In the second half of 1991 the thing NATO had been built to oppose simply stopped existing. The Warsaw Pact, the military alliance of the Soviet bloc, was formally dissolved in July. The Soviet Union itself was gone by the end of December, broken into fifteen separate states, its empire in central Europe already dismantled. For forty-two years the North Atlantic Treaty Organization had had one organizing purpose, the containment of Soviet power in Europe, and now the Soviet power was not contained but abolished. By the plain logic of its own founding, the job was done. The enemy had not been defeated in battle. It had dissolved on its own, which is rarer and more complete. And the alliance built against it faced the question that every alliance eventually faces and that almost none survive: what do you do when the reason you exist disappears?

The answer NATO gave is the subject of this chapter, and it is more interesting than either of the usual stories about it. One story, the official one, says NATO is a defensive alliance that keeps the peace, and that its persistence simply reflects the persistence of danger. The other, the polemical one, says NATO is a war machine that invents enemies to justify itself. Both contain something true and both miss the actual mechanism, which is neither noble nor sinister but structural, and which applies not only to NATO but to every large institution that has ever outlived its original task. The mechanism is this: an institution, once built, develops an interest in its own continuation that is separate from the purpose it was built to serve, and when that purpose vanishes, the institution does not vanish with it. It looks for a new one.

The determining variable, in other words, is not whether NATO faces a real threat. Sometimes it does and sometimes it has not. The determining variable is what an institution does when its founding threat is removed, and the answer, visible across the three decades since 1991, is that it adapts, enlarges, and finds fresh missions, because the alternative to adapting is dissolving, and institutions are built, above all else, not to dissolve.

What an Alliance Is For

To see why NATO's survival is remarkable, you have to remember what alliances normally are, because NATO has made us forget.

Through most of history, a military alliance is a temporary thing. States band together against a specific danger, fight or deter it, and when the danger passes the alliance dissolves, because its members no longer have a reason to subordinate their sovereignty to a common command. The coalitions that defeated Napoleon dissolved. The alliance that won the First World War did not survive the peace. The grand alliance of the Second World War, which united the Western powers with the Soviet Union against Hitler, came apart within a few years of victory, its members turning from partners into the two camps of the Cold War. This is the normal life cycle of an alliance: it is summoned by a threat and released when the threat is gone. An alliance is a means, and means are supposed to be retired when the end is reached.

NATO was founded inside this logic, on the fourth of April 1949, by twelve states binding themselves under the famous Article 5, the pledge that an attack on one would be treated as an attack on all. Its purpose was specific and of its moment: to deter the Soviet Union from pressing westward into a Europe still in ruins. Its first secretary general, Lord Ismay, is supposed to have summed up its aims in a line that has outlived him, that NATO existed to keep the Russians out, the Americans in, and the Germans down. The line is worth dwelling on, because two of its three purposes had nothing to do with the Soviet threat at all. Keeping the Americans committed to Europe and keeping a defeated Germany contained were aims about the internal architecture of the West, not about the enemy to the east. Even at its founding, in other words, NATO was already more than a response to a single threat. It was a structure for organizing the Western world, and structures for organizing the world do not have a natural expiry date the way a response to a threat does.

That double nature is the seed of everything that followed. An alliance that exists only to counter an enemy ends when the enemy ends. An institution that also organizes a civilization has reasons to persist that the disappearance of any particular enemy does not touch.

The Enemy Disappears

So return to 1991, and to the moment that should, by the older logic of alliances, have been NATO's dissolution and instead became its reinvention.

The case for winding the alliance down was not a fringe position at the time. Serious people made it. If NATO was a defensive pact against the Soviet Union, and the Soviet Union no longer existed, then the pact had succeeded and could be honorably retired, perhaps folded into some broader European security arrangement that included Russia rather than excluding it. The Warsaw Pact had done exactly this, dissolving itself when its reason vanished. The symmetry would have been clean. But symmetry is not how institutions behave, and NATO did not dissolve. It did not even seriously consider dissolving. Instead it began, almost immediately, the work of redefining what it was for.

