Last Sunday, I wrote that the Taliban was beginning to test whether Afghanistan’s government could defend major cities without large numbers of foreign troops.
Seven days later, that test is already producing answers.
And they are not reassuring.
On Friday, the Taliban captured Zaranj, the capital of Nimroz province.
It was the first provincial capital to fall to the group during its current offensive.
On Saturday, Sheberghan, the capital of Jowzjan province, fell.
Today, the situation accelerated dramatically.
The Taliban captured Kunduz, one of northern Afghanistan’s most important cities. It also took Sar-e-Pul and Taloqan.
Five provincial capitals have now fallen in three days.
Fighting continues elsewhere, including around Herat, Kandahar and Lashkar Gah.
A week ago, it was still possible to describe the Taliban’s extraordinary territorial expansion mainly as a rural offensive.
That description is becoming impossible.
The question is no longer whether the Taliban can threaten Afghanistan’s cities.
It can.
The question is whether the Afghan government can stop individual defeats from becoming something much more dangerous:
the expectation that it is going to lose.
Because wars are fought with weapons.
But they are also fought with expectations.
And expectations can collapse much faster than armies.
Kunduz changes the picture
Zaranj matters because it was the first provincial capital to fall.
Kunduz matters for a different reason.
It is a major city.
It sits in a strategically important part of northern Afghanistan and has previously been contested by the Taliban. Its capture demonstrates that the group’s offensive is not confined to its traditional southern heartlands.
That geographical expansion is politically significant.
For years, the Taliban has been associated particularly strongly with southern and eastern Afghanistan.
If it can sustain military operations and seize important cities in the north as well, the government faces a national rather than regional challenge.
There is another reason cities matter.
Territory is not equal.
Capturing an enormous rural district containing relatively few people is not politically equivalent to capturing a provincial capital containing government offices, infrastructure, businesses and hundreds of thousands of residents.
Maps can therefore be misleading.
Colouring a district Taliban-controlled tells us something.
Controlling Kunduz tells us considerably more.
The Afghan military should be stronger
This remains the central contradiction.
The Afghan security forces have received twenty years of American and NATO training.
The United States has spent enormous amounts equipping them.
Afghanistan possesses an air force.
The Taliban does not.
The government has armoured vehicles, artillery and other equipment far more sophisticated than much of what Taliban fighters possess.
On paper, the state should have overwhelming advantages.
But wars are not fought on paper.
The Afghan military has weaknesses that equipment cannot automatically solve.
Units need ammunition.
Aircraft need maintenance.
Soldiers need salaries.
Outposts need resupply.
Commanders need reliable information.
Wounded soldiers need evacuation.
And everybody needs to believe that the institution behind them will continue functioning.
The American military did not simply provide soldiers.
It provided an enormous system surrounding Afghan forces.
Intelligence.
Air support.
Logistics.
Maintenance.
Planning.
Contractors.
Transport.
When that system is removed, giving somebody an advanced piece of equipment does not necessarily mean they retain the ability to use it effectively.
Perhaps Afghanistan’s military dependence on America ran much deeper than the number of American soldiers physically fighting beside Afghan troops.
But logistics still doesn’t explain everything
Something psychological is happening too.
Consider the decision facing an Afghan soldier defending an isolated position.
He knows the Taliban is approaching.
He hears that another district surrendered.
Then another.
Perhaps his unit has not been paid properly.
Perhaps ammunition is running low.
Perhaps reinforcements have been promised before and never arrived.
Perhaps local leaders are already negotiating with the Taliban.
Now he has to decide whether to fight.
From outside Afghanistan, we can say:
He is a soldier. His duty is to defend the government.
From his perspective, the calculation may look different.
If he believes Kabul will ultimately win, fighting makes sense.
If he believes Kabul is going to lose, continuing to fight could mean dying for a government that will not exist afterwards.
That is where expectations become self-fulfilling.
A soldier expects defeat, so he surrenders.
His surrender makes defeat more likely.
The next soldier sees that surrender and updates his own expectations.
He surrenders too.
Eventually, a military force can begin collapsing faster than its actual battlefield losses would predict.
The Taliban appears to understand this extremely well
The Taliban does not need to physically destroy every Afghan military unit.
It can persuade some to stop fighting.
Reports throughout the offensive have described local surrender agreements in which Afghan forces hand over positions and weapons.
