By Robbin Laird
It is no accident that I am writing this essay early in the morning in Canberra, Australia. There is no government which uses the term “rules based order” more than the Australian one does.
For most of the postwar period, the phrase “rules-based order” carried a single, largely unexamined meaning: a body of rules invented by the democratic powers, principally the United States and its allies, and enforced by them, sometimes through institutions, sometimes through raw preponderance of power.
The phrase did real work. It distinguished a world governed by agreed norms from one governed by naked coercion, and it gave Western policymakers a vocabulary for criticizing rivals who violated agreements they had ostensibly joined.
That vocabulary has stopped doing its job cleanly. The rising multipolar authoritarian powers — China, Russia, and to a lesser degree a widening circle of states hedging between blocs — have not simply rejected the rules-based order. They have done something more complicated: they leverage its rules selectively, when doing so serves their interests, while simultaneously constructing parallel codes of conduct, institutions, and normative languages of their own. Meanwhile, a genuinely different kind of rules-based order, the one that keeps aircraft from colliding, telecommunications networks interoperable, and financial messaging systems functional, persists largely untouched by any of this contestation, because everyone, regardless of regime type, needs it to work.
The result is a term that is asked to describe three fundamentally different phenomena at once, and it collapses under the weight.
The confusion dissolves, however, once the phrase is unbundled into its constituent layers: a functional order that is genuinely universal, an institutional order that was never as neutral as it claimed to be, and an emerging contest over who gets to write that institutional order’s successor. Understanding defense architecture, alliance cohesion, and the broader strategic environment in the years ahead depends on keeping these three layers analytically distinct.
Layer One: The Functional Order
The first layer is the one people invoke least often when they argue about “the rules-based order,” yet it is the layer with the strongest claim to universality. This is the world of the International Civil Aviation Organization’s flight rules, the International Telecommunication Union’s spectrum allocations, the International Maritime Organization’s collision-avoidance regulations, SWIFT’s financial messaging standards, and the technical protocols that route traffic across the internet. It also includes identification-friend-or-foe conventions, air traffic separation standards, and the increasingly dense web of data and communications standards that make modern shipping, aviation, and finance possible at all.
What distinguishes this layer is the nature of the incentive to comply. States and firms follow these rules not because a hegemon polices them and not because they reflect a shared political philosophy, but because the rules solve a coordination problem that every actor needs solved, regardless of ideology. An aircraft that ignores separation standards does not strike a blow against Western hegemony; it risks a midair collision with a Chinese, Russian, or Emirati aircraft as readily as an American one. A state that abandons agreed maritime signaling conventions degrades its own navy’s ability to operate safely as much as anyone else’s. A bank that departs from SWIFT messaging standards cannot transact with the rest of the system at all.
This is why authoritarian powers that reject, undermine, or openly flout the liberal institutional order described below remain scrupulous,sometimes more scrupulous than democracies, about compliance with Layer One. Russian and Chinese airliners still fly ICAO-compliant routes and squawk standard transponder codes. Chinese shipping still observes IMO collision regulations. Even amid full-scale war in Ukraine, both belligerents have continued to rely on internationally standardized communications and navigation protocols for the vast majority of their civil and much of their military traffic, because the alternative is not defiance of an adversary’s system but the sacrifice of one’s own operational competence.
Layer One rules, in other words, are closer to engineering standards than to law in the political sense.
Their legitimacy does not derive from the moral or political authority of whoever wrote them; it derives from the fact that they work, and that defection carries an immediate, self-inflicted, technical cost rather than a diffuse, negotiable, political one.
This is the part of “rules-based order” that genuinely deserves to be called near-universal but calling it an “order” at all is slightly misleading. It is better understood as a stack of interoperability standards that any actor, however revisionist its politics, has a rational interest in preserving.
Layer Two: The Liberal Institutional Order
The second layer is what most participants in the current debate actually mean when they say “rules-based order,” and it is the layer under genuine, accelerating contestation. This is the architecture built primarily by the United States and its allies after 1945, the Bretton Woods institutions, the General Agreement on Tariffs and Trade and its successor the World Trade Organization, the United Nations Charter system, the web of human-rights instruments, and the broader normative claim that sovereign states, however unequal in power, are bound by a common set of legal and political expectations.