This is the hinge of the whole story, and it is where the Italian political theorist Niccolò Machiavelli is a better guide than any strategist. Machiavelli's enduring insight was that institutions, once they exist, behave so as to perpetuate themselves, and that this behavior requires no conspiracy and no malice, only the ordinary incentives of the people whose careers, budgets, status, and identity are bound up in the institution's continuation. A vast organization with headquarters and a command structure and tens of thousands of staff and an integrated military bureaucracy does not contain within it any mechanism for recommending its own abolition. Every incentive inside it points toward continuation. When such an institution loses its purpose, the people within it do not conclude that they should disband. They conclude, sincerely, that the institution is still needed, and they begin to articulate the reasons why. The search for a new mission is not cynical. It is what it feels like, from the inside, to be an institution that cannot imagine its own end.

The cleanest single fact in the entire history of NATO is this: the adversary it was created to oppose ceased to exist in 1991, and the alliance not only continued but grew. Everything else is commentary on that fact. An organization whose founding enemy has vanished, and which responds by enlarging, is an organization telling you that its existence is no longer primarily a function of that enemy. Something else is keeping it alive.

The Other Side of the Curtain

There is a control case for this argument, and it is the alliance on the other side of the Cold War, because the Warsaw Pact was also an institution built against an enemy, and it did dissolve. Asking why one survived and the other did not is the sharpest test of whether the institutional-survival drive is real or imagined.

The Warsaw Pact was founded in 1955, in part as a response to West Germany's admission to NATO, and for thirty-six years it was the formal military structure of the Soviet bloc. It too had headquarters, a command, a treaty, and a roster of member states. By the logic of self-preservation it should have been as reluctant to dissolve as NATO proved to be, and yet in July 1991 it wound itself up and ceased to exist. If institutions always perpetuate themselves, the Warsaw Pact is a counterexample, and a serious one.

But look at why it dissolved, and the counterexample turns into a confirmation. The Warsaw Pact was never an alliance of sovereign equals. It was an instrument of a single hegemon, the Soviet Union, which had used it to enforce its control over its satellites, sending its tanks into Hungary in 1956 and Czechoslovakia in 1968 to crush members that strayed. Its continuation depended entirely on the power and will of Moscow. When that power collapsed, the Pact had no independent constituency to keep it alive, because its members had never wanted it for themselves; they had been compelled into it. The moment the compulsion lifted, they fled it. The institution dissolved not because institutions naturally dissolve but because this particular institution had only ever had one stakeholder, and that stakeholder was disintegrating.

NATO was the opposite kind of structure, and the contrast is the whole point. Its members were sovereign states that had, for the most part, chosen it and valued it, each with its own bureaucracy, its own defense establishment, its own domestic constituency invested in the alliance. NATO had not one stakeholder but dozens, distributed across many capitals, none of whom wanted it to end. An institution with a single master dies when the master dies. An institution with many stakeholders, each with reasons to keep it, is nearly impossible to kill, because dissolution would require all of them to agree to it at once, and they never will. The Warsaw Pact and NATO ran the same experiment from opposite starting conditions, and the results map exactly onto the structure of who had an interest in continuation. That is not a coincidence. It is the mechanism, shown twice.

The Identity Problem

It is worth being precise about what the survival drive actually is, because the word makes it sound like greed, and greed is the wrong picture.

The people who staffed NATO in 1991 were not cynics plotting to manufacture a reason to keep their jobs. They were officials, officers, and analysts who had spent entire careers inside the alliance, for whom it was not merely an employer but a professional world, a community, a shared understanding of how European security worked and what their own lives had been for. To such people the proposition that the alliance should now dissolve was not a threat to be resisted so much as a thought that could barely be entertained, because it asked them to conclude that the structure they had served and believed in had become unnecessary. Almost no one, in any institution, reaches that conclusion about their own life's work. They reach the opposite conclusion, sincerely, that the work remains vital and the structure is still needed, and they begin, in good faith, to find the reasons.

