12.2 — The Mechanics of the Cold War

On 27 October 1962, at the height of the Cuban Missile Crisis, an American destroyer dropped practice depth charges on a Soviet submarine to force it to surface.

The submarine, B-59, had been out of radio contact for days. Its crew did not know whether war had started. It carried a nuclear torpedo, and the captain concluded that war had begun and ordered it prepared.

Launch required the agreement of three officers. Two agreed. The third, Vasili Arkhipov, refused, and argued until the captain relented and surfaced.

On the same day, an American U-2 strayed into Soviet airspace over Siberia and Soviet fighters scrambled to intercept it. Another U-2 was shot down over Cuba, killing its pilot, by a local commander acting without Moscow's authorisation.

Any one of those could have started a nuclear exchange. All three happened in a single day.

This chapter is about the system in which that was possible.

Deterrence

The logic is simple and its requirements are demanding.

Deterrence works if an attacker believes that attacking will cost more than it gains. With nuclear weapons, that means being certain the target can retaliate after absorbing a first strike.

Which requires a survivable second-strike capability. If your weapons can be destroyed in a first strike, you have an incentive to launch them the moment you think an attack is coming — which makes crises unstable.

So the arsenals were made survivable in three ways — the triad. Land-based missiles in hardened silos. Bombers that can be kept airborne. And submarines, which are the crucial leg: a submarine at sea cannot be found and therefore cannot be destroyed pre-emptively.

Mutual assured destruction is the resulting equilibrium. Both sides can destroy the other regardless of who strikes first, so neither gains from striking first.

And the counter-intuitive conclusions that follow are worth stating because they explain policies that look mad.

Defence is destabilising. A missile defence system that works would let its owner strike first and survive the reply, which restores the incentive to strike first. This is why the Anti-Ballistic Missile Treaty of 1972 limited defences on purpose, and why its abandonment in 2002 was consequential.

Vulnerability of one's own cities is stabilising. A leader whose population is hostage is a leader who will not gamble.

And accuracy is destabilising. A missile accurate enough to destroy a hardened silo is a first-strike weapon, whereas one accurate only enough to hit a city is a retaliation weapon.

The whole doctrine depends on both sides being rational, informed and in control of their forces, and the near-misses above are what happens when one of those fails.

The crises

Berlin, 1948–49 (Chapter 12.1) and 1961, when the building of the Berlin Wall stopped the flow of East Germans westward — around 3.5 million had left since 1949, disproportionately the educated — and produced a standoff with American and Soviet tanks facing each other at Checkpoint Charlie.

Cuba, October 1962, is the closest the world came.

The sequence. The United States had missiles in Turkey, minutes from Soviet cities. The Soviet Union placed medium-range missiles in Cuba, which had had a revolution in 1959 and had survived an American-backed invasion attempt at the Bay of Pigs in 1961.

American reconnaissance found the sites. Kennedy's advisers were divided between an air strike, an invasion and a blockade. He chose a blockade — called a quarantine, because a blockade is legally an act of war — which bought time.

The resolution. Public agreement that the Soviet Union would remove the missiles and the United States would pledge not to invade Cuba. And a secret agreement that American missiles in Turkey would be withdrawn within months, which was concealed for decades so that Kennedy would not appear to have traded.

What was learned. That neither side had adequate information about the other, that communication was too slow, and that local commanders had more discretion than anyone realised. The Moscow-Washington hotline was established in 1963 as a direct result, and the Partial Test Ban Treaty followed the same year.

Kennedy had recently read Barbara Tuchman's account of 1914 and referred to it during the crisis (Chapter 11.2), saying he did not want a comparable book written about him. That is a case of history being used correctly.

Able Archer, 1983, is the least known and possibly the most dangerous. A NATO command exercise simulating a nuclear release, conducted with unusual realism, during a period when the Soviet leadership genuinely believed a surprise first strike was possible. Soviet intelligence had a standing programme, Operation RYaN, looking for indicators of an imminent attack, and the exercise matched several of them. Some Soviet forces were put on alert. The West did not understand at the time how the exercise was being read.

And Stanislav Petrov, in September 1983, saw his early warning system report an American launch of five missiles and judged it a false alarm — on the reasoning that a real attack would involve hundreds, not five — and did not pass it up the chain. It was sunlight reflecting off high-altitude clouds.

Twice in one year, the system was saved by an individual's judgement.

