Eleven days into Russia’s invasion of Ukraine, the war already looks different from many expectations before it began. Russian forces have captured territory and continue attacking from several directions, but they have not produced the rapid collapse of the Ukrainian government that seemed possible during the invasion’s first hours. Kyiv remains under Ukrainian control. President Volodymyr Zelensky remains in the capital. Ukrainian forces continue resisting, while Russia’s enormous military convoy north of Kyiv appears to have made limited progress.
The human consequences, however, are expanding much faster than the front lines. More than 1.5 million people have already fled Ukraine, according to the United Nations refugee agency. Most are women and children because Ukrainian men aged 18 to 60 have generally been required to remain in the country. Poland and other neighbouring states are receiving enormous numbers of refugees, while millions more people remain displaced or trapped inside Ukraine. Russian attacks are increasingly affecting urban areas, and attempts to establish humanitarian corridors have become part of negotiations between the two sides.
President Zelensky has repeatedly asked NATO to do something that, at first, sounds both straightforward and morally compelling: establish a no-fly zone over Ukraine.
NATO has refused.
That decision has produced understandable anger in Ukraine. If Russian aircraft are attacking a sovereign country and NATO possesses some of the world’s most capable air forces, why not prevent Russian aircraft from entering Ukrainian airspace?
The answer is that a no-fly zone is not a force field.
It is a war.
To establish one, NATO would have to tell Russian military aircraft that they are prohibited from flying over Ukraine. If Russian pilots ignored the order, NATO aircraft would have to intercept them and potentially shoot them down. Russian air-defence systems capable of threatening NATO aircraft might also have to be attacked, including systems operating from territory outside Ukraine.
At that point American, British, French or other NATO forces would be deliberately attacking Russian forces.
Russia could retaliate against NATO aircraft, bases or other military assets.
The alliance would no longer merely be supporting Ukraine.
It would have entered direct combat with Russia.
That distinction explains almost everything about Western strategy since the invasion began. NATO governments are trying simultaneously to accomplish two objectives: make Russian aggression fail and avoid a direct NATO-Russia war.
The problem is that policies which advance the first objective can increase risks to the second.
Send more weapons to Ukraine and Ukrainian resistance becomes stronger, but Russia can argue that NATO involvement is deepening. Impose more severe sanctions and Russia pays a larger economic price, but confrontation intensifies. Share intelligence and Ukrainian forces become more effective, but the boundary between assistance and participation becomes less clear. Establish a no-fly zone and the boundary largely disappears.
This is escalation management.
And because Russia possesses nuclear weapons, escalation management is not cowardice. It is one of the central responsibilities of every government involved.
That can sound terribly unsatisfying when Ukrainians are being killed.
It is much easier to discuss restraint from London, Washington or Delhi than from Kharkiv while explosions are occurring nearby. Zelensky’s frustration is therefore completely understandable. His responsibility is to protect Ukrainians. If NATO air power could reduce Russia’s ability to attack his country, he has every reason to demand it.
Joe Biden and other NATO leaders have a different responsibility. They have to consider not only whether intervention would help Ukraine today but what happens after the first Russian aircraft is shot down by an American pilot.
Political arguments often stop one step too early.
Do X because X would help.
Policy requires another question:
And then what?
Suppose NATO declares a no-fly zone tomorrow. Russia ignores it. An American fighter shoots down a Russian aircraft. Russia then attacks the NATO aircraft or the base supporting it. NATO retaliates. Russia escalates again.
At which stage does the confrontation stop?
Nobody can answer confidently.
That uncertainty is precisely why the first step matters so much.
Escalation is not necessarily automatic. Governments can choose restraint at every stage. But once two nuclear powers begin directly killing each other’s military personnel, every subsequent decision occurs under greater pressure, fear and domestic demand for retaliation.
The probability of catastrophic escalation may still be low.
The consequence is so enormous that even a low probability matters.
This is one of the uncomfortable lessons nuclear weapons impose on international politics. They do not make conventional aggression impossible. They change what other states are willing to do in response.
Russia’s nuclear arsenal therefore protects Russia in a way that extends beyond preventing a nuclear attack on Russian territory. It raises the potential cost of direct military intervention against Russian forces elsewhere.
That is an extraordinary strategic advantage.
It is also deeply troubling for the non-proliferation system.
