The Nuclear Promise Gap

Would America trade New York for Paris?

Charles de Gaulle posed it in 1963. His conclusion — that the answer was probably no, and certainly unknowable — led directly to the Force de Frappe, France's withdrawal from NATO's integrated command structure, and six decades of argument about what exactly NATO's nuclear umbrella covers and whether it actually covers anything.

We're excavating that layer again.

The Foreign Affairs case for a coming deterrence crisis frames this as a novel emergency: what happens to NATO's eastern flank if American nuclear guarantees lose credibility? Can European powers fill the gap? What does deterrence look like without reliable American commitment?

These are the right questions. They have been since the alliance was founded. The stratum is very thick.

i · the architecture that was always unstable

Extended deterrence is a political promise wearing military clothing.

The mechanics are simple: a nuclear power tells allies it will defend them with nuclear weapons if necessary. Allies don't build their own weapons. Potential aggressors are deterred by the prospect of nuclear war. Everyone benefits from not having dozens of countries with atomic bombs.

The structural problem is obvious once you name it. Nuclear weapons are useful precisely because using them would be catastrophic. A commitment to use them on behalf of another country requires accepting catastrophic risk for a non-survival interest. Any adversary with functioning pattern recognition will notice this commitment might not hold under pressure.

NATO's founders noticed immediately. That's why they built the tripwire: American troops on German soil. Not primarily to fight the Red Army — both sides knew the numbers were unfavorable — but to make American deaths an automatic political trigger. Kill enough Americans and public opinion forecloses options regardless of what any president prefers. The tripwire converts "would the US really fight for Germany" into "the US is already fighting because Germans killed Americans."

This is the key structural insight: extended deterrence without forward presence is rhetoric. The nuclear umbrella requires physical fabric underneath it.

The 1960s were the decade Europeans learned to inhabit this instability. De Gaulle's solution was exit — build French weapons, leave NATO's command structure, never have to trust American judgment on French survival. Everyone else developed sophisticated rituals of shared commitment: the Nuclear Planning Group, dual-key arrangements, American weapons hosted in allied countries, giving allies enough ownership to feel invested without actually holding the trigger.

The rituals worked because American domestic politics remained stable on NATO for five decades. Even during rocky intervals — Nixon's détente, Reagan's SDI, the nuclear freeze movement that nearly unraveled missile deployments in 1983 — the commitment held. The adversary knew it held.

ii · the 1983 parallel that didn't kill nato

The closest historical match to the current moment isn't de Gaulle's 1963 departure. It's 1983.

The Soviet Union had deployed SS-20 intermediate-range ballistic missiles targeting Western Europe throughout the late 1970s. NATO's response — the dual-track decision of 1979 — was to simultaneously negotiate removal of the SS-20s and prepare to deploy Pershing II missiles and ground-launched cruise missiles in West Germany, the Netherlands, Belgium, Italy, and the United Kingdom.

The European left erupted. Hundreds of thousands marched in Bonn, Amsterdam, London. Helmut Schmidt's government nearly collapsed. The debate wasn't only about these specific weapons — it was about whether American weapons in Europe made Europe safer or simply a target, and whether European survival was being subordinated to American nuclear strategy that Europeans didn't control and weren't certain they trusted.

This was the de Gaulle anxiety in mass-movement form.

NATO held. The missiles went in. The Soviets came to the table. The INF Treaty in 1987 eliminated an entire class of weapons. The crisis ended up strengthening alliance credibility precisely because it tested solidarity and solidarity held.

The current crisis has one structural difference that makes it harder to resolve the same way. The 1983 pressure came from outside NATO — Soviet deployments creating the dilemma, American commitment reasserting itself against external pressure. The current pressure comes from inside. When the destabilizing actor is the United States itself, collective solidarity cannot solve the problem the way it solved the Euromissile crisis.

iii · the substitution problem

The Foreign Affairs analysis argues European allies cannot substitute for American nuclear guarantees. This is largely correct, and the reasons are instructive.

France holds approximately 290 nuclear warheads. The United Kingdom around 225. Combined, roughly 515 weapons — against Russian nuclear capacity this is manageable for survival-level deterrence. The problem is not size, doctrine, or delivery capability for national purposes. The problem is extending that guarantee to thirty-one other nations with different threat perceptions and different red lines.

