The Soviet collapse is usually pictured as a defeat. A system that had competed with the West for four decades is imagined to have been beaten: outspent in an arms race, outproduced in its factories, outargued in the contest of ideas, and finally abandoned by peoples who rose against it. The picture is natural, and it is wrong in the one respect that matters most for explanation. The union was not overthrown by mass revolt, and it was not destroyed by military defeat or foreign invasion. It was dismantled through a process that began with reforms launched by its own leadership, pursued by its own ruling stratum, and completed when the men at the summit of the system concluded that the arrangements over which they presided were no longer worth defending. Armies stayed in their barracks. The crowds that gathered in the capitals gathered late, and mostly to ratify changes already set in motion above them. The state that had survived civil war, famine, and the largest land invasion in history ended in committee rooms, in republican parliaments, and in a hunting lodge where three republican leaders signed its dissolution over the head of its own president.

A reading desk with the explanations tested table, archival files and a map of the union republics, Insight Crunch

That single fact reorganises the whole explanatory problem, and this article is built around it. Explanations of the collapse are plentiful, and most of them answer the question of why the system was in difficulty. Far fewer answer the question of who ended it, and by what route. Yet the route is evidence. A system destroyed by revolt from below testifies to one kind of failure; a system dismantled by its own governors testifies to another. The Marxist explanations of the collapse, which this article sets against each other and against their non-Marxist rivals, differ precisely in what they say the system had been, and therefore in what they must say about the stratum that ran it. One analysis had argued for half a century that the bureaucracy was a caste balanced on property it did not own, and that it would either be swept away by political revolution or would convert its administrative power into property of its own. A second held that the bureaucracy was already a ruling class of a capitalist type, so that nothing fundamental in class relations changed when the forms changed. A third held that a new class society had been built which simply proved the weaker competitor. A fourth, shared by Marxists and non-Marxists alike, locates the cause in the exhaustion of an economic mechanism. Each of these accommodates the internal initiation with a different degree of comfort, and grading them by that comfort is more informative than surveying them one by one.

Two disciplines govern everything that follows. The first concerns figures. The collapse attracts numbers the way battlefields attract legends, and the most contested quantities in the field, the true rate of Soviet growth, the real burden of military spending, the price movements of the nineteen eighties, are exactly the quantities on which the surviving evidence is most disputed. This article therefore asserts no growth rates, no oil prices, no military spending shares, and no output figures at all. Where the argument needs the economy’s direction, it describes direction: a deceleration visible on every series anyone has constructed, official or reconstructed, however sharply the series disagree about levels. The second discipline concerns comparison. Where Soviet performance is set against Western performance, the comparison states the economies, the period, the measure, and the accounting concept, because the Soviet statistical system counted a different object from the Western one, and a comparison that ignores that difference is not a comparison but a pun.

The Soviet collapse as an analytical problem

Begin by separating three questions that popular discussion habitually fuses. Why was the system in difficulty by the decade’s middle years? Why did the attempt to reform it begin when it began, and from where it began? And why did reform end in dissolution rather than in a reformed union? The first question is economic and has been studied the longest. The second is political, and its answer is the one this article treats as decisive. The third is the question of sequence: of why political liberalisation ran ahead of economic restructuring, why the union’s federal structure became the instrument of its own dismantling, and why no constituency, inside the apparatus or outside it, organised to defend the arrangements when they were challenged. An explanation that answers the first question brilliantly and the other two not at all has explained a stagnation, not a collapse. Stagnant systems can persist for generations, as many did before and since; the Soviet system had itself persisted through difficulties in earlier decades that were, by any measure of output or welfare, far more severe than those of the nineteen eighties. Difficulty is a condition. Collapse is an event. The distance between them is where explanation must work.

The prior question, whether the system that ended was socialist at all, belongs to a companion article on whether the USSR was socialist, and this article deliberately does not settle it. The refusal is not evasion. The Marxist explanations disagree about the answer, and the disagreement is the substance: an account that calls the system a degenerated workers’ state, an account that calls it state capitalist, and an account that calls it a new class formation cannot all be right about what ended, and each derives a different collapse from its different premise. To settle the classification in advance would be to award the argument to one contestant before the evidence is heard. What can be done without prejudice is to fix the facts that any classification must accommodate, and chief among them the internal initiation: the ending was begun by the ruling stratum’s own reforms. Whatever the stratum was, caste, class, or something for which the tradition had no ready word, it behaved in a way that some of the tradition’s analyses had described in advance and others strain to describe at all.

What makes the manner of an ending evidence about the system that ended?

Because endings are produced by the relations they terminate. A regime destroyed by invasion testifies about armies; a regime destroyed by revolt testifies about the governed and their grievances; a regime dismantled by its own officeholders testifies about the officeholders, about what they held, what they expected to keep, and what they calculated they could convert.

Hold that answer against the standard images and their inadequacy becomes plain. The image of a population rising explains the Eastern European endings of the same period better than the Soviet one, and even there the crowds followed openings created by decisions in Moscow. The image of economic collapse forcing political surrender inverts the actual order: the Soviet economy of the late nineteen eighties was stagnating, contracting in places, and generating shortages that angered consumers, but it was not in the condition of collapse that followed the political dissolution rather than preceding it. The deep disintegration of supply, output, and public finance belongs to the years after the union ended, when the successor arrangements were being improvised. Whoever wishes to explain the ending by economic breakdown must explain why the breakdown trailed the political event it is summoned to cause. The image of a system defeated in ideological combat, finally persuaded of its rival’s superiority, mistakes the vocabulary of the victors for a mechanism. Ideas mattered, as this article will argue when it reaches legitimacy, but ideas acted through the decisions of identifiable men in identifiable offices, and the explanation must pass through those decisions rather than around them.

There is a further reason the manner matters, and it is methodological. The collapse is one of the rare events on which rival theories of a whole social order can be tested against a single outcome, because the theories were stated in advance, in print, over decades, with their expectations attached. The tradition this article examines had been arguing about the Soviet system since the nineteen twenties, and its schools had committed themselves to incompatible descriptions of what the bureaucracy was and what it could do. When the end came, it came in a specific shape, and the shape fits some descriptions and strains others. That is the closest thing to an experiment that the study of social systems provides, and the article exploits it deliberately. The explanations tested table, presented after the sequence has been established, is the instrument: one row per explanation, its mechanism, its expectation for the ending’s initiator, its expectation for the class result, and how the record bears on each. Readers who keep the table in view will find the long discussions that follow easier to grade, because every school is asked the same three questions in the same order.

The sequence: reform from above, dissolution from within

The sequence matters more than any single episode in it, so it is set out here as a chain before any school is allowed to interpret it. Through the nineteen seventies and into the nineteen eighties the leadership itself acknowledged, in its own guarded vocabulary, that growth was faltering and that the economy’s methods were exhausting themselves. Successive general secretaries promised intensification, discipline, and technological renewal, and the promises produced campaigns rather than transformation. When a new leadership took office in the middle of the nineteen eighties it inherited that acknowledged problem and chose, from within the summit of the party, a programme of restructuring and openness: perestroika for the economy, glasnost for public life. The choice was made by the ruling stratum’s own leading faction, against no mass pressure that the record can find, and initially with the intention of strengthening the system it governed. Whatever else the collapse was, it began as an act of government.

The second link in the chain is the asymmetry between the two halves of the programme. Political liberalisation moved faster and further than economic restructuring, and the gap between them is where the ending was decided. Openness released criticism of the system’s history, its privileges, and its founding myths into public circulation, while the economy continued to be run by administrative methods that the criticism was busy discrediting. Competitive elections to a new legislature brought men into public life whose authority did not derive from the apparatus. The party’s constitutional monopoly on power was surrendered. Republican institutions, which the union’s structure had maintained as formal shells around a unitary reality, acquired elected occupants with mandates of their own, and the republics began to assert sovereignty over their laws, their resources, and their taxes. Economic restructuring, by contrast, produced half-measures that weakened the old coordinating mechanisms without building new ones: enterprises were granted autonomy while prices, supplies, and credit remained administered, a combination that loosened discipline before it created incentives. The system was thus being delegitimised politically at speed while being disorganised economically at leisure, and the two processes fed each other.

The third link is the failure of defence. When conservative figures within the leadership attempted, in the late summer of the union’s final year, to reverse the process by emergency rule, their attempt collapsed within days for want of the very things the sequence had dissolved: reliable obedience down the chain of command, a party able to mobilise, and any organised social force willing to fight for the arrangements as they stood. The attempted reversal then accelerated what it meant to stop, transferring authority to the republican leaders who had resisted it. Within months the presidents of three republics declared the union dissolved and a commonwealth in its place, the union president resigned his office on the ground that there was no longer a state for him to preside over, and the union’s flag came down. A system that had commanded the obedience of hundreds of millions ended without a battle, a siege, or a general strike in its defence. Its own stratum had either converted its hopes to the successor arrangements, retreated into its republics, or simply gone home.

Why does the order of political and economic change matter to the explanation?

Because reversing the order reverses the causal story. Had economic breakdown preceded political opening, the collapse could be read as a system overwhelmed by its failures. The record shows opening first and breakdown after, which means the explanation must run through choices made at the summit, and through what the summit’s members believed about their own position.

Notice what the sequence excludes. It excludes the explanation by immiseration, since the population’s material condition, though deteriorating in the final years, was not the trigger and produced no insurrection. It excludes the explanation by military overstretch acting directly, since no army was defeated and no front collapsed; whatever the military burden contributed, it contributed as a weight on the economy and on the leadership’s calculations, not as a catastrophe. It excludes, above all, the explanation by external imposition: foreign governments welcomed the ending, and some had worked for it, but no foreign power dissolved the union, and the decisive signatures were internal. What the sequence includes, at every link, is the ruling stratum itself: choosing reform, choosing the order of reform, failing to defend the unreformed remainder, and finally choosing dissolution over a union it no longer believed it could hold. Any explanation of the collapse is therefore, whether it admits it or not, a theory of that stratum’s nature and interests. The Marxist schools understood this earliest because they had been theorising the stratum for fifty years, and it is to their theories, graded against the sequence, that the article now turns, by way of the claim that organises the grading.

The internal initiation: the claim and its evidence

The claim can be stated in one sentence, and the rest of this article is its elaboration and its testing. The internal initiation: the collapse was begun by reforms from within the ruling stratum rather than by revolt from below or defeat from outside, and this fact is the single most discriminating piece of evidence among the competing explanations of what the system had been. A discriminating fact is not the most important fact, nor the most dramatic. It is the fact on which rival accounts differ in their ability to accommodate it without strain. Many facts about the collapse are common property: the stagnation, the shortages, the corruption, the cynicism of the final decades. Every explanation absorbs those easily, because every explanation was built partly from them. The internal initiation is different. It is a fact about agency, about whose hands were on the lever when the machine stopped, and the schools had committed themselves, long before the event, to incompatible accounts of those hands.

Consider what each account must say. If the bureaucracy was a parasitic caste resting on property forms it administered but did not own, then its members had no secure possession to defend, only positions, and positions can be traded. Such a stratum, facing a system in decay, might rationally prefer to convert its administrative power into ownership in a successor order, and an analysis built on that premise can describe the stratum initiating its own system’s end as the caste’s final act of self-preservation by other means. If, by contrast, the bureaucracy was already a ruling class in full possession of the means of production through the state, then the ending requires that class to have dismantled the very arrangements that constituted its rule, exchanging secure collective possession for an uncertain scramble, and the account owes an explanation of why a class in possession would do such a thing. If the bureaucracy was a new class of a novel type, the ending must be explained by that class losing a competition it could not win and its members deserting a formation that had stopped delivering their collective future. And if the cause was systemic and economic, the stratum’s initiative is the transmission belt rather than the engine: the leadership reformed because the growth model had exhausted itself, and dissolved the union because reform destroyed the mechanisms that had held it together. None of these is impossible to state. They differ in how much auxiliary argument each needs, and auxiliary argument is the currency in which explanatory strain is paid.

Who initiated the Soviet collapse?

the leadership itself. The process began with reforms launched from within the ruling stratum, not with mass revolt or military defeat, and political liberalisation preceded effective economic restructuring. That fact is the most discriminating evidence between competing explanations of what the system had been.

The evidence for the claim is the sequence itself, but two features of it deserve emphasis because they are the features rival readings most often blur. The first is the direction of the initiative at each turning point. The reform programme was designed inside the leadership, staffed by its own appointees, and legislated through its own institutions. The elections that displaced the apparatus were conceded by the apparatus. The surrender of the party’s monopoly was voted by the party’s own parliamentarians. The republican sovereignty declarations were issued by republican leaderships that had themselves been formed inside the system, in several cases by men who had been senior figures of the union apparatus before re-founding themselves as national leaders. Even the dissolution instrument was signed by republican heads who were products of the order they dissolved. At no point in the chain did an organised force external to the stratum seize an initiative the stratum was defending. The second feature is the absence at the end. When the emergency committee attempted its reversal, the stratum did not rally, the army did not move as one, the security organs split, and the industrial ministries calculated their futures in the successor order rather than in the union’s defence. A ruling group that believes its arrangements embody its interests does not behave this way. A stratum whose members individually expect to do better under new arrangements, or fear to do worse by resisting them, behaves exactly this way. The internal initiation is thus not merely a fact about who signed what. It is a revelation about the stratum’s relation to the system it ran, disclosed at the moment when concealment stopped paying.

What the internal initiation discriminates

A discriminating fact earns its keep by the work it does, so it is worth pausing over the work before the schools are examined in detail. The internal initiation does not, by itself, refute any explanation. A determined defender of any school can absorb it, and the better writers of each school did absorb it, usually within a few years of the event. What the fact does is impose costs, unevenly distributed. For the caste analysis, the cost is lowest, because that analysis had named the stratum’s conversion into owners as one of the two futures open to it, and the stratum’s behaviour in initiating and then profiting from the transition reads like the named future arriving by an unnamed route. For bureaucratic collectivism the cost is middling: a new class that loses a systemic competition and deserts its own formation is describable, but the analysis must explain why the class’s collective organisation, the party, proved so much weaker than its members’ individual adaptability. For state capitalism the cost is highest, because the account must explain a ruling class presiding over the liquidation of the institutional form through which it ruled, and must locate the compulsion to do so in pressures that the class had survived, in harsher forms, for decades. For the systemic-economic account the initiation is neither predicted nor an embarrassment but a mechanism: the account explains why reform was attempted and why it failed, and treats the identity of the reformers as a consequence of where power sat in a centralised system. Only the centre could initiate change in a system that had suppressed every other centre, so the initiation’s internal character is, for this account, almost a tautology. Grading by strain therefore yields an ordering, not a verdict, and the ordering is one of this article’s central results.

Which explanations does the internal initiation embarrass most?

Those that describe the bureaucracy as a settled ruling class in secure possession of the means of production. A class so described should have defended the arrangements that embodied its possession, or been dispossessed by force. Instead the stratum traded its collective position for individual prospects, which possession language describes awkwardly.

