When the Shadow State Wins
Afghanistan and the Politics of Administrative Collapse
Afghanistan’s collapse is usually remembered as a sudden event: provincial capitals falling in sequence, security forces dissolving, and Kabul surrendering without a battle. On paper, the Afghan state had more than 300,000 soldiers and police, two decades of international investment, and the formal architecture of sovereignty: ministries, courts, elections, budgets, security forces, and diplomatic recognition. In practice, much of the country had already been moving through a different system of authority.
The speed of the collapse was misleading. The Afghan state did not lose control only when the Taliban entered Kabul. It had been losing authority for years in smaller transactions: a land dispute taken to a Taliban court, a trucker paying at a Taliban checkpoint, an aid organization negotiating access, a district official hedging against the future, a commander accepting terms before the final offensive began. By August 2021, the formal state still held the title. The Taliban increasingly held the claim.
That is what made the end look less like a conventional military defeat than a transfer of title. The fall of Kabul did not create Taliban rule. It formalized a shift in authority that had already occurred across much of the country.
The Making of a Shadow State
The Taliban’s shadow government was not propaganda or aspiration. It was a functioning parallel administration, and it mirrored the state it was displacing. The movement appointed shadow governors for nearly every province and shadow district officials beneath them, and ran commissions for justice, finance, education, and health: a second institutional map laid over the first.
The courts were the system’s center of gravity. Taliban judges, often mobile, heard land disputes, inheritance claims, and criminal complaints, and delivered verdicts in days. The verdicts were harsh and ideological, but they were final and they were enforced. Researchers who studied wartime Afghanistan found the same pattern again and again: Afghans, including many who feared the movement, took their disputes to Taliban courts by preference. The government alternative required bribes at every stage, took months or years, and produced rulings that often went to the highest payer and could not be enforced in any case. A farmer with a land dispute did not face a choice between justice and brutality. He faced a choice between a system that worked and one that did not.
Taxation told the same story. The Taliban collected ushr on harvests and zakat on wealth. It taxed trucking at checkpoints, mining operations, fuel shipments, the drug trade, and, in one of the war’s quieter absurdities, the very aid organizations and contractors funded to strengthen the government, which paid Taliban fees to operate in Taliban-controlled districts. The revenue mattered, but the reach mattered more. Every tax paid was an acknowledgment of who actually governed the field, the road, the district. The movement regulated daily life the same way: warning letters, night visits, negotiated arrangements with communities over whether schools would open and what they would teach.
None of this was legitimacy in any liberal or democratic sense. It was not justice, and it was not consent. Fear organized much of the compliance, and coercion did the rest. But it was authority. A system can govern brutally and still govern. The Taliban did not need to persuade Afghans that it was benevolent. It needed to make itself unavoidable. And by the late war years, across much of rural Afghanistan, it had. Little of this was visible in the capital’s institutional charts, which is part of why it was so consistently undercounted. But by the time of the handoff, the movement had crossed a threshold: it was not yet the state, but it was already a governing authority. The Taliban was governing before victory, not waiting to govern after it.
The Limits of Formal State-Building
The United States and its partners helped build an Afghan state with ministries, elections, courts, security forces, donor programs, and international recognition. These were not trivial achievements, and many Afghans risked and gave their lives trying to make them work.
But the state’s institutions were built to be legible to Kabul and to donors, more than to be usable at the point of need. Formal courts were slow, distant, expensive, and corrupt. Government officials appeared intermittently in districts they nominally administered. Police were often predatory: in many communities, the arrival of the Afghan Local Police meant extortion, not protection. Development projects arrived without durable authority behind them and became resources to be captured rather than services to be relied on. The state existed. It was recognized, funded, and measured. It was just not, in much of the country, the system people could use when something went wrong.
Strategy struggled to see this because it measured the wrong layer. It counted districts held, forces generated, budgets executed, schools built, elections conducted: inputs and territory. Those indicators answered the question of whether the state existed. They did not answer whether anyone could use it. At the village and district level, many American soldiers, diplomats, and aid workers, and far more Afghans, understood the real ledger with painful clarity: they could see who settled disputes and who collected the tax. The failure was that this local knowledge never translated into strategic design. It did not fit the reporting formats, so it did not shape the campaign.