That can be much more efficient than fighting.
It preserves Taliban fighters.
It provides equipment.
It reduces destruction.
Most importantly, every negotiated surrender sends information to the next district.
Other people have already decided resistance is pointless.
This is political warfare as much as military warfare.
The objective is not simply to defeat your opponent.
It is to convince your opponent that defeating you is impossible.
Once enough people believe that, military power can snowball.
States depend on coordination
This has made me think differently about what a state actually is.
We normally imagine government as something physical.
A presidential palace.
Government buildings.
Police stations.
Military bases.
Flags.
But a state is also an enormous coordination system.
Millions of people behave every day as though its authority will continue tomorrow.
Civil servants go to work because they expect their ministry to exist.
Police enforce laws because they expect courts and government to support them.
Soldiers fight because they expect other soldiers to fight.
Citizens pay taxes because they expect everybody else to remain inside the same political system.
The government therefore possesses power partly because everybody expects everybody else to recognise its power.
That makes political authority strangely circular.
The government is powerful because people believe it is powerful.
Usually, that belief is extremely stable.
During state collapse, it can reverse.
Once enough people begin believing the government may disappear, their behaviour makes its disappearance more likely.
That may be what Afghanistan is beginning to experience.
This doesn’t mean Kabul is about to fall
We need to be careful here.
Five provincial capitals have fallen.
Afghanistan has 34 provinces.
The government still controls Kabul and many major population centres.
Its forces are still fighting.
The United States is still conducting some air strikes.
There is a very large difference between capturing provincial capitals and capturing the national capital.
It would therefore be irresponsible, on 8 August, to announce that Taliban victory is inevitable.
Wars change.
Offensives stall.
Governments reorganise.
Foreign governments alter policy.
Cities are recaptured.
The Afghan government still possesses significant military resources.
But something has clearly changed.
The burden of proof is shifting.
Several months ago, the question was whether the Taliban could survive after America left.
Then it became whether the Afghan government could prevent Taliban territorial gains.
Now it is increasingly whether the Afghan government can demonstrate that it is capable of reversing them.
That is an extremely dangerous progression for Kabul.
What was America actually training?
This also forces another uncomfortable question about the twenty-year intervention.
Last week I compared the Afghan state to a bridge that had remained standing while scaffolding supported it.
This week makes me wonder whether that analogy was still too generous.
Perhaps the problem was not simply dependence.
Perhaps America helped construct Afghan institutions according to an American model that Afghanistan could not independently sustain.
The United States built an Afghan military heavily dependent on technology, air power and sophisticated logistics.
That model works extraordinarily well when the world’s richest military is supplying the logistics.
It becomes harder when Afghanistan has to maintain the system itself.
This is a broader problem with foreign state-building.
A powerful country arrives with enormous resources and asks:
How would we build an army?
Perhaps the better question is:
What kind of army can this country realistically maintain after we leave?
Those answers may be very different.
An institution should not be judged only by how effective it is while receiving outside assistance.
It should also be judged by whether the country inheriting it can afford and operate it afterwards.
Corruption becomes a national-security problem
There is another part of state-building that receives less attention than weapons.
Corruption.
For years there have been reports of corruption inside Afghanistan’s government and security forces, including problems involving salaries, procurement and so-called ghost soldiers who exist on official records but not necessarily on battlefields.
Corruption is often discussed as though it is principally about morality.
Politician steals money.
Official accepts bribe.
Public becomes angry.
But in wartime, corruption becomes something more serious.
Money stolen from a military budget can mean ammunition that never arrives.
A nonexistent soldier cannot defend an outpost.
A commander appointed because of political connections rather than competence can lose a battle.
And every instance of corruption damages trust.
Why should a soldier risk his life for a political elite he believes is stealing from the state?
That means corruption can eventually become military weakness.
Governance and warfare are not separate systems.
One determines the capacity of the other.
The Taliban has problems of its own
None of this means Taliban control will produce effective government.
Capturing territory and governing it are different achievements.
The Taliban can exploit dissatisfaction with Kabul.
Once it controls a city, it becomes responsible for that city.
People need electricity.
Food needs to reach markets.
Hospitals need supplies.
Civil servants need salaries.
Businesses need predictable rules.
Crime needs to be controlled.
