Wallerstein vs. Zeihan: is globalisation dying, or changing its state of matter?

(Marxism versus geographical determinism)

The discrete nature of globalisation

For all of humanity’s obvious and undeniable progress, the world of the mid-2020s is still astonishingly contradictory.

On the one hand, it seems that cooperation and interdependence are only deepening. Raw materials, energy and foreign markets still play a decisive role in whether a country functions properly. In this respect there is no difference between, say, Europe, the United States and China. The sheer durability of global production chains is demonstrated vividly enough by the US itself, which is trying to bring manufacturing back with the help of often non-market mechanisms. It turns out, though, that political will alone (or, in Trump’s case, hysterics) is not enough, because a globalised market gravitates towards more comfortable — above all cheaper — jurisdictions. Mexico, for instance.

On the other, there are signs that in decision-making the market is losing its controlling stake. First and foremost to political security, whose primacy came to the fore after Russia’s attack on Ukraine / Donald Trump’s arrival in power. The events accompanying both tragedies point to exponential growth in the political-security factor, which often wins out in argument with economic reasoning.

Everyone wants to buy cheap energy, build factories and sell technology. But now they do it less riskily than before, on the principle that it is better to lose out on profit than to get it in the neck. There are, of course, still enthusiasts for fishing in troubled waters — buying sanctioned Russian energy, for example. But that is a matter of an economy adapting slowly rather than of a conscious ideological choice.

Analysing the geopolitical and geoeconomic events of this decade brings us to two diametrically opposed approaches, both suited to making sense of the current historical rupture.

The author of the first is the Marxist Immanuel Wallerstein, a sociologist and the author of world-systems analysis.

The second is Peter Zeihan, an analyst and forecaster who held senior positions at Stratfor.

The difference between them is that the first looks at states as elements of a global capitalist system. The second, conversely, starts from the state and its assets.

They have something in common too: the rejection of the “end of history” thesis. Particularly where globalisation is concerned, which Fukuyama regarded as the factor behind final planetary unification. Wallerstein and Zeihan, by contrast, both call it into question.

Wallerstein is the more academic: for him this is merely another malfunction in a system at least 500 years old.

Zeihan is the more radical. He works with a short chronological lever and predicts the death of precisely the “American model of globalisation” — the one built after 1945.

Uneven accumulation as the norm

To understand the difference between their approaches, let us look at Wallerstein’s view of history. One of the cardinal errors of traditional social science, he believed, was the axiom by which the individual state was treated as the principal unit of development.

Every question was therefore posed through that prism:

• why did one country grow rich and another not?
• why did one industrialise while another remained a source of raw materials?

The most widespread answer suggests looking for the causes inside the country itself, in its institutions: culture, education, religion, governance and so on.

Wallerstein changes the scale and takes the international division of labour as his unit of analysis. In his view the modern capitalist world-economy emerged in the sixteenth century and spread across the entire globe. Its task, its logic and its imperative is the permanent accumulation of capital. That is precisely what produces a structure of core, periphery and semi-periphery:

• The core concentrates complex production, capital, financial institutions, technology and political power.
• The periphery supplies cheap labour, raw materials and goods with lower added value.
• The semi-periphery occupies the middle ground. It can exploit weaker regions and at the same time remain dependent on stronger ones.

The crucial point is that this inequality is not temporary, since peripheral countries will not necessarily “catch up” with the core. Time does not heal the wounds of “belated modernisation”. What is more, the exploitation of their resources may be one of the conditions of the core’s prosperity — a core that, like an antlion, draws the weaker and the careless down into its funnel.

This, in fact, is one reason why world-systems analysis is not about modernisation happening “sooner or later”. Modernisation’s advocates hold that there is one road and merely different speeds. Wallerstein, by contrast, is convinced — and sets out to convince — that unevenness itself is the heart pumping the system’s blood. Which is why some regions concentrate the profits and others the costs.

His understanding of hegemony follows from this. In theory the world-system has no permanent political centre. In practice, in particular historical periods one state or another gains an advantage in production, trade, finance and arms, and therefore becomes the hegemon. First the Netherlands, then Great Britain, now the United States.

