The Victors and the Vacuum

The first essay in this series mourned the inadequacy of moral clarity — the gap between what Chaplin's speech could achieve and what the world it addressed required. This essay moves from mourning to mechanism, and from the present to the decade that made the present possible. The vacuum that produced 2026 was not an accident of history or an inevitable consequence of forces beyond political control. It was built, decision by decision, by people who believed they were managing a transition rather than designing an order — and this is not an argument that those decisions caused Russian aggression, but that they shaped the cost environment in which Russian aggression evolved. The difference between managing a transition and designing an order is the subject of what follows.

I. The Operating Doctrine

In the months after the Soviet Union's formal dissolution in December 1991, the dominant intellectual atmosphere in Western capitals was one of measured but genuine confidence. Francis Fukuyama's 1989 essay "The End of History?" had not functioned as a policy document — most serious policymakers read it with scepticism — but it had created a climate.¹ The sense that liberal democratic capitalism had defeated its last systemic challenger shaped the questions Western governments were willing to ask and, more importantly, the ones they were not.

The operating doctrine was named more precisely in September 1993, when National Security Advisor Anthony Lake delivered a speech at Johns Hopkins University's School of Advanced International Studies.² The strategy he outlined — "democratic enlargement" — proposed expanding and strengthening the community of market democracies as the successor framework to Cold War containment. The logic was not unreasonable given 1993 conditions. The Soviet Union had collapsed. Democratic transitions were proceeding across Central and Eastern Europe. The direction of history appeared, if not settled, at least directional.

E.H. Carr, writing in 1939, had identified the structural weakness in this kind of thinking: liberal internationalism mistakes the interests of dominant powers for universal norms, and the preferences of winners for the settled values of the world.³ The post-1991 order produced genuine public goods — reduced nuclear risk, expanded trade, democratic transitions in Central Europe that were real and durable. The Carr critique does not dissolve these. It identifies where the framework strains: in the assumption that states benefiting most from the new order shared their preferences with states that had not designed it. The post-1991 order is where Carr's critique becomes useful — not as a total indictment, but as a warning about the distance between universal language and institutions designed by the winners of the previous order. The parallel is specific: as in the interwar order Carr diagnosed, the post-1991 framework produced declarations of principle without enforcement mechanisms for the states that most needed them.

II. The Structural Constraint and the Available Choice

Before the failure of the 1990s is examined, the strongest counterargument to examining it at all must be stated directly.

John Mearsheimer's The Tragedy of Great Power Politics (2001) and Kenneth Waltz's Theory of International Politics (1979) together make a claim that the following section implicitly disputes: that the post-1991 outcome was largely overdetermined by the distribution of power, and that attributing it to decision-making failures overstates the causal weight of contingent human choices.⁴ On the structural realist account, a unipolar moment produces specific pressures regardless of who manages it. The hegemon expands its sphere. Regional powers resist. The form and timing vary; the trajectory does not. Better decisions by Western policymakers in the 1990s would have altered the specifics while leaving the structural conflict intact.

The structural realist account is not wrong about the landscape. The distribution of power after 1991 did create specific pressures, and those pressures were real. The argument fails at the level of specificity. The 2022 invasion was not structurally inevitable in February 2021. The 2008 Bucharest Summit declaration was not structurally required in April 2008. The Nord Stream 2 approval — the Russian gas pipeline to Germany, built despite explicit warnings from Baltic, Polish, and American allies — was not structurally necessary in any year it was granted. M.E. Sarotte's archival work in Not One Inch (2021), drawing on thousands of previously classified documents from multiple national archives, shows specific moments at which specific actors made specific choices that were genuinely available in more than one form.⁵ The record is specific. The structural constraints shaped the landscape of available choices. They did not determine which choice was made. That distinction carries thirty years of consequences.

III. Alliance Without Architecture

NATO expansion into Central and Eastern Europe was not a Western strategic blunder dressed as liberation. For Poland and the Baltic states, membership was a rational response to a specific and historically grounded threat assessment — one the subsequent thirty years have largely confirmed. Poland had been partitioned out of existence twice in living historical memory. Estonia, Latvia, and Lithuania had been occupied, subjected to mass deportations, had their populations partially replaced by Russian settlers, and in January 1991 had watched Soviet OMON special forces and paratroopers drive armoured vehicles into their capitals — killing fourteen people at the Vilnius TV Tower and five more at the Interior Ministry building on Raina bulvaris in Riga⁶ — while the Baltic states' declared independence remained unrecognised by Moscow, and before the USSR had formally dissolved. Their application to join NATO was not a provocation of Russia. It was a lock on a door Russia had been trying to keep open since the Baltic independence movements began, and suggesting they should have remained in strategic ambiguity for the comfort of the power that had occupied them belongs to the vocabulary of colonial administration, not security analysis.

