The idea that the world is becoming multipolar has acquired the status of conventional wisdom. It is repeated by governments, international organizations, business leaders, and much of the academic and policy community. The argument appears straightforward. American primacy is declining. China has risen. India is growing. Russia has returned as a major military power. BRICS is expanding. Middle powers such as Turkey, Saudi Arabia, Brazil, and Indonesia increasingly refuse to align with either Washington or Beijing. The unipolar moment is over; therefore, we are told, the world must be multipolar.
The first proposition is correct. The second does not follow from it. The end of unipolarity does not necessarily produce multipolarity. It can produce bipolarity. In fact, this is what has been happening during the past quarter-century.
Two transformations have occurred simultaneously. China has rapidly converged with the US across the principal dimensions of national capability. At the same time, the US and China have increasingly separated themselves from the states beneath them. The first process destroyed unipolarity. The second prevented the emergence of multipolarity. The result is not merely a bipolar international system. The more important point is that the international system is becoming more bipolar, not less.
The logic is simple. The gap between the US and China has narrowed. The gap between the US and China on the one hand and the rest on the other has increased. What we are witnessing is not the multiplication of poles but the formation of a second pole and the increasing concentration of comprehensive power at the summit of the international system.
Stay ahead of the geopolitical week.
MD Briefing delivers expert analysis across five global fronts — the Indo-Pacific, energy, geoeconomics, European security, and the Middle East — every Monday morning. Free.
What is a pole?
Much of the confusion begins with a loose use of the concepts of power and polarity. In the realist tradition, power ultimately rests on the capabilities and resources controlled by states, and a pole is distinguished not by excellence in a single category but by the possession of comprehensive power. For contemporary purposes, this requires a broader definition. A twenty-first-century pole must possess what I would call full-spectrum, full-stack power.
Full-spectrum power refers to breadth: substantial capabilities across the economic, industrial, technological, financial, military, nuclear, demographic, maritime, and network dimensions of power. Full-stack power refers to depth: the capacity to connect the different layers through which resources are transformed into usable power.
The stack begins with underlying resources: population, territory, geography, natural resources, and energy. These provide the material foundations of potential power. Above them lies latent power: wealth, industrial capacity, technological sophistication, and human capital. Latent power is converted into manifest capabilities: conventional and nuclear military forces, maritime capability, financial power, and positions of centrality within international trade, industrial production, logistics, communications, and other networks. Finally comes strategic conversion: the political and institutional capacity to mobilize these resources, combine them, and translate them into strategic effects. Political and institutional capacity operates throughout the chain, determining how efficiently resources and latent power become actual power.
This is in tune with the traditional realist conception of power and polarity but updates it for the twenty-first century. Financial networks can now be weaponized. Industrial supply chains can generate coercive leverage. Maritime logistics determine access to markets and resources. Technological ecosystems increasingly shape both economic productivity and military effectiveness. These have become important components of national power and must therefore be incorporated into any contemporary assessment of polarity.
A state does not need to rank first in every category. Neither the US nor China does. Nor does full-stack power imply autarky. The US, for example, generates part of its power through a network of alliances and technologically advanced partners. But a pole must possess substantial capabilities across the full spectrum and sufficient depth through the stack to transform resources into strategic effects at a systemic scale.
This immediately separates poles from merely important powers. Russia possesses enormous nuclear and important military capabilities. Japan and South Korea possess advanced industrial and technological capabilities. India possesses continental-scale nuclear weapons, a huge population, and rapidly growing economic resources. Britain and France possess nuclear weapons and advanced technology. European countries collectively command enormous economic and technological resources. They all possess important parts of the spectrum or stack. Only two states presently possess something approaching the whole: the US and China.
Myth one: Western decline means multipolarity
The first myth is perhaps the most consequential. We are repeatedly told that the declining share of world output accounted for by the G7 demonstrates the emergence of a multipolar economy. The old Western industrial powers are losing relative weight. Asia is rising. Production has spread across the developing world. The Global South has become more important. This is all correct. But it tells us only that the old distribution of power has changed. It tells us nothing about what has replaced it.