This is why the survival drive needs no conspiracy and produces no villain. It is the aggregate of thousands of ordinary people, each behaving reasonably, each unable to see the institution they belong to as expendable, and the sum of all that sincerity is an organization that cannot recommend its own end. The search for new missions after 1991 was not a plot hatched in a room. It was the natural output of an institution full of people who believed in it, asking what it should do next, because what it should do next was a question they could ask and whether it should exist at all was a question they could not. The mechanism is banal in the precise sense: it runs on normal people doing normal things, and it produces, without anyone deciding it, an institution that outlives its purpose. The danger in the polemical account is that by insisting on villains it makes the structural truth invisible, because if there are no villains to find, the reader concludes there is nothing to see. There is plenty to see. It is just not a conspiracy. It is the ordinary physics of institutions.

The Search for a Mission

What kept it alive, through the 1990s and 2000s, was a sequence of new roles, each of which gave the alliance a fresh reason to exist, and the sequence is worth tracing because it is the institutional survival drive made visible.

The first new role was crisis management beyond NATO's own borders, the so-called out-of-area mission that the Cold War alliance had never contemplated. In Bosnia in 1995 NATO conducted its first major combat operation, and in 1999 it fought a seventy-eight-day air war over Kosovo, bombing Serbia to force a withdrawal from the province. The Kosovo campaign was conducted without the authorization of the United Nations Security Council, where Russia would have vetoed it, and it generated a lasting legal debate that has never been settled: defenders called it legitimate even if not strictly legal, a humanitarian necessity; critics called it a violation of the United Nations Charter by the very powers that wrote it. For the present argument the legal question matters less than the institutional one. An alliance founded for the territorial defense of its members was now fighting a war over a province in a non-member state, on humanitarian grounds, outside the framework of international law that was supposed to govern such things. It had found a new kind of mission, and the new mission justified the old institution.

After the attacks of the eleventh of September 2001, NATO invoked Article 5 for the first and only time in its history, and the new mission became Afghanistan, where the alliance took command of an international force in 2003 and remained, in one form or another, until the chaotic Western withdrawal of 2021. In 2011 it ran the air campaign over Libya, authorized this time by the Security Council under the banner of a responsibility to protect civilians, an operation that ended with the fall and death of Gaddafi and the collapse of the Libyan state into a decade of civil war. The missions differed in their legality, their success, and their justification. Some were UN-mandated and some were not; some were defensible and some were disasters. But viewed structurally they have a common feature that is independent of their merits: each gave an alliance without its founding enemy a reason to continue, a theatre in which to be necessary, a purpose to replace the one that had dissolved in 1991. The institution did not sit idle waiting for the Soviet threat to return. It went looking for things to do, and it found them.

The search for missions is not free, and the structural lens explains a pattern in how the post-Cold War operations went that the official and polemical accounts both struggle with.

An institution that needs a mission to justify its continuation has a particular relationship to that mission. It can enter one readily, because a new mission answers a real institutional need, the need to be relevant and necessary. But it has trouble exiting one, and even more trouble judging one a failure, because to admit that a mission was a mistake is to weaken the justification that the mission was supposed to provide. The result is a tendency to stay too long and to declare success too easily, not from stupidity but from the institution's interest in the mission's value being affirmed rather than questioned. Afghanistan is the clearest case. The intervention lasted twenty years, far longer than the threat that prompted it required, expanding from a narrow counterterrorism operation into an open-ended project of nation-building whose goals receded as the years passed, and it ended in 2021 in a withdrawal so abrupt and disordered that it read as a collapse. Libya is the other face of the same problem. The 2011 intervention was declared a success at the moment the old regime fell, and the institution moved on, even as the country slid into a decade of civil war and fragmentation that the success narrative had no room to acknowledge.

None of this means the operations were undertaken in bad faith, and the evenhanded point is not that NATO is incompetent or malign. Plenty of institutions with clear purposes also botch difficult operations. The structural observation is narrower and more specific: an institution for which missions are also justifications will find it unusually hard to end a mission cleanly or to assess it honestly, because the mission is doing double duty, accomplishing some external goal and at the same time validating the institution's continued existence, and those two jobs can pull in opposite directions. When they do, the institutional job quietly wins, and the result is the war that runs too long and the victory declared too soon. This is not a charge unique to NATO. It is the characteristic failure mode of any organization that has come to need its activities to prove that it should still exist, and recognizing it is part of seeing the institution clearly rather than either celebrating or denouncing it.

