Last Sunday, I wrote that the Taliban might be capturing something more important than territory:
expectations.
I did not expect to find out how important that was this quickly.
One week ago, five Afghan provincial capitals had fallen.
Tonight, the Taliban is in Kabul.
President Ashraf Ghani has left the country.
Taliban fighters have entered the presidential palace.
Afghans have rushed towards Kabul airport hoping to leave.
And the United States is evacuating its embassy as the political order it spent twenty years helping to construct collapses around it.
The speed is almost impossible to comprehend.
Herat has fallen.
Kandahar has fallen.
Mazar-i-Sharif has fallen.
Jalalabad has fallen.
Provincial capitals that the Afghan government was expected to defend for months disappeared in days, sometimes with remarkably little fighting.
Then Kabul itself was surrounded.
There was no enormous final battle for the capital.
The government simply ceased to possess a credible path forward.
Twenty years after the United States invaded Afghanistan and removed the Taliban from power, the Taliban has returned.
There will be enormous arguments over who lost Afghanistan.
Biden.
Trump.
Ghani.
The Afghan military.
Pakistan.
American generals.
Corrupt Afghan politicians.
Perhaps that is the wrong place to begin.
Before deciding who deserves blame, we need to understand what actually failed.
Because an army did not merely lose a war this week.
A state disintegrated.
The collapse became a cascade
At the beginning of this week, the Afghan government still controlled many of the country’s largest cities.
Then the map changed almost hourly.
On Thursday, the Taliban captured Ghazni, placing it only around 150 kilometres from Kabul.
That same day, Herat fell in the west.
Then came Kandahar, Afghanistan’s second-largest city and the birthplace of the Taliban movement.
On Saturday, Mazar-i-Sharif, the government’s principal stronghold in the north, fell.
Today, Jalalabad was taken.
That effectively left Kabul surrounded.
President Ghani departed Afghanistan.
And Taliban fighters entered the capital.
The government did not lose every Afghan city through enormous battles.
In many places, local officials, commanders and security forces surrendered or negotiated arrangements with the Taliban.
That distinction matters.
If the Taliban had physically destroyed the entire Afghan military, we could explain the outcome principally through battlefield superiority.
That is not what happened.
The Afghan state lost the willingness of enough people to continue defending it.
Once that process began, collapse accelerated.
Last week’s expectation problem became real
Suppose you are a provincial governor on Monday.
You believe Kabul might survive.
Fighting therefore makes sense.
By Thursday, Kandahar and Herat are falling.
By Saturday, Mazar-i-Sharif is gone.
Now imagine making the same decision.
If you believe the Taliban will control the country within days, negotiating your surrender may suddenly appear rational.
The same calculation applies to soldiers.
Police officers.
Local politicians.
Tribal leaders.
Civil servants.
Each individual may decide that survival requires accommodating whoever is going to govern tomorrow.
But every person who makes that individually rational decision makes Taliban victory more likely.
That is how expectations become political power.
The Taliban’s most effective weapon this week may not have been a rifle.
It may have been inevitability.
Once enough Afghans believed the Taliban would win, the belief helped produce the outcome.
Numbers concealed weakness
The Afghan security forces were frequently described as numbering roughly 300,000 personnel.
The Taliban was estimated to have far fewer fighters.
That comparison now looks almost meaningless.
A military is not simply the number of names on a payroll.
It is an organisation capable of turning people, equipment and information into coordinated force.
How many soldiers are actually present?
How many are trained?
How many are willing to fight?
Can they be supplied?
Do their commanders function?
Can aircraft be maintained?
Does intelligence reach the people who need it?
Do soldiers believe reinforcements will arrive?
Do they believe the government deserves defending?
Those questions determine military capability much more than a headline number.
This is something governments repeatedly do with statistics.
Numbers that are easy to measure become substitutes for things that actually matter.
Number of soldiers trained.
Money spent.
Schools constructed.
Elections held.
Districts controlled.
Those figures can demonstrate activity.
They do not necessarily demonstrate resilience.
The real test arrived when American support was removed.
Afghanistan’s institutions failed that test catastrophically.
$2 trillion cannot purchase legitimacy
The United States has spent an extraordinary amount on the war and reconstruction of Afghanistan over twenty years.
That makes today’s outcome instinctively feel impossible.
Surely enough money can build a functioning state.
But money can purchase equipment.
It can construct government buildings.
It can pay salaries.
It can train soldiers.
It can build roads.
What it cannot directly purchase is legitimacy.
People have to believe that the political system represents something worth preserving.
That seems to have been one of the Afghan government’s deepest weaknesses.
Afghanistan held elections, but those elections were repeatedly disputed.
Political elites were accused of corruption.
Government authority outside major urban areas was often limited.
Local relationships could matter more than institutions in Kabul.
And the state remained heavily dependent on foreign financial and military support.
The Taliban could therefore tell a powerful story:
The Kabul government exists because foreigners created and sustain it.
That narrative was obviously self-serving.
The Taliban itself has benefited from outside relationships and sanctuary.
But narratives do not need to be completely fair to become politically effective.
They need enough people to believe them.