Disputes need to be resolved.
An insurgent movement can blame the government for everything that goes wrong.
A government eventually has to provide alternatives.
That transition from insurgency to administration may expose completely different weaknesses.
Military momentum therefore tells us who may control territory.
It does not tell us whether they can govern it successfully.
And what do the people in these cities want?
This is another point that risks disappearing beneath military maps.
A city “falling to the Taliban” does not mean everyone inside it suddenly supports Taliban government.
Nor does government control mean everyone supports President Ashraf Ghani.
Afghans are not merely territory being transferred between two coloured areas on a map.
They have preferences.
Some may support the government.
Some may support the Taliban.
Some may distrust both.
Many probably care most immediately about surviving whichever force controls their neighbourhood tomorrow.
That makes measuring political legitimacy extremely difficult during war.
If somebody obeys an armed group because the alternative is punishment, that is not necessarily consent.
If somebody remains silent because expressing opposition is dangerous, silence is not support.
This seems obvious.
Yet once an armed organisation controls territory, outsiders often begin treating control as evidence of popularity.
We should not.
The peace negotiations now look even weaker
Taliban representatives can watch the same developments we are watching.
Their bargaining position is improving.
That makes compromise less attractive.
Suppose you are negotiating over how much political power you will receive in a future government.
Then your forces capture five provincial capitals.
Why accept yesterday’s offer?
Wait.
Perhaps tomorrow you will control six.
This creates one of the cruelest dynamics in peace negotiations.
Military success can make peace less likely precisely because the successful side begins believing it can obtain more through continued fighting.
Diplomacy works best when both sides believe the costs of continuing exceed the likely gains.
Right now, the Taliban appears to have reasons to believe the opposite.
Is this Biden’s responsibility?
Partly.
But “Biden lost Afghanistan” would be far too simple.
The Trump administration negotiated the withdrawal agreement with the Taliban.
George W. Bush launched the war.
Barack Obama escalated and later reduced the American military presence.
Afghan political leaders have governed the country throughout much of the intervention.
Pakistan has pursued its own interests.
The Taliban made strategic choices.
Thousands of local commanders, officials and politicians shaped events.
A twenty-year war cannot reasonably be explained by the decisions of whichever American president happens to occupy the White House when the final troops leave.
At the same time, Biden made a choice.
He could have attempted to maintain an American military presence.
He decided not to.
Presidents inherit constraints, but inheriting a problem does not eliminate responsibility for what they choose within those constraints.
The intellectually easy positions are therefore both wrong.
Everything happening is Biden’s fault.
No.
Nothing happening is Biden’s responsibility because the war began twenty years ago.
Also no.
Political responsibility can be shared without being equal.
There may never have been a good ending
This is the possibility I find most uncomfortable.
Maybe we keep searching for the correct Afghanistan policy because we assume one must have existed.
Stay longer.
Leave earlier.
Send more troops.
Send fewer troops.
Negotiate sooner.
Build different institutions.
Pressure Pakistan harder.
Each might have improved some outcomes.
But perhaps by a certain point, every available option was bad.
Leaving risks Taliban victory.
Staying postpones withdrawal while Americans continue paying the costs of a war with no obvious endpoint.
Negotiating gives legitimacy to the Taliban.
Refusing to negotiate leaves military force as the alternative.
Supporting Kabul perpetuates dependence.
Removing support risks collapse.
Sometimes political failure is not choosing the wrong option from a menu containing a correct one.
Sometimes previous decisions have narrowed the menu until only bad options remain.
That doesn’t eliminate responsibility.
It means responsibility extends backwards.
The next week matters enormously
Five capitals in three days creates momentum.
The Afghan government now has to break it.
Not simply militarily.
Psychologically.
It needs to demonstrate to soldiers, governors, local leaders and civilians that surrender is not inevitable.
Because if everyone begins making individual decisions based on the assumption that the Taliban will eventually win, those individually rational decisions may collectively ensure that it does.
This may be the most important difference between the Afghanistan I wrote about last Sunday and the Afghanistan I am writing about tonight.
A week ago, the Taliban was capturing territory.
Now it may be capturing something much more valuable:
expectations.
And once a government loses people’s expectation that it will still be governing tomorrow, no amount of military equipment can easily restore it.
That is when a military crisis becomes a crisis of the state itself.