There is a second discrimination the fact performs, quieter but as valuable, between explanations of the collapse and explanations of the stagnation. The literature is full of accounts that demonstrate, convincingly, why the Soviet economy slowed, why its technology lagged, why its consumers queued. Those accounts explain a condition that obtained for at least two decades before the ending, during which the system did not collapse but endured, geriatric and stable. The internal initiation forces the further question: what changed, in the stratum’s own assessment of its future, that converted an endured condition into a terminal one? Explanations that cannot pose that question are incomplete however accurate their economics. The systemic account can pose it too: a growth model visibly exhausting itself alters the expectations of those who run it. The degenerated workers’ state account can pose it, because a caste’s privileges depend on the health of the property forms beneath it. The question is hardest for purely external accounts, for the arms race as master key and the oil price as assassin, since external pressures had been severe before without producing dissolution, and something internal must mediate between pressure and surrender. That mediation is the stratum’s calculation, and the calculation is what the internal initiation names.

The explanations tested

The table below is the article’s findable artifact, and it is designed to be used rather than admired. Each row states an explanation’s mechanism in a phrase, then what the explanation leads one to expect about who initiates an ending and about the class outcome, and finally how the record bears on those expectations. The rows are ordered not by plausibility but by the age of the analysis, since the predictive credit an explanation can claim depends on when its commitments were made. Discussions in the sections that follow expand each row; the table is the map, and readers who disagree with the article’s grading can locate their disagreement in a specific cell.

Explanation Mechanism What it predicts about who initiates What it predicts about the class outcome How the record bears on it
The degenerated workers’ state A bureaucratic caste administers nationalised property it does not own, and must either fall to political revolution or convert its power into ownership The caste itself, once conversion looks safer than defence, since its members hold positions rather than property Restoration of private property, with former officials prominent among the new owners Closest fit on initiation and outcome; wrong on the route, since conversion came through reform from above rather than through the crisis the analysis sketched
State capitalism The bureaucracy is a collective capitalist class driven by accumulation and military competition with rival capitals No fundamental ending is expected at all, only changes of form imposed by competitive pressure Continuity of class relations beneath a change in the form of capital Strong on continuity of personnel and privilege; strained by the initiation, since the class dismantled the arrangements that constituted it
Bureaucratic collectivism A new ruling class governs a novel social formation, neither capitalist nor socialist, and competes with capitalism as a system The new class defends its formation until defeat in systemic competition makes desertion rational Dissolution of the new class as a collective, its members scattering into successor arrangements Fits the desertion and the absence of defence; weakest on mechanism, since the competition it cites had been running for decades without dissolution
The systemic-economic account The planning mechanism exhausts extensive growth and cannot make the transition to intensive growth, so reform is forced and then fails The leadership, since only the centre can initiate change in a centralised system, acting under economic compulsion Open, since the account is about the mechanism rather than the class character of its operators Strongest on why reform was attempted and why it failed; silent by design on what the system was, which the other rows treat as the question

What does each row of the table commit its explanation to?

To a claim about the stratum that can be checked against the stratum’s behaviour. Every row must say what the men who ran the system held, what they stood to lose, and what they could carry across the ending. Rows that cannot fill those cells are describing an economy, not explaining a collapse.

Two cautions belong with the table. The first is that the rows are ideal types. The actual literature mixes them: writers combined the caste analysis with the systemic economics, or held state capitalism with a theory of competitive pressure borrowed from the collectivists, and the hybrid positions often read better than the pure ones. The table separates what the literature combines, because grading requires separable commitments. The second caution is that the fourth row is not a Marxist possession. The systemic-economic account was developed by scholars inside the Marxist tradition and by scholars far outside it, and its presence in the table is the article’s first acknowledgement of a convergence examined at length later: on the mechanism of the ending, as distinct from the classification of the system, the Marxist and mainstream literatures substantially agree, and the agreement is itself a finding. What divides the schools is not chiefly why the system failed but what the system was, and the collapse discriminates between those classifications only through facts like the internal initiation, which is why this article gives that fact its prominence.

A word on how the verdicts in the final column were reached, since a table that hides its method invites the suspicion it exists to dispel. Each cell was filled from the school’s own texts before it was filled from the record: the expectation attributed to a row is one its founders or its leading custodians stated, or one that follows from their stated premises without charitable reconstruction, and the record’s bearing is stated in the vocabulary of fit, strain, and silence rather than of proof, because a single case, however rich, grades theories the way a single examination grades a candidate, decisively about performance on the day and provisionally about capacity in general. Readers who fill a cell differently are doing the exercise the table exists to provoke, and the article’s grading stands or falls by whether its cells survive that exercise, not by whether its conclusions are adopted.

The degenerated workers’ state: the caste that could convert itself

The oldest of the analyses, and the one with the longest paper trail, begins from a refusal. Writing in the nineteen thirties, as the purges consumed the generation that had made the revolution, Leon Trotsky refused both of the descriptions on offer: that the Soviet Union remained a workers’ state in health, and that it had become simply a new tyranny of a familiar kind. In The Revolution Betrayed he argued that the revolution’s economic conquests, nationalised property and planning, survived, but that political power had been usurped by a bureaucratic stratum which he called a caste, deliberately choosing a word that denied it the dignity and the solidity of a class. A caste, in this usage, is a privileged layer that reproduces itself through its control of administration and distribution, that draws its income from its position within a system of property it did not create and does not own, and whose privileges are therefore insecure in a specific way: they depend entirely on the survival of the arrangements that allocate them. The bureaucrat owned no factory he could bequeath. He enjoyed the use of goods, residences, and deference that his office carried, and lost them with his office. His wealth was his place in a queue he did not control.

From that premise the analysis derived its fork, and the fork is the reason this row of the table exists. Trotsky argued that the bureaucracy’s position was untenable in the long run and would resolve in one of two directions. Either the working class would remove the caste by political revolution, restoring soviet democracy while preserving nationalised property, or the bureaucracy, seeking to make its privileges secure and hereditary, would convert itself into a new possessing class by restoring private property, a process he described as the caste’s necessary search for supports in property relations of its own. The bureaucracy, he wrote in substance, would not willingly surrender its position without a fight if the challenge came from below; but the same stratum, or its decisive sections, could sponsor the transformation from above if that route promised to convert administrative privilege into owned wealth. The analysis thus contained, from the nineteen thirties, a named outcome in which the system’s end was initiated by its own governors for reasons internal to their situation as a caste. Ernest Mandel and the tradition that followed Trotsky maintained this framework through the postwar decades, applying it to the Eastern European states and insisting, against the state capitalists, that a social formation be classified by its property relations rather than by the political character of its rulers or the hardships of its workers.

What did the degenerated workers’ state analysis claim the bureaucracy was?

A caste, not a class: a privileged stratum administering nationalised property it did not own, drawing income and status from office rather than from possession, and therefore insecure in a way a possessing class is not. Its privileges could not be inherited as property, only transmitted through continued control of the allocating apparatus.

The analysis had genuine strengths beyond the fork. It explained the bureaucracy’s conservatism, its terror of independent working class organisation, and its simultaneous fear of capitalist restoration, as the behaviour of a layer balanced between two property orders and committed to neither. It explained the stratum’s ideological elasticity, its ability to wrap itself in whatever doctrine the moment required, as the reflex of a group whose real constitution was positional rather than principled. And it made a prediction about the Eastern European regimes that events partially bore out: that their bureaucracies, installed from outside, rested on the same unstable foundation. Where the analysis was weakest, its own adherents came to admit, was in its account of the working class that was supposed to supply the fork’s first prong. The political revolution never came, in the Soviet Union or anywhere else in the bloc, as an independent movement of workers against the caste; the crowds of the final years demanded many things, but the restoration of soviet power in Trotsky’s sense was not among them. The fork’s second prong arrived instead, by a route the analysis had named as an outcome but not described as a process: not a counter-revolutionary seizure but a reforming leadership, a negotiated dissolution, and then a scramble in which former officials, enterprise directors, and party functionaries were conspicuously placed to become owners. The caste converted itself. It simply did so holding the pen of reform rather than the sword of reaction, and the distinction matters for grading the prediction, a task reserved for the sections on the predictive record below.

Trotsky’s fork: political revolution or restoration

The fork deserves a section of its own because it is the one place in the entire literature where the collapse’s shape was anticipated in structural terms, and precision about what was and was not specified is the difference between honest grading and ancestor worship. What was specified: the agent of a possible ending from within, namely the bureaucracy itself; the motive, namely the conversion of insecure administrative privilege into secure property; the direction, namely toward private property and market relations; and the alternative that would have prevented it, namely a political revolution from below that the analysis expected and that never occurred. What was not specified: the timing, across a margin of half a century; the mechanism, since the conversion was imagined as a political act of a frightened caste rather than as the unintended destination of a modernising reform; the survival of the union as the unit in question, since the analysis treated the Soviet state as a whole and did not foresee its dissolution along republican lines; and the passivity of the working class in whose name the entire framework was constructed. A prediction that names who will act, why they will act, and what they will build, while missing the date, the route, and the map, is a substantial intellectual achievement and a failed prophecy at the same time, and the article will return to that double character when the strongest objection is heard.

What were the two futures the fork allowed, and which arrived?

Political revolution from below, removing the caste while keeping nationalised property, or restoration from above, in which the caste converted its administrative power into private ownership. The second arrived, by the route of reform and dissolution rather than counter-revolutionary seizure, and the first was never attempted by an independent movement.

It is also worth recording what the fork’s survival cost its custodians. Through the decades when the Soviet system appeared permanent, the analysis required its holders to describe as transitional a formation that outlasted most of its critics, and the word transitional stretched across three generations became an embarrassment that hostile critics exploited and friendly ones felt. The tradition’s journals filled with refinements: the caste was hardening, the caste was consolidating, the transition was entering its final phase, in essays whose dates span forty years. A theory that predicts an outcome without a horizon has insured itself against falsification at the price of predictive content, and that price is part of the honest assessment this article owes. The fork named the destination. It did not, and in its custodians’ hands could not, name the decade, and a reader deciding how much credit the anticipation deserves should hold both halves of that sentence together, because the strongest objection, stated later in this article, is built precisely from the second half.

State capitalism: a change of form, not of mode

The second analysis rejects the first at its root. Where Trotsky saw nationalised property as the surviving conquest that made the bureaucracy a mere caste, the state capitalist school saw the wage relation and the compulsion to accumulate as the surviving realities that made the bureaucracy a class. Tony Cliff, whose State Capitalism in Russia, written at the end of the nineteen forties, became the school’s founding text, argued that the Soviet worker stood in the same relation to the means of production as the worker in the West: separated from them, selling labour power, subject to a management he did not control, and producing a surplus disposed of by others. The bureaucracy, collectively, performed the function of capital, and the drive that governed its decisions was the drive that governs capital everywhere, accumulation for accumulation’s sake, imposed in the Soviet case less by market rivalry between firms than by military rivalry between states. The competition with Western capitalism, above all its military form, played the role that competition between capitals plays in Marx’s analysis of the market economies: the external whip that forced each unit to accumulate or fall behind. Raya Dunayevskaya reached a version of the same conclusion from the study of the Soviet economy’s own published figures and debates, and the school thereafter treated the Soviet Union not as a new world to be classified but as a variant of the existing one to be explained by the existing theory.

If the system was already capitalist, what exactly ended?

A form of capital, not a mode of production. On this account the ending replaced bureaucratic state capitalism, in which one collective owner accumulated under military compulsion, with market capitalism, in which many owners accumulate under commercial compulsion. The wage relation, the school insists, survived the transition uninterrupted, which is the point of the analysis.

The school’s explanation of the collapse follows from its premise with real elegance. If the system was capitalist, then its crisis was a capitalist crisis of a recognisable kind: an accumulation regime that had exhausted its extensive phase, in which growth came from drawing ever more labour and resources into production, and that faced the transition to intensive accumulation under conditions of intensifying international competition. The turn to the market then reads as the bureaucracy’s collective decision to change the form of its rule in order to preserve its substance, integrating itself into the world economy on terms that its old autarkic form could no longer sustain. The school points, with justification, to the continuity of the transition: the same directors managing the same enterprises, the same networks allocating the same advantages, the personnel of the old order reappearing as the proprietors of the new. No other explanation makes that continuity so central or explains it so directly. Where the degenerated workers’ state analysis sees a caste converting itself, the state capitalist sees a class repositioning itself, and the difference between conversion and repositioning is precisely the classification dispute in miniature.

The school also handles the economic evidence with fewer auxiliaries than its rivals. The deceleration is, for it, the familiar tendency of an accumulation regime to run into its limits; the innovation failure is the familiar behaviour of monopolised capital sheltered from commercial competition and driven instead by a single customer’s military demands; the consumer shortages are the familiar subordination of consumption to accumulation that the system’s own priorities announced. A reader who finds the Soviet economy’s pathologies uncannily similar to the pathologies Marx described in market economies will find this school waiting with an explanation of the resemblance. Its difficulties lie elsewhere, and they are serious enough to deserve their own section, because the brief of this article is grading, and the state capitalist row of the table carries the heaviest strain at exactly the point this article has made central: the initiation.

The state capitalist difficulty: a class dismantling itself

State the difficulty plainly. A ruling class, in the Marxist tradition’s own theory, is defined by holding the means of production and by the arrangements that reproduce that hold. Classes in possession do not, in that theory, voluntarily liquidate the institutional form of their possession. They are dispossessed by other classes, by conquering states, by the slow erosion of their economic function, but the active dismantling of their own arrangements by their own leading members, in the absence of defeat or revolt, is an event the theory of class interest does not generate. Yet on the state capitalist account, that is what the record shows: the collective capitalist of the Soviet Union, through its own reforming leadership, surrendered its monopoly of political power, dissolved the planning apparatus through which it had allocated the surplus, broke up the union that formed its economic space, and exposed itself to a transition in which many of its members were ruined even as others prospered. The school must explain why a class would do this, and its explanations, though not empty, carry a weight of auxiliary assumption that the rival rows do not.

Why is a ruling class dismantling its own arrangements a problem for class analysis?

Because class analysis explains action by interest in the reproduction of a position. A stratum that holds the means of production through a state has its interest embodied in that state’s arrangements, and should defend them as classes have defended theirs throughout the record. Voluntary self-liquidation requires a motive the possession itself cannot supply.