Yet hegemony, both as an option and as a phenomenon, is always temporary. Sooner or later competitors take over the technologies, production moves elsewhere, the productivity advantage narrows, and the cost of maintaining global order rises.

In Giovanni Arrighi’s theory, close to Wallerstein’s, financialisation becomes one of the principal symptoms of hegemony’s late phase. At first the centre earns from production and trade, then increasingly from finance. What looks like the summit of power may at the same time be a sign that the cycle is closing.

The Netherlands lost hegemony — the system did not disappear. Great Britain lost global primacy, but capitalism did not disappear. In this logic, the decline of American hegemony does not mean the end of the global economy. The centre of accumulation will change.

And here the first fundamental difference with Zeihan appears. For Zeihan, American hegemony is one of the principal conditions of modern globalisation. For Wallerstein it is only one historical phase of a far older system. So if the US gradually gives up the role of global guarantor, that does not yet mean global economic ties have to fall apart. They may be rebuilt.

At the same time, it would be a mistake to imagine Wallerstein as a theorist of capitalism’s endless stability. On the contrary, in his later work he wrote more and more often about a structural crisis of the world-system itself. His argument was that capitalism can exist for as long as sufficiently profitable accumulation remains possible. For centuries the system compensated for its internal contradictions through geographical expansion, the incorporation of new cheap labour, the displacement of ecological costs and the use of new resource zones.

But the world has limits. When the cost of labour, material costs, the ecological price of production and the tax burden all rise, the scope for endless cheapening shrinks. At a certain point the system begins to reproduce itself less well. This is why Wallerstein spoke no longer simply of a crisis of hegemony but of a systemic bifurcation — the moment when the old system stops returning to equilibrium while the new one has not yet taken shape.

And here his thesis is far more radical than the usual story about American decline. He did not predict an automatic transition to something better. What comes after capitalism may be a more authoritarian, hierarchical and unequal order. That is, the crisis itself opens up not guaranteed progress but a struggle over the shape of the next system.

It is this part of Wallerstein that makes him unexpectedly close to the present. Wars, climate change, pandemics, demographic shifts, financial instability and political fragmentation really do look less like separate crises than like interconnected symptoms of a deeper transformation.

The world-systems tradition, moreover, did not disappear with Wallerstein’s death.

1. Christopher Chase-Dunn continued the analysis of hegemonic cycles and extended the world-systems approach to older historical systems.
2. Beverly Silver applied it to global labour history, showing how capital relocates production in search of cheaper labour — but how, along with the factories, labour conflict gradually migrates to the new regions too.
3. Ho-fung Hung studies China and American hegemony, asking whether China is really capable of becoming the system’s new centre when its own development was so deeply embedded in an America-centred order.
4. Jason W. Moore developed world-systems analysis in the direction of “world-ecology”, treating capitalism as a system that sustains accumulation through the constant reproduction of the “Four Cheaps” — cheap labour, food, energy and raw materials.

The return of geographical determinism

Zeihan moves from the opposite direction. Where Wallerstein advises us to rise above states, Zeihan appeals to the physical map.

What interests him is mountains, plains, ports, rivers, arable land, oil and gas deposits, the distances between industrial regions and sea lanes.

For him geography is not scenery but the foundation on which the political house stands. His favourite example is the US, which has an enormous domestic market, fertile land, an extraordinarily convenient system of navigable rivers, access to two oceans, large energy resources and no strong hostile neighbours along its land borders. Seen through distilled geographical determinism, then, America power is not only the product of well-designed institutions but, to a large extent, geographical luck.

The second key element of his model is demography. And this is precisely where his framework acquires particular weight. The point is that geography changes slowly, whereas demography makes it possible to explain why the break may be happening right now.

His logic is simple.

1. An agrarian society encourages high birth rates, because a child becomes an economic resource for the family relatively quickly.
2. Urbanisation changes household economics. In the city a child is an expense, so the birth rate falls.
3. A generation later, society has large cohorts of working-age people and relatively few children.

But in time those cohorts age. There are no longer enough young workers. The burden on the pension system, on medicine, on the labour market grows. Savings and consumption change. The country has fewer people able to work in industry, construction, logistics and energy.

