Why wars happen

The historian Thucydides, writing 2,400 years ago about the Peloponnesian War, identified three motives: fear, honor, and interest. Modern political scientists have refined this into longer lists, but the basic insight survives. Most wars are launched because some combination of leaders believed the war would, on expected value, leave their faction better off than not fighting. They are usually wrong about this — the empirical record on whether the initiator wins is dismal — but the calculations are real, and they're worth understanding.

This lesson treats war as something with causes you can analyze, not as either a moral imperative or an inexplicable atrocity. Both of those views obscure the question that matters: what conditions make war more or less likely, and which of those conditions can be changed?

The five structural drivers

Resources. Direct seizure of land, minerals, oil, water, or food. The textbook example, and still operative: the 1990 Iraqi invasion of Kuwait was about oil reserves; the 2003 US invasion of Iraq, whatever else it was, removed the geopolitical autonomy of a major oil producer; Russia's 2022 invasion of Ukraine included the deliberate seizure of Ukrainian grain-belt agricultural land and Black Sea ports. Even when resources aren't the stated cause, the maps of contested territory and the maps of valuable resources overlap suspiciously often.

Debt and economic crisis. Mancur Olson's "rise and decline of nations" thesis suggests that distributional coalitions accrete until they paralyze states, after which war or revolution clears the table. More concretely: governments under domestic economic stress sometimes seek external conflict as a unifying distraction (the "rally around the flag" effect). Argentina's 1982 Falklands War, launched amid hyperinflation and dictatorship crisis, is the often-cited example. Whether this dynamic is operative in any current case is hotly debated, but it is operative often enough to belong on the list.

Prestige and credibility. States, like people, care about reputation. Once a state has made a commitment ("we will defend Taiwan," "we will respond if you cross this line"), backing down has costs beyond the specific case — it signals weakness to other adversaries and other commitments. This logic, called "audience cost theory," explains why states sometimes fight wars they don't want over stakes that seem disproportionate. Vietnam (US credibility in containment) and Afghanistan (US credibility post-9/11) both ran on prestige logic long after their original strategic rationale had collapsed.

Security competition (the "security dilemma"). When state A builds military capacity for defensive reasons, state B sees a threat and builds in response, which makes A more anxious, etc. The dynamics can produce wars neither side wanted. The pre-1914 European alliance system is the canonical case. Current US-China dynamics in the Pacific, NATO-Russia dynamics in Eastern Europe, and Iran-Israel dynamics across the Levant all have security-dilemma structure.

Domestic political economy. Who profits from war? Defense contractors, energy producers (in some configurations), commodity speculators, and increasingly cybersecurity and surveillance vendors. Eisenhower's farewell warning about the "military-industrial complex" identified a structural interest group that grows during war and lobbies hard against the conditions for peace. In its modern form, the complex includes defense primes (Lockheed, RTX, BAE, Northrop, General Dynamics), the network of think tanks and consultancies that staff and recycle senior personnel, and the journalists and pundits who graduated from one and write about the other. This isn't a conspiracy theory; it's a description of a documented and self-aware coalition.

The proxy war pattern

Direct great-power war has been rare since 1945 because nuclear weapons make it catastrophic. What has emerged instead is the proxy war: great powers compete by funding, arming, and advising third-party combatants in someone else's territory.

The mechanism is consistent across cases. Power A (the patron) has strategic interests in region X. Local actor B has its own grievances or ambitions. A supplies B with weapons, intelligence, money, training, and sometimes diplomatic cover. B does the fighting and the dying. A's costs are financial and reputational, not in body bags. If B wins, A has projected power without direct commitment. If B loses, A walks away with limited consequence.

The Cold War was the golden age of this pattern: Korea, Vietnam, Afghanistan (1980s), Angola, Nicaragua, El Salvador, and dozens of smaller theaters were proxy wars between US and Soviet patrons. The post-Cold-War era saw a partial pause, then a return: Syria (US and Russia and Iran and Turkey all backing different factions), Yemen (Saudi vs. Iranian proxies), Ukraine (a NATO-vs-Russia proxy war by every standard analytical definition, though one with unusual openness on the Western side), Sudan (Russia, UAE, others), and the various Israel-vs-Iranian-proxy fights across Lebanon, Gaza, and the Red Sea.

The math of proxy war is brutal: the patron gets influence and information at the cost of money and reputational hedging; the local population gets the casualties, the destruction, and the refugee flows. The pattern is operative right now in multiple theaters. Reading any major conflict, the question "who benefits if this continues" is usually answered by mapping the proxy structure.

The "merchant of death" problem

Arms exports are a structural feature of major economies, including democracies. The US is the world's largest arms exporter ($200B+/year recently). Russia, France, China, Germany, the UK, Israel, South Korea, and Turkey all rank in the top ten. The combined effect is that almost every modern conflict is fought with weapons sold by countries that publicly deplore the conflict.

The mechanism: defense industries are politically powerful in their home countries (jobs, political contributions, lobbying). Sales require export licenses, which require political approval. Sales sometimes get approved to regimes the same government is publicly criticizing on human rights. Saudi Arabian weapons in Yemen, UAE weapons in Sudan, Israeli weapons in conflicts across the global south — none of these are accidents. They're the predictable output of a defense industry that needs scale to survive and a foreign policy apparatus that uses arms sales as leverage.

The peace mechanism that has worked

Through all this, the most striking finding in conflict research is that economic interdependence reduces (though does not eliminate) the likelihood of war between states. The post-1945 European peace was constructed deliberately on this principle — first coal and steel, then the common market, then the EU. The US-China relationship, despite all its tensions, has been moderated by their economic entanglement. The Russia-Europe gas relationship was supposed to do the same; the 2022 invasion of Ukraine showed that this can fail when one side prioritizes other goals.

The corollary: decoupling raises the risk of war. The current US-China decoupling — semiconductor controls, supply chain re-shoring, financial decoupling — is reducing the economic cost of conflict for both sides. This may be wise security policy or unwise economic policy or both. It is unambiguously raising the risk that a Taiwan crisis escalates to something neither side would have allowed if interdependence had been preserved.