The school’s defenders offer three replies, and each deserves its strength. The first is that the class did not act as one: a reforming faction initiated changes whose destination it did not intend, and the class as a whole was dragged through a door opened by its own modernisers, which is true of the sequence but relocates the puzzle rather than solving it, since the faction was the class’s own leading section and its intentions were formed inside the class’s own institutions. The second reply is that the old form had become a fetter on accumulation itself, so that the class’s deepest interest, accumulation, overrode its interest in the form, which is the school’s strongest move and amounts to saying that capital’s loyalty is to accumulation rather than to any of its vessels. The third is that individual members calculated individually, each preferring a chance at personal ownership to a share in a decaying collective possession, which explains the desertion elegantly but concedes the premise of the rival analysis: a possession that its holders experience as decaying, collective, and individually inalienable is functionally close to what Trotsky meant by the position of a caste. The replies, taken together, save the account at the price of narrowing its distance from the account it was founded to refute. That narrowing is a genuine finding of this article, and it generalises: the collapse pushed the rival classifications toward each other, because the event itself had the mixed character, conversion and continuity together, that each pure classification captures only in part.

A variant within the broader anti-Trotskyist camp should be noted here, since it escapes some of the difficulty by refusing both classifications. Hillel Ticktin and the journal Critique argued across the same decades that the Soviet system was neither capitalist nor socialist nor a stable new mode, but a contradictory formation in permanent decline, characterised by waste as a systemic property and by a ruling group that was not a class because it could not reproduce the system it ruled. On that view the ending needs no conversion and no self-liquidation: a formation that was never viable in the long run finally stopped. The view’s strength is its fit with the absence of defence, since a stratum that never constituted a class had no class position to defend. Its weakness is the obverse of the fork’s weakness: a formation described as declining for half a century, without a specified mechanism of termination, predicts an ending the way a stopped clock predicts the hour.

Bureaucratic collectivism: the new class that lost the race

The third analysis took the bureaucracy’s novelty as its starting point and refused to assimilate the system to any known category. Bruno Rizzi, writing on the eve of the Second World War, argued that a bureaucratic collectivism was emerging in the Soviet Union and, in different dress, elsewhere; James Burnham’s The Managerial Revolution gave the thought its widest audience by proposing that managers, in all the advanced societies, were becoming the ruling stratum of a new order; Max Shachtman carried the analysis into the Marxist movement, breaking with Trotsky over the classification of the Soviet Union and concluding that a new class society, bureaucratic collectivism, had arisen which was neither capitalist nor socialist and which represented, in his bleak formulation, a historically regressive alternative to both. Milovan Djilas, writing from inside a Yugoslav leadership he had helped to build and then denouncing it in The New Class, supplied the analysis with its most authoritative witness: a stratum that monopolised administration monopolised, in effect, ownership, and constituted a class by its collective control of the nationalised economy and its distribution of privilege through political loyalty.

What did the bureaucratic collectivists claim had been built?

A new class society: neither capitalist, since private owners had been expropriated, nor socialist, since the producers controlled nothing, but a formation ruled by a bureaucratic class whose collective possession ran through the party and the plan. The claim made the Soviet Union a specimen of a new historical type rather than a deformation of an old one.

For the collapse, this analysis supplies a distinctive mechanism: systemic competition lost. A new class society, the argument runs, proved less dynamic than its capitalist rival, less capable of innovation, less able to satisfy consumers, less able to renew its own legitimacy, and over four decades the gap became undeniable even to its rulers. The ending then reads as the new class’s recognition of defeat and its members’ desertion of a formation that could no longer guarantee their collective future. The analysis fits several features of the record well. It fits the absence of any defence of the system as a system, since a class whose formation has failed has nothing left to defend but its members’ individual prospects. It fits the speed of the ideological evaporation, since a doctrine that had served as a class’s charter rather than as a science could be discarded without intellectual remainder. And it fits the behaviour of the republican elites, who abandoned the union formation for national formations in which their class position could be re-founded on a smaller stage. Where the analysis strains is on timing and mediation. That competition had been under way since the nineteen forties, through decades in which the Soviet formation gained ground, matched its rival weapon for weapon, and terrified its strategists; a competition lost so slowly explains a decline, and the analysis must still find, inside the class’s own deliberations, the moment and the men who converted decline into dissolution. It finds them, as every account must, in the reforming leadership, and at that point its mechanism converges with the systemic-economic account it would prefer to keep at a distance: the class acted because the economy it administered had ceased to grow in the manner its rule required.

The collectivist analysis also carries a burden its rivals do not: it predicted, in its founding texts, that the new formation was the wave of the future, the form toward which all advanced societies were travelling. Burnham expected managerial regimes to succeed capitalism across the world; Rizzi expected bureaucratisation to generalise. The collapse therefore refutes this school not on a detail but on its founding wager, and the school’s survivors have had to argue that a formation they described as history’s destination was, in the event, history’s dead end. That reversal is survivable, since the rival schools carry reversals of their own, but it disciplines any claim the collectivists might make to have understood the system’s trajectory. They named its class character with a clarity the others lacked, and misjudged its future completely.

The systemic-economic account: the growth model that ran out

The fourth row of the table differs from the other three in kind. It does not begin from a classification of the bureaucracy and derive a collapse; it begins from the operating mechanism of the economy and derives a necessity for reform, leaving the classification to others. Its central claim is that the Soviet growth model was extensive: it grew by mobilising inputs, drawing peasants into factories, women into the workforce, and investment into heavy industry, at rates that a command system is uniquely able to enforce. Such a model has a built-in terminus. Labour reserves run out, the capital stock’s maintenance claims grow, each increment of investment yields less than the last, and continued growth requires the opposite method, intensive growth, in which output rises because each worker and each machine produces more, which in turn requires innovation to be generated and diffused through the economy continuously. The planning mechanism, superb at mobilisation, proved structurally poor at diffusion: it could order a technology into existence at a priority site and could not make ten thousand enterprises adopt, adapt, and improve it. By the time the extensive reserves were spent, the economy faced a transition its institutions were not built to perform, and the reform era was the attempt, from the only place it could be attempted, to perform it anyway.

Which mechanism does the systemic account treat as decisive?

The exhaustion of extensive growth and the planning system’s inability to replace it with intensive growth. Mobilisation had delivered decades of expansion by adding labour and capital; when the additions ran out, growth depended on productivity, and productivity depended on innovation diffusing through enterprises whose incentives punished the risks innovation requires.

This account was developed, in overlapping versions, by scholars who agreed about little else. Alec Nove, the Glasgow economist whose studies of the Soviet economy set the standard for Western scholarship, argued across a long career that the system’s centralised methods suited an earlier stage of development and misfitted an advanced economy of interdependent products, where the centre could no longer know enough to command well. János Kornai, the Hungarian economist whose Economics of Shortage anatomised the system from inside the bloc, supplied its micro-foundations: the soft budget constraint, under which enterprises that failed were rescued rather than closed, dissolved the discipline that forces adaptation, and chronic shortage, the system’s normal state, taught every manager to hoard inputs and labour rather than to economise on them. Marxist writers reached parallel conclusions from their own premises: the tradition’s economic journals carried, through the nineteen seventies and eighties, analyses of the Soviet slowdown that attributed it to the same exhaustion of the extensive model and the same blockage of intensive methods, differing from Nove and Kornai chiefly in vocabulary and in what they hoped reform might preserve. The systemic account is thus the one explanation of the collapse that commands assent across the classification dispute, and its presence inside both camps is the seed of the convergence examined later in this article.

Its explanatory reach is real. It explains why reform was attempted when it was: the deceleration was visible to the leadership in its own statistics before it was visible to anyone else, and the succession of intensification campaigns under successive leaders shows the problem being recognised at the summit for a generation before the final attempt. It explains why reform took the form it did, loosening administrative control in search of initiative, and why the loosening disorganised before it invigorated, since a command economy’s coordination runs through the very instructions that reform suspended. It explains the timing of the ending better than any rival: the collapse came when the extensive model’s exhaustion coincided with a leadership willing to gamble on transformation, and the gamble destroyed the mechanisms of the old order faster than it built those of the new. What it does not explain, by its own honest admission, is everything political that made the ending a dissolution rather than a reformation: why the union broke along republican lines, why the party’s authority evaporated rather than being renegotiated, why no reformed socialism of any description emerged from a process begun by men who described themselves as socialists. For those facts the systemic account needs the political and national analyses it brackets, and its best practitioners have always said so. An account that knows its own boundary is not thereby refuted; it is thereby usable, and the article’s grading treats the systemic explanation as the strongest account of the collapse’s economic premise and an incomplete account of the collapse itself.

Extensive and intensive growth: the distinction underneath

Because the distinction does so much work in the systemic account and in the tradition’s economics generally, it deserves a plain statement before the evidence is weighed. Extensive growth increases output by increasing inputs: more workers, more machines, more land under cultivation, more tonnes of fuel extracted. It is the growth of a frontier economy, and it can be very rapid while reserves last, because it requires no improvement in how inputs are used, only their mobilisation. Intensive growth increases output by increasing the product of each input: the same workforce, better organised, better equipped, and more skilled, producing more and better goods. Intensive growth is the mature economy’s kind, and it depends on a continuous flow of innovation, not merely inventions but their diffusion, the thousand adaptations by which a new method becomes a normal practice across an industry. The two kinds of growth make opposite demands on institutions. Mobilisation rewards central command, which can concentrate resources on priorities and compel the movement of labour. Diffusion rewards decentralised initiative, tolerance of failure, and the freedom of producers to copy, modify, and compete, since no centre can order an economy to be inventive in ten thousand places at once.

Why did mobilising more inputs stop producing growth?

Because the inputs ran out and the returns diminished together. The countryside’s surplus labour was absorbed, the workforce’s growth slowed as the population aged, new resource deposits lay in harsher and more distant regions, and each successive increment of investment had first to maintain an ageing capital stock before it could extend it.

The Soviet record, described directionally as this article’s protocol requires, is the biography of that distinction. The early decades achieved extensive growth on a heroic scale, industrialising a peasant country by mobilisation, at costs this article does not minimise but also does not need to enumerate for its argument. The postwar decades continued the method against diminishing returns, and by the nineteen seventies the leadership’s own economists were recording the exhaustion in their internal literature: the working-age population’s growth was slowing, the easy deposits were worked out and the new ones lay in Siberia’s cold and distance, the capital stock accumulated by decades of investment now demanded a rising share of each year’s investment merely to be maintained and replaced. The system’s answer, intensification, was proclaimed at congress after congress, and the proclamation is itself evidence: the men running the economy knew which transition was required. What they could not do was make their own institutions perform it. An enterprise director rewarded for fulfilling a plan measured in gross output does not volunteer for the disruption of installing new equipment, which endangers this quarter’s fulfilment for the sake of a productivity that will be credited to a successor. A ministry judged by its sector’s tonnage does not close its obsolete plants. The transition to intensive growth was blocked not by ignorance but by the incentive structure of the mechanism itself, which is why the systemic account treats the blockage as structural rather than as a failure of personnel, and why replacing leaders, as the reform era did, changed so much less than the reformers hoped. The mechanics of the planning system that generated these incentives, its targets, its supply organs, and its bargaining between centre and enterprise, are treated in the companion article on how Soviet planning worked, and are assumed rather than repeated here.

The deceleration and the two sets of books

Every explanation in this article leans on one empirical claim, that Soviet growth decelerated steadily from the nineteen fifties through the nineteen eighties, and the claim must now be examined rather than assumed, because the evidence for it has a peculiar structure that the figures protocol exists to respect. The deceleration appears in the official Soviet series. It appears in the reconstructions prepared by Western agencies and scholars, who distrusted the official figures and rebuilt the accounts from physical indicators and price data. And it appears, most dramatically, in the alternative reconstructions prepared inside the Soviet Union itself, above all by the statistician Grigory Khanin, whose recalculations, first circulated among specialists and then published in the glasnost years, argued that the official series had overstated growth so persistently that the true performance of the economy, especially in its later decades, was far weaker than either the official figures or the Western estimates suggested. The three families of series disagree about levels so sharply that quoting any of their numbers against the others is meaningless without a chapter of accounting. They agree about direction. On every construction, the economy of the nineteen eighties grew more slowly than the economy of the nineteen seventies, and the seventies in turn trailed the sixties, which had themselves trailed the fifties. A deceleration attested by hostile witnesses using rival methods is as close to a settled fact as this disputed field provides, and it is on the direction, never on any contested level, that this article builds.

Why do the official and reconstructed series disagree about the same economy?

Because they count different objects with different prices. The official accounts measured net material product, excluding most services, at administered prices that concealed inflation in the guise of new products. Reconstructions revalued output on Western accounting concepts, or rebuilt it from physical series, and each correction moved the result in the same downward direction.

The comparison discipline follows from the same point, and it is worth stating as a rule because the collapse literature breaks it constantly. To say that the Soviet economy grew more slowly than the American, Japanese, or West German economies in a given period is to compare a series built on net material product, at administered prices, for a command economy, with series built on gross national product, at market prices, for market economies, and the comparison is legitimate only when the accounting concepts have been reconciled, which is what the Western reconstruction agencies spent their budgets doing. To say that the Soviet economy decelerated is a comparison within one accounting tradition across time, and it is robust in a way the cross-system comparison can never be. This article therefore makes the strong claim about deceleration and the weak claim about relative performance, and readers should treat any source that reverses that allocation of confidence as unreliable however impressive its tables. The measurement problem, the rival reconstructions, and the archival revisions that followed the opening of the Soviet statistical record are treated at length in the companion article on the economic record of the socialist states, which this article draws on for its evidence and to which readers who want the numbers, with their health warnings attached, are directed.

One further consequence of the measurement dispute matters for grading the explanations. The systemic account is sometimes accused of resting on Western estimates that the alternative reconstructions show to have been too kind, as if the kindness weakened the argument. The opposite holds. If Khanin’s reconstructions are nearer the truth, the extensive model exhausted itself earlier and more completely than the systemic account’s standard versions supposed, and the leadership’s turn to desperate reform becomes more intelligible, not less. The measurement dispute thus discriminates among explanations hardly at all, which is precisely why this article refuses to build on any figure: the figures are the disputed territory, and an argument that needs a particular growth rate to be true is an argument hostage to the next archival revision. Direction is enough. An economy whose growth fell decade by decade on every honest count, whose leadership knew it, and whose institutions could not perform the transition the leadership proclaimed, is an economy whose rulers will eventually attempt something drastic. The explanations divide over what they attempted and why it ended as it did, not over whether the pressure existed.

Innovation, incentives, and the soft budget constraint

The systemic account’s mechanism becomes concrete at the level of the enterprise, and the enterprise evidence is the strongest part of its case. The American scholar Joseph Berliner, studying Soviet industrial management from interviews with emigrated managers and from the Soviet managerial literature itself, documented the pattern a generation before the end: the enterprise director’s career depended on fulfilling the plan for the current period, innovation threatened fulfilment by disrupting production, and the rational director therefore resisted new technology unless it was imposed with priority backing, preferring proven methods, padded targets, and quiet arrangements with suppliers and ministries. The system did not lack inventions. Its research institutes were large, its scientists numerous, its priority sectors, weapons and space above all, capable of work at the world’s frontier. What it lacked was the connective tissue by which an invention becomes an industry: the supplier who profits from offering it, the competitor whose adoption forces imitation, the manager rewarded for risking a quarter’s output on a better process. Innovation in the civilian economy arrived by administrative campaign, which is to say episodically, at chosen sites, without the follow-through of diffusion, and each campaign’s showcase coexisted with an unchanged hinterland of obsolete plant.