The Surrender Cascade
The clearest evidence for what had actually happened in Afghanistan is how it ended. The cascade of 2021 was not primarily a military event. It began in the districts in May, months before Kabul, and most districts did not fall to assaults. They were handed over. In the final ten days, the provincial capitals fell in waves; besieged cities like Kandahar and Herat and cities that fell in a day, like Mazar-i-Sharif, collapsed into the same wave. The surrenders had been arranged in advance, some of them years in advance: deals negotiated between the Taliban and local commanders, elders, and officials, sealed with money, safe-passage guarantees, and the simple, unanswerable argument that the Americans were leaving and the Taliban was not. When the moment came, positions flipped by phone call. Soldiers who had not been paid in months, defending a government that could not reach them, accepted terms from the authority that had been taxing their districts and judging their neighbors’ disputes for a decade.
The speed shocked the world, yet it should not have. It was not military momentum at all, but the simultaneous public acknowledgment of a transfer of authority that had already occurred privately, transaction by transaction, over years. In August 2021, the state discovered that the ground had already changed hands.
When the Taliban entered Kabul, it inherited the shell of the state and nationalized the practices it had developed in the shadows. The regime that resulted governs through coercion, exclusion, surveillance, social control, and above all, through the systematic removal of women from education, employment, and public life. That repression deserves moral condemnation, and it should also be understood analytically for what it is: an administrative instrument. It orders society, narrows participation, makes compliance visible, and makes defection costly. Repression is not incidental to how the Taliban governs. It is how the Taliban governs.
The Warning Beyond Afghanistan
Afghanistan’s relevance lies less in its resemblance to any one conflict than in the pattern it reveals. States can lose practical authority before they lose formal control. Armed movements can become necessary before they become sovereign. And outside powers can mistake the survival of institutions for the survival of political order.
That warning matters in the Sahel. In parts of Mali, Burkina Faso, and Niger, armed groups are no longer only attacking the state. They are moving into the daily functions through which authority is experienced: settling disputes, taxing roads and markets, regulating movement, punishing noncompliance, and providing rough order where state authority is absent, predatory, or feared. These functions do not make such groups legitimate, but they do make them durable.
The relevant question is therefore not whether the Sahel is “becoming Afghanistan.” It is whether outside actors are again recognizing the contest too late. By the time an armed group threatens a capital, the more important struggle may already be far advanced in the places where daily life is actually governed. Military pressure can disrupt that system, but it cannot automatically replace it. Killing leaders, clearing territory, and training forces may create tactical openings, but those activities do not settle who people trust, fear, pay, obey, or expect to survive.
This was the Taliban’s central advantage. It grasped what coalition metrics often obscured: people do not experience sovereignty as an abstraction. They experience it through the minutiae of daily life filtered through one singular question: who will answer when something goes wrong. The Sahel is still in that contest. That is precisely why Afghanistan matters now.
The Limits of a Violence-First Lens
Counterterrorism is built to see violence. It tracks attacks, leaders, cells, financing, recruitment, and territorial control, and those indicators matter because the violence is real. But violence is not always the most important signal. Sometimes violence announces the contest, while administration decides it.
The more revealing indicators are quieter. Who settles the land dispute? Who collects the fee at the checkpoint? Who decides whether the road is safe, whether the school opens, whether the market functions? Who punishes theft? Who do people bargain with before moving goods? Who do communities expect to still be there next year? These questions can look secondary from a capital, but they are central to the people living inside the conflict. They are the daily mechanics of power, and they are strategic questions. This is why governance cannot be treated as a stabilization add-on or a development concern that comes after security. In conflicts like Afghanistan and the Sahel, governance is one of the ways security is produced, contested, and lost. An armed group does not have to be popular, competent by Western standards, or just. It only has to become more usable, more feared, more present, or more durable than the state.
Afghanistan’s lesson is that an administrative system cannot be defeated by building a formal state people cannot reliably use. That conclusion cuts across the familiar policy divides: withdrawal versus persistence, counterterrorism versus counterinsurgency, democracy promotion versus restraint. The problem sits beneath all of them. The Afghan state lost the daily contest over authority long before it lost the capital, and no amount of activity at the formal layer could compensate for what had already been decided at the practical one.
The same pattern runs through the broader set of cases: armed actors fill vacancies, parallel orders harden, political architectures fracture, and authority migrates to systems outsiders are slow to recognize because they do not look like the state. Afghanistan is the case where that transfer ran to completion. By the time Kabul fell, the title simply caught up with the claim. The shadow state had already won.