For the senior men of the Soviet Politburo — Brezhnev, Andropov at the KGB, Ustinov at Defence — the late 1960s and early '70s felt like an arrival.
They had survived the nightmare of Stalin and the bumbling of Khrushchev. There was finally stability. The Eastern European satellites were docile, particularly after Czechoslovakia was crushed in 1968. The Americans had taken a black eye in Vietnam. And most promising of all, Soviet influence was expanding steadily in the Third World.1
The future looked bright.1
Afghanistan was a key country in those plans: on the southern border, rich in natural gas and minerals, with ports on the Indian Ocean.
to make it a Soviet satellite would be a dream come true.1
They had been insinuating themselves since the 1950s — training the army, building the Salang Highway from Kabul north to the Soviet Union, modernising a backward nation. All going according to plan.1
Then Islamic fundamentalism became a political force across Afghanistan in the early-to-mid 1970s, and the Russians saw two problems:
first, that the fundamentalists would come to power and, seeing communism as godless and loathsome, would cut off ties with the Soviets; and second, that fundamentalist unrest would spill over from Afghanistan into the southern Soviet Union, which had a large Islamic population.1
In 1978, to prevent that, Brezhnev secretly supported a coup that brought the Afghan Communist Party to power.1
It made everything worse:
By December 1979 the Communist government was on the verge of collapse.
The Politburo met. Losing Afghanistan would be a devastating blow after so much progress, and they blamed Amin. He had to go.
Ustinov proposed a plan modelled on Eastern Europe:
a lightning strike by a relatively small Soviet force that would secure Kabul and the Salang Highway. Amin would then be ousted, and a Communist named Babrak Karmal would take his place. The Soviet army would assume a low profile, and the Afghan army would be beefed up to take over from it. During the course of some ten years, Afghanistan would be modernized and would slowly become a stable member of the Soviet Bloc.1
Blessed with peace and prosperity, the Afghan people would see the great benefits of socialism and embrace it.1
Then Ustinov presented it to the army's chief of staff, Nikolai Orgakov, and got an answer he did not want.
Told the invading force would not exceed 75,000 men, Orgakov was shocked. That force, he said, was far too small to secure the large, mountainous expanses of Afghanistan — a very different world from Eastern Europe.1
Ustinov countered that a giant invading force would generate bad publicity in the Third World and give the insurgents a rich target.
Orgakov responded with the thing that turned out to be exactly right:
the fractious Afghans had a tradition of suddenly uniting to throw out an invader—and that they were fierce fighters. Calling the plan reckless, he said it would be better to attempt a political solution to the problem.1
His warnings were ignored.1
On 24 December the plan went into motion. Some Red Army forces flew into Kabul while others marched down the Salang Highway. Amin was quietly taken away and killed. Karmal was shuffled into power.
Condemnation poured in from around the world, and the Soviets figured it would eventually die down — it usually did.1
In February 1980 Andropov met Karmal and instructed him on winning the support of the Afghan masses, presenting a plan and promising money and expertise. Once the borders were secure, the army built up and the people reasonably satisfied, Karmal should politely ask the Soviets to leave.1
The invasion itself went more easily than the Soviets had expected, and for this military phase their leaders could confidently declare "mission accomplished."1
Within weeks of Andropov's visit they had to adjust that assessment.
the mujahideen were not intimidated by the Soviet army, as the Eastern Europeans had been. In fact, since the invasion their power only seemed to grow, their ranks swelling with both Afghan recruits and outsiders.1
Ustinov funnelled in more soldiers and ordered offensives into the sheltering regions. The first major operation, that spring, went into the Kunar Valley with heavy weaponry — levelling entire villages and forcing the inhabitants to flee to refugee camps in Pakistan. Having cleared the area of rebels, they withdrew.1
A few weeks later, reports came in that the mujahideen had quietly returned.
All the Soviets had done was leave the Afghans more embittered and enraged, making it easier for the mujahideen to recruit.1
And then the trap, stated as a set of options with no good branch:
But what could the Soviets do? To let the rebels alone was to give the mujahideen the time and space to grow more dangerous, yet the army was too small to occupy whole regions. Its answer was to repeat its police operations again and again, but with more violence, hoping to intimidate the Afghans—but, as Orgakov had predicted, this only emboldened them.1
Karmal ran programmes for literacy, women's rights, development and modernisation, all designed to peel support from the rebels. But the Afghans preferred their traditional way of life by a vast majority, and Communist outreach had the opposite effect.1
Most ominous of all, Afghanistan became a magnet for anyone with a reason to bleed the Soviets:
Ustinov, with Soviet armies stretched thin worldwide, refused to send more men — instead arming his soldiers with the latest weaponry and enlarging the Afghan army. None of it translated into progress. The mujahideen improved their ambushes and used Stinger missiles from the Americans to great effect.1
Years passed, and morale in the Soviet army dropped precipitously: the soldiers felt the hatred of the local population and were stuck guarding static positions, never knowing when the next ambush would come. Abuse of drugs and alcohol became widespread.1
This is the part that makes the case a lesson rather than a tragedy.