The Ratchet

Alongside the new missions ran two other movements that point the same way, and both have the quality of a ratchet, a mechanism that turns in only one direction.

The first is enlargement. NATO had grown during the Cold War, taking in Greece and Turkey, West Germany, and Spain, reaching sixteen members by 1991. But enlargement did not stop when the threat that justified a Western defensive bloc disappeared. It accelerated. In 1999 the alliance admitted Poland, Hungary, and the Czech Republic, states that had until a decade earlier been on the other side. In 2004 it took in seven more, including the three Baltic republics that had been part of the Soviet Union itself. Further rounds followed through the 2000s and 2010s, and by 2024 the alliance numbered thirty-two members, twice its Cold War size. An alliance that doubles in membership in the decades after its enemy disappears is not contracting toward obsolescence. It is expanding, and expansion creates new commitments, new borders to defend, new reasons the institution is indispensable.

The second ratchet is money. For years NATO members were urged toward a guideline of spending two percent of their economic output on defense, a target formalized at a summit in 2014 in the wake of Russia's annexation of Crimea, when only a handful of members were meeting it. By the middle of the 2020s most were. And the target did not stop at two percent. At a summit in The Hague in 2025 the allies agreed to a new and far higher goal, a headline figure of five percent of national output by 2035, structured as three and a half percent on core defense and a further percentage and a half on related security spending. The number that was a stretch in 2014 had, within a decade, become a floor on the way to something more than double. Behind these figures stands a permanent economic constituency, the defense industries whose order books are filled by every increase, for whom the alliance's expansion and the climbing targets are not abstractions but revenue. This need not be described as a racket, and the polemical version that does so overreaches. It is enough to observe, factually, that a large and well-organized set of interests benefits from the institution's growth and would be harmed by its contraction, which adds one more force pulling steadily in the direction the institution was already inclined to go.

The Strongest Objection

At this point an honest account has to stop and give the other side its full weight, because there is a powerful objection to everything said so far, and the events of recent years have sharpened it rather than blunting it.

The objection is that NATO's persistence and growth are not driven by the institution's hunger to survive but by genuine demand from sovereign states that want its protection. NATO cannot annex anyone. Under its own treaty, states apply to join, and the alliance admits them only by unanimous consent. The enlargement of the 1990s and 2000s was not NATO reaching out to swallow territory to justify itself. It was a queue of newly free central and eastern European states, fresh from decades under Soviet domination and frightened of a Russia they expected to recover, lobbying hard and persistently for the security that membership offered. From this angle the alliance did not perpetuate itself by seeking members; it grew because nervous neighbors demanded entry, and the persistence of their fear, not any institutional survival drive, explains the persistence of the alliance.

And then came the strongest evidence of all. On the twenty-fourth of February 2022, Russia launched a full-scale invasion of Ukraine, the largest war in Europe since 1945. Whatever one believes about the long history that led to it, the invasion demonstrated that the danger NATO's eastern members had feared was not imaginary. The response was not manufactured by Brussels. It came from the members and from states outside the alliance entirely. Finland and Sweden, two countries that had stayed officially neutral or non-aligned for generations, Sweden for two centuries, applied to join NATO specifically because of the invasion, abandoning a defining national tradition to seek the alliance's protection. A neutral state of two hundred years giving up its neutrality to get inside the tent is about the cleanest refutation imaginable of the claim that NATO survives by inventing its own necessity. Here the necessity arrived, unmistakable and unwanted, and drove a sovereign people to NATO's door. The threat, in 2022, was real, and the demand for the alliance was real, and no theory of institutional self-interest is needed to explain why NATO exists when a nuclear-armed neighbor is invading the country next door.

The Synthesis

So which is it? Is NATO an institution perpetuating itself past its purpose, or a genuine response to a genuine and recurring danger? The honest answer, and the one this chapter is built to reach, is that the question contains a false choice, and seeing why is the whole point.