This is why Zeihan sees demography not as social background but as one of the main driving forces of future geopolitics. And here his approach has real academic parallels. Charles Goodhart and Manoj Pradhan, for example, also write about how the ageing of the large economies and the shrinking reserve of cheap global labour are changing fundamental macroeconomic conditions.

But Zeihan’s most contested thesis concerns the origin of modern globalisation itself. For him the post-1945 world economy is not simply the natural result of developments in transport and trade. It is an “American geopolitical project”.

In his view, after the Second World War the US created a system in which

1. the American navy guaranteed the security of maritime routes,
2. the American market was open to allies,
3. the dollar became the foundation of the international financial system,
4. Washington’s security umbrella allowed Japan, Germany and other countries to concentrate on economic development.

In his optics, globalisation was part of Cold War strategy. America was in effect offering its allies an algorithm in which it underwrote global trade and access to its own market. In return, the allies helped contain the USSR.

It is precisely this process that gives Zeihan grounds to speak of the break-up of a single global economy, to be replaced by several regional systems.

In North America, for instance: the US, Canada, Mexico.

In Asia — Japan.

Europe will have structures of its own.

The states to suffer most, then, will be those critically dependent on long maritime chains, imported energy and a demographically ageing workforce.

Zeihan is especially pessimistic about China. In his framework China became the world’s manufacturing centre under conditions of safe sea lanes and access to global markets. Yet at the same time it is ageing, it imports a significant share of its energy and it depends on maritime routes. In this model China may turn out to be not the heir of American globalisation but one of its greatest casualties.

• Goodhart and Pradhan converge with him on the demographic argument.
• Dani Rodrik has long written about the crisis of hyperglobalisation and the return of space for national economic policy.
• Rana Foroohar criticises excessively long supply chains and talks about a shift from efficiency to resilience.
• George Friedman, like Zeihan, attaches great weight to geography and to the long-term structural advantages of the US.

This does not make them “Zeihanites”. What is at issue is a broader consensus on several trends: the ageing of populations, the return of industrial policy, rising political risk in world trade, the regionalisation of parts of supply chains, and weakening faith in endless hyperglobalisation.

The state versus the world-system

The main difference between Wallerstein and Zeihan can be put very simply: Zeihan sees the world through states, Wallerstein sees states through the world-system.

1. For Zeihan the US is strong because of geography, energy, food, demography, internal transport routes and naval capacity. For Wallerstein American power is the product of a specific historical phase in the development of the capitalist world-economy.
2. Zeihan inclines towards geographical and demographic determinism. Wallerstein towards structural economic determinism.
3. In Zeihan the state remains the principal “actor”. In Wallerstein it is only part of a larger system of capital accumulation.

China shows this especially well.

Zeihan asks whether China can physically sustain its industrial system amid an ageing population, import dependence and vulnerable sea lanes.

The world-systems researcher asks the opposite: are the functions of the core moving to China, where are profits concentrating, and can it become a new centre of accumulation?

Yet for all their divergences, Wallerstein and Zeihan have a great deal in common.

1. Both regard the global order as historically temporary.
2. Both are sceptical of the idea that American hegemony can last forever.
3. Both reject the notion of the world economy as a neutral market detached from power.
4. And both, in effect, see the 1990s not as the beginning of the “end of history” but as a brief and exceptional phase.

They are equally sceptical of the scenario in which the US gradually weakens and China painlessly takes its place.

For Wallerstein, transitions between systemic cycles are always bound up with chaos and struggle. For Zeihan, the very mechanism that allowed China to rise is on a downward track.

The weak points of both models

Zeihan’s weakness is that his models are sometimes too beautiful. Geography, demography, energy and logistics line up into a very clear causal chain. But real societies are capable of adapting.

For instance,

1. ageing is offset by migration and automation;
2. resource shortages by technological change;
3. dependency risk by diversifying suppliers.

And in general, the “skeleton” of any system — its political institutions — sometimes corrects geographical influences more forcefully than a determinist model allows.

