the number of deployed intercontinental ballistic missiles, submarine-launched ballistic missiles, and strategic bombers at 700, and the number of deployed warheads on those platforms at 1,550 (with one bomber counting as one warhead). Under a new framework, Russia, China, and the United States could each agree to similar limits on the number of strategic nuclear weapons platforms and warheads they deploy.

The second feature of a new deal could be agreements or understandings on the development of homeland ballistic missile defenses, theater nuclear weapons, fractional orbital bombardment systems, or hypersonic glide vehicles on nuclear-capable systems, among other technologies. For example, Chinese nuclear scholars have long complained that qualitative improvements in U.S. homeland ballistic missile defenses threaten China’s nuclear retaliatory capability. The United States could consider trading some concessions on its ballistic missile defense capabilities for a Chinese agreement not to field a nuclear-armed fractional orbital bombardment system.

To monitor compliance with the agreement, Russia, China, and the United States could each rely on a mix of satellite imagery, data exchanges, and on-site inspections. If the last is too great an obstacle to a deal, then the parties could place more reliance on the other two tools. For example, the first Strategic Arms Limitation Talks agreement, signed in 1972, included both the Interim Agreement, which limited the number of nuclear weapons launchers the Soviet Union and the United States could deploy, and the Anti-Ballistic Missile Treaty, which capped the number of ballistic missile interceptors each side could field. These agreements primarily relied on what historian John Lewis Gaddis called the “reconnaissance revolution” of satellite imagery and communications intelligence intercepts for verification. A contemporary verification regime without inspections could rely on satellite imagery paired with cooperative measures by each side to uncover missile silos or warhead bases as satellites passed over.

To encourage compliance with a new deal, each side could claim a right to expand its arsenal beyond the limits of the treaty should it detect cheating by any of the other parties. The overall goal of an arms control accord for the United States would be to put a cap on the arms race and push geopolitical competition into other areas, such as AI and cyber technologies, where the United States currently has a lead and may hold a comparative advantage.

Hedging Strategy: Refurbish and Manufacture

Of course, it will likely be difficult to reach such a deal with Russia and China. To account for this possibility or that Russia and China fail to show similar restraint in their nuclear strategies, the United States should focus on recapitalizing and strengthening its nuclear industrial enterprise as part of a hedging strategy.

It is no secret that the U.S. nuclear enterprise has atrophied since the Cold War. The most cited shortcomings in today’s enterprise are delays in building new nuclear weapons platforms. Shortages of skilled workers and shipbuilding capacity will cause the delivery of the first Columbia-class submarine to slip to 2028 or 2029, roughly 18 months behind schedule. Initial delivery of the new Sentinel intercontinental ballistic missile is now four years behind schedule, and the United States may need to operate its current intercontinental ballistic missile, the Minuteman III, through 2050, 14 years longer than expected.

Another shortcoming is in warhead refurbishment and production, exemplified by the difficulties in plutonium pit production. The United States would want to produce new plutonium pits to refurbish existing warheads or build new ones. The National Nuclear Security Administration is currently struggling to reach its goal of producing 80 pits per year by 2030. That was supposed to be achieved by having two sites for pit production — PF-4 at Los Alamos, New Mexico, and the Savannah River site in South Carolina — operational by 2030. However, the latter, which was itself supposed to produce 50 pits per year, is behind schedule and unlikely to start production until 2035. In the meantime, PF-4 is being pushed to ramp up production from its goal of 30 pits per year by 2028 to 60 pits per year. While PF-4 did produce its first “diamond-stamped” pit in 2024, the facility is aging and crowded due to other activities at the site, raising doubts about its ability to meet the higher 60 pit-per-year production standard.

Investing in a nuclear industrial hedge would involve making greater investments in overcoming these difficulties in platform and warhead production that the United States is already undertaking. This could entail even more funding for upgrades to aging National Nuclear Security Administration facilities to enable faster pit production, tritium reprocessing, and warhead assembly and refurbishment. To speed up ballistic missile submarine production, the United States could fund more projects like the recently opened Hadrian plant in Alabama, which produces individual submarine parts and allows shipyards to focus on hull production. To improve the rate at which the new Sentinel missile is deployed, Congress could fund ways to speed up the construction of new missile silos.