As the costs of the war rose, the Russian public began to turn against it. But the Soviet leaders could not afford to pull out: besides creating a dangerous power vacuum in Afghanistan, that would deliver a sharp blow to their global reputation as a superpower.1
And so they stayed, each year supposedly the last.1
The senior members of the Politburo died off without seeing the slightest progress — Brezhnev in 1982, Andropov and Ustinov in 1984.1
Gorbachev, who had opposed the war from the beginning, became general secretary in 1985 and started phased withdrawals. The last soldiers left early in 1989.
In all, over 14,000 Soviet soldiers died in the conflict, but the hidden costs—to the delicate Russian economy, to the people's slender faith in their government—were far greater. Only a few years later, the entire system would come tumbling down.1
Greene reads the whole thing through a distinction he attributes to Rommel, and it is the most portable idea in the chapter.
A risk and a gamble both involve an action with only a chance of success, heightened by boldness.
The difference is that with a risk, if you lose, you can recover: your reputation will suffer no long-term damage, your resources will not be depleted, and you can return to your original position with acceptable losses.1
With a gamble, on the other hand, defeat can lead to a slew of problems that are likely to spiral out of control. With a gamble there tend to be too many variables to complicate the picture down the road if things go wrong.1
And the trap that makes gambles distinctively bad:
if you encounter difficulties in a gamble, it becomes harder to pull out—you realize that the stakes are too high; you cannot afford to lose. So you try harder to rescue the situation, often making it worse and sinking deeper into a hole that you cannot get out of.1
People are drawn into gambles by their emotions: they see only the glittering prospects if they win and ignore the ominous consequences if they lose.1
Taking risks is essential; gambling is foolhardy.1
Applied here:
The Soviets were drawn in by the irresistible lure of possessing a client state in the region. Dazzled by that prospect, they ignored the reality: the mujahideen and outside powers had too much at stake to ever allow the Soviets to leave behind a secure Afghanistan. There were too many variables beyond their control.1
And the structural trap named exactly:
An occupying army in Afghanistan involved a double bind: the larger the military presence, the more it would be hated, and the more it was hated, the larger it would have to be to protect itself, and so on indefinitely.1
Now, too late, they realized that the stakes had been raised: to pull out—to lose—would be a devastating blow to their prestige. . . . Since they should never have invaded in the first place, they had no rational exit strategy.1
Greene's prescription for someone already inside a gamble is unusually concrete:
The worst way to end anything—a war, a conflict, a relationship—is slowly and painfully. The costs of such an ending run deep: loss of self-confidence, unconscious avoidance of conflict the next time around, the bitterness and animosity left breeding.1
Before entering anything:
you must calculate in precise terms your exit strategy. How exactly will the engagement end, and where will it leave you? If the answers to those questions seem vague and full of speculation, if success seems all too alluring and failure somewhat dangerous, you are more than likely taking a gamble.1
And if you are already in:
you have only two rational solutions: either end the conflict as quickly as you can, with a strong, violent blow aimed to win, accepting the costs and knowing they are better than a slow and painful death, or cut your losses and quit without delay.1
Never let pride or concern for your reputation pull you farther into the morass; both will suffer far greater blows by your persistence. Short-term defeat is better than long-term disaster.1
You are considering something that could go very well and might not.
Run the risk/gamble test first, and run it as a question about recovery rather than about odds. If this fails: can you return to roughly your current position with acceptable losses? Or does failure open a set of problems you cannot enumerate?
The tell is the enumeration. A risk has a knowable downside. A gamble has a downside you cannot list, because it depends on how other parties respond, and those parties have their own reasons.
Then check for the specific feature that made Afghanistan unrecoverable: do other people have an interest in your failure that is independent of you? The Soviets were not beaten by Afghans alone. The CIA, Zia and Sadat each had their own reasons to keep the thing running, and none of those reasons were about Afghanistan. If your project would create an opportunity for people who dislike you, you are not managing one variable, you are managing several you cannot see.
Then look for the double bind before you're in it. The Soviet version was that presence generated hatred, hatred required more presence. Any structure where the remedy is also the cause cannot be solved by more effort, and more effort is what everyone applies.
Then write the exit, specifically, before entering. How does this end, and where does it leave you? Vagueness there is the diagnostic — if the success case is vivid and the ending is fuzzy, you are looking at the glittering prospect and not at the consequence.
And if you are already inside one: take one of the two rational options rather than the third, which is the one everyone takes. Finish it fast, or quit now. The Soviets did neither for nine years, and the reason was pride and reputation — which is exactly the thing that suffered most from the persistence.