His thesis about the role of the American navy is especially debatable. The security of sea lanes really is critically important, but global trade does not rest on the American military presence alone. There is international law, insurance, private shipping companies, port infrastructure, the navies of other states and, not least, the colossal interest that trading partners themselves have in keeping the system running.

America’s withdrawal from part of its global commitments therefore does not necessarily mean the automatic destruction of world trade. It may mean trade that is more expensive, more dangerous and more regionalised.

With Wallerstein the problem is the reverse. His system is too differentiated. Core, semi-periphery and periphery work well over long historical distances. Up close it is short-sighted. When, for example, one has to classify today’s China, South Korea or Singapore precisely.

China is at once an industrial centre, a capital exporter, a technological power and a country with incomes far below those of the old Western core.

The state, moreover, has turned out to be more autonomous than rigid world-systems schemes sometimes assumed. Chinese industrialisation is impossible without the Chinese state. American technology policy without the American one. The EU is not merely a market but a regulatory and political centre. In this sense Zeihan is right to bring the state back into the analysis.

Alternative theories

But the argument need not be reduced to two authors alone. There is a whole range of alternative theories that allow us to see the contemporary world differently.

Classical and structural realism need neither a capitalist world-system nor demographic collapse to explain the US–China confrontation.

1. For Kenneth Waltz or John Mearsheimer it is enough that China is growing stronger and the US fears losing its advantage. In a world without a supreme authority, great powers are always forced to balance one another. In this optics, trade wars, technology restrictions and military alliances are consequences of the struggle for security.
2. Liberal institutionalism gives the opposite answer. Robert Keohane and Joseph Nye show that international cooperation can survive even without an absolute hegemon. Institutions reduce uncertainty, rules make behaviour predictable, and economic interdependence raises the price of conflict. A weakening US therefore does not necessarily destroy the global system. It may become less American but remain global.
3. An even more interesting approach is the theory of global value chains. It shows how badly the modern economy fits into a simple geography of states. A smartphone may be designed in California, made from components from Taiwan, South Korea, the Netherlands and China, assembled in Vietnam, sold in Europe — while the bulk of the profit stays in the US. This is not simply trade between states but a complex transnational network.

Real deglobalisation, therefore, may not look like factories coming home. An American company leaves China, but moves production to Mexico. A Chinese corporation opens a plant in South-East Asia. Formally, direct ties between the US and China weaken, but the global network does not disappear — it becomes more complicated.

1. Dani Rodrik proposes another very useful frame: the end not of globalisation but of hyperglobalisation. His famous trilemma holds that it is impossible to maximise national sovereignty, democracy and extremely deep economic integration all at once. If the international market starts dictating too many domestic rules, states will inevitably try to win back room for policy.

In this logic, today’s protectionism is not necessarily the beginning of a catastrophe. It may be a correction. Trade will survive, but states will defend strategic sectors more actively, support their own production and limit dependencies.

1. Giovanni Arrighi allows the question to be posed differently. Perhaps what we are observing is not the end of globalisation but the completion of the American systemic cycle of accumulation? In that case the central problem becomes not the destruction of the world economy but the search for a new centre. Will China be it? Or, more broadly, East Asia? Or, conversely, will the system for the first time prove unable to produce a new hegemon?
2. William Robinson and the theory of transnational capitalism move in the opposite direction. They argue that capital is already so globalised that national states cannot be considered its only containers. Transnational corporations, financial networks and global elites cross borders so deeply that even the geopolitical conflict between the US and China takes place inside an already global capitalism.
3. Manuel Castells adds one more dimension: perhaps there is no single globalisation at all. Material production may regionalise while information globalises. The energy system may become more local while financial flows remain planetary. The internet may fragment politically while scientific knowledge goes on circulating around the world.

In that case the question “is globalisation ending or not?” loses part of its meaning. It is far more precise to ask: which flows exactly are deglobalising, and which are continuing to globalise?

Not deglobalisation but geoeconomic fragmentation

This is why today the word “deglobalisation” is increasingly giving way to the concept of geoeconomic fragmentation. It describes reality far better. World trade is not disappearing, but it is ever more strongly subordinated to political logic. Production networks are being rebuilt in line with alliances, sanctions and security risks. Goods may travel not by the shortest route but by the politically more reliable one.