Last month I wrote about the signal Ukraine’s experience may send to other countries considering whether nuclear weapons improve their security. This week’s no-fly-zone debate strengthens that concern. NATO is willing to impose enormous economic costs on Russia and provide Ukraine with extensive military assistance, but nuclear escalation risk places a clear ceiling on direct intervention.
Other governments will notice.
Again, this does not mean every country should seek nuclear weapons. Nuclear arsenals are enormously expensive and dangerous, and proliferation itself increases the probability of catastrophic accidents or conflict.
But non-proliferation becomes harder when states observe that nuclear weapons appear to provide powerful insurance against foreign military intervention.
The international system therefore needs ways to make aggression costly enough that nuclear weapons do not become a shield behind which conventional conquest becomes profitable.
That is where sanctions and Ukrainian resistance become strategically important.
The economic response to Russia has developed with remarkable speed. Western governments have imposed severe restrictions on Russian banks and individuals, limited the Russian central bank’s access to reserves held abroad, introduced export controls and restricted selected banks’ access to SWIFT. Companies are also suspending operations or leaving Russia independently of formal government sanctions.
The rouble has fallen sharply, financial restrictions are affecting ordinary economic activity, and the Russian government has introduced measures intended to stabilise the economy.
This is economic warfare in everything except the formal terminology.
Yet the same question applies here as with the no-fly zone: what is the objective?
If the purpose is punishment, sanctions are clearly imposing costs.
If the purpose is deterrence, they failed to prevent the invasion because they were mostly threatened before the war and imposed afterwards.
If the purpose is now to persuade Putin to end the war, their success remains uncertain.
These objectives should not be conflated.
Sanctions can still matter even if they do not produce immediate Russian withdrawal. They can reduce Russia’s long-term economic and technological capacity, increase the cost of continuing the war and signal to other states that territorial aggression produces serious consequences.
But broad sanctions also impose costs on ordinary Russians, many of whom neither chose the war nor possess realistic mechanisms for changing Putin’s decision.
That creates a moral problem.
It also creates a strategic one.
Economic pain can produce political pressure against a government.
It can also produce nationalism and allow the government to blame foreigners for hardship.
Which mechanism dominates depends on institutions, information and public beliefs.
Russia’s increasingly restricted media environment matters here because citizens do not necessarily receive the same account of the war that people outside Russia do. The Kremlin continues calling the invasion a “special military operation”, and Russian authorities have moved against independent reporting that contradicts the official narrative.
Information control therefore becomes part of wartime strategy.
Governments need citizens to understand why they are being asked to bear costs. Russia needs its population to interpret sanctions as Western hostility rather than consequences of the invasion. Western governments want Russians to interpret those same costs as consequences of Putin’s decision.
The economic effect can be identical while the political effect depends on the story attached to it.
This is another reason sanctions cannot be evaluated through GDP alone.
Political behaviour is mediated by interpretation.
Ukraine’s military resistance is also changing international calculations. Before the invasion, Western governments were trying to deter Russia from attacking. Now they are trying to make the attack less successful without entering the war themselves.
Weapons transfers have therefore accelerated dramatically. Even countries historically reluctant to supply weapons into conflicts have changed policy. Germany’s reversal last weekend is particularly significant. The European Union itself has agreed to finance military assistance to Ukraine, an extraordinary development for an institution originally built principally around economic integration.
Russia’s invasion is therefore changing Europe faster than years of debate over strategic autonomy managed to do.
Threats can solve coordination problems.
European governments previously disagreed over defence spending, relations with Russia and how much strategic responsibility Europe should assume. An external shock has suddenly increased the cost of disagreement.
This does not mean European unity will necessarily survive indefinitely. Sanctions will create economic costs. Refugee flows will create political pressures. Energy prices may rise. Different governments will experience these burdens differently.
Unity is easiest during the first emotional response to a crisis.
The real test comes when solidarity becomes expensive.
Poland provides an interesting example. Only months ago, Warsaw and Brussels were engaged in a serious confrontation over judicial independence and EU law. Today Poland is one of the principal destinations for Ukrainians fleeing the war and one of the strongest advocates of support for Kyiv.
Geopolitical threats can reorder political priorities extraordinarily quickly.
The EU’s internal disputes have not disappeared.
They have become less important relative to the immediate external threat.
This is another reason institutions can survive serious internal disagreement. Countries do not need to agree about everything to conclude that cooperation remains preferable to separation.