Why did enterprises hoard labour and resist new machinery?

Because shortage made hoarding rational and the plan made innovation dangerous. A manager who could not trust supplies kept reserves of materials and workers against the storming periods; a manager whose bonus and career hung on the current plan avoided the disruption of retooling, whose gains would arrive after the evaluation that decided his fate.

Kornai’s concepts organise this evidence into a mechanism. The soft budget constraint meant that the enterprise’s survival did not depend on its performance: losses were covered, taxes adjusted, credits extended, because the state that owned the enterprise also employed its workers and answered for its town. A firm that cannot die need not adapt, and a manager who knows rescue is certain prices risk accordingly. Chronic shortage was the system’s other signature: because prices did not clear markets and plans allocated by quantity, demand chronically exceeded supply, sellers faced queues rather than competitors, and the producer’s problem was never finding a buyer but always finding inputs. In a shortage economy the incentives that drive intensive growth in a buyer’s market are inverted. Quality is unrewarded because everything sells; economising is unrewarded because the plan measures gross output; innovation is punished because it interrupts the flow on which evaluation depends. The Marxist versions of this analysis, developed by economists working in the tradition’s idiom, reached the same junction by a different road, attributing the blockage to the bureaucracy’s interest in administering scarcity, since scarcity is what makes administrators powerful, or to the absence of the market discipline that the tradition’s own theory assigns to the law of value. The vocabularies differ and the mechanisms converge, and the convergence, once noticed, raises the question this article keeps returning to: how much of the Marxist explanation of the collapse is specifically Marxist, and how much is a general theory of the command economy that Marxists and their opponents built together. The answer matters less for the collapse itself than for the objection considered later, since a tradition is not diminished by sharing a mechanism with its rivals, though it is diminished in its claim to have seen what no one else could.

External pressure: the military burden weighed

No explanation of the collapse is complete without the weight the system carried in the world, and no part of the subject is more encrusted with legend. The Soviet Union maintained military forces and a defence industry on a scale that its economy, smaller and less productive than its principal rival’s, could sustain only at a cost its consumers and its civilian industries paid visibly. The burden was real, and its reality is not in dispute among serious writers of any school. What is disputed is its size, with estimates constructed by rival methods diverging so widely that this article, under its protocol, asserts none of them, and its causal weight, with the triumphalist version, in which an American president deliberately spent the Soviet Union into bankruptcy, competing against versions in which the burden was a chronic drag that the system had carried for decades without dissolving. The weighing must therefore be done with the sequence in view. The burden was heaviest, relative to the economy, in periods when the system did not collapse; it pressed on a decelerating economy in the nineteen eighties, when a renewed arms competition across the Atlantic raised the prospective cost of staying in the race at the very moment the civilian economy’s stagnation was becoming undeniable to its rulers. A weight that a growing economy carries as a burden, a stagnant economy experiences as a sentence, and the burden’s contribution to the collapse runs through that change of experience rather than through any threshold of expenditure crossed in a particular year.

How should a burden be weighed when its size is itself disputed?

By its direction and its opportunity cost rather than by a contested share. The defence sector absorbed a disproportionate share of the economy’s best engineers, materials, and machine tools across the whole period; whatever the aggregate figure, those resources were unavailable to the civilian modernisation the leadership knew it needed, and the rivalry’s renewal raised the price of continuing.

The Marxist schools divide over the burden in an instructive way. For the state capitalist analysis, military competition is not an external pressure at all but the system’s internal motor: the bureaucracy accumulated under the whip of rivalry with Western capital, and the defence sector’s priority was the form the law of accumulation took in a state capitalist economy. On that reading the burden cannot be a cause of collapse separable from the system itself, since it was the system working as designed, and the collapse must be explained by the accumulation regime’s exhaustion rather than by the weight of its characteristic product. For the degenerated workers’ state analysis, the burden was a parasitic drain imposed by encirclement on a post-capitalist economy, and its growth testified to the bureaucracy’s integration into a world order it could neither defeat nor join. For the systemic account, the burden enters as a misallocation compounding the growth problem: the sector that received the best of everything returned its products to no one’s consumption and few industries’ improvement, since secrecy walled its innovations off from civilian diffusion. The three readings share the fact and divide over its meaning, and the reader grading them should notice that the burden explains the stagnation more readily than the dissolution. Armies did not dissolve the union; at most, the cost of matching them persuaded its rulers that their model had failed. Persuasion, once again, is an event inside the stratum, and the external explanations, however weighty their facts, must pass through the internal initiation to reach the ending they claim to explain.

Oil, grain, and hard currency

The second external pressure operated through the budget and the balance of payments, and it is the pressure with the most precise timing in the whole story. From the nineteen seventies the Soviet Union earned a large part of its hard currency from energy exports, above all oil and gas sold to the West, and it spent that currency on imports the domestic economy could not supply: grain for a livestock programme whose promise of more meat had become a pillar of the leadership’s bargain with its citizens, and machinery and technology for the modernisation that domestic innovation was failing to deliver. The arrangement concealed the deceleration for a decade. Energy earnings paid for the imports that filled the gaps the planning system left, and the leadership could postpone the reckoning with intensive growth while the wells paid for its avoidance. In the mid nineteen eighties the world oil price fell sharply, and the concealment ended. Hard currency earnings contracted at the moment the import bill did not, foreign borrowing filled the interval, and the leadership that took office in those years inherited an external position that turned an acknowledged slowdown into a fiscal emergency. This article asserts no prices and no quantities, under its protocol; the direction and the timing are what the argument needs, and both are attested by every account of the period whatever its politics.

What did the hard currency from energy exports conceal, and for how long?

It concealed the deceleration itself. Energy earnings paid for grain and machinery imports that filled the planning system’s gaps, letting the leadership postpone the transition to intensive growth through the decade when the reserves for that transition were still available. When prices fell, postponement and the emergency arrived together.

The strongest statement of the oil argument comes from inside the reform camp itself. Yegor Gaidar, who served as acting prime minister in the Russian government that succeeded the union and who wrote the collapse’s most influential insider anatomy in Collapse of an Empire, argued that the Soviet leadership was trapped by the grain and oil nexus: an agricultural system that could not feed the cities without imports, an import bill payable only in hard currency, a currency earned overwhelmingly from a commodity whose price the Soviet Union did not control. On Gaidar’s account the leadership’s choices in the final years were made inside a narrowing corridor, and political liberalisation itself was partly a search for Western credit and goodwill by men who knew the treasury’s position. The argument is powerful and this article gives it its full weight, while noting the two respects in which it cannot be the whole story. The first is that other states have survived commodity price collapses without dissolving, by borrowing, retrenching, or repressing, and the Soviet leadership’s failure to choose any of those roads returns the explanation to the stratum’s choices and cohesion. The second is that Gaidar writes as a participant with a thesis about his own government’s inheritance, and his corridor, though real, is drawn by a man justifying the speed of what followed. Weighed rather than worshipped, the oil shock stands as the trigger that converted chronic stagnation into acute crisis at the summit, the event that made the internal initiation happen when it did. Triggers are not causes of the structure they ignite, and a price movement in a commodity market cannot explain why a union dissolved along national lines or why its ruling stratum declined to defend it. The oil explains the timing of the desperation. It does not explain the form of the surrender.

The war in Afghanistan and the costs of holding an empire

The brief of this article assigns the external pressures a fifth of its space, and the war that began as the nineteen seventies closed belongs among them, though its weight is of a different kind from the budgetary pressures. The Soviet intervention in Afghanistan, undertaken by the leadership of the pre-reform period to sustain a friendly regime, became a decade-long war that the army could not win and the leadership could not initially abandon. Its material cost, whatever its contested size, was smaller than the defence burden as a whole; its political cost was larger. The war consumed the lives of conscripts from every republic and returned their bodies to communities that the official press could not indefinitely deceive, corroding the regime’s claim to competence and benevolence in the years just before glasnost made such corrosion speakable. It taught a generation of officers and officials that the system could commit itself to unwinnable enterprises and sustain them by inertia, a lesson in institutional self-knowledge that mattered when the same men later assessed the union’s own prospects. And it formed part of the inheritance that the reforming leadership cited, in its internal deliberations and its public diplomacy, as evidence that the old course was unsustainable: the war was wound down and the forces withdrawn in the reform period’s first years, an act of the same internal initiative that this article tracks, since ending the war was among the earliest exercises of the new leadership’s authority.

The empire’s wider costs compounded the lesson. The Soviet Union sustained client states and allied parties across the world, subsidised friendly governments with energy and arms, and held an Eastern European bloc whose loyalty, as the events at the decade’s end showed, depended on a guarantee of Soviet force that the reforming leadership withdrew. The bloc’s dissolution, examined in its own right in the companion article on actually existing socialism in Europe, preceded the union’s and instructed it: the Eastern European regimes fell, in the main, when their populations moved and their rulers, denied the prospect of Soviet rescue, declined to shoot. That sequence, peoples first and rulers yielding, is nearly the reverse of the Soviet one, and the contrast is analytically precious. In Eastern Europe the internal initiation belonged to crowds and the stratum’s role was abdication; in the Soviet Union the initiation belonged to the stratum and the crowds arrived to find the work begun. A single explanation stretched across both endings must accommodate the reversal, and the systemic account does so most easily, since the guarantee’s withdrawal was itself a decision of the Soviet centre, taken for reasons of the centre’s own exhaustion: the empire was released because holding it had become one burden too many for an economy and a leadership in acknowledged crisis. The war, the subsidies, and the bloc thus enter the explanation as the leadership experienced them, as evidence accumulating at the summit that the old course could not be financed, fought, or believed in indefinitely. External pressure, in every one of its forms, reached the ending by the same road: through the calculations of the men at the top, whose conversion from custodians to liquidators is the event to be explained.

The non-Marxist explanations at full strength

The Marxist schools did not hold the field alone, and this article’s brief requires their rivals to be stated as strongly as their proponents state them, not as foils. Four non-Marxist explanations command the literature. The first is the knowledge and calculation objection, the oldest of all, raised against socialist planning in principle decades before the Soviet collapse and treated at length in the companion article on whether planning can beat the market. Ludwig von Mises argued in nineteen twenty that without market prices for the means of production, a planning authority cannot calculate: it cannot compare the costs of alternative uses of resources, because the numbers it needs are generated only by exchange among owners. Friedrich Hayek restated the objection as a problem of knowledge rather than arithmetic: the information an economy needs is dispersed among millions of minds, much of it tacit and local, and no centre can gather it, because gathering changes it and the gathering apparatus would need to know in advance what it was looking for. On this account the Soviet economy did not merely perform badly; it performed blindly, substituting the planner’s guess and the enterprise’s bargaining for the price signals it had abolished, and its survival for seven decades is explained by its ruthless concentration on a few priorities and its borrowing of the world market’s prices as an external compass. The collapse, for this school, is the long argument arriving at its verdict: the calculation problem was never solved, only postponed, and the deceleration is the postponement running out.

What separates the calculation objection from the incentive objection?

Their diagnoses of where the failure sits. The calculation objection says the centre could not know what to order, lacking prices that only exchange generates. The incentive objection says the orders, whatever their quality, would not be executed well, because no one in the chain gained from executing them well or lost from executing them badly.

The second explanation is the incentive objection in its general form, of which Kornai’s shortage economics is the most sophisticated version and the property-rights school the most ideological. Its claim is that a system without residual claimants, without anyone whose wealth rises and falls with an enterprise’s performance, will systematically misdirect effort: managers will serve the indicators by which they are judged, workers will calibrate effort to a wage disconnected from output, and officials will convert discretion into income. The objection explains the texture of late Soviet life, the universal second economy, the blat networks of favour and exchange, the moonlight repairs and the traded privileges, better than any rival, because it treats those phenomena not as moral decay but as the rational behaviour of people inside an incentive structure, rebuilding privately the signals and rewards the official economy had abolished. The third explanation is legitimacy, treated in the next section. The fourth is the national question, treated after it, and it is the explanation whose neglect by the Marxist schools this article has already flagged as their most consistent blind spot.

Stated at full strength, the non-Marxist explanations share a common architecture that should be made visible. Each identifies a feature the system could not reform away without ceasing to be itself: it could not introduce real prices without dissolving the plan, could not create residual claimants without creating owners, could not manufacture belief by decree, and could not federalise in earnest without arming its own dissolution. Reform therefore faced a trap that the reformers themselves came to describe: every measure sufficient to cure the identified defect was sufficient to dissolve the system, and every measure safe for the system was insufficient for the defect. That trap, and not any single pressure, is the strongest non-Marxist candidate for the collapse’s deep cause, and it converges with the Marxist systemic account at the level of mechanism while diverging from it at the level of moral: where the systemic Marxist sees a historically specific formation exhausting its possibilities, the liberal critic sees a refuted idea receiving its sentence. The article’s grading must record both the convergence and the divergence, because collapsing them into each other would flatter both camps at the expense of the record. The record shows a mechanism failing. What the failure means is the part the camps will never share.

Legitimacy: the belief that drained away

The third non-Marxist explanation locates the collapse in the evaporation of belief, and it explains features of the ending that the economic accounts leave untouched: above all, the silence. A system that had demanded, and for decades received, public professions of faith ended without defenders willing to argue for it in the open, and its own cadres, when the emergency committee called, behaved like men who had stopped believing the words they had been paid to repeat. The anthropologist Alexei Yurchak, studying the final Soviet generation, described the late system’s peculiar condition: its ideological forms were reproduced with perfect fidelity and decreasing conviction, in rituals whose performance mattered and whose content had drifted free, so that the system appeared immutable to everyone inside it precisely because no one any longer expected the forms to mean what they said. Such a structure is stable until the moment it is not: because the forms were sustained by performance rather than belief, a shift in what the performers expected each other to perform could dissolve the whole edifice without anyone having changed their mind in public. The explanation illuminates the ending’s eerie quiet, the absence of a Thermidorian defence or a Bonapartist saviour, and the speed with which symbols that had organised a civilisation became memorabilia.

The legitimacy account also explains the specific lethality of glasnost in the sequence. A regime whose authority rests partly on a founding narrative, of revolution, encirclement, and victory in the great war, is vulnerable to its own archives in a way a merely coercive regime is not. Openness licensed the public recovery of everything the narrative had buried: the terror, the famine years, the deportations, the secret protocols, the privileges of the apparatus itself. Each revelation was survivable alone; cumulatively they withdrew the past’s warrant for the present’s sacrifices, and they did so through the regime’s own newspapers and television, under a leadership that believed exposure would purify rather than dissolve. Here the internal initiation and the legitimacy mechanism interlock exactly: only the leadership could license the exposure, the exposure dissolved the belief that made the leadership’s authority something more than office, and officeholders whose authority has thinned to office are precisely the men who calculate conversion rather than resistance when the crisis comes. The Marxist schools have generally treated legitimacy as superstructure, a reflection to be explained by deeper movements, and the collapse is among the events that rebuke the habit. Belief was not the deepest cause of the ending, but it was a load-bearing one, and the schools that underweighted it were surprised by the quiet in the streets when the end arrived, having expected either defence or revolt and prepared their theories for both.