This layer was never neutral, and pretending otherwise has done real damage to its credibility.
It encoded a specific set of assumptions that happened to serve the interests of the powers who wrote it: open market access on terms that favored economies already industrialized; a doctrine of sovereign equality that existed on paper in the UN General Assembly but coexisted with a Security Council structure that gave five states — four of them Western or Western-aligned at the order’s founding — a permanent veto; and a human-rights regime that its principal authors were prepared to invoke against rivals while exempting themselves, their client states, and their own historical conduct from equivalent scrutiny.
Bretton Woods institutions were headquartered in Washington and governed by voting shares that reflected 1944 economic realities long after those realities had changed. The order’s authors retained, and still retain, disproportionate power to enforce its provisions against others, waive them for themselves, or rewrite them unilaterally when convenient, sanctions regimes, extraterritorial financial reach through dollar clearing, and selective invocation of “rules-based” language against adversarial conduct while tolerating comparable conduct from allies are all symptoms of the same underlying asymmetry.
None of this means the order was worthless or purely self-serving. It provided real public goods: a framework for trade liberalization that lifted hundreds of millions out of poverty, a forum for dispute resolution that substituted adjudication for war in a great many cases, and a normative vocabulary that genuinely constrained even powerful states some of the time.
But its legitimacy always rested on a bargain: rising and non-Western powers would accept a system whose founding terms they had no hand in writing, in exchange for access to its benefits and the promise, frequently deferred, sometimes betrayed, that the system’s governance would eventually be rebalanced to reflect changing realities.
That bargain is now visibly failing.
Rising powers no longer accept that the terms set in 1944 or 1945, or even those set amid the unipolar triumphalism of the 1990s, should continue to govern institutions in which their economic and demographic weight has grown enormously.
This is the layer where the phrase “rules-based order” has become genuinely contested, not because rules as such are rejected, but because the specific rules, and more importantly the specific distribution of authority to write, interpret, and enforce them, are seen by an increasing number of major states as illegitimate on their face.
Layer Three: Counter-Order Building
The third layer is the newest and the least understood, and it is where the user’s original observation that authoritarian powers “leverage” existing rules “when useful to them” while “developing their own code of conduct” is precisely correct. What China, Russia, and an expanding set of hedging middle powers are doing is neither simple compliance with Layer Two nor simple rejection of it. It is selective extraction combined with parallel construction.
The selective extraction is visible everywhere.
China remains a WTO member and uses its dispute-settlement mechanisms when doing so serves Chinese commercial interests, even as it violates the spirit and often the letter of WTO commitments on subsidies, market access, and intellectual property.
Russia retains its permanent Security Council seat and veto, one of the load-bearing asymmetries of Layer Two and uses it to block action against itself while invoking the UN Charter’s language on sovereignty and non-interference when it suits Moscow’s purposes, including, with no apparent sense of contradiction, in relation to its own invasion of Ukraine.
Both states participate in international financial institutions, human-rights forums, and standard-setting bodies precisely where doing so preserves useful access or influence, while treating the same institutions as instruments of Western control when their verdicts go the other way.
The parallel construction is the more consequential long-term development.
BRICS, expanded to include Iran, the UAE, Egypt, Ethiopia, and Saudi Arabia’s eventual participation under discussion, is explicitly framed by its members as a vehicle for building institutions, a New Development Bank, currency-settlement mechanisms outside SWIFT, credit-rating agencies independent of the Western “big three”, that reduce dependence on Layer Two architecture.
The Shanghai Cooperation Organisation performs a similar function on the security and diplomatic side for Eurasia.
China’s Cross-Border Interbank Payment System, still modest relative to SWIFT but growing, is a direct hedge against the demonstrated willingness of the United States to weaponize dollar-clearing access, a lesson reinforced for every non-aligned finance ministry by the sanctions imposed on Russia after 2022.