The refugee response is similarly revealing. More than 1.5 million people have left Ukraine in less than two weeks, creating one of Europe’s fastest refugee movements in decades. European governments have responded by opening borders and preparing temporary protection rather than forcing every Ukrainian through ordinary asylum procedures.
The generosity is necessary and impressive.
It also raises questions about consistency. Europe has spent years arguing bitterly over refugees and migrants arriving from Syria, Afghanistan, Africa and elsewhere. Governments that previously resisted accepting large numbers of refugees are now welcoming Ukrainians.
There are practical differences. Ukraine directly borders the EU. Most people fleeing are expected to seek temporary protection relatively close to home. Cultural and historical relationships differ.
But those explanations should not prevent examination of whether race, religion and identity also influence who Europeans perceive as deserving refugees.
Humanitarian principles are strongest when they apply beyond people who look culturally familiar.
This does not reduce the case for welcoming Ukrainians.
It strengthens the case for asking whether previous refugees deserved more of the same empathy.
The war is already forcing Europe to reconsider several assumptions simultaneously: defence spending, energy dependence, refugees, sanctions, NATO and the relationship between economic and military power.
That is what genuine geopolitical shocks do. They do not simply create one new policy. They alter the relative importance of existing policies.
Energy security is now national security.
Refugee policy is now security policy.
Banking infrastructure is now foreign policy.
Technology exports are now strategic assets.
The boundary between economics and geopolitics is disappearing because it was probably never as clear as we pretended.
Yet the most difficult boundary remains the military one.
How much can NATO do for Ukraine without becoming a combatant?
There is no perfectly objective answer. International law provides some guidance, but escalation is partly about perception. Russia may interpret an action as participation even when NATO considers it assistance.
The West therefore needs to calibrate support continuously.
Weapons, humanitarian aid, intelligence and economic pressure can substantially strengthen Ukraine while keeping NATO forces outside direct combat. A no-fly zone would cross that boundary much more clearly because enforcement requires direct military action against Russia.
For that reason, I think NATO is right to reject it.
That conclusion is uncomfortable precisely because the case for intervention is morally powerful.
Ukrainian civilians are being killed. Cities are being attacked. Russia initiated the war. Ukraine is asking for help.
But foreign policy cannot evaluate an action only through the suffering it might reduce immediately. It has to consider the suffering the action could create through escalation.
Restraint is not indifference when the alternative carries a meaningful risk of a much larger war.
At the same time, fear of escalation cannot become an excuse for doing almost nothing. If nuclear powers learn that possessing nuclear weapons allows them to attack non-nuclear neighbours while the rest of the world remains paralysed, the international system becomes substantially more dangerous.
The strategy therefore has to occupy an uncomfortable middle: support Ukraine strongly enough that aggression becomes costly, while maintaining enough separation that NATO and Russia do not begin directly fighting one another.
There will be constant arguments over where that line lies.
Those arguments are not evidence that the strategy is incoherent.
They are the strategy.
Escalation management is a process of repeatedly deciding which additional risk is justified by which additional benefit.
Perhaps this is the most important difference between political analysis before and after a war begins. Before 24 February, the central objective was preventing the invasion. That objective failed.
Policy now operates inside the consequences of that failure.
The relevant question is no longer how to create a world in which Russia never invaded Ukraine. That counterfactual is gone.
Governments have to choose among the options that remain.
A no-fly zone might help Ukraine militarily while dramatically increasing the risk of direct NATO-Russia conflict.
Doing nothing would reduce escalation risk while making Russian military success easier.
The current strategy—sanctions, weapons, intelligence, humanitarian support and no direct NATO combat—tries to occupy the space between them.
It may fail.
It may prove insufficient.
It may need to change as the war changes.
But the fact that an option is morally imperfect does not mean a morally perfect option exists somewhere else.
That may be the hardest political habit to develop: comparing policies with their actual alternatives rather than with the outcome we wish were available.
Tonight, Ukrainians understandably want NATO aircraft protecting their skies.
NATO governments are refusing because protecting those skies would require being willing to shoot at Russians.
In another war, that might be a risk worth taking.
In a confrontation involving the world’s largest nuclear arsenals, the threshold should be extraordinarily high.
Sometimes refusing to intervene more directly looks like weakness because the suffering we can see is immediate while the catastrophe we are trying to avoid is hypothetical.
The hypothetical still matters.
Especially when the reason it remains hypothetical is that governments continue choosing not to make it real.