The national question: the dimension the tradition underweights

The union did not dissolve into its citizens; it dissolved into its republics, and that geographical fact is the standing rebuke to every explanation that treats the collapse as a purely economic or ideological event. The Soviet state was constitutionally a federation of national republics, each with its territory, its titular language, its party and academy and writers’ union, and a formal right of secession written into the constitution it never expected to be used. For most of the union’s life the federation was a façade over unitary rule from Moscow, and the national institutions were transmission belts. But a façade maintained for seventy years builds real institutions behind it: republican elites with careers to make, cultural establishments with grievances to curate, and populations for whom the republic, not the union, was the frame in which a political future could be imagined. When the centre loosened its grip, the republics were the only alternative sovereignties ready to hand, complete with flags, borders, and governments in waiting. The union’s end was therefore negotiated among republics, declared by republican presidents, and ratified along republican borders, and the map of the successor states is the map of the federation’s internal divisions, almost without amendment. Any explanation of the collapse that cannot say why the ending took that shape has left the event’s most visible feature unexplained.

Why did the union’s federal form become the instrument of its dissolution?

Because the federation had spent decades building the successor sovereignties inside itself. Each republic possessed borders, institutions, an official language, and a political class whose careers ran through republican office. When the centre’s authority failed, those ready-made states were the only vessels available, and ambition and fear alike flowed into them.

The Marxist tradition’s handling of this dimension is the weakest part of its record on the collapse, and the weakness has a genealogy. The tradition’s classical writings on the national question were substantial: the early Soviet state was built partly on a theory of national self-determination, and its nationalities policy, in its first decades, actively cultivated the republican institutions that later dissolved it. But the schools that theorised the mature system subordinated the national dimension to the class one as a matter of method. For the degenerated workers’ state analysis, the bureaucracy was a union-wide caste and the republics administrative conveniences; for state capitalism, the relevant unit was the accumulation process, which respected no internal borders; for the systemic account, the mechanism was the plan, and the plan was one. Each framework therefore treated the federation’s form as decoration, and each was caught unprepared when the decoration turned out to be the load-bearing structure of the ending. The non-Marxist literature has its own anticipations to be proud of here. Hélène Carrère d’Encausse, in L’Empire éclaté, published in nineteen seventy-eight, argued from the demographic growth of the Central Asian republics and the persistence of national identity that the union would break along its national seams, a prediction made when the system’s permanence was the consensus of specialists. Andrei Amalrik, the dissident historian whose essay asked whether the Soviet Union would survive until nineteen eighty-four, likewise foresaw disintegration with the nationalities at its centre, though he wrapped the insight in a mechanism, war with China and a collapse of the centre’s authority, that events did not oblige. The anticipations were partial and their authors were not Marxists; they are recorded here because the predictive ledger, when the article reaches it, must be kept honestly across all camps, and on the national dimension the camps’ scores are not close.

Why did the stratum itself split along national lines? The internal initiation, pressed one level deeper, requires the question, because the initiators were not a unified caste acting on a single calculation. The republican leaderships who dissolved the union were men of the same apparatus as the union leadership that tried to preserve it, and their divergent choices tracked their institutional positions: those whose careers were housed in republican institutions could convert their authority into national presidencies, while those housed in union institutions faced extinction with the union. Conversion, the process Trotsky’s fork named, thus had a geography. The caste did not convert as a body; it converted where it stood, into owners where ownership was on offer and into national rulers where sovereignty was, and the union died of the divergence between its servants’ available futures. This refinement strengthens the fork’s core insight, that the stratum’s behaviour followed from the insecurity and convertibility of its position, while correcting its union-wide framing, and it suggests the general lesson that the article’s grading will formalise: the explanations succeed in proportion as they model the stratum not as a character in a morality play, defender or traitor, but as a population of officeholders whose options were set by the institutions that housed them.

The reformers’ intentions: what the initiators thought they were doing

An explanation centred on the stratum’s initiative owes the stratum the courtesy of its own self-understanding, because the distance between what the initiators intended and what they produced is itself evidence about the system. The leadership that launched reform in the mid nineteen eighties did not intend dissolution, privatisation, or the end of the party’s rule. Its leading figures described their project, in public and in the internal record that has since been published, as the system’s renewal: the removal of bureaucratic obstruction, the release of initiative, the return to what they presented as the revolution’s authentic principles after decades of stagnation. The general secretary who embodied the reform had been formed entirely inside the apparatus, had risen through its agricultural administration and its central secretariat, and spoke the language of its ideology with evident sincerity. His closest collaborators included the party’s ideology secretary, who had spent a career policing the doctrine he now proposed to open, and economists who had spent the stagnation years drafting the memoranda that the new leadership finally read. The reform was, in its origins, an inside job in the fullest sense: the system’s own diagnosis of itself, written by its own servants, enacted by its own summit. Nothing in the record supports the retrospective fable that the reformers were secret liberals executing a concealed design, and nothing supports the opposite fable that they were traitors to a healthy order. They were custodians who concluded that custody required transformation, and who discovered, link by link, that the transformation had a logic that consumed the custody.

That discovery is the human form of the trap the non-Marxist explanations describe. Each step of the reform was defensible in the reformers’ own terms and corrosive in its effects. Openness was to discipline the bureaucracy by exposure; it dissolved the narrative that made the bureaucracy’s rule legitimate. Enterprise autonomy was to release initiative; it released the directors from the plan before subjecting them to any market, and they used the interval to convert administrative control over assets into something approaching ownership, through cooperatives, joint ventures, and the leasing arrangements that the reform’s own legislation created. Republican elections were to renew the soviets; they created rival sovereignties. The surrender of the party’s constitutional monopoly was to democratise the system; it removed the one institution that had coordinated the whole, since the party, whatever else it was, had been the union’s actual constitution in working form, the body through which the centre’s will reached every ministry and republic. When the party’s authority failed, there was no second skeleton inside the state to hold it upright, because the system had spent seventy years ensuring there would not be. The reformers thus enacted, in sequence, the discovery that their system was indivisible: it could be run as it was, decaying, or it could be opened, whereupon it came apart along seams they had not known were structural. An explanation of the collapse that omits this discovery mistakes the ending for a plan, and an explanation that rests on it alone mistakes a plan’s failure for the whole cause. The systemic account supplies the reason the reform was attempted; the intentions supply the reason it was attempted in that order and that spirit; the stratum’s conversion supplies the reason no one stopped it. The complete explanation is the three together, and the schools’ partiality consists in each mistaking its own third for the whole.

The conversion in detail: how office became ownership

The fork’s second prong deserves to be traced in the mechanism, because the conversion of a caste into owners is easy to assert and harder to picture. The stratum’s members held, at the moment reform began, three assets that were not property: authority over enterprises and ministries, which decided what was produced and who received it; membership of networks, the nomenklatura system of appointment and favour, which decided careers; and information, the insider’s knowledge of where value lay in an economy whose official prices concealed it. The reform’s legislation, designed for other purposes, created the instruments by which each asset could be capitalised. Laws permitting cooperatives allowed officials and directors to found nominally collective firms that traded with their own state enterprises on terms the founders set. The loosening of foreign trade monopolies allowed those positioned at the gates to earn hard currency from the price differences between the administered domestic economy and the world market. Enterprise autonomy and leasing allowed directors to detach the profitable parts of their plants into structures they controlled. The youth organisations of the party, licensed to experiment with market forms, incubated banks. By the time formal privatisation arrived under the successor governments, a substantial part of the conversion was already accomplished in substance: the men who emerged as owners were, in striking proportion, the men who had held the relevant offices, directorships, and connections, joined by a smaller number of outsiders whose talents the new game rewarded. The process had no central direction and needed none. It was the aggregate of ten thousand individual conversions, each rational, each exploiting a position the old system had conferred, and its aggregate effect was to give the stratum, as a population, a stake in the old system’s non-return that no counter-revolutionary conspiracy could have organised more effectively.

This mechanism matters for the grading because it discriminates finely between the first two rows of the table. The state capitalist reading of the same facts stresses continuity: the same people, the same enterprises, the same subordination of the workers, under new legal dress. The degenerated workers’ state reading stresses the change of form in the stratum’s own position: what the officials held before was office, revocable and non-hereditary, and what they held after was property, saleable and bequeathable, and the difference between office and property is not a detail of legal dress but the difference the entire classification dispute was about. The record of the conversion supports the second reading’s emphasis more than the first’s. The officials themselves behaved as men acquiring something they had not had: they scrambled, they competed, many failed and fell, and the winners guarded their new possessions with a jealousy they had never needed for their offices. A class merely changing its dress does not scramble. A caste converting its perquisites into property does, because the conversion is a race in which position must be exchanged for assets before the position expires. The emergency committee’s failure belongs in this ledger too: by the union’s final summer, enough of the stratum had converted, or saw conversion within reach, that the defence of the old arrangements promised them less than their abandonment. The internal initiation thus has an economic micro-foundation, and the fork, alone among the anticipations, describes its motive in advance.

The working class that did not move

Every Marxist explanation of the collapse must eventually face the empty chair at the table: the class in whose name the system claimed to rule, and in whose name the tradition’s oldest analysis expected the caste to be removed, played almost no independent part in the ending. There were strikes in the final years, above all by miners whose grievances over pay, supplies, and safety were real and whose organisation briefly promised an independent labour politics. The miners’ movement was courted by the republican challengers of the union leadership, lent its strength to the politics of dissolution, and was then discarded by the successor governments whose market reforms fell on the miners first. Beyond that episode, the industrial working class, the largest in the world, watched its system end with a passivity that contemporaries recorded with astonishment. It did not defend the arrangements, which it experienced as stagnation and privilege; it did not storm them, having no organisation, programme, or memory of independent action on which to draw; it conserved its energy for the catastrophe of the transition itself, in which its wages, savings, and industries were consumed. The tradition that had theorised this class as history’s protagonist must explain its absence from the decisive scene, and the explanations on offer are more damaging to the tradition’s self-image than any of the classification disputes.

The schools’ answers divide by their premises. The degenerated workers’ state tradition answers that seventy years of bureaucratic rule had atomised the class, destroyed its independent organisations, and corrupted the very language of its emancipation into the property of its oppressors, so that when the crisis came the workers possessed grievances without instruments; the answer is historically plausible and concedes the essential point, that the political revolution the fork required was beyond the class’s actual capacities by the time it was needed. The state capitalist school answers that the class behaved as a working class under capitalism behaves when a regime changes form: it pursued its immediate interests, struck over wages and supplies, and did not mistake a change in the form of its exploitation for its liberation; the answer fits the record’s surface and is the school’s quiet triumph in this section of the argument. The systemic account, being about mechanisms, has least to say, and says it honestly: a system that had abolished independent unions and folded representation into the apparatus had deprived itself of the feedback and the defence that an organised citizenry might have supplied, so that when the summit faltered there was no constituted public to renegotiate the order with. Underneath the three answers sits a common admission that none of the schools enjoys making: the system’s subjects were its audience, not its authors, at the end as at the beginning, and a theory of the collapse is a theory of elites and institutions or it is nothing. The tradition’s critics have made much of this admission, and the strongest objection, to which the article now approaches through the predictive record, draws its force from it.

The predictive record: timing, manner, speed

The ledger must now be opened, and it must be opened without favour, because the tradition whose explanations this article has graded claimed, as no rival tradition claimed in the same way, to possess a science of historical development. The entry on the debit side is long. No Marxist account predicted the collapse’s timing: no school named the decade, let alone the year, and the tradition’s journals in the middle nineteen eighties were filled, as everyone else’s were, with analyses premised on the system’s continuation. No account predicted the manner: the fork expected either revolution from below or a restoration imposed as counter-revolution, and what occurred, a reforming leadership dissolving its own system through legislation and negotiation while the population watched, appears in no school’s scenario. No account predicted the speed: from the first reform decree to the lowering of the flag took under seven years, for a system that had been expected by its critics to decay across generations and by its defenders to endure indefinitely. And no account predicted the dissolution’s geography, the union breaking into its republics, since the schools had filed the federation’s form under decoration, a point the national discussion above established. On timing, manner, speed, and map, the predictive record of the Marxist tradition on the largest event in its own territory is a blank, and the article states the blank plainly because the tradition’s strongest critics state it, and because a grading that flinches here forfeits its authority everywhere else.

What precisely did every school fail to foresee?

The conjunction, and each element of it. That the end would come within the decade in which reform began; that it would be initiated by the leadership as an act of renewal; that it would proceed without civil war, invasion, or insurrection; and that the state would finish as fifteen successor states drawn on its internal borders.

The blank is not softened by noticing that nearly everyone else failed too, though the fact belongs in the ledger for fairness. Western Sovietology, a profession with institutes, journals, and intelligence budgets devoted to the subject, did not predict the collapse either; its mainstream expected continuity, its debate concerned the pace of decline, and the few who foresaw an ending were regarded, until the ending, as excitable. The intelligence services of the system’s principal rival, with every incentive to detect terminal weakness, reported a stagnating but durable adversary into the final years. The dissident and outsider anticipations that did exist were partial in instructive ways. Amalrik foresaw disintegration and misdated it by the better part of a decade while assigning it to a war that never came. Carrère d’Encausse foresaw the national fracture and located its engine in Central Asian demography, whereas the fracture was opened by the Baltic republics and completed by the Slavic core. Emmanuel Todd, the French demographer whose La chute finale appeared in nineteen seventy-six, inferred the system’s terminal decline from its infant mortality statistics, a genuinely independent anticipation, and he foresaw a long decomposition rather than the negotiated end that occurred. The pattern across all camps is consistent and worth stating as a finding: the system’s fragility was visible to scattered observers for decades, and its actual ending, in its timing, manner, and speed, was foreseen by no one, because the ending depended on the choices of a handful of men in a specific conjuncture, and conjunctures are the part of history that theories of structure are worst at pricing. That finding excuses no one. It locates the failure where it belongs, in the gap between structural diagnosis, which several traditions achieved, and event prediction, which none did, and it sets up the one credit entry the Marxist ledger can honestly claim.

The one anticipation: what was named in advance

Against the blank stands a single entry, and this article has been building toward the precision it requires. In the nineteen thirties, Trotsky’s analysis of the bureaucracy specified four things about a possible ending from within: the agent, the ruling stratum itself; the motive, the turning of insecure administrative privilege into secure and hereditary property; the destination, the restoration of private property in the means of production; and the condition, the failure of the alternative, a political revolution from below, to materialise. Half a century later, the ending arrived with that agent, that motive legible in the conversion mechanism traced above, that destination, and that condition satisfied by default. No other analysis in any tradition specified as much. The liberal critics foresaw a refutation, not a process; the demographers foresaw a decline, not an agent; the nationalities scholars foresaw a fracture, not a conversion. The fork foresaw the stratum’s self-transformation as one of two named futures, in print, with the reasoning attached, two generations before the event. That is a genuine anticipation, and a grading that denied it would be as dishonest as one that inflated it.