The crucial question is not whether economic and industrial capabilities moved away from the traditional Western powers. They did. The question is where they went. A disproportionate amount went to China.
At the beginning of this century, the US and China together accounted for roughly one-third of world GDP at market exchange rates. Today they account for 43 percent. In other words, during a quarter-century supposedly characterized by the diffusion of economic power, the share of world output concentrated in the two states at the summit actually increased.
Purchasing-power measures increase the relative economic weight of India and Russia, but they do not alter the larger picture. Even on PPP measures, India remains a distant third, while neither India nor Russia approaches the US and China across the wider spectrum of economic, industrial, technological, and financial capabilities.
The transformation is even clearer in research and development, an important indicator of future power. In 2000, the US accounted for 38.9 percent of global R&D expenditure. China accounted for only 4.8 percent. Together they represented 43.7 percent of the world total. At that time other centers of technological capability were formidable: the European Union accounted for 23.6 percent and Japan for 14.3 percent.
By 2024 the distribution had been transformed. China’s share had risen to 29.6 percent, and America’s stood at 29 percent. Together they accounted for 58.6 percent of global R&D. Meanwhile, the European Union’s share had fallen to 17.6 percent and Japan’s to only 6.7 percent. This is not multipolarization. It is bipolarization.
Manufacturing tells an even more dramatic story. China now accounts for approximately 32 percent of global manufacturing value added, and the US another 15 percent. Together they account for almost 47 percent of world manufacturing production. Japan, Germany, India, and South Korea combined account for little more than 17 percent. Again, this is not the emergence of several comparable industrial poles. It is concentration.
For twenty-five years analysts have watched the relative decline of American predominance and assumed that power was dispersing. They were watching only half of the transformation. China was not merely closing the gap with the US. It was widening the gap with almost everyone else. This is the central fact that the multipolarity narrative misses. China’s convergence with the US ended unipolarity. China’s divergence from the rest prevented multipolarity.
There has been another concentration at the same time. Within the traditional West, the US has proved substantially more dynamic than Japan and the major European economies. Treating the G7 as a single unit conceals this redistribution. The relative decline of the G7 does not amount to an equivalent decline of the US.
Two processes have therefore operated simultaneously: the concentration of Western power around the US and the concentration of Asian power around China. Together they have produced increasing concentration at the global summit.
Myth two: more autonomous states mean more poles
The second myth confuses strategic autonomy with polarity. India refuses alignment. Turkey maneuvers continuously among great powers. Saudi Arabia diversifies its partnerships. Brazil seeks greater autonomy. Indonesia is becoming more
consequential in its region. BRICS has expanded substantially. Countries across the Global South increasingly resist Western preferences. This is constantly offered as proof of multipolarity. It proves absolutely nothing!
Strategic autonomy is not polarity. Diplomatic insubordination is not polarity. Regional influence is not polarity. A country can become considerably more important in its region without becoming a pole. In fact, the growing autonomy of secondary states may itself be a consequence of bipolar competition. When the two superpowers compete for markets, resources, access, bases, technologies, and diplomatic support, states positioned between them acquire bargaining leverage. They can hedge, extract concessions, and play one side against the other, as Turkey frequently does.
None of this is new. Tito’s Yugoslavia pursued considerable autonomy during the Cold War. Nasser’s Egypt maneuvered between the superpowers. Nehru’s India became a leader of the Non-Aligned Movement. Nobody concludes from this that the Cold War international system was multipolar. Why, then, should the expansion of BRICS demonstrate multipolarity today? Analysts should count power capabilities, not rhetoric.
Myth three: Russia makes the system tripolar.