Institutional self-preservation and genuine demand are not opposites. They coexist, and the most durable institutions are precisely the ones in which they reinforce each other. An institution can serve a real function and, at the very same time, develop an interest in its own continuation that outlasts and reshapes that function. The two are not in tension; the real function is what makes the self-preservation respectable, and the self-preservation is what keeps the institution standing during the intervals when the real function is in doubt. NATO's behavior across the three decades since 1991 shows both forces at work, and shows them in sequence. In the long interval from 1991 to 2021, when no existential threat to its members was visible, the institution displayed the marks of self-preservation: it did not dissolve, it went looking for new missions in the Balkans and Afghanistan and Libya, it enlarged into a gray zone of contested assurances about how far east it would go, and it kept its spending targets climbing. Then in 2022 the real function returned with brutal clarity, and the alliance that the survival drive had kept intact through the lean decades was there, ready, when the danger came back.

The determining variable, then, is not whether the threat is real. It is the asymmetry between how institutions are born and how they end. An institution is born for a purpose, but it does not die when the purpose does. It persists, by the ordinary force of the interests inside it, through the periods when its purpose is absent, and that persistence is neither heroic foresight nor cynical fraud. It is simply what institutions are: structures built to continue, and continue they do, finding new reasons when the old ones lapse, until a returning danger makes the new reasons unnecessary and the institution can point to the danger and say, you see, we were needed all along. Both things are true. The survival drive kept NATO alive through the years it had no clear enemy. The real threat then vindicated the survival. An institution that had dissolved in 1992 on the clean logic of mission accomplished would not have been there in 2022, and that counterfactual is the quiet case for the survival drive that the polemics and the official line both miss.

The Contested Ledger

Because this is a subject that attracts overstatement from every direction, it is worth setting down plainly what is established and what is not, since the difference is exactly where most writing on NATO goes wrong.

Take the claim, repeated as settled fact in much anti-NATO argument, that the West made a binding promise not to expand the alliance eastward and then broke it. The documented record is narrower and more interesting than that. In February 1990, in conversations about the reunification of Germany, the American secretary of state James Baker did tell the Soviet leader that NATO would not extend one inch to the east, and similar assurances came from other Western figures. But no treaty was signed, no binding commitment was ever written down, and the immediate context was the stationing of NATO forces on former East German soil, not the future admission of sovereign central European states that did not yet exist as independent actors. Historians genuinely disagree about what was promised and how far it bound anyone. The honest statement is that verbal assurances were given, that Russia later read them as a promise against enlargement, that most Western scholars read them more narrowly, and that whether a promise was broken remains contested. That is less satisfying than a clean betrayal, and it is the truth.

Take, too, the darker claim that NATO orchestrated terrorism against the citizens of its own member states. The factual core is real: clandestine stay-behind networks, the best known of them the Italian Gladio, did exist across Western Europe, set up during the Cold War to resist a feared Soviet occupation, and their existence was officially admitted by the Italian prime minister in 1990. But the leap from the existence of these networks to the assertion that NATO staged the bombings of Italy's years of lead to manipulate its own public is not established. Italian courts convicted figures from the neofascist milieu and condemned intelligence officers for diverting investigations, and individual operatives gave testimony pointing to a strategy of tension. That is serious, and it is also not the same as a proven chain of command running from Brussels to a bomb in a piazza. The networks were real and admitted. The staged-massacre attribution is contested and unproven, and a structural account of NATO does not need it and is weakened by leaning on it.

And take the example of Libya, often invoked to show what NATO destroyed. By 2010 Libya did have the highest human development ranking in Africa, with oil revenues funding free schooling and healthcare and a relatively high standard of living. It was also a closed dictatorship that one man had ruled for four decades, without a free press or political freedom, sustained by repression. The welfare and the repression were two faces of the same oil-funded state, and an account that mentions only the first is propaganda in the same way that an account mentioning only the second is. The honest ledger keeps both columns open. The discipline of doing so is what separates a structural analysis from a polemic, and it is the discipline the subject most often lacks.