What did the one anticipation specify, and what did it leave blank?

It specified who could end the system from within, why they would wish to, and what they would build instead. It left blank the date across half a century, the route, which ran through reform rather than reaction, the dissolution of the union itself, and the conduct of the working class, whose assigned role in the alternative future never materialised.

The inflation to resist is as important as the credit to record. The fork was a disjunction, and a disjunction that covers two opposite futures purchases its accuracy cheaply unless the conditions discriminating between them are specified, which in the original they largely were not: the analysis could not say which prong would arrive, or when, or by what trigger, and its custodians spent fifty years expecting the prong that never came. A weather forecaster who predicts, each year for fifty years, that the dam will either hold or be removed by its keepers, and is then vindicated when the keepers remove it, has seen something real about dams and keepers and has not forecast the flood. The anticipation’s true value lies in its anatomy rather than its prophecy: it identified the stratum’s structural insecurity, the non-hereditary, revocable character of bureaucratic privilege, as the system’s deepest internal fault line, and the collapse did run along that line. Structural insight of that order is what the tradition at its best produced, and the honest verdict on the predictive record is therefore double, and must be quoted whole or not at all: the tradition failed as prophecy and succeeded, in one analysis, as anatomy; it foresaw the fault and neither the earthquake nor its date. The strongest objection takes the first half of that verdict and presses it as the whole, and it is time to let the objection speak at the strength its authors gave it.

The strongest objection: the tradition indicted on its own claim

The objection is stated here at full strength, attributed to the named critics who made its components, and answered only after it has been allowed its full force. Karl Popper devoted The Poverty of Historicism and long stretches of The Open Society and Its Enemies to the claim that Marxism’s predictive pretensions were its defining vice: a doctrine that presents itself as a science of history’s laws must stand or fall by its forecasts, and its forecasts, Popper argued, had either failed or been rendered unfalsifiable by the reinterpretations with which its adherents protected them. Leszek Kołakowski, whose Main Currents of Marxism remains the most formidable internal history of the doctrine, closed his account with the judgement that the tradition’s prophecies had been refuted by the century they claimed to interpret, and that its survival as a faith, after its failure as a science, was the phenomenon requiring explanation. Applied to the Soviet collapse, the objection assembles itself from the record this article has already conceded, in three counts. First, a tradition claiming to explain the laws of motion of social systems failed to predict the dissolution of the state built in its name: not the date only, but the manner, the speed, and the map, on the largest event in the territory the tradition had studied longest and claimed to understand best. Second, the tradition produced, after the fact, incompatible explanations of the same event, caste conversion, class repositioning, new class defeat, systemic exhaustion, each derived with confidence from the same doctrine, which suggests that the doctrine constrains its adherents’ conclusions so loosely that any outcome could have been explained with equal facility, and a framework that explains everything predicts nothing. Third, and most pointedly, the tradition’s most widely held account of the system’s nature, the analysis of the bureaucracy as a stratum whose behaviour follows from its position, requires as its central event something that no theory of class interest anticipates: a ruling stratum dismantling, by its own initiative, the arrangements that constituted its power, voting away its monopoly, dissolving its state, and scattering its own collective position to the winds of individual conversion. Classes, on the tradition’s own teaching, fight for the arrangements that embody their rule; they do not convene to abolish them. An event that the theory’s core mechanism cannot generate, occurring at the centre of the theory’s home territory, is not an anomaly to be absorbed but a refutation to be answered.

How does the objection turn the tradition’s own standard against it?

By accepting the tradition’s self-description and grading it there. The doctrine claimed scientific standing through its grasp of history’s laws and its capacity to foresee the fate of social formations. The objection simply presents the record: the formation built in the doctrine’s name ended unforeseeen in manner, timing, and shape, and was explained afterwards in mutually incompatible ways.

The answer this article gives is the one its brief prescribes, and it is an answer that concedes before it replies. The predictive record is poor: conceded, without reserve, in the ledger above. The retrospective explanations were adjusted to fit: conceded in part, since each school did revise its account in the event’s light, softening the prong that failed, discovering that the route had been implicit, re-describing conversions as repositionings and repositionings as conversions, and the adjustments are visible in the literatures of the nineteen nineties to anyone who reads them against the literatures of the nineteen eighties. The incompatibility of the explanations is likewise conceded as a fact about the tradition’s condition, though with one discrimination: the incompatibility is chiefly about classification, about what the system was, while on the mechanism of the ending the accounts converge more than the objection allows, since caste theorists, systemic economists, and even state capitalists all route the ending through the growth model’s exhaustion and the stratum’s response to it. What the answer claims, against the objection’s first and third counts, is the one anticipation. A single analysis within the tradition did specify, in advance, the agent of an internal ending, its motive, and its destination, and that specification is the direct reply to the charge that class interest theory nowhere anticipates a stratum dismantling its own arrangements. Trotsky’s fork anticipates exactly that, and anticipates it from a theory of the stratum’s interest: the caste dismantled arrangements that constituted its power because those arrangements did not constitute property, and a stratum whose power cannot be bequeathed has an interest, legible in advance, in exchanging the arrangements for ones in which its gains become its own. The objection’s third count therefore lands on the state capitalist and collectivist rows, whose strata were classes in possession, and misses the first row, whose stratum was defined by the insecurity that made self-dismantling rational. The objection’s first count lands on the whole tradition and is answered only by the admission that one structural anticipation is not a predictive record. The verdict the article returns is therefore mixed and precise: the tradition stands indicted as a predictive science and acquitted, in one of its schools, of the charge that the event was unintelligible to it in advance. Whether that mixed verdict is enough is a judgement the reader must make with the ledger open, and the schools’ subsequent conduct, their rush to explain what none had foreseen, remains the objection’s best evidence that the indictment stung because it was partly just.

Inevitability, contingency, and the road not taken

Every explanation of the collapse is finally graded by a question it cannot evade: did it have to happen? The question splits into three, and the article’s grading depends on keeping them apart. Was the stagnation inevitable, given the system? Was some terminal crisis inevitable, given the stagnation? Was this ending, dissolution along republican lines under a reforming leadership, inevitable, given a crisis? The systemic account answers the first confidently: a mobilisation model exhausts itself by its own logic, and the incentive structure that blocked the intensive transition was constitutive, not accidental, so the deceleration was built into the design. The second it answers more cautiously: exhausted models can be endured, as the nineteen seventies were endured, and the step from stagnation to crisis required triggers, the oil price collapse chief among them, and a leadership’s decisions. The third it declines, and the declension is the honest position. The dissolution’s specific form ran through a chain of choices each of which had alternatives: the choice to liberalise politically before restructuring economically, the choice to concede republican elections, the choice by the emergency committee to act and to act feebly, the choice by three republican presidents to sign a dissolution rather than a renegotiated federation. A union treaty was being negotiated in the union’s final year and had initialled drafts; the state that ended was, months before its end, the subject of a plausible re-foundation. Contingency is not the opposite of explanation. It is the name for the level at which the structural pressures, which the explanations genuinely identify, were converted into one outcome among several by identifiable decisions, and a grading that demands inevitability from an explanation demands the one thing the record refuses to supply.

Which parts of the outcome were contingent on decisions?

The form and the timing, as distinct from the pressure. Stagnation was structural and a crisis of the model was probable across the medium term, but the dissolution’s route, its republican geography, its peacefulness, and its completion within one year turned on choices by the reforming leadership, the emergency committee, and the republican presidents, each of which had alternatives.

The comparison that tests inevitability hardest sits at the other end of the socialist world, and it must be handled as history, which it is. In the same years, another large party state faced the exhaustion of its own extensive model and chose a different sequence: economic reform first, sustained across a decade, with political monopoly not conceded but violently defended when challenged, and with the reform’s beneficiaries inside the party recruited as the order’s defenders rather than its liquidators. That state did not collapse. The comparison does not show that the Soviet Union could have been saved by imitation, since the two systems differed in their national structure, their level of development, their place in the world economy, and the composition of their ruling strata, and the Soviet leadership’s room for manoeuvre was narrower in every dimension. What the comparison shows is narrower and sufficient: the sequence was a variable, not a fate. Political liberalisation preceding economic restructuring was chosen in Moscow and not chosen elsewhere, and the outcomes diverged with the choice. Explanations that treat the Soviet sequence as the system’s necessary unfolding must explain why the unfolding was optional in the other case; explanations that treat it as pure accident must explain why the pressures that produced it were so general. The article’s grading rewards the accounts that hold both ends: structure supplied the crisis, decisions supplied the ending, and the internal initiation is the hinge where the two meet, since the decisions were the stratum’s and the stratum’s situation was the structure’s product.

Convergence: where the explanations meet

The article has noted the convergence repeatedly in passing, and it deserves a direct statement, because it is among the most underreported findings in the whole literature. On the mechanism of the collapse, the systemic-economic account developed by Marxist economists and the mainstream account developed by Western economic historians and the more reflective Sovietologists are substantially the same account. Both locate the deep cause in the exhaustion of extensive growth. Both treat the innovation blockage and the soft budget constraint, or their equivalents in other vocabularies, as the mechanism that prevented the intensive transition. Both treat the oil shock as the trigger that converted slowdown into fiscal crisis. Both treat the reform’s asymmetry, political opening outpacing economic reconstruction, as the proximate cause of dissolution rather than renewal. Both arrive at the ending by way of the leadership’s decisions, and the most admired mainstream narrative of the collapse, Stephen Kotkin’s Armageddon Averted, explains the ending precisely as an establishment destroying itself, an internal initiation in everything but this article’s name, argued by a historian with no Marxist commitments from the archival record of the elite’s own behaviour. When a Marxist systemic economist and a mainstream archival historian, starting from opposed premises about what the system was, converge on the same mechanism of its ending, the convergence is evidence that the mechanism is where the record is strongest, and the classification dispute, on which they continue to differ, is where the record underdetermines the answer.

Where do a Marxist systemic account and a mainstream account become indistinguishable?

In their mechanism. Both derive the crisis from the growth model’s exhaustion, the innovation blockage, and the fiscal trigger, and both route the ending through the leadership’s choices under those pressures. They diverge only on what the failing system should be called and on what its failure proves about the alternatives.

The divergence that remains is not trivial, and convergence should not be inflated into consensus. The camps differ on whether the mechanism’s failure indicts planning as such or a specific historical formation; on whether the market transition that followed was a restoration, a liberation, or a catastrophe, a question on which the transition’s own record of output collapse and demographic harm, described directionally here as the protocol requires, is Exhibit A for one side and a transitional cost for the other; and on the classification questions this article has kept open. They differ, above all, in what they think the collapse was an experiment about. For the liberal tradition it was the decisive test of an idea, and the idea failed. For the Marxist systemic writers it was the biography of a formation, interesting for what it reveals about formations generally, including the one that succeeded it. The article’s position is the grading position it has occupied throughout: the experiment discriminates among claims about the Soviet system finely, through discriminating facts of the initiation kind, and discriminates among claims about socialism, capitalism, and planning in general hardly at all, because a single case with a specific history cannot bear the weight of universal verdicts in either direction. Readers who came for a universal verdict have been given, instead, a table, a sequence, and a ledger, which is what the evidence supports.

Grading the explanations: a balance sheet

The article’s work can now be drawn together into the grading its title promises, stated as a balance sheet rather than a coronation, since no row of the table emerges solvent in every column. The degenerated workers’ state analysis holds the strongest position on the initiation and the outcome: it alone anticipated, structurally, that the stratum might end the system from within in order to convert its position, and the conversion’s mechanism, traced in the offices and cooperatives of the reform years, matches its anatomy of bureaucratic insecurity with uncomfortable fidelity. It is weakest on the agency it expected from below, on timing, and on the union’s national dissolution, which its framework had no resources to foresee. State capitalism is strongest on continuity: the personnel, the networks, and the subordination of labour crossed the ending intact, and no rival explains that crossing as directly. It is weakest at the initiation itself, where it must explain a possessing class’s self-liquidation, and its best replies narrow its distance from the caste analysis it opposes. Bureaucratic collectivism speaks best to the desertion: a class abandoning a formation that had failed it describes the absence of defence better than any rival’s vocabulary. It is weakest on its founding wager, that the new formation was history’s destination, and on the mechanism by which slow competitive decline became sudden dissolution. The systemic-economic account is strongest on the premise: why reform was forced, why it took its ruinous asymmetric form, and why it came when it came. It is weakest, by its own design, on everything that made the ending a political and national event rather than an economic one, and it requires supplementation from the legitimacy and national analyses that sit outside the Marxist schools’ traditional strengths. The non-Marxist explanations, finally, hold the strongest positions on legitimacy and on the federation, the two dimensions the Marxist accounts underweighted, and the weakest position on the stratum’s interior, the conversion mechanism, which the oldest Marxist analysis had mapped before any of them.

Two conclusions follow for the reader who has stayed with the grading. The first is that the explanations are strongest exactly where their founding premises directed their attention, and weakest where their premises told them not to look, which is a general lesson about theories of society that this case teaches with unusual clarity because the outcome is known and the commitments were in print. The second is that the internal initiation, this article’s organising fact, has done the work claimed for it: it is the one feature of the event on which the rows’ performances differ most sharply, and a reader who remembers nothing else should remember the grading it produces. The stratum ended the system. The analysis that had described the stratum as a caste balanced over property it did not own found that behaviour most intelligible; the analyses that had described it as a class in possession found it most embarrassing; and the account that declined to classify it explained the pressure under which it acted while declining to explain the act. Between those three sentences lies the whole Marxist debate on the collapse, and the debate, graded honestly, is a better monument to the tradition’s powers and limits than either the triumphalism of its enemies or the defensiveness of its friends.

The party as constitution: what actually held the union together

An explanation of a dissolution must identify what was dissolved, and in the Soviet case the answer is less obvious than the flag suggests. The union’s formal constitution described a federation of sovereign republics with a right of secession, a supreme soviet, and a government of ministries. Its working constitution was the Communist Party: a single hierarchy of committees paralleling every state organ, deciding appointments through the nomenklatura, transmitting decisions downward, and supplying the only channel through which a career, a grievance, or an instruction could travel between the centre and the periphery. The state administration executed; the party decided, staffed, and coordinated. This dual structure explains a feature of the ending that purely economic accounts pass over. When the reform surrendered the party’s constitutional monopoly and then watched the party itself split into platforms, factions, and finally republican parties, the union lost not an ornament but its nervous system. The ministries remained, the supreme soviet remained, the presidency was invented and filled, but the tissue that had connected them, and through which the centre had known and commanded the whole, was gone. Republican governments that wished to defy the centre no longer faced a hierarchy that appointed their subordinates; union ministers who wished to enforce a decision no longer commanded a party cell inside the enterprise that was to execute it. The dissolution of the state followed the dissolution of the party’s authority with the regularity of an anatomy lesson, and the lesson generalises the article’s central claim: the system was ended from within because its connective institution was dismantled from within, by a leadership that had intended to democratise the party and discovered that the party, undemocratised, was the union.