The most serious argument against my interpretation is that Russia constitutes a third pole. Russia is, after all, one of the world’s leading military powers and possesses one of its two largest nuclear arsenals. On a definition of polarity that places the decisive threshold on the ability to wage a major conventional war and maintain a survivable nuclear deterrent, Russia clearly qualifies. This is essentially the position taken by John Mearsheimer, an important international relations theorist and fellow realist, who consequently describes the contemporary system as tripolar.
But the conclusion changes when power is measured comprehensively. Russia possesses important conventional military forces, enormous territory, substantial natural resources, and a demonstrated capacity to sustain prolonged warfare. Its nuclear capabilities alone ensure that Russia will remain one of the most important states in the international system. But importance is not polarity.
Russia does not remotely approach the US or China in economic scale, manufacturing output, global trade, financial power, R&D expenditure, technological ecosystems, commercial network centrality, or maritime-industrial capacity. Its formidable nuclear and military capabilities sit on top of a much narrower economic and technological resource base.
Russia is a nuclear superpower. It is not a full-spectrum, full-stack superpower. Nuclear weapons provide Russia with the capacity to inflict catastrophic destruction, deter attacks, and manipulate escalation risks. But destructive capacity cannot by itself define polarity. If it did, the distinction between a pole and other nuclear-armed states possessing secure second-strike capabilities would become increasingly difficult to sustain.
The relevant question is therefore not simply which states can wage a serious war. It is which states control the range of resources and capabilities from which systemic power can be generated and sustained. On that measure, the discontinuity between the US and China and everyone else becomes much clearer.
The politics of multipolarity
There is another reason why the idea of multipolarity has become so difficult to dislodge. Almost everybody has an interest in believing in it, or more precisely in saying that they believe in it.
For Russia, multipolarity is the language of great power status. A multipolar world is, almost by definition, a world in which Russia remains one of the poles. The concept therefore elevates Russia’s exceptional nuclear and substantial military capabilities into a claim to comprehensive great-power status.
India has different reasons for embracing the same language. India possesses enormous potential power. Its population, economic growth, geography, and nuclear capabilities make it the most plausible candidate eventually to become another full-spectrum pole. Describing the present world as multipolar effectively grants India today some of the status its leaders expect it to acquire in the future.
Europe also finds the language of multipolarity extremely convenient. Collectively, European states possess enormous wealth, technology, and military resources. But resources distributed among sovereign states are not equivalent to resources controlled by a single strategic actor. Europe lacks unified political authority over taxation, military mobilization, nuclear strategy, and the crucial decisions of war and peace. Calling the world multipolar allows Europeans to avoid confronting an uncomfortable reality: Europe has moved from being the center of the international system to becoming a collection of wealthy but strategically weak powers.
A significant increase in European military expenditure does not solve this problem. Resources become systemic power only when a political unit can mobilize and convert them strategically. Without political unity, aggregation on a spreadsheet does not create another pole.
Middle powers have still another reason to welcome the narrative of multipolarity. The language magnifies their growing autonomy. Turkey, Saudi Arabia, Brazil, Indonesia, and others understandably prefer an international order described as containing multiple centers of power rather than one structured primarily around two superpowers.
China presents the greatest irony. Beijing has long promoted the language of multipolarity even though China’s own extraordinary rise has been the principal force producing bipolarization. There are good strategic reasons for this. Openly accepting bipolarity would mean acknowledging that China has become America’s systemic competitor. It could reinforce perceptions of a new Cold War and sit uneasily with Beijing’s continuing presentation of itself as part of the developing world.
Multipolarity is therefore an unusually useful concept. Russia gains status. India anticipates status. Europe disguises its relative decline. Middle powers magnify their autonomy. China avoids formally accepting the role of America’s systemic rival. But international structure is not determined by diplomatic preference. It is determined by the distribution of capabilities in the international system.