The Soviet–Afghan War is extensively documented, and the outline — the 1978 coup, Amin's removal, the December 1979 invasion, the escalating counterinsurgency, external support to the mujahideen, and the 1989 withdrawal — is uncontroversial. [POPULAR SOURCE]
The Orgakov exchange is the load-bearing detail and it is unusually well sourced, having emerged in the post-Soviet record. The chapter uses it correctly: this was not a failure that could only be seen afterwards, it was predicted by the chief of the general staff, in the room, and overruled.
The final causal claim is much stronger than the evidence supports. "Only a few years later, the entire system would come tumbling down" implies a directness that most historians would resist — the Soviet collapse had deep economic, structural and political causes, of which Afghanistan was an accelerant and a symptom rather than a cause. Greene needs the war to have destroyed the state because it makes the exit-strategy lesson maximal.
And the human cost is again accounted asymmetrically. The chapter records 14,000 Soviet dead and the damage to the Russian economy and to public faith. Afghan deaths in the same period ran to somewhere between several hundred thousand and over a million, with millions more displaced into the refugee camps the Kunar Valley operation is described as producing. Those camps appear here as a tactical detail in a paragraph about why an operation failed.
This is the running pattern of the whole book at its most quantitatively stark: the costs that are counted are the ones borne by the party whose strategy is being assessed.
Against Macnaghten and the Mirrored Enemy — built earlier in this same cluster — the pairing is the strongest thing this build has found, and Greene never makes it.
Same country. A hundred and forty years apart. An outside power installs a client ruler it distrusts. It keeps an army in to protect him. The presence generates the hatred that requires more presence. Local grievance is fed by the occupier's own reforms. It cannot leave without a catastrophic loss of prestige. It eventually leaves having achieved nothing, at enormous cost, with one survivor's ride or fourteen thousand coffins as the image that lasts.
Greene uses the first as a lesson in reading people and the second as a lesson in exit strategy. The British error was psychological and the Soviet error was structural, and both produced an identical campaign — which suggests neither diagnosis is complete, and that something about that specific undertaking generates the same shape regardless of which mistake you arrive with. Orgakov's warning names it better than either chapter does: the fractious Afghans had a tradition of suddenly uniting to throw out an invader.
Against Iran-Contra as Envelopment, the same decade and the same region from the other superpower's side — and the American money going to the mujahideen was, by Greene's own account, revenge for Soviet support of North Vietnam. Both are ratchets: arrangements that could not be exited without an admission nobody could afford.
Against Strategy 11: The Nonengagement Strategy, the mujahideen ran that chapter's playbook — trade space for time, refuse the decisive engagement, let the aggressor overextend — and this chapter never notes that the Soviets were defeated by the strategy the book taught eleven chapters earlier.
Business → The Initiative With No Kill Criteria. Organisations enter projects with success criteria and almost never with exit criteria — the specific, pre-agreed conditions under which the thing stops. The result is the corporate version of this war: an initiative that everyone privately knows has failed, that cannot be killed because killing it requires someone to own the failure, and that consumes resources for years while each year is supposedly the last.
The insight neither domain gives alone: the binding constraint on stopping is never the sunk cost, it's the absence of anyone who can stop it without personally absorbing the admission. The Politburo's problem in 1985 was not that withdrawal was strategically wrong — Gorbachev did it — but that it required someone whose reputation was not attached to the decision to invade, and the men whose reputations were attached died in office first. The practical consequence is that kill criteria must be written before entry and by someone senior enough that invoking them is not a confession, because after entry the only people with authority to stop are the people who would be admitting something by doing so.
Psychology → Escalation of Commitment, and the Variable That Makes It Fatal. The Soviet pattern — increasing investment in a failing course specifically because so much has already been invested — is the textbook case, and the chapter's version of it is precise: you realize that the stakes are too high; you cannot afford to lose.
The insight neither field gives alone: escalation is driven far more by the anticipated cost of the admission than by the sunk resources themselves, which is why it is so much worse for reputations than for budgets. A private investor writes off a bad position more readily than a government abandons a war, and the difference is not the money. This also identifies the only reliable countermeasure, which is structural rather than psychological: decide the exit conditions while the decision is still cheap and record them publicly, so that stopping later is the execution of a plan rather than the confession of an error. Greene's advice to "calculate in precise terms your exit strategy" before entering is usually read as prudence. It is better read as insurance against your own future inability to admit anything.
The chief of the general staff said the force was too small, that Afghanistan was not Eastern Europe, and that the Afghans had a tradition of uniting against invaders — before a single soldier moved. Being right in advance changed nothing, because the people who could stop it were the people who would have had to admit something. Which means the useful moment to write down what would make you quit is while quitting is still free, and the useful person to write it is someone who won't have to confess anything by invoking it.