The point discriminates among the explanations once more. The degenerated workers’ state analysis had always treated the party as the caste’s instrument and the union as its shell, and it accommodates the party’s primacy naturally. The systemic account, focused on the plan’s information flows, tends to treat the party as the plan’s political wrapper and underweights its coordinating role, which is one reason the systemic account explains the economy’s disorganisation better than the state’s. The state capitalist school, for which the party was the collective capitalist’s board of directors, must explain why a board would dissolve its own company, and finds itself again at its characteristic difficulty. The national analysis gains from the point as well: the party was the one union-wide career structure, and its failure is what released republican elites from a ladder that had led to Moscow into ladders that now led only to their own capitals. When the all-union ladder broke, every ambitious official in every republic recalculated his future within a national frame, and the union lost its servants before it lost its territory. Sovereignty declarations followed the career logic as faithfully as they followed any national awakening, and the two motives, conviction and calculation, are not rivals in the record but partners, varying in proportion from republic to republic in ways the next section traces.

The republics: three roads out of the union

The dissolution’s geography repays a closer look, because the republics did not leave for the same reasons and the differences grade the explanations once more. The Baltic republics travelled the road of national restoration: their incorporation into the union within living memory, under the shadow of the secret protocols whose existence glasnost finally forced into print, gave their movements a claim of interrupted statehood that no other republic could match, and their popular fronts mobilised mass publics earlier and more completely than anywhere else. In the Baltics, alone in the union, the ending really was driven substantially from below, by movements that the republican party leaderships first resisted and then, calculating as officeholders do, joined and led. The Slavic core travelled a different road, that of elite rivalry within the system: the Russian republic’s leadership, having lost the contest for the union’s summit, built the Russian institutions into an alternative summit, declared Russian sovereignty as a weapon in a Moscow power struggle, and discovered that the weapon dissolved the union it was meant to reform. Russia’s role is the internal initiation’s purest case, since the largest republic’s defection was initiated neither by national oppression nor by mass revolt but by a faction of the ruling stratum that found a republican throne more available than the union’s. Central Asia travelled the third road, reluctance: its leaderships, the most conservative in the union and the most dependent on union transfers, defended the union longest, voted for its preservation in the union-wide referendum of the final spring, and accepted dissolution only when the Slavic republics had made it a fact, converting themselves thereafter into national rulers of states they had not sought. Three roads, three different mixtures of mass movement, elite calculation, and institutional collapse, and a single explanation that flattens them into one mechanism, whether national awakening or caste conversion, will misdescribe at least two of the three.

Which single-cause story does the record refuse most plainly?

Every one of them, and the refusal is the record’s most consistent feature. Defeat, revolt, bankruptcy, betrayal, and national awakening each name a real element and fail as a sovereign explanation, because the ending was produced by their interaction inside a stratum whose own choices connected them, and removing any single element leaves the others insufficient.

The referendum deserves its sentence in this section, because it is the fact most often deployed against the internal initiation and it deserves an honest answer. In the spring of the union’s final year, the union government held a referendum on preserving a renewed federation, and the participating republics voted, by a large majority of those who voted, for preservation. The union was nonetheless dissolved within the year, by its republican leaderships, over the expressed preference of its citizens for some form of its continuation. The sequence is sometimes offered as proof that the ending was an elite coup against the popular will, and sometimes as proof that the population’s attachment was too shallow to matter. For this article’s argument it is evidence of something more precise: the dissolution was an act of the stratum even against the recorded preference of the governed, which is the internal initiation in its strongest form, and the population’s failure to defend its own expressed preference returns the argument to legitimacy, to a citizenry that preferred the union in the abstract and would not organise for the system in the concrete. A preference without an organisation is a sentiment, and sentiments do not hold states together when their elites defect. The Marxist schools, which had always distinguished between a class’s sentiments and its organised capacity, might have been expected to make exactly this point; that so few of them did, in the event’s immediate literature, measures how thoroughly the ending surprised the tradition’s habits as well as its forecasts.

Why the ending was peaceful: the coercion that was not used

The collapse of a nuclear superpower, armed to a degree without precedent, controlling internal security forces of great size, was accomplished almost without bloodshed on its core territory, and the peacefulness is a fact as discriminating as the initiation. It was not for want of occasions. There were violent episodes at the union’s edges in the final years, in the Caucasus and in Central Asia and, in the union’s last January, in the Baltic capitals, where Soviet forces killed civilians defending broadcast facilities; the episodes are part of the record and this article does not write them out. But the centre never attempted the systematic repression that its instruments made physically possible and that comparable regimes, facing comparable challenges, have attempted. The emergency committee of the final August, commanding on paper the army, the security police, and the interior forces, failed to arrest its principal opponent promptly, failed to cut his communications, failed to order units whose commanders it trusted, and dissolved when its orders were refused by the formations sent into the capital. An apparatus of coercion built across seventy years discovered, at the moment of use, that it would not obey. The explanations must account for that refusal, and their accounts are among their most revealing products.

The legitimacy analysis accounts for it most directly: orders to fire on crowds require men who believe the order defends something, and the chain of belief had rotted from the summit down, so that commanders calculated disgrace and prosecution rather than duty. The caste analysis accounts for it through interest: the officers and officials who would have had to execute a repression were the same population then converting its positions into property and presidencies, and a bloodbath in the capital threatened every conversion in progress, uniting the stratum’s factions, for once, in refusal. The systemic account treats the refusal as a symptom of the general institutional decay it diagnoses, true but shallow at this point, since armies have obeyed decaying states before when their corporate interests were engaged, and the interesting question is why this army’s were not. The honest synthesis is that coercion failed because all three conditions coincided: the summit was divided against itself, the stratum’s members had futures elsewhere, and the ideology that had once made obedience a virtue had thinned into a performance no one would kill to sustain. Remove any one condition and the ending plausibly turns violent, as it did turn violent, later and locally, in several successor states where one of the conditions was absent. The peacefulness, so often credited to the wisdom or restraint of individuals, is better explained structurally, and the structural explanation is the article’s whole argument in miniature: a system whose governors have ceased to believe, ceased to agree, and acquired exits does not get defended, by anyone, at any price. It gets dismantled, in committee rooms, by the men who used to run it.

The West’s role: pressure, example, and the pull of the world market

Between the internal initiation and the external pressures stands a factor that fits neither category cleanly, and the grading would be incomplete without it: the West as example and destination. The Soviet system’s citizens and, more consequentially, its elites lived for decades in enforced comparison with a wealthier rival whose consumer abundance was visible through every crack in the information wall: through the testimony of travellers, the smuggled catalogue, the foreign broadcasts, and finally through the leadership’s own admissions. The comparison worked differently on different floors of the society. For ordinary citizens it set the standard by which the system’s promises were audited and found wanting, the queues measured against the supermarkets of report. For the technical intelligentsia it set the professional standard, the computers and laboratories against which their own were judged. For the ruling stratum it did something more dangerous to the system: it supplied the destination of conversion. A caste contemplating the exchange of office for property needs a world in which property is honoured, markets function, and fortunes are enjoyed openly, and the West’s existence, its banks, its resorts, its universities for officials’ children, made the post-Soviet future imaginable as enrichment rather than exile. The stratum did not defect to an abstraction. It defected toward a visible civilisation whose goods its members had been sampling, as privileged travellers and hard currency shoppers, for decades. Explanations that ignore the pull, treating the stratum’s choices as made in a vacuum of pure interest, miss the direction the interest pointed, and the direction was set by the rival system’s demonstrated success at exactly the thing the home system had stopped delivering: rising consumption without queues.

The triumphalist version of the West’s role, in which a deliberate strategy of pressure and spending forced the surrender, must nonetheless be weighed and found partial, as this article weighed the arms burden earlier. Western pressure raised the cost of the old course; Western example raised the attractiveness of abandoning it; Western credit, in the final years, was actively sought by the reforming leadership and extended, hesitantly, in amounts that postponed rather than decided. But no Western government planned the dissolution, several feared it, and the union’s principal adversary spent the final summer urging the republics to preserve a reformed union, from a well founded dread of loose nuclear weapons and a vast territory in disorder. The ending surprised its supposed engineers as completely as it surprised its victims, and a grading that credits the West with the initiation contradicts the record of Western diplomacy in the decisive months. The accurate formula is the one this article has used for every external factor: pressure and example worked on the stratum, and the stratum acted. The internal initiation is not diminished by acknowledging the weather in which it occurred; it is located.

The speed: six years from reform to dissolution

The tempo of the ending is the feature contemporaries found most uncanny, and it is the feature the explanations handle worst after timing, so it deserves direct treatment. Systems of the Soviet type were supposed to change slowly when they changed at all; the literature of the stagnation years is a literature of immobility, of a gerontocracy recycling its formulas while the society aged beneath it. Yet from the reform’s launch to the dissolution took less than seven years, and the final unravelling, from the monopoly’s surrender to the flag’s lowering, took under two. The speed has a mechanism, and the mechanism is the same one Yurchak’s account of legitimacy implies. A system held together by performed unanimity, in which every institution’s authority depends on every other institution continuing to perform, has no partial equilibria: it cannot be a little bit dissolved. Each defection alters every other actor’s expectation of the whole, and the alteration compounds. The republics declared sovereignty in a cascade because each declaration lowered the cost of the next; enterprises stopped obeying ministries because other enterprises had; officials converted their positions in a rush because conversion is a race against the position’s expiry. Cascades are fast because they are made of expectations, and expectations move at the speed of news. The explanations that model the system as a structure of interests predict the cascade’s possibility, since interests recalculate instantly when the future changes; the explanations that model it as a structure of coercion predict stability until the coercion is defeated, which never happened; and the explanations that model it as an economy predict a decline at the economy’s pace, which is the pace of plant and demography, far slower than the pace the event displayed. The speed thus grades the field once more, in favour of the accounts centred on the stratum’s expectations, and against any account that treats the collapse as the economy’s slow verdict finally arriving. The economy’s verdict had been arriving for twenty years. What arrived in the final two was the stratum’s conclusion, drawn all at once, that the future lay elsewhere, and conclusions of that kind, when a whole governing population draws them together, are the fastest force in politics.

The debate inside the tradition after the event

The tradition’s response to its own surprise is part of the record this article grades, and it unfolded in three movements that repay summary. The first movement, in the immediate aftermath, was salvage by reclassification. Schools that had defended the system’s post-capitalist character discovered that its fall confirmed their analysis of its instability; schools that had called it capitalist discovered that its fall changed nothing essential; schools that had called it a new formation discovered that its fall proved its inferiority. Each salvage operation is documented in the journals of the early nineteen nineties, and each is evidence for the objection’s second count, the retrospective adjustment of explanations to fit, which this article has already conceded. The second movement was more honest and more interesting: a reckoning literature, in which the tradition’s serious figures audited the failure. Ernest Mandel, the fork’s most distinguished custodian, wrote in the collapse’s aftermath that the bureaucracy’s self-conversion had vindicated the analysis’s core while its route had refuted the movement’s expectations, and he drew the conclusion his tradition found hardest, that the working class of the Soviet Union had been so thoroughly atomised by the caste’s rule that the political revolution his school had awaited for fifty years had never been a live possibility in the form imagined. Hillel Ticktin argued that the ending confirmed his formation’s non-viability while insisting that the market restoration succeeding it was no more viable, a position whose second half the transition’s miseries made less eccentric than it first appeared. The state capitalist school divided between those who treated the transition as their analysis’s coronation and those troubled, in print, by the initiation problem this article has pressed. The third movement, still unfolding in the tradition’s journals, is absorption: the collapse has become a case inside larger arguments about bureaucracy, accumulation, and the state, cited by writers who were children when it happened, and the schools’ founding texts are read as classics rather than as programmes. Absorption is what traditions do with their refutations when the refutations cannot be answered and need not be, the living questions having moved elsewhere.

The reckoning literature matters to the grading because it supplies the tradition’s own verdict on the objection, and the verdict is closer to this article’s than to either pole of the shouting. The serious Marxist writers on the collapse concede the predictive failure, concede the retrospective adjustment, and rest their defence on the anatomical claim: that the tradition’s best analysis of the stratum identified, in advance, the fault line along which the system broke. Their critics within the wider left reply that an anatomy vindicated once, on the tradition’s home case, at the cost of fifty years of false expectation about the agent of change, is a poor return on a science of history. Both positions are represented in the literature, both are honestly held, and the article’s balance sheet, drawn in an earlier section, is an attempt to state the terms on which a neutral reader can hold them together. What neither side disputes is the fact with which this article began and on which its grading rests. The system was ended by its own governors, in pursuit of futures they judged better than its defence, and the theories are graded by how nearly they saw those governors plain.

The archival turn: what the opened record changed

The explanations graded in this article were formed, most of them, before the evidence that now dominates the subject existed, and the account would be incomplete without reckoning with what the opened record did to them. In the years after the dissolution, the Soviet archives were partially opened, the diaries and notes of participants were published, and the memoir literature of the reform era accumulated into a documentary base of unusual richness for an event so close in time. The notes of Anatoly Chernyaev, the reforming general secretary’s principal foreign policy aide, kept across the whole period and published after it, record the summit’s deliberations season by season. The Politburo’s discussions, the diaries of ambassadors and ministers, the memoirs of the general secretary himself and of his rivals and successors, and the stenographic remains of the union’s final negotiations together allow the sequence to be reconstructed from inside the rooms where it was decided, a luxury historians of collapses rarely enjoy. What the opened record changed can be stated under three heads. It confirmed the internal initiation beyond serious dispute: the initiative at every turning point lay with the leadership and the republican elites, and the documentary trace of mass pressure forcing a decision is, with the Baltic exception noted earlier, essentially absent. It confirmed the leadership’s early and accurate knowledge of the deceleration: the summit was not surprised by its economists, having commissioned their diagnoses for a generation, and the reform was a considered response to a known condition rather than a panic. And it revised the picture of the reformers’ radicalism downward: the documents show men improvising within a vocabulary of renewal long after their measures had escaped that vocabulary’s control, surprised by each consequence, and divided among themselves about ends almost until the end. The archival turn thus strengthened the explanations centred on the stratum and weakened every explanation, heroic or demonic, that requires the initiators to have known what they were doing.