And it is here that it becomes clear why Wallerstein and Zeihan are more usefully overlaid on one another than set against each other.

Zeihan sees globalisation’s material body better. Every good has weight. It has to be transported. Transport requires fuel. Production requires energy. Agriculture requires fertiliser. A factory requires workers. Export requires a port. World trade requires a safe sea.

The COVID-19 pandemic demonstrated that physicality very well. A world economy that for decades had seemed an almost bodiless network of finance and data suddenly turned out to depend on containers, ports, drivers, mask factories and a handful of semiconductor manufacturers.

Wallerstein, for his part, forces the next question: why do some control design, patents, finance and brands while others merely assemble the product? Why can a factory move from China to Vietnam while the bulk of the profit stays in California? Why do some countries remain raw-material exporters for decades?

This is precisely why the two optics complement each other especially well in the case of China. Zeihan is right that China is ageing and depends on imported energy and sea lanes. The world-systems approach is right that over recent decades China has acquired an enormous share of the world’s industrial base and ever more of the functions of the core.

But China has not become a fully autonomous centre. Its rise took place inside an America-centred system. It is strong enough to weaken American unipolarity but perhaps not autonomous enough to take the place of the US.

Globalisation is changing its state of matter

Here it is worth returning to the question we began with: is globalisation over or not?

If globalisation means the specific model of the 1990s–2010s, the answer is most probably yes. The era in which a corporation could build a supply chain almost exclusively on the criterion of the lowest price is ending. Alongside price there are now questions of war, sanctions, technological dependence, political loyalty and national security.

Semiconductors have stopped being merely a commodity. Energy has stopped being merely a market. Lithium, batteries, artificial intelligence, pharmaceuticals, telecommunications — more and more sectors are becoming strategic.

But if globalisation means a planetary system of flows of capital, goods, information, technology and labour, then talk of its death is premature. Even leaving China often means not bringing production home but moving it to another country.

And perhaps the best metaphor for this process is a change in the state of matter.

The globalisation of the 1990s resembled a liquid. Capital sought to flow to where profit was higher. Production, to where labour was cheaper. Goods, along the shortest routes. Political borders were perceived as obstacles to be reduced as far as possible.

Harder regional blocs are forming. North America. The European Union. The Chinese industrial zone. India. South-East Asia. The in-between countries trying to work with several centres at once.

The enormous flows of goods, capital and information between them do not stop. But crossing from one space into another becomes more expensive and more difficult. Tariffs appear, along with sanctions, export licences, investment screening, localisation requirements, technology restrictions.

And here Wallerstein probably offers the stronger long-term frame. The world-system has already changed hegemons, trade routes, technologies and production centres more than once. A change in the American order does not necessarily mean the end of the global system.

But Wallerstein also leaves a more radical question. Perhaps this time it is not simply another change of centre. Perhaps the capitalist system itself is entering a phase in which its old mechanisms of accumulation work ever less well?

If so, then today’s geopolitical fragmentation is only one part of a far deeper crisis.

Climate, demography, automation, artificial intelligence, housing costs, sovereign debt, political polarisation and the crisis of trust in institutions — all of this may turn out to be not a set of separate problems but symptoms of a systemic transition.

Zeihan explains better why the particular infrastructural form of globalisation created by American hegemony after 1945 is breaking down. Wallerstein explains better the longer historical system inside which that order arose and is now being rebuilt.

Globalisation in its late-twentieth-century form does indeed seem to be coming to an end. But the global world is not disappearing. It is regionalising, politicising and hardening.

The world is not returning to fully national economies. It is turning into an archipelago of large interdependent blocs.

So perhaps the most precise answer sounds like this: globalisation is not dying. It is changing its state of matter.

And the main intrigue of the coming decades lies not even in whether China becomes the new hegemon or the US retains its primacy. Something else is far more interesting.

Will the world-system manage to rebuild itself once again, changing the centre of accumulation and the form of globalisation — or are we this time really observing the moment when it is no longer the American cycle that is ending, but the very historical logic of a system that has existed for some five hundred years?

When capitalism dies.