The Cold War paradox
There is also a revealing contradiction in contemporary language. Many of the same observers who describe the international system as multipolar increasingly speak of a new Cold War. Yet the rivalry they have in mind is overwhelmingly between the US and China. An intense strategic rivalry between two states can also exist inside a multipolar system. Anglo-German rivalry before 1914 is an obvious example. But the widespread Cold War analogy is nevertheless revealing because the Sino-American competition is becoming increasingly system-organizing.
The principal contest over advanced semiconductors is Sino-American. The military balance in East Asia is increasingly organized around Chinese capabilities and American countervailing power. Alliance adaptation in the Indo-Pacific is driven principally by the same competition. Trade controls, industrial policy, critical minerals, supply-chain security, maritime power, and advanced technologies are increasingly viewed through the interaction between Washington and Beijing.
Mearsheimer himself, despite describing the system as multipolar, calls the US-China competition “the most important facet” of the new structure. He identifies China, not Russia, as America’s peer competitor and describes their competition over artificial intelligence, quantum computing, biotechnology, microchips, and robotics as central to the future balance of power. The terminology says multipolarity. Much of the actual analysis describes an increasingly bipolar competition.
A quarter-century of bipolarization
The argument becomes clearest when polarity is treated not as a snapshot but as a historical process. Over the past quarter-century, two changes have occurred simultaneously: China has moved closer to the US, while the distance separating both from the next tier has increased.
The numbers already discussed capture the scale of the transformation. The significance of these figures lies not in any single measure but in their cumulative direction. Two things are happening simultaneously: concentration, measured by the growing share of important global capabilities controlled by the US and China; and discontinuity, measured by the distance separating them from the powers below. The first tells us that power is clustering at the top. The second tells us why the resulting structure is bipolar rather than multipolar. The great geopolitical transformation of the past quarter-century has therefore not been the multiplication of poles. It has been the formation of the second pole.
And then there is artificial intelligence
There is one final reason to suspect that bipolarization may have considerably further to run.
Artificial intelligence is surrounded by extraordinary claims about its potential effects on productivity, scientific discovery, technological innovation, and warfare. Some may prove exaggerated. We do not need to resolve that debate here because the argument for increasing bipolarity does not depend on AI.
What matters is that the resources required to develop and deploy advanced AI are themselves highly concentrated. Frontier AI requires enormous capital investment, advanced semiconductors, computing infrastructure, data centers, energy, highly skilled human capital, and sophisticated technological and industrial ecosystems. The US and China possess these complementary capabilities and are committing resources to AI on a scale that greatly separates them from the rest of the world. Whether American investment is presently larger than Chinese investment is secondary. The relevant fact is the distance separating both from everyone else.
Three scenarios follow.
If the more ambitious claims made for AI prove largely exaggerated, nothing in the preceding argument changes. The movement toward increased bipolarity can already be clearly demonstrated over the past quarter-century using conventional measures of national power.
If AI becomes a significant source of economic productivity, technological innovation, and military capability, which is the more plausible scenario, the consequences for the distribution of global power become more important. Because the two states best positioned to exploit the technology at scale are already the two full-spectrum powers, AI would add another layer to their latent and manifest capabilities. The gap separating the US and China from the rest would widen further.
The third possibility is more consequential. If the most bullish expectations surrounding AI prove broadly correct, if it produces very large productivity gains, dramatically accelerates scientific and technological innovation, and transforms important dimensions of military power, then its geopolitical effects could be profound. The benefits would not begin from an equal distribution of capabilities. They would be superimposed on a system in which the resources necessary to exploit AI at scale are already heavily concentrated. The distance separating the two poles from everyone else could consequently become enormous.
AI is therefore not the foundation of my argument for bipolarity. It is a potential accelerator of a historical process already well under way. If AI disappoints, the existing argument stands. If AI matters significantly, bipolarity deepens, and the gap between the two poles and the rest widens further. If its most ambitious promises materialize, that gap could become enormous.
The irony is difficult to miss. Multipolarity has become conventional wisdom at precisely the moment when the material distribution of comprehensive power is moving in the opposite direction.