Size matters less than it appears — you don't need thousands of warheads to deter, you need credible second-strike capability. France and Britain have that for their own defense. What they lack is the institutional architecture, the targeting doctrine, and above all the domestic political authorization to extend that guarantee to Estonia or Poland.

De Gaulle's insight runs in reverse here. France never declared clearly what would trigger French nuclear use, even for France — the ambiguity was deliberate, making adversary calculations harder. Extending that ambiguity to thirty-one nations with different interests and different threat perceptions requires a political commitment French voters have never been asked to authorize and French governments have never sought to make.

There's also the trigger question. American nuclear doctrine, whatever its imperfections, has decades of institutional development behind it. Hypothetical European extended deterrence would need to answer: when German calculations about Russian provocations diverge from French ones, who decides? That's not a military problem. It's a governance problem that the EU doesn't currently have the machinery to resolve and won't develop quickly.

None of this is permanently impossible. It's very slow, very expensive, and involves political negotiations that European governments haven't started in earnest.

iv · what actually deters

Here is the structural fact underneath the nuclear debate: the most stable period in European security was not when nuclear guarantees were most credible. It was when conventional deterrence was most robust.

American and NATO conventional forces in Europe made conflict too costly regardless of what any president might decide about nuclear escalation. The nuclear umbrella was backstop; the actual deterrent was the certainty of a prolonged conventional war that neither side could win quickly or cleanly. Conventional force credibility is what made the nuclear question theoretical rather than urgent.

This is why the current gap is more dangerous in the conventional dimension than the nuclear one. The real question is not whether Putin believes Trump would authorize a nuclear strike. It's whether Russian conventional forces would face credible resistance in the first seventy-two hours of a conflict.

If they do, the nuclear question probably never arises. If they don't, the nuclear question is already moot.

European conventional rearmament — higher defense budgets, rebuilt stockpiles, genuine readiness rather than paper readiness — is the tractable answer to the deterrence gap. It doesn't require solving the extraordinarily difficult political problem of European extended nuclear deterrence. It requires spending money that European governments had reasonable justifications not to spend for thirty years and now have better reasons to spend.

The nuclear debate is real. The credibility gap is genuine. But it is partly a distraction from the more immediate problem that has a more tractable solution.

v · the cycle that keeps completing

This crisis has a shape. European security follows the same pattern roughly every generation:

1. American commitment appears robust. European defense spending declines.
2. American domestic politics shift. Commitment becomes questionable.
3. Europeans rediscover defense spending. Agonized debate about strategic autonomy.
4. American commitment is reaffirmed or the threat environment changes. Cycle resets.

1970s. Early 1980s. Post-Cold War drawdown. Post-2014 adjustment. Now.

Each iteration leaves Europe somewhat better defended because each crisis produces real learning. NATO's eastern flank is stronger now than it was before 2014. It will be stronger after this crisis than before it. The pattern isn't pure failure — it's expensive, lurching progress toward a posture European governments keep deferring.

What makes the current iteration more dangerous is the uncertain timeline. Previous cycles had identifiable resolution mechanisms: a new administration, a treaty, a shift in the threat environment. The current uncertainty has no obvious endpoint.

De Gaulle's question — would America trade New York for Paris — was never really answered. It was deferred by constructing enough physical commitment that the question became theoretical rather than urgent. The question remained unanswerable; the answering was made unnecessary.

The deferral mechanism is weakening. The question is becoming less theoretical.

Europe is moving, with notable speed for the continent, to construct its own answer. Whether that answer gets built fast enough is the actual problem. Not the nuclear guarantee debate — real but glacial. Not the question of French extended deterrence — real but harder still. The immediate problem is whether conventional deterrence can be rebuilt quickly enough that the nuclear question stays theoretical while the political architecture catches up.

History suggests probably yes. The cycles have resolved before. What history also says is that the cost of resolution scales with the duration of the deferral.

The bill comes due eventually. It always has.

vi · sources

• The Coming Crisis of NATO Deterrence — Foreign Affairs, 2026
• Nuclear Weapons: Who Has What at a Glance — Arms Control Association, 2024
• NATO's Nuclear Sharing Arrangements — NATO
• The INF Treaty at a Glance — Arms Control Association, 2019