Addressing Counterarguments

There are three principal counterarguments to a strategy of nuclear restraint. First, proponents of pursuing damage limitation say that it increases the credibility of extended nuclear deterrence. This claim is questionable and much debated by nuclear scholars, mostly because damage limitation is distinct from deterrence. The former would be attempted only in an extreme scenario where deterrence has largely already failed, and the United States thinks the adversary is about to launch a nuclear strike. It is in that scenario that the United States would strike preemptively to try to achieve damage limitation to a meaningful extent, versus not striking at all.

Achieving meaningful damage limitation could therefore still entail several warheads falling on American cities. The prospect of receiving that level of damage could very well deter a U.S. leader from deliberately launching a nuclear strike or taking actions that they know would raise the risk of escalation to a large-scale nuclear exchange, especially on behalf of an ally.

Second, one could argue that Russia and China are unlikely to reciprocate any restraint the United States shows because their nuclear decisions are not much related to U.S. actions. This perspective ignores that current nuclear modernization efforts by Russia and China could have been spurred by the United States’ pursuit of damage limitation. For instance, the United States withdrew from the Anti-Ballistic Missile Treaty in 2002 to break free from constraints on its ballistic missile defense capabilities. Russia’s novel nuclear weapons systems, such as the Tsirkon and Avangard hypersonic missiles and the Poseidon unmanned underwater vehicle, were likely developed to provide Russian leaders with nuclear options that could evade U.S. defenses. If the pursuit of meaningful damage limitation sparked greater nuclear innovation by U.S. adversaries, then why can U.S. nuclear restraint not engender restraint in kind?

Third, one could point out that while the United States was abiding by New START limits and pursuing a new arms control framework, it would have no way of knowing if Russia expanded its arsenal, given that Russia ended its participation in the treaty’s system of on-site inspections in 2023. It is not clear, however, why a Russian decision to put more warheads on its existing stock of intercontinental ballistic missiles, strategic bombers, or ballistic missile submarines would necessitate an increase in U.S. nuclear warheads. By the proponents of expansion’s own logic, such a Russian decision would not increase the number of targets that U.S. nuclear forces need to hold at risk.

What would be a problem is if Russia decided to build more strategic nuclear launchers beyond the New START limit of 700 deployed intercontinental ballistic missiles, submarine-launched ballistic missiles, and strategic bombers. A Russian effort to expand the number of nuclear platforms appears unlikely because of its own economic issues and the strains of funding the war in Ukraine. Besides, the production and deployment of new strategic nuclear platforms is something that can be detected with high confidence by U.S. intelligence. Thus, if the United States holds off on expanding its deployed nuclear arsenal in response to China’s nuclear expansion, Russia is likely to hold off as well, because a more intense strategic nuclear arms race may not be in its self-interest.

Buying Time

Scholars and analysts claim that the world is entering a new nuclear age, and there is much debate on how the United States should approach it. Some argue that it should repeat the Cold War playbook and engage in a renewed arms race to pursue nuclear superiority over its rivals. However, today’s geopolitical environment is different: The United States is not in a position to succeed in a trilateral nuclear arms race in which China — with its combination of industrial scale and technological sophistication — is one of the main adversaries.

Thus, a more prudent approach is to adopt a restrained nuclear strategy and an arms control policy that aims to make it less dangerous, while preparing for the possibility that such an attempt fails. This entails deemphasizing damage limitation in U.S. nuclear strategy and pursuing a new strategic arms control deal with Russia and China. If successful, this approach could put a lid on the emerging trilateral nuclear arms race, push geopolitical competition into areas such as AI and other cyber technologies, and free up resources for the United States to spend on other domestic and foreign policy priorities.

If, however, China continues to expand its arsenal of nuclear weapons platforms at a frenetic pace after 2030 and Russia decides to expand its nuclear arsenal, the United States may have no choice but to deploy more nuclear warheads. In this pessimistic scenario, by investing in its nuclear enterprise, the United States will at least have bought itself time to prepare to accept the strategic costs of engaging in a more intense nuclear arms race.

Tyler Bowen, Ph.D., is an assistant professor in the Deterrence Studies Institute at the U.S. Naval War College. His research focuses on issues related to nuclear deterrence, escalation, and U.S. nuclear strategy, and his work has appeared in outlets such as the Journal of Politics and the Texas National Security Review. He received his Ph.D. in political science from Yale University.

The views expressed here are the author’s alone and do not necessarily represent the views, policies, or positions of the U.S. Department of Defense or its components, including the Department of the Navy or the U.S. Naval War College.

**Please note, as a matter of house style, War on the Rocks will not use a different name for the U.S. Department of Defense until and unless the name is changed by statute by the U.S. Congress.