Into this picture enters George Kennan's 1997 warning, which has been treated as wisdom for a quarter century and deserves to be examined more carefully.⁷ He correctly predicted Russian hostility toward expansion. Predictive accuracy and diagnostic accuracy are not the same thing. Kennan's formative strategic framework was built — the Long Telegram, the X Article, the containment doctrine — inside a world where Soviet control of Eastern Europe was a geopolitical fact to be managed rather than a crime to be reversed. When he wrote that NATO expansion would be a "tragic mistake" that would inflame Russian nationalism, he was treating the sovereign choices of Baltic and Polish citizens as strategic provocations requiring Western accounting. That framing, whatever Kennan intended, implicitly validates Russian interest in the orientation of states it had occupied by force under the terms of a secret protocol to a pact with Nazi Germany. That pact was an act of aggression, illegitimate under the international law of its own time, and never recognised as legitimate by any democratic government. The occupation that followed it was equally illegitimate. A strategic framework that treats the victims' exit from that arrangement as a problem to be managed has confused the predator's anger for a legitimate security concern. The prey escaping is not an act of aggression against the predator.

Some will read Kennan's warning not as a claim about Russian legitimacy but about Western strategic prudence — arguing that it does not matter whether Russian anger is justified if it is predictable and consequential. That version of the argument fails on its own terms. In the Russian case, accommodated challenges repeatedly produced larger challenges rather than stable equilibria. Accommodation that makes further accommodation more necessary is not strategic prudence; it is a liability that compounds with each concession, dressed in realist language.

The political effect of Kennan's framing was significant regardless of his intentions. His argument — historically grounded, authored by the architect of postwar containment strategy, carrying the weight of genuine scholarly authority — gave Russian diplomatic and information operations something they could not generate themselves: a usable propaganda instrument with respectable Western packaging. Versions of that argument circulated through American and European policy debate, shaping the terms in which expansion was discussed across two decades.⁸ Russian officials quoted it as evidence that even Western analysts recognised Russia's legitimate grievances. It gave intellectual cover to a Western response at each critical juncture — Georgia 2008, Crimea 2014, the full invasion 2022 — that was slower and smaller than the threat warranted. Sarotte's archival finding — that the 1990 negotiations produced "common complicity in deliberate ambiguity" rather than a broken promise — fed the same machine.⁵ It gave the Russian narrative enough factual scaffolding to survive formal debunking. The propaganda instrument did not require the original claim to be true. It required only that the question remain open long enough to paralyse decision.

Estonian security institutions later documented in public annual reviews what Baltic policymakers had long treated as operational reality: Russia viewed the region through retained-sphere logic, not through ordinary defensive anxiety.⁹ Russia maintained military infrastructure in Estonia until 1994 and resisted withdrawal at every stage. It used Russian-speaking minorities as levers of influence rather than as populations requiring protection. Russia's pattern of behaviour belongs to a state managing a temporarily lost asset, not to one managing a security dilemma. The anger that followed Baltic NATO membership was not a security response to encirclement. It was the anger of a former occupying power watching a restoration window close.

The actual architectural failure of the 1990s was not that NATO expanded. It was that the alliance created no framework for the states it could not or would not admit. Ukraine, Georgia, Moldova, and Belarus were left without membership, but also without an enforceable alternative. The Partnership for Peace programme, established in 1994, had the institutional form that could have carried such an architecture. Allowing it to become a waiting room rather than a security structure in its own right was a choice, made by governments that could not agree on what to offer states they were not prepared to admit. A framework explicitly designed as temporary produces the political dynamics of a waiting room, not a security community. The states left in that waiting room did not stand still. They made calculations of their own, in an environment that offered them no structural anchor either way. The Baltic experience offers no consolation here. If membership was the only guarantee that held — and the evidence suggests it was — then the architectural failure of the 1990s was not the failure to build a better Partnership for Peace. It was the failure to extend membership to states that needed it, against alliance consensus that made extension politically impossible. That is a failure without a clean fix, and naming it honestly is more useful than implying a better intermediate framework would have been sufficient.