The Asian Infrastructure Investment Bank and the Belt and Road Initiative’s financing architecture perform an equivalent function for development lending, offering terms and conditionality explicitly marketed as free of the governance strings attached to World Bank and IMF lending.
Critically, this parallel construction comes wrapped in its own normative vocabulary, not merely its own institutions. “Non-interference in internal affairs” is asserted as a superior organizing principle to the human-rights conditionality embedded in Western aid and lending. The language of “civilizational states”, China and Russia both use versions of this framing, asserts a right to define governance and rights on terms rooted in a state’s own historical and cultural trajectory rather than a universalist standard set elsewhere.
And “multipolarity” itself has been elevated from a description of the distribution of power into a positive value: a multipolar world is presented not merely as an emerging fact but as more legitimate, more representative, and more just than a world organized around a single set of institutions and norms.
This is not rule-rejection.
It is counter-rule-making: a deliberate, well-resourced effort to write Layer Two’s successor on different terms, while continuing to extract whatever benefits remain available from the existing system and never once questioning the necessity of Layer One.
The Mechanism: Unbundling Layer Two from Layer One
Once the three layers are separated, the apparent paradox that motivates this article, powers that flagrantly violate “the rules-based order” while remaining, in other respects, meticulous rule-followers, resolves into a single underlying mechanism: Layer Two is being unbundled from Layer One.
For most of the postwar period, these two layers were treated, especially in Western policy discourse, as a single seamless fabric. Adherence to technical and functional standards was implicitly bundled together with adherence to the broader liberal institutional and normative order, as though a state that flew ICAO-compliant routes and cleared payments through SWIFT was, by that same behavior, endorsing the WTO’s trade disciplines, the UN’s human-rights architecture, and the broader claim of Western institutional leadership. That bundling was always more assumption than logical necessity, and revisionist powers have spent the past decade demonstrating precisely how separable the two layers are.
The mechanism becomes visible in each domain where it operates. In trade, China exploits WTO market access, a Layer Two benefit, while running an industrial-subsidy regime that violates WTO disciplines, secure in the knowledge that dispute-settlement remedies are slow, partial, and rarely reverse the underlying advantage gained.
In finance, Russia continues to rely on Layer One payment-messaging protocols wherever sanctions permit, while treating the Layer Two institutions that impose those sanctions and the normative claims used to justify them as illegitimate instruments of a hostile bloc.
In maritime affairs, China simultaneously observes IMO safety and navigation standards for its commercial fleet while rejecting, in the South China Sea, the Permanent Court of Arbitration’s 2016 ruling under the UN Convention on the Law of the Sea, accepting Layer One’s collision-avoidance mechanics while rejecting Layer Two’s dispute-adjudication authority over the same waters.
The reason the unbundling is possible, and the reason it will likely deepen rather than reverse, is that the two layers rest on different foundations of legitimacy. Layer One is a coordination good: its rules are, in game-theoretic terms, close to a focal-point equilibrium that every rational actor wants regardless of preferences over outcomes, because the alternative, genuine fragmentation of air-traffic control, payment messaging, or maritime signaling, imposes costs on everyone roughly in proportion to their participation in the global system. Layer Two, by contrast, was always partly a hegemonic good: a system whose rules, however useful in aggregate, also functioned to entrench the relative power and preferences of the states that wrote them. A coordination good does not lose legitimacy when the balance of power shifts; a hegemonic good does, because the shifting balance of power is precisely what its critics point to as evidence the rules no longer reflect who actually holds the pen.
For analysts and policymakers, the practical implication is that continued compliance with Layer One should not be read as continued buy-in to Layer Two, and defection from Layer Two should not be expected to produce defection from Layer One. These are separable choices with separable logics, and treating them as a single bundle either by assuming that technical cooperation implies normative alignment, or by assuming that normative defection threatens technical cooperation will consistently produce bad predictions about how revisionist powers actually behave.
