Why the Same Geography Produces Opposite Political Outcomes

In 1648, two peace treaties signed at Osnabrück and Münster ended the Thirty Years War and drew the political map of central Europe in ways that would persist for centuries. The negotiators carved German-speaking Europe into hundreds of principalities, free cities, and ecclesiastical territories, creating a patchwork of micro-states in the heart of the continent. Meanwhile, across the English Channel, a single kingdom controlled the entirety of an island of comparable geographic complexity. Both landscapes — the Rhine basin and the British Isles — consist of navigable rivers, defensible upland zones, productive agricultural lowlands, and accessible coastlines. The geographic inputs were similar. The political outputs were radically different: fragmentation and particularism on one side of the Channel, centralization and proto-nationalism on the other. If geography determined political outcomes, we would have to explain why it apparently determined opposite ones within a day’s sail of each other.

This is the central embarrassment of deterministic geographical theories of politics. The correlation between geography and political form is real and meaningful — but it is not the tight, predictive relationship that its strongest advocates claim. The same rivers that unify one polity fragment another. The same mountains that isolate a culture and preserve its distinctiveness in one case become trade corridors in another. The same coastal geography that produces maritime commerce and cosmopolitan liberalism here produces piracy and tribalism there. Geography is necessary but never sufficient, and the gap between necessary and sufficient is where all the interesting political history happens.

What Geography Actually Determines

To understand why geography produces opposite outcomes, you first need to be precise about what it actually determines and what it merely influences.

Geography determines the cost structure of violence and trade. This is its fundamental political contribution. Mountain ranges increase the cost of military projection, making it expensive for lowland states to conquer highland peoples and giving those peoples a natural defensive advantage. River networks decrease the cost of trade, making commercial integration natural and organic along river corridors. Coastlines with navigable harbors decrease the cost of seaborne commerce and communication, giving coastal communities access to larger markets and broader information networks than their inland counterparts. Deserts and dense forests increase the cost of political administration over dispersed populations, making it difficult to maintain centralized control.

These cost structures create political tendencies, not political outcomes. A mountain people is harder to conquer, which creates a tendency toward political independence. But a mountain people that builds roads and develops a commercial culture can become an imperial center — the Romans built their empire from a city on seven hills. A river basin creates natural incentives for commercial integration, which creates a tendency toward political cooperation. But the Rhine basin, as the Holy Roman Empire demonstrated for most of a millennium, can also be a theater of perpetual political fragmentation if the commercial interests of river cities conflict sharply enough with the military interests of territorial princes.

The confusion arises because geographic cost structures make some political outcomes more probable than others — meaningfully more probable, in ways that leave statistical traces across large samples of cases. Jared Diamond’s argument in Guns, Germs, and Steel that the east-west orientation of the Eurasian continent gave it agricultural and epidemiological advantages over the north-south orientation of Africa and the Americas is a real argument about geographic cost structures, and it has real explanatory power at the longest time scales. But it cannot tell you why Switzerland became a confederal democracy and Germany became a military autocracy, or why South Korea developed into a liberal democracy while North Korea developed into the most complete totalitarian state in modern history. For those questions, geography is at best a weak constraint, and the explanatory work is done by institutions, contingent historical choices, and the agency of specific actors at specific moments.

The Island Illusion

The most seductive geographic theory of political form is the island theory: that islands develop distinctive political cultures because insularity reduces the threat of land invasion, allows the state to substitute naval power for land armies, and orients the population toward commerce rather than military organization. The theory has obvious applications to Britain, Japan, and Taiwan. It generates testable predictions.

It fails spectacularly on inspection. Britain is the obvious evidence for the theory. An island at the edge of the European continent developed a parliamentary tradition, common law, naval supremacy, and commercial empire that fit the island theory perfectly. But Japan — an island of similar size at the edge of the Asian continent — developed a military aristocracy, a rigid feudal social order, and a political culture that was, for most of its history, the precise opposite of liberal and commercial. Both islands are islands. Their political histories are almost diametrically opposed.

The difference lies not in the geography but in the contingent institutions that established themselves on each island at formative moments. England’s constitutional evolution — from Magna Carta through the Glorious Revolution of 1688 to the gradual extension of parliamentary sovereignty — established a set of property rights and legal institutions that were favorable to commercial development and that, crucially, survived the transition from one ruling dynasty to another. Japan’s institutions — the samurai military caste, the shogunate system that concentrated political power while formally preserving imperial legitimacy — evolved in a completely different direction, reflecting the different balance of power between its early institutional actors.

Cuba and Taiwan are another instructive pairing. Both are large islands off the coast of major continental powers. Both spent the nineteenth and early twentieth centuries under colonial governance — Cuba under Spain, then informally under the United States; Taiwan under first China, then Japan. Both achieved meaningful sovereignty in the mid-twentieth century. Cuba became a one-party communist state; Taiwan became, after a period of authoritarian rule, one of the most dynamic democracies in Asia. The geographic situations are nearly identical. The political outcomes are opposite. The explanatory variables are entirely institutional and historical: the Cold War alignment of the governing regimes, the specific character of the authoritarian parties that took power, the presence or absence of a commercially dynamic ethnic Chinese population with incentives to demand property rights protection.

The River Paradox

Rivers are geography’s most direct contribution to political development. The correlation between navigable river networks and early state formation is among the most robust findings in comparative historical analysis. The Nile, the Tigris-Euphrates, the Indus, the Yellow River — the earliest complex states all developed in river valleys where intensive agriculture was possible and where the river provided the infrastructure for commercial exchange and administrative control.