The arms race and its limits

Arsenals peaked at over 60,000 warheads globally in the 1980s — enough to destroy every city on Earth many times over, which is a fact about procurement dynamics rather than about strategy.

Why they grew so large. Each side had to plan against the other's worst case. Each service branch wanted its own weapons. Contractors and constituencies developed around production. And action-reaction dynamics: a Soviet deployment justified an American one, which justified another Soviet one, with intelligence estimates systematically erring on the high side.

The "missile gap" of the 1960 American election is the standing example: it did not exist, and the United States was substantially ahead.

Arms control eventually constrained it. The Nuclear Non-Proliferation Treaty of 1968. SALT I in 1972, including the ABM Treaty. SALT II in 1979, signed and never ratified. The Intermediate-Range Nuclear Forces Treaty of 1987, which eliminated an entire category of weapons and included on-site inspection. And START, which cut strategic arsenals substantially from 1991.

Most of that architecture has since lapsed. The ABM Treaty ended in 2002, the INF Treaty in 2019, and the remaining strategic treaty framework is under strain.

Espionage

Enormous effort, and its actual value is more mixed than fiction suggests.

Where it clearly mattered. Soviet atomic espionage — Klaus Fuchs and others at Los Alamos — accelerated the Soviet bomb by an estimated one to several years (Chapter 11.10). Oleg Penkovsky, a Soviet colonel working for Britain and the United States, supplied information on Soviet missile capabilities that was directly useful during the Cuban crisis. And satellite reconnaissance, from 1960, transformed the situation by making it possible to count the other side's missiles — which is what made arms control verifiable and therefore possible.

Where it produced disasters. Coups and covert operations with long consequences: Iran in 1953, Guatemala in 1954, Chile in 1973 (Chapter 12.4). And in the Soviet bloc, the internal security apparatus — the Stasi in East Germany had files on a very large share of the population and an extensive informer network, which is documented because the files were opened in 1990.

And the general finding. Both sides consistently overestimated the other's capabilities and hostility. Intelligence services have an institutional incentive to report threats, and the cost of understating one is career-ending while the cost of overstating one is not.

How it ended

Chapter 12.6 covers the collapse. The system's end had specific features.

The Soviet economy stagnated from the 1970s — growth slowed, agriculture required grain imports, and consumer goods were poor and scarce (Chapter 9.6).

Afghanistan, from 1979, was expensive and unwinnable and delegitimising at home (Chapter 12.4).

Military spending consumed an estimated 15 to 25 percent of Soviet output, against around 5 to 6 percent for the United States — which is the asymmetry that mattered most. The same absolute spending was a far heavier burden on a much smaller economy.

And Gorbachev, from 1985, chose reform and then chose not to use force to hold the empire together, which is a decision that could have gone otherwise and did not.

What the Cold War cost

Directly, in the proxy wars (Chapter 12.4): millions of deaths in Korea, Vietnam, Afghanistan, Angola, Ethiopia, Central America and elsewhere.

In distorted development. Countries that received aid and weapons for alignment rather than for anything productive, and dictatorships sustained by superpower support.

In resources. Trillions of dollars on arsenals never used.

And in risk. The near-misses above, several of which the public learned about decades later.

And what it delivered, honestly. No war between the great powers. The space programme, the internet's origins in defence-funded research, and a large part of post-war science funding. And the competition arguably drove domestic reform in the West — the American civil rights movement was pressed partly by the embarrassment of segregation in a country claiming to lead the free world, and Soviet propaganda used it constantly.

Where this shows up in your life

Nuclear weapons and their doctrines have not gone away (Chapter 11.10), and India and Pakistan are the pair for whom the stability-instability paradox is currently most acute (Chapter 6.26).

The internet exists because of ARPANET, funded by the American defence department, and GPS is a military system opened to civilian use.

And the specific lesson worth carrying is Arkhipov's and Petrov's. Two of the twentieth century's most consequential decisions were made by mid-ranking officers who judged that the system in front of them was wrong. Every account of deterrence assumes rational actors in control of their forces, and on at least two documented occasions what actually held was an individual's refusal.

What the next page covers

Chapter 12.3 covers China from 1949 — how the Communists won, what the Great Leap Forward actually was and how it killed tens of millions, what the Cultural Revolution did to a generation, and how a country that had done all of that became, within twenty-five years of Mao's death, the world's second economy.