Implications for Allied Defense Architecture
The same unbundling logic operates inside the alliance system, and it bears directly on the kill web and interoperability frameworks this analysis has applied to allied maritime and air power. Military interoperability standards — NATO STANAGs, Link 16 and its successors, IFF protocols, common data formats for sensor fusion across a kill web — function structurally like Layer One. They are coordination goods: a data link standard that lets an Australian frigate, a Japanese destroyer, and a U.S. carrier strike group share a common tactical picture has value that is largely independent of the political temperature between Washington, Canberra, and Tokyo on any given day. States adopt and maintain these standards because the alternative, operating blind alongside a partner force, or rebuilding a unique interface for every coalition, imposes an immediate operational cost that has nothing to do with underlying political alignment.
Strategic and political alignment among allies, the willingness to commit forces, share the most sensitive intelligence, extend nuclear guarantees, or accept the political risk of forward basing, is closer to Layer Two: a normative and institutional commitment that rests on shared threat perception, domestic political consensus, and trust accumulated (or eroded) over time.
And exactly as with the global order, there is no guarantee these two layers move together. A coalition partner’s technical interoperability with the United States can persist, and even deepen even as that same partner hedges its strategic alignment, expanding economic ties with China, questioning the durability of U.S. security guarantees, or declining to align rhetorically with Washington on issues where its own regional interests diverge.
This has a genuinely double-edged implication for allied force design.
On one hand, it is good news: interoperability investments are not wasted the moment political consensus frays, because the technical layer has its own independent logic and its own independent payoff. A kill web architecture built around common standards retains value as an option even during a period of strategic drift, in the same way ICAO compliance retains value for a state that otherwise rejects the broader Western institutional order. This argues for continuing to invest heavily in Layer One-equivalent interoperability, data standards, common training pipelines like the CNATRA “Street to Fleet” model, shared sustainment ecosystems precisely because these investments are more durable and less politically fragile than the alliance commitments that sit above them.
On the other hand, it is a warning against a complacent assumption that technical interoperability is a reliable proxy for, or guarantor of, strategic reliability. A partner can be fully plugged into the kill web’s data architecture and still decline to commit forces, share the most sensitive targeting data, or stand alongside its ostensible ally in a crisis, exactly as a state can be fully compliant with ICAO and SWIFT while actively working to dismantle the institutional order those systems are conventionally assumed to support. Middle powers navigating a multipolar order, the audience for much of the Australian, Brazilian, and Indian hedging analysis this framework connects to, are likely to become more, not less, adept at exactly this kind of unbundling: maximizing the technical benefits of interoperability with multiple partners, including partners who do not trust one another strategically, while carefully preserving political and strategic flexibility above that technical layer.
Allied planners should expect, and design for, a world in which the kill web’s technical architecture is shared more widely than the strategic commitments it is meant to serve.
Conclusion: Precision Over the Blanket Term
“Rules-based order” will continue to be used as a single phrase in political speeches, and there is no realistic prospect of retiring it from the vocabulary of diplomacy.
But analysts, and the policymakers who rely on their analysis, gain little by using the term as loosely as public discourse does. The confusion the user identified that today’s authoritarian powers seem simultaneously to defy and depend on a rules-based order is not a genuine paradox. It is the predictable consequence of collapsing three distinct things into one word: a functional order that is close to universal because it is a coordination good nobody has an interest in destroying; an institutional order that was always partly a hegemonic good, and whose legitimacy is now eroding precisely because the distribution of power that once justified its terms has shifted; and an emerging, well-resourced project of counter-order building that extracts what remains useful from the old institutional layer while constructing an alternative on different normative foundations.
The real contest of this era is not over whether a rules-based order should exist, every serious power, revisionist or status quo, needs and wants the functional layer to keep working.
The real contest is over who writes the institutional layer’s successor, on whose terms, and with whose interests embedded in its architecture. Seeing that contest clearly, rather than through the flattening lens of a single undifferentiated phrase, is the necessary first step to responding to it, in trade policy, in alliance management, and in the design of the interoperable defense architectures that will have to function across an increasingly contested and unbundled global order.
Note: I would like to thank my Australian colleague John Blaxland for insisting on a more nuanced discussion of the rules based order.