But rivers also fragment. The Rhine-Main-Danube basin in central Europe has more navigable waterway than almost any comparable area in the world. It should, by the river-based theory of state formation, have produced a unified commercial empire. Instead it produced the Holy Roman Empire — that extraordinary anti-state whose very name, as Voltaire observed, was a contradiction on three counts. The explanation for this paradox reveals something important about how geography and institutions interact.

The German river system did produce commercial integration — the Hanseatic League, the great trading cities of Cologne, Frankfurt, Hamburg, Lübeck, and Nuremberg. But the same geography that facilitated trade also created a distribution of economic power across dozens of nodes rather than concentrating it in a single dominant center. No single German river city was powerful enough to subordinate the others. The political result was a negotiated equilibrium of mutual weakness rather than a unified commercial republic. England’s river geography, by contrast, concentrated economic and administrative weight on the Thames basin: London was not one city among many but the overwhelming center of a national economy. When English merchants wanted to influence political decisions, they had one place to go and one institution to lobby. German merchants had to negotiate with fifty different princes.

The geographic inputs — rivers — were similar. But the specific topology of those rivers generated different distributions of commercial power, which generated different political bargaining structures, which generated different institutional outcomes. This is geography influencing politics through an institutional mechanism, not determining it directly.

Colonial Imposition and Artificial Geography

The modern world contains many states whose boundaries were drawn not by any geographic or political logic internal to the territory but by the administrative convenience of colonial powers. The straight-line borders of Africa are the most famous example — vast stretches of the continent’s political boundaries follow lines of latitude and longitude that bear no relationship to any physical feature, ethnic community, economic region, or historical polity. These boundaries divided peoples who shared languages and cultures and lumped together peoples who were historical enemies. They are, in the most literal sense, arbitrary.

The conventional expectation, derived from geographic determinism, was that these arbitrary boundaries would quickly give way to more “natural” ones — that states built on colonial geography would fragment, reorganize, and redraw themselves along the ethnic, economic, or geographic lines that colonial partition had ignored. This has not happened. With remarkably few exceptions, the boundaries drawn by European colonial administrations at the Berlin Conference of 1884–1885 remain the boundaries of African states today. The colonial geography became the institutional reality, defended by the elites who inherited power within those boundaries because changing the boundaries would threaten their own positions.

This is the institutional capture of geography — the reverse of the usual relationship. Instead of geography shaping institutions, institutions have calcified around an arbitrary geography and prevented it from changing. The Organization of African Unity’s founding principle of uti possidetis — respect the boundaries you inherited from colonialism, however irrational — was not a geographic judgment. It was a political judgment by post-colonial elites who understood that any legitimization of boundary change would threaten everyone’s boundaries. The result is a continent where the geographic logic of state boundaries is often weak and the political durability of those boundaries is paradoxically strong.

This African case generalizes. Institutional path dependence is frequently more powerful than geographic determinism. Once a political form establishes itself — even if it establishes itself for contingent, historical reasons that have nothing to do with the underlying geography — it creates vested interests in its own perpetuation that can resist geographic logic almost indefinitely. The Holy Roman Empire persisted in its fragmented form for eight centuries. The Ottoman Empire retained control of the Balkans, a region “naturally” suited by its geography to fragmented small states, for over four centuries. North Korea has maintained a hermit kingdom on a peninsula whose geographic situation otherwise suggests deep integration with the surrounding regional economy.

The Uncomfortable Synthesis

The honest intellectual position on geography and political outcomes is uncomfortable because it denies two popular simplifications: that geography is destiny (the strong determinist view) and that geography doesn’t matter and everything is contingent (the strong institutionalist view). Both are wrong.

Geography matters enormously. The cost structures it creates — for violence, trade, administration, and communication — shape the range of political possibilities available to any society. A landlocked country surrounded by hostile neighbors faces different political constraints than a small island nation with a protected harbor. An agricultural society in a disease-rich tropical environment faces different developmental challenges than a temperate-zone population with dense, navigable river networks. These differences create real and measurable effects on political and economic outcomes at civilizational time scales.

But within the range of possibilities that geography permits, institutional choices, contingent historical events, and the agency of specific actors at critical junctures determine which possibilities are realized. Geography sets the menu. Institutions choose from it. And institutions, once established, develop their own path dependencies that can override geographic logic for centuries.

The practical implication is sobering. Attempts to derive predictions about political development from geographic data alone will systematically fail at the cases that matter most — the cases where a society managed to escape the political form that its geography would have suggested, or got stuck in one it could have escaped. The exceptions are not noise to be explained away. They are the signal. Switzerland should not, by its geography, be a prosperous democracy — it is landlocked, mountainous, linguistically fragmented, and resource-poor. It is a prosperous democracy because specific institutional choices made in the medieval period, consolidated during the Napoleonic era, and reinforced by subsequent generations created a system with strong incentives for cooperation and strong mechanisms for managing diversity. That institutional achievement is the product of geography in the sense that Switzerland’s mountain geography made it defensible and its lack of resources made commercial cooperation necessary. But it is not reducible to geography. It required people making choices. And the fact that those choices were made, and maintained, and built upon over eight centuries, is what matters.

Geography, in the end, is not the author of political history. It is the paper on which that history is written. The words on the page are the product of the writers — the rulers, the merchants, the soldiers, the reformers — who made decisions within geographic constraints. Some of those decisions were good. Some were catastrophic. Most were contingent. And the contingency is not a problem to be engineered away by finding better geographic determinants. It is the essential fact about political life, the irreducible remainder that separates history from geology.