What Institutions Do

Step back from NATO and the pattern is general, which is why it is worth the trouble. The deepest fact about institutions is that they are built to persist, and persistence outlives purpose. A bureaucracy created for an emergency survives the emergency. An agency established to fight a particular problem is still there, looking for problems, long after the original one is solved. The test of an institution is not how it behaves while its purpose is clear and urgent. It is what it does in the interval when its purpose has lapsed, and the answer, almost always, is that it does not conclude it should cease to exist. It adapts, it diversifies, it finds the next thing, because every incentive of every person inside it points toward continuation and none points toward dissolution.

NATO is the most consequential example of this in the modern world, but it is an example, not an exception. An alliance built to contain an empire watched that empire dissolve and did not stand down. It found new wars, took in new members, raised its budgets, and persisted through three decades in which its founding enemy was gone, until a new war on its eastern edge made its persistence look like wisdom. Perhaps it was wisdom. Perhaps the institution's refusal to die in 1992 is exactly why it was standing in 2022. But it is worth seeing clearly that the refusal came first and the vindication came after, and that an institution which can only be vindicated in retrospect was, in the long interval before the vindication, running on something other than necessity. It was running on the oldest engine an institution has, the one that needs no enemy at all, the simple and relentless imperative not to end.

There is a portable rule in this, worth carrying to any institution one wants to understand. Do not judge it by how it behaves when its purpose is urgent and clear, because in that condition every institution looks necessary. Judge it by what it does in the interval when its purpose has lapsed, because that is when its true nature shows: whether it can contemplate its own end, or whether it can only contemplate its own continuation. The ones that can imagine ending are rare and usually have a single master whose fate decides theirs. The ones that cannot are the norm, and they persist by the quiet arithmetic of distributed interest, finding new reasons across the lean years until a returning danger hands them an old one. Watch what an institution does when it is not needed, and you learn what it is. NATO, watched across the thirty years when it was not obviously needed, told you exactly what it was, and then 2022 told you it had been, all along, also something more.

The enemy gave the alliance its purpose. The alliance outlived the purpose and kept the enemy's shape, waiting, until an enemy came to fit it again.

Evidence Map

Facts, interpretations, forecasts, and disconfirming signals.

Core claim. NATO is best understood not only as a defense alliance responding to threats but through the imperative of any large institution to perpetuate itself. The decisive evidence is its behavior after 1991, when the adversary it was founded to oppose dissolved and the alliance, rather than disbanding, enlarged, sought new out-of-area missions, and raised its spending targets. Institutional self-preservation and genuine security demand coexist: the survival drive kept the institution intact through the decades without a clear enemy, and the return of a real threat in 2022 then vindicated its continuation.

Evidence level. Facts: high. Documented: NATO's founding in 1949 and the single invocation of Article 5 after 2001; the dissolution of the Warsaw Pact and the USSR in 1991; the post-1991 enlargement from 16 to 32 members (waves in 1999, 2004, and after, including Finland in 2023 and Sweden in 2024); the out-of-area operations in Bosnia, Kosovo (1999, without Security Council authorization), Afghanistan, and Libya (2011, under Resolution 1973); the 2 percent spending guideline formalized in 2014 and the higher target agreed at The Hague in 2025; Russia's invasion of Ukraine in 2022. Interpretation: medium, marked. Reading these facts through institutional self-preservation, and the synthesis that it coexists with genuine demand, is an analytical conclusion, not a claim of conspiracy.

What would confirm this. Continued institutional persistence and mission-seeking during any future period when no clear threat to members is present; further enlargement or budget growth disconnected from an active danger.

What would disprove this. The 2022 revival and the voluntary accession of long-neutral Finland and Sweden are the strongest disconfirming signals for the strong form of the claim: they show enlargement is substantially demand-driven and the threat real, not manufactured. The thesis advanced here is therefore the weak form, that institutions develop a survival interest coexisting with real function, and it would be falsified by evidence that NATO has in fact dissolved or contracted when its purpose lapsed, or that its post-1991 missions and enlargement were purely reactive with no institutional momentum of their own.

Watchlist. Whether the 5 percent spending trajectory holds or reverses; whether the alliance contracts or seeks new missions in any future interval of reduced threat; the long-run relationship between member demand and institutional momentum.

Related from The Manifest Archive