The archives also disciplined the economic debate in a way the participants had not expected. The internal statistical record, once accessible, showed the leadership’s own agencies reporting the slowdown, the shortages, and the fiscal deterioration in terms close to the alternative reconstructions and far from the published figures, which means the measurement dispute, examined earlier in this article, was never a dispute inside the summit: the men who ran the economy knew which set of books was honest. That knowledge matters for grading the systemic account, since it converts the reform from a response to published propaganda into a response to privately acknowledged reality, and it matters for the caste analysis, since a stratum that knows its system’s true condition and chooses conversion over defence behaves exactly as the analysis of its insecurity predicts. The archives are not complete, and the security organs’ records in particular remain substantially closed, so the documentary base has a shape: fullest on deliberation, thinnest on coercion’s inner workings. The grading in this article has respected that shape by resting its claims about the stratum’s calculations on the deliberative record and its claims about coercion’s failure on the public sequence of orders given and refused, which no archive is needed to establish.

The final years: disorganisation after the initiative

The sequence’s last phase requires separate treatment because it is where the popular image of economic collapse belongs, displaced in time. Once the party’s coordinating authority failed and the republics began their sovereignty cascade, the economy’s connective mechanisms failed with them, and the failure fed back into the politics that had caused it. Republican governments withheld taxes from the union budget, asserting control over the revenues collected on their territory, and the union government, responsible for the army, the debt, and the subsidies that held the whole together, found its income confiscated by its own constituents. Enterprises that had taken their instructions and their supplies through union ministries now faced competing sovereignties and chose, or were captured by, the nearer one. Inter-republican supply chains, designed on the assumption of a single economic space, were interrupted by republican export bans as each government hoarded food and materials against its neighbours. The monetary system, a single currency issued by a union bank, was undermined as republican authorities extended credits in the common currency for their own purposes, exporting the inflation to the whole. Shortages that had been chronic became acute; the rationing of basic goods by coupon appeared in city after city; the union’s final year presented the spectacle of a state whose economy was disintegrating beneath it, and that spectacle is the source of the retrospective belief that economic collapse caused the political end. The chronology refutes the belief while explaining it. The disorganisation was real, severe, and visible to every citizen, and it followed, link by link, the political dissolution it is summoned to explain. Cause and effect in the final years run from the stratum’s defection to the economy’s breakdown, and the explanations that respect the direction are the ones this article has graded highest throughout.

The fiscal detail carries one more discrimination worth recording. The union government’s bankruptcy in its final year was not the bankruptcy of an economy that had ceased to produce; it was the bankruptcy of a centre whose constituents had stopped remitting, in a federation whose fiscal constitution had always depended on the party’s discipline to function. When the discipline went, the constitution’s fiction, republican sovereignty over republican resources, became practice, and the centre starved in the midst of an economy that, however stagnant, still fed, clothed, and employed its population at the levels of the preceding decade. The distinction matters because it separates the collapse of a state from the collapse of an economy, two events that coincided in time and are persistently confused in the literature. The Soviet state collapsed in nineteen ninety-one. The Soviet economy’s deepest contraction came after, under its successors, as the transition’s own disorganisation compounded the inheritance. An explanation of the first event that borrows its evidence from the second will misdate its causes by years, and the misdating is the single most common error in the popular literature, which is why the misreadings section that follows gives it pride of place.

Could reform have worked? The alternatives audited

The inevitability section separated structure from decision; this section audits the decisions that were not taken, because an explanation’s quality shows in how it handles the counterfactuals. Four alternatives were live in the debates of the time and have been defended in the literature since. The first is discipline without liberalisation: the path associated with the short tenure of the leadership that preceded the reform, which proposed to restore growth by tightening labour discipline, fighting corruption, and modernising within the command framework. Its audit is brief. Discipline campaigns had been tried repeatedly across the system’s history and produced brief exertions followed by relapse, because the blockage was not effort but the incentive structure, and a campaign cannot repeal a structure. The second alternative is the sequence chosen elsewhere in the socialist world: market reform under an intact political monopoly, introduced gradually, with the party recruiting the reform’s winners. Its audit occupied the inevitability section: the sequence was a genuine variable, but the Soviet version faced obstacles the other case lacked, a federal structure of national republics waiting to be activated, a far heavier military burden, an economy already urbanised and industrialised so that reform threatened existing jobs rather than releasing peasants into new ones, and a stratum whose conversion options, as the West’s proximity and the federation’s structure supplied them, pointed toward dissolution rather than toward managed transition. The third alternative is conservative conservation: no reform, the stagnation endured, the lid held by the security organs and the party’s unbroken monopoly. Its audit is the hardest, because the nineteen seventies prove that endurance was possible for a long time, and its defenders can fairly ask what, in the structure, forbade another twenty years of decay. The answer the record supports is that conservation was available and was rejected, not by the population but by the stratum’s own leading faction, which had read the internal diagnoses and concluded that decay was cumulative and the terms of endurance worsening: the oil income that had financed the nineteen seventies was gone, the technological gap was widening in the industries that would decide the next military competition, and the succession of aged leaders had itself become a symptom the elite was ashamed of. Conservation failed, in other words, at the level of the custodians’ belief in it, which returns the counterfactual to the article’s centre: the alternatives were closed less by the structure than by the stratum’s unwillingness to inhabit them, and a theory of the stratum is therefore a theory of the collapse. The fourth alternative, a negotiated democratic federation, was attempted in the union’s final year and failed on the republics’ defection and the emergency committee’s intervention, which destroyed the centre’s remaining credibility at the moment the treaty awaited signature. Its audit is the referendum’s: the preference existed, the organisation and the trust did not, and by the time the treaty was ready the men who would have had to operate it had acquired other futures.

The counterfactual audit yields the article’s most compact verdict on inevitability. The stagnation was overdetermined; the crisis was probable; the dissolution was chosen, in the specific sense that at every node where it advanced, identifiable men preferred it, or preferred the step that entailed it, to the available alternatives, and preferred it for reasons their situations made intelligible. Explanations that erase the choosing flatter the structure; explanations that erase the situations flatter the men. The Marxist explanations at their best refuse both flatteries, since their founding method is precisely the location of choices inside structures, and the collapse literature’s Marxist entries are strongest exactly where they practice that method on their own tradition’s subject. Where they fail, they fail by exempting the stratum from the method, treating it as the bearer of a historical mission or a historical crime rather than as officeholders in the mass, whose options it is the analyst’s business to reconstruct.

The event’s aftershock: the tradition beyond the Soviet Union

The collapse was a Soviet event with a world subject, since the tradition it indicted was a world tradition, and a complete grading should follow the shock outward before returning to close. The communist parties of Western Europe, already transformed or declining, drew their conclusions at different speeds: the largest of them dissolved itself into a post-communist successor within weeks of the union’s final crisis, its leadership judging that the name and the model had become unusable together; others retained the name and shrank into custodianship of a memory. The parties of the developing world, many of which had oriented themselves by Moscow’s compass for two generations, lost funding, doctrine, and bearings in a single season, and the guerrilla movements that had survived the Cold War’s end found their logistics and their legitimacy withdrawn together, several of them negotiating settlements within the decade on terms their founders would have refused. The intellectual aftershock was deeper. A generation of scholars who had built careers inside the tradition’s questions migrated to new subjects, and the questions themselves, about planning, bureaucracy, and the possibility of a post-capitalist order, passed out of the universities’ centre for a generation, to return later in altered forms. The tradition’s opponents drew the largest conclusion available, that history’s direction had been settled, and the subsequent decades’ financial crises, wars of choice, and returns of state direction have subjected that conclusion to the same audit this article has applied to the tradition’s own prophecies, with results the reader can grade by the same ledger: a prediction of permanence, made at the moment of victory, that the record has not obliged. The symmetry is not a consolation prize for the tradition. It is a methodological warning that the collapse teaches impartially: the end of a system is the worst moment to theorise the permanence of its successor, because the men drawing the conclusion are standing in the cascade, mistaking its momentum for a law.

For the Marxist tradition itself, the aftershock’s most durable product was a change in what its serious work is about. The classification disputes that fill this article’s table belonged to a world in which the Soviet formation was a standing fact to be explained; with the fact gone, the disputes survive as instruments, sharpened on the case, for analysing bureaucracies, party states, and accumulation regimes wherever they appear, including inside the market economies whose own administrative strata the collectivists had always insisted on comparing. The tradition’s centre of gravity moved from the defence or indictment of a state to the analysis of mechanisms: soft budget constraints and their analogues in subsidised industries everywhere, the conversion of office into assets as a general pathology of transitions, the cascade dynamics of regimes held together by performed unanimity. A reader who asks what the tradition’s explanation of the collapse finally contributed can be answered without reference to the schools’ survival. It contributed the habit of asking, about any ending, who initiated it and what the initiators held, and that habit, applied to the tradition’s own founding state, produced the fork, the conversion mechanism, and the internal initiation on which this article is built. Traditions are judged by the questions they teach their successors to ask, and by that standard the aftershock, for all its wreckage, left the workshop better equipped than the triumphalism of the victors left theirs.

The inside critics: warnings the system filed away

One population sits awkwardly in every school, and the grading should acknowledge it before the misreadings are corrected: the critics inside the system who described its condition accurately, in advance, from within its own institutions and its own vocabulary. The sociologist Tatyana Zaslavskaya, working at the Novosibirsk institute that sheltered the system’s most honest social science, drafted in nineteen eighty-three a memorandum for a closed seminar that named the crisis in terms the systemic account would later make canonical: the economic mechanism corresponded to an earlier stage of the productive forces, the management of the economy by administrative command had exhausted its possibilities, and the interests of the groups administering the system had become the obstacle to its renewal. The memorandum leaked, was published abroad, and was treated by the authorities as a security matter rather than as the diagnosis it was; its author became, in the reform years, one of the leadership’s advisers, carrying her own filed warning into the office that finally acted on it. Around her stood a whole stratum of inside critics: the economists of the Novosibirsk and Moscow institutes who had computed the deceleration in the internal series, the historian Roy Medvedev, whose Let History Judge documented the terror from within a loyal Marxist framework and circulated in the West while its author remained a Soviet citizen arguing for a reformed socialism, and Andrei Sakharov, the physicist whose nineteen sixty-eight essay on progress and coexistence proposed the convergence of the two systems and who was exiled internally for his later defence of human rights. The inside critics matter to the grading for two reasons. Their existence refutes the lazy claim that the system’s condition was invisible until the archives opened: it was visible, documented, and argued, in Russian, by the system’s own servants, decades before the end. And their fate refutes the opposite claim, that knowledge would have sufficed: the diagnoses were filed, leaked, punished, and shelved, because the institutions that received them had no procedure by which a true report could compel a change of course. The reform, when it came, was the moment the filed warnings reached a general secretary willing to act on them, and the internal initiation is therefore also a story about information: a system that could generate its own accurate diagnosis and could not use it, until the diagnosis and the power met in one office, too late for the gradualism the diagnosticians had hoped for.

The inside critics also complicate the predictive ledger in a way fairness requires. None of them predicted the dissolution; Zaslavskaya’s memorandum proposed renewal, Medvedev proposed reform, Sakharov proposed convergence, and the dissident tradition generally imagined either a democratised socialism or a long moral resistance, not the flag coming down within the decade. But their work establishes that the structural diagnosis on which the systemic account rests was not a retrospective construction of the nineteen nineties, assembled to explain an unexpected corpse. It was the common property of the system’s best servants by the early nineteen eighties, in documents whose dates cannot be revised. When the objection charges the tradition with explaining after the fact, the inside critics are the documentary answer for the mechanism, though not for the event: the exhaustion of the growth model was foreseen, in writing, by the people running the model’s own institutes. What no one foresaw, insider or outsider, Marxist or liberal, was that the cure would kill the patient in six years and divide the estate among fifteen heirs. The distinction between foreseeing a condition and foreseeing an event is the distinction this article’s predictive sections have insisted on, and the inside critics are its clearest illustration: diagnosis, in this case, was abundant, accurate, and useless as prophecy.

Three generations of writing about the collapse

The literature on the collapse has itself acquired a history, and a reader navigating it benefits from knowing the strata. The first generation, written in the event’s immediate aftermath, divides between triumph and autopsy. The triumphalist works read the ending as a verdict on an idea and a vindication of its rival, and their characteristic error is the one this article flagged in the aftershock section: theorising the successor’s permanence from inside the cascade. The autopsy literature, much of it by participants, established the narrative sequence on which everyone since has relied, and its characteristic error is the participant’s: each memoirist’s corridor of the labyrinth is lit, and the corridors do not always join. The second generation is the archival one, built on the opened record and the published diaries, and its achievement was to replace the era’s journalistic certainties with documented deliberation: the summit’s knowledge, the reformers’ improvisation, the republics’ calculations, all recovered from paper. Its characteristic temptation is the opposite of the first generation’s, a narrative so attentive to the rooms where decisions were taken that the structure outside the rooms, the growth model, the incentive system, the federation’s loaded architecture, recedes into scenery. The third generation, the one now forming, treats the collapse as a case among cases: of regime dissolution without war, of elite defection and cascade, of the conversion of office into property, and it reads the Soviet ending alongside the endings of other orders with a comparativist’s cold eye. Its temptation is abstraction, the case dissolving into the model. This article has tried to write with all three generations open on the desk: the sequence from the autopsies, the deliberations from the archives, the mechanisms from the comparativists, and the grading from the schools whose commitments predated them all. A reader who wants to go further should know which stratum any given book belongs to, since the strata disagree less about facts than about what a fact is evidence of, and the disagreement about evidence is the permanent condition of this subject.

One further feature of the literature deserves note, because it affects every source a reader will meet. The collapse produced almost no literature of defence from the system’s former custodians. The union’s last leadership wrote memoirs of justification, the general secretary’s account of his intentions being the most substantial, but the stratum as a whole did not produce its apologia, for the reason the conversion mechanism supplies: its members were busy, and then they were owners, national officials, or pensioners, and the system they might have defended had no institution left to publish the defence. The archive of the collapse is therefore asymmetric, fullest where its critics and its liquidators wrote, emptiest where its believers would have. Historians call this the victors’ archive problem, and it should discipline every confident sentence in the literature, including the ones in this article that rest on the deliberative record of men who had already decided, by the time they wrote, how they wished to be remembered. The grading has preferred documentary traces made before the outcome was known, the filed memoranda, the internal series, the dated predictions of the schools, precisely because such traces cannot have been adjusted to the ending. Where the article has relied on memoirs, it has used them for the sequence of decisions rather than for the motives their authors assign themselves, motives being the one commodity in which every memoirist is long.

The sequel as evidence: reading the transition backward

Every camp in the debate recruits the transition that followed the collapse as evidence about the system that preceded it, and the recruitment deserves scrutiny, because the sequel is evidence of a treacherous kind. The transition’s first decade brought, on any accounting, a contraction of output deeper than the stagnation that preceded it, a collapse of public health and male life expectancy without peacetime precedent in an industrial country, the transfer of the economy’s commanding assets to a narrow circle of owners through privatisations whose mechanics confirmed every cynical expectation, and the conversion of the former stratum’s networks into the new order’s skeleton. The tradition’s defenders read the sequel as the restoration’s true face: this, they argue