The failure completed itself on 3 April 2008, when NATO's Bucharest Summit communiqué declared that Ukraine and Georgia "will become members of NATO" — with no pathway, no timeline, and no genuine alliance consensus on what the declaration meant.¹⁰ It alarmed Russia without providing Ukraine any actionable guarantee. It gave Moscow another instrument for propaganda. It was the worst available option: a commitment too vague to deter and too provocative to ignore.

The shock therapy that hollowed out Russian and Ukrainian institutions, the Chechen wars that built the permission structure of impunity, and the specific vulnerability of Ukraine that those conditions created — these are the subject of the next essay. They do not change the architectural failure described here. They compound it.

Sources

1. Francis Fukuyama, "The End of History?" The National Interest, Summer 1989. The essay as intellectual climate rather than operating doctrine; most senior policymakers engaged it critically, but it shaped the questions the post-Cold War order was willing to ask. ↩
2. Anthony Lake, "From Containment to Enlargement," address at the School of Advanced International Studies, Johns Hopkins University, 21 September 1993. The actual operating doctrine of the Clinton administration's foreign policy, not the Fukuyama essay more frequently cited as the period's intellectual foundation. ↩
3. E.H. Carr, The Twenty Years' Crisis, 1919–1939 (Macmillan, 1939). Written on the eve of the Second World War as a critique of interwar liberal internationalism; its central argument maps onto the post-1991 period with considerable precision. ↩
4. John J. Mearsheimer, The Tragedy of Great Power Politics (W.W. Norton, 2001, updated 2014); Kenneth N. Waltz, Theory of International Politics (Addison-Wesley, 1979). The structural realist counterargument to the series' decision-centred framework, engaged directly in Section II and returned to in later essays. ↩
5. M.E. Sarotte, Not One Inch: America, Russia, and the Making of Post-Cold War Stalemate (Yale University Press, 2021). Sarotte's "common complicity in deliberate ambiguity" formulation is the most precise characterisation of the 1990 negotiating record currently available in the scholarly literature. ↩
6. January 1991 Baltic events: Anatol Lieven, The Baltic Revolution: Estonia, Latvia, Lithuania and the Path to Independence (Yale University Press, 1993), pp. 230–242. The Vilnius TV Tower attack of 13 January 1991 killed fourteen civilians; the Riga Interior Ministry attack of 20 January 1991 killed five. Both attacks were carried out by Soviet special forces before the formal dissolution of the USSR. ↩
7. George Kennan, "A Fateful Error," New York Times, 5 February 1997. Kennan correctly predicted Russian hostility toward expansion. This essay argues he was wrong about its cause — treating Russian anger as a legitimate security response rather than as a former occupying power's reaction to losing control of territories it had not strategically relinquished. ↩
8. The circulation of Kennan-aligned arguments in American policy debate is documented in James Goldgeier, Not Whether But When: The U.S. Decision to Enlarge NATO (Brookings Institution Press, 1999), which traces the internal debate between expansion advocates and sceptics through the Clinton administration. For the European policy context, the argument is treated here as a pattern visible in post-Ostpolitik German debate and wider European accommodation-oriented approaches to Russia, rather than as a single traceable doctrinal line. ↩
9. Estonian security institutions produced public annual reviews through two separate organisations with distinct mandates: the Kaitsepolitseiamet (Estonian Internal Security Service, kapo.ee), whose annual reviews document Russian active measures targeting Estonian domestic institutions and predate 2016; and the Välisluureamet (Estonian Foreign Intelligence Service, valisluureamet.ee), whose first public annual report was published in 2016 and covers Russian external behaviour and strategy. Claims in the body text about Russian retained-sphere logic draw on the pattern documented across both organisations' public outputs. Where specific assessments are cited elsewhere in this series, the relevant organisation and publication year are identified. ↩
10. NATO, Bucharest Summit Declaration, 3 April 2008, paragraph 23: "NATO welcomes Ukraine's and Georgia's Euro-Atlantic aspirations for membership in NATO. We agreed today that these countries will become members of NATO." Available at nato.int. ↩