Corruption wasn’t a side issue
I wrote last week that corruption can become a national-security problem.
This week demonstrates why.
If money intended for soldiers disappears, the military becomes weaker.
If officials obtain positions through patronage rather than competence, administration becomes weaker.
If citizens have to pay bribes for basic services, the government becomes less legitimate.
If political elites become wealthy while ordinary people remain poor, claims about democracy begin sounding hollow.
Corruption therefore attacks a state twice.
First materially.
Resources disappear.
Then psychologically.
Trust disappears.
The second may be even more damaging.
A soldier can tolerate bad equipment if he believes he is defending something important.
It becomes harder if he believes the people asking him to die have already stolen from the country he is supposed to defend.
But blaming Afghans would be extraordinarily convenient
There will undoubtedly be an argument that America gave Afghanistan twenty years, enormous amounts of money and a military, and Afghans simply failed to use them.
That explanation absolves the foreign power almost perfectly.
It is also inadequate.
The United States helped design the system that failed.
It selected partners.
It supported political elites.
It shaped the Afghan military.
It chose where money went.
It tolerated some forms of corruption because particular individuals were useful allies.
It repeatedly changed strategy.
If you spend twenty years constructing a machine and the machine stops functioning the moment you remove your hand from it, you cannot simply blame the machine.
The architect deserves scrutiny too.
This does not mean Afghans possess no agency.
President Ghani and other Afghan leaders made decisions.
Commanders made decisions.
Officials stole money.
Political factions prioritised their own interests.
Responsibility is distributed.
But distributed responsibility does not mean nobody is responsible.
Biden’s prediction aged terribly
Only last month, President Biden publicly rejected comparisons between Afghanistan and Vietnam and said a Taliban takeover of the entire country was not inevitable.
Technically, he was correct that it was not inevitable.
Politically, today’s scenes make that statement look disastrously optimistic.
The administration appears to have badly underestimated the speed at which the Afghan government could collapse.
That matters independently of whether withdrawal itself was correct.
These are two separate questions:
Should America have withdrawn?
and
Was the withdrawal planned around an accurate understanding of what would happen afterwards?
I still think there is a strong argument for withdrawal.
I am increasingly convinced the second answer is no.
A government can make the correct strategic decision and execute it badly.
It can also make a bad decision efficiently.
We should not combine those questions simply because politics prefers one-word verdicts.
The embassy evacuation matters symbolically
American personnel are now being evacuated from Kabul.
Helicopters above an embassy inevitably create comparisons with Saigon in 1975.
Biden explicitly rejected that comparison only weeks ago.
The historical situations are obviously different.
Afghanistan is not Vietnam.
The Taliban is not North Vietnam.
The geopolitical contexts are completely different.
But the image matters because political events are interpreted through symbols.
For twenty years, the American embassy represented the permanence of America’s commitment to the Afghan government.
Now its evacuation communicates the opposite.
Again, expectations.
If you are an Afghan official watching foreign diplomats leave, what does that tell you?
If you are a Taliban commander, what does it tell you?
Military withdrawals communicate information even before the final soldier leaves.
Ghani’s departure may have prevented a battle — but destroyed what remained of political authority
President Ashraf Ghani has left Afghanistan.
He says he did so to avoid bloodshed.
If remaining in Kabul would genuinely have produced a destructive battle through a city containing millions of civilians, avoiding that battle matters.
But the president leaving also destroyed whatever remained of the government’s political centre.
A state needs somebody capable of issuing orders that other people expect to be followed.
Once the president departed, that expectation effectively disappeared.
This demonstrates again why political authority is partly psychological.
Ghani may still have possessed the legal title of president when he left.
Legal authority became irrelevant once nobody believed his government could exercise it.
The Taliban possessed less constitutional legitimacy.
It possessed more effective power.
Today, effective power won.
And now comes the question of women
For millions of Afghan women and girls, today’s political change is not theoretical.
Under Taliban rule before 2001, girls were largely prevented from attending school and women faced severe restrictions on employment, movement and public life.
The Taliban now says it has changed.
It says women’s rights will be respected within Islamic law.
The world will have to judge those claims by behaviour rather than statements.
Twenty years created a generation of Afghan women whose expectations are completely different from those of 2001.
They attended universities.
They became journalists.
Doctors.
Politicians.
Judges.
Entrepreneurs.
Teachers.
That social change cannot simply be erased from history.
But political power can constrain it.
This is perhaps the most painful aspect of America’s withdrawal.
The United States cannot remain militarily in Afghanistan forever simply to impose social policies it prefers.
But the rights people acquired under the political order America helped protect are now vulnerable partly because that protection has disappeared.
Again, responsibility survives withdrawal.
What happens to the people who trusted America?
There is a more immediate obligation.
Afghan interpreters, embassy employees and others who worked directly with foreign governments now face potentially enormous danger.
Their relationship with America is not abstract.
They made specific decisions because the United States asked for their assistance.
That creates a specific obligation in return.
Evacuating them should therefore not be described as charity.
It is the completion of an agreement.
If powerful countries expect local people to assist them during future wars, those people need to believe they will not be abandoned when political priorities change.
Foreign policy has memory.
Credibility is built from what happens to people after they stop being strategically useful.
Pakistan will face consequences too
A Taliban victory may appear strategically beneficial to some in Pakistan who have long sought influence in Afghanistan and feared an Afghan government aligned too closely with India.
But victory creates new problems.
A Taliban-controlled Afghanistan could embolden Islamist militant movements elsewhere.
Refugees may cross borders.
Instability does not remain neatly contained.
Pakistan therefore faces a familiar problem in foreign policy:
an actor can be useful while opposing your rival and dangerous once it becomes powerful enough to pursue its own objectives.
Proxies are not puppets.
They have interests too.
That lesson appears repeatedly in international relations.
Governments support armed groups because their objectives temporarily overlap.
Then the armed group survives long enough that the overlap disappears.
China, Russia and Iran now have to deal with reality
America’s withdrawal does not remove geopolitics from Afghanistan.
It changes who manages it.
China worries about instability and Islamist militancy near its western frontier but may see opportunities for economic and political influence.
Russia fears instability spreading towards Central Asia.
Iran shares a long border with Afghanistan and has religious, economic and security interests there.
Pakistan has perhaps the deepest relationship with the Taliban.
India has invested heavily in the Afghan government that is now disappearing.
None of these countries can simply watch from outside.
This is another reason describing the withdrawal as America “ending the war” was always misleading.
America is ending its military role.
Afghanistan remains located at the intersection of several states’ interests.
Geography does not withdraw.
Was the entire war pointless?
I don’t think so.
That conclusion is emotionally understandable tonight.
But it would erase twenty years of Afghan lives.
Girls who received an education did not receive an imaginary education because the Taliban later returned.
A journalist who spent twenty years reporting freely did not somehow lose those twenty years.
A child who survived because healthcare improved does not become irrelevant because the political project eventually failed.
Political outcomes do not retroactively erase every benefit produced along the way.
But neither can those benefits erase the central strategic failure.
America’s ultimate project was supposed to produce an Afghan state capable of surviving without permanent American military protection.
Tonight, that state no longer meaningfully exists.
That is failure.
The question is what kind.
Twenty years, and then ten days
There is something deeply unsettling about the timescale.
The United States spent almost twenty years constructing Afghanistan’s post-Taliban political order.
The Taliban dismantled most of its territorial control in little more than a week.
That does not mean twenty years of work literally disappeared in ten days.
But it reveals how differently we should measure political strength.
A building can look impressive until an earthquake tests its foundations.
Then the architecture matters less than what was underneath it.
Afghanistan had ministries.
An army.
A constitution.
Elections.
Foreign embassies.
Billions of dollars in assistance.
All the visible architecture of a modern state.
This week tested the foundations.
Too many of them depended on the assumption that the United States would remain.
So was withdrawal wrong?
After watching Kabul fall, it would be easy for me to reverse what I wrote in April.
I don’t think the evidence allows such a simple conclusion.
If America had maintained troops, perhaps the Afghan government would still control Kabul tonight.
That is a powerful argument.
But then the question returns:
For how long?
One year?
Five?
Twenty?
If the Afghan state could survive only while foreign troops guaranteed its survival, maintaining that arrangement indefinitely was not necessarily victory.
It was postponement.
But Biden’s administration also cannot hide behind that argument.
If withdrawal was necessary, the government still had an obligation to understand the consequences, prepare evacuations, protect people who had assisted America and plan for the possibility of rapid collapse.
Today’s events suggest those preparations were nowhere near sufficient.
Two propositions can therefore be true:
America could not remain in Afghanistan indefinitely.
America could have left Afghanistan much better than this.
Politics becomes easier when we insist on choosing only one.
Reality does not.
The question I asked in April has an answer now
When Biden announced the withdrawal, I wrote:
“Ending a war and ending responsibility for it may not be the same thing.”
Four months later, I think that sentence matters more than I realised.
America’s war is ending.
Its responsibility is not.
There are people to evacuate.
Refugees to assist.
Diplomatic commitments to reconsider.
A Taliban government whose behaviour must be judged.
Regional instability to manage.
And eventually, an enormous accounting of how a twenty-year state-building project collapsed so quickly.
That investigation should not begin with the objective of finding one person to blame.
It should ask something more useful:
Which assumptions were wrong?
Because Afghanistan’s collapse was built from assumptions.
That money could create institutions.
That training could create loyalty.
That military superiority could create political legitimacy.
That elections could create a state.
That the Afghan government would survive long enough for withdrawal to be orderly.
And perhaps the most consequential assumption of all:
that because a political system had survived for twenty years, it had become capable of surviving on its own.
Tonight, Taliban fighters are inside Kabul.
The American flag is coming down.
Ashraf Ghani is gone.
And twenty years after the United States removed the Taliban from Afghanistan’s capital, the Taliban has returned without having to fight the enormous final battle almost everyone expected.
Last week I wrote that governments can collapse when people stop believing they will still be governing tomorrow.
Today Afghanistan demonstrated the terrifying speed with which tomorrow can arrive.
