Nuclear defense in a tripolar world would be complex and expensive. For this reason, the United States, Russia, and China could still have a shared interest in arms control agreements.

The New START arms control treaty between the United States and Russia is scheduled to expire in February 2026, unless it is extended by mutual agreement. Conventional wisdom is that a revived agreement is dead on arrival, mired in Russian disagreements with the United States and NATO over the war in Ukraine. The following discussion argues for a more optimistic outlook. A post-New START agreement that includes three instead of two nuclear superpowers is not impossible, although it will be technologically and politically challenging. 

A future strategic nuclear arms control regime must satisfy the criteria of strategic stability. These criteria include deterrence stability, crisis stability, and arms race stability. Deterrence stability means that each of the three nuclear superpowers (the United States, Russia, and China) must have forces sufficiently large, diverse, and competent enough to survive any conceivable nuclear first strike and retaliate, inflicting unacceptable damage on the attacker. 

Crisis stability implies that policymakers are disinclined to choose preemption over retaliation during a military confrontation and are fully in control of their diplomatic outreach and military operations. Both deterrence and crisis stability are also based on the timely availability of accurate intelligence about other states’ strategic thinking and operational maneuvers, especially with regard to indicators of possible attack. 

The third component, arms race stability, emphasizes the preferential deployment of weapons and technologies that are well understood by all sides and are not potential game changers with respect to the vulnerabilities of other states. In this regard, the trajectory of future versions of drones, hypersonic weapons, artificial intelligence, and missile defenses must be subjected to cooperative management among the great powers.

Future American, Russian, and Chinese strategic nuclear forces will maintain the following characteristics for the period between 2030 and 2035. First, each state will continue to deploy a triad of forces including land-based intercontinental ballistic missiles (ICBMs), submarine-launched ballistic missiles (SLBMs), and long-range bombers. 

Second, their preferred mix of launchers will not be identical. Russia and China will emphasize ICBMs compared to SLBMs and bombers. In contrast, the United States will place greater relative emphasis on SLBMs and bombers while maintaining a modernized ICBM leg of the triad. 

Third, Russia or China may be more reliant on “launch on warning” (LOW) than the United States, due to differences in their force structures and strategic doctrines. For example, some US assessments express concern that China is working to implement a launch-on-warning posture called “early warning counterstrike,” in which early warning of a missile attack leads to a counterattack before the enemy’s first strikes can detonate. 

On the other hand, China may perceive this LOW option as consistent with its “no first use” declaratory policy, as a form of retaliation after a reliable warning of attack. Doubtless, Russia and the United States also have this option available for anticipatory attack in the last resort, provided the warning of attack is unambiguous. Some US analysts have distinguished between “launch on warning” and “launch under attack” (LUA): the implication is that LUA demands more convincing evidence than LOW. However, neither involves “riding out the attack” (ROA) and then retaliating.

The irony of tripolar strategic nuclear competition, compared to the bipolar deterrence situation of the Cold War, is that the complexity of intelligence estimation in the former case is considerably greater. State “A” contemplating an attack on State “B” must not only take into account the intentions and capabilities of State B, but also those of State C and how they interact. 

This problem has two sides: (1) how will State C behave during a crisis between A and B; and (2), if deterrence fails and war breaks out between A and B, what will State C do? Can Russia assume, for example, that a nuclear first strike against the United States will motivate China to engage against the United States either simultaneously or sequentially? 

Russia would be foolish to make such an assumption. China might act to support Russia diplomatically during a crisis with the United States. Still, it does not follow automatically that China will take part in a nuclear war, given the unavoidable costs of doing so. On the other hand, cautious US planners might have to assume that China could enter the fray in order to minimize its otherwise acute vulnerability after the attack. It might also matter how two powers in conflict climbed the escalation ladder from nuclear first use of tactical or less than strategic weapons to the first strike of one side’s strategic nuclear forces. 

What difference would the deployment of defenses make with respect to the likelihood of deterrence or crisis stability in a more crowded nuclear superpower world? This question is especially significant in view of the Trump administration’s commitment to develop and deploy the “Golden Dome” comprehensive missile defense system, including technologies for space-based intercept. 

In general, strategic missile defenses will have to be perfect or nearly so in order to prevent unacceptable levels of nuclear retaliation or, even more daunting, to forestall a nuclear first strike. This remains true even under the assumption of agreed-upon three-sided limitations on the deployment of offensive strategic nuclear forces. 

In the absence of agreed-upon limitations on offensive delivery systems and warhead deployments, missile defenses will be on a treadmill, chasing ever-increasing and improving offenses. Could any of the nuclear superpowers, within their existing and foreseeable budgetary constraints and domestic policy challenges, endure open-ended arms races? In the case of the United States, projections by the US Congressional Budget Office of the costs of US nuclear forces from 2025 to 2034, as specified in Department of Defense and Department of Energy 2025 budget requests, would total $946 billion. That’s almost a trillion dollars: even for a superpower, that’s a Big Mac.

Strategic nuclear arms control among the three nuclear superpowers is possible, but it is becoming increasingly challenging and complex. Three nuclear superpowers will be deploying weapons terrestrially and, presumably, in space. Meanwhile, the “action” for military planners will shift toward long-range, conventional precision strike systems, including drones for warfare in the air, on land, and at sea.

Additionally, the United States, Russia, and China may conclude that building additional layers of strategic nuclear forces beyond the minimum necessary for deterrence is a wasting asset, compared to diversifying in less than strategic or tactical nuclear weapons and in high-end conventional forces.