What is it about?
This paper examines the structural conditions under which arms control agreements emerge and assesses their implications for U.S.-China strategic stability. It advances the argument that two variables capability parity and preference alignment largely determine whether arms control becomes feasible. Drawing on a new dataset of nearly 300 formal agreements signed between 1815 and 2020, the study develops a typology of symmetric, agreed asymmetric, and imposed asymmetric agreements. Empirical patterns show that symmetric agreements are rare, arising mainly during periods of power parity and simplified rivalry structures, as in U.S.-Soviet détente. Case comparisons further illustrate how different constellations of parity and rivalry shape the form and durability of restraint. Applying this framework to the contemporary U.S.-China relationship suggests that the structural conditions for symmetric agreements are currently absent, yet history points to pathways for limited, domain-specific cooperation that may reduce risks of escalation. By linking long-term empirical trends with present-day great power rivalry, the paper contributes both a theoretical framework and new evidence to debates on arms control and strategic stability.
Featured Image
Photo by Markus Winkler on Unsplash
Why is it important?
My analysis of nearly two centuries of arms control agreements reveals that symmetric restraint where all parties mutually limit capabilities is historically rare, emerging only under specific structural alignments. Agreements of this type cluster during periods of relative capability parity and simplified dyadic rivalries, as seen in the U.S. Soviet détente of the 1970s. In contrast agreed asymmetric and imposed asymmetric arrangements dominate in postwar or multipolar settings, reflecting unequal power and fragmented preferences. Quantitative trends demonstrate a strong correlation between declining capability inequality (low Gini values) and the frequency of symmetric agreements. We argue higher rivalry complexity (high Hamming distance) corresponds to their decline & these findings suggest that arms control is not the product of goodwill but a rational response to strategic equilibrium. Applied to the U.S.-China relationship this pattern indicates that current conditions marked by asymmetric capabilities, entangled rivalries, and mutual suspicion are unfavourable to traditional arms control. Yet the diffusion of AI-enabled military technologies, autonomous systems, and cyber operations amplifies crisis instability, creating new incentives for limited, domain-specific cooperation. Incremental measures such as data sharing protocols, crisis communication channels, and transparency in AI decision systems could mitigate escalation risks. Our findings underscore that structural parity, simplified rivalry, and technological transparency remain the key prerequisites for sustainable strategic restraint in the evolving geopolitics of the digital age.
Perspectives
As Artificial Intelligence, cyber capabilities, and autonomous weapons systems redefine deterrence and escalation, the logic of arms control must evolve from rigid treaty frameworks toward adaptive, technology-sensitive mechanisms of restraint. First, policymakers must acknowledge that AI introduces opacity into strategic decision-making. Algorithmic unpredictability and data dependence can accelerate crises or trigger inadvertent escalation. Establishing AI transparency protocols such as joint risk assessments and incident notification systems could serve as a functional equivalent to traditional verification measures. Second, confidence-building measures remain essential where formal treaties are unlikely. Renewing military-to-military hotlines, pre-notification of missile tests, and agreements on cyber non-interference in critical command-and-control systems can stabilize relations even in the absence of parity. Third, a multilateral framework for emerging technologies linking the U.S, China, India and other major powers should integrate AI safety norms, export control coordination, and ethical standards into global governance mechanisms. We believe the challenge is not to replicate Cold War models of arms control but to adapt their underlying logic mutual vulnerability, predictability, and communication to the technological realities of the twenty-first century. In this sense, arms control in the AI age must function less as a static agreement and more as a dynamic process of strategic adaptation to preserve international stability.
Biswayan Bhandari
Jadavpur University
Read the Original
This page is a summary of: Strategic Restraint and Rivalry: Structural Conditions for Arms Control in U.S.-China Relations, January 2025, Elsevier,
DOI: 10.2139/ssrn.5583131.
You can read the full text:
Contributors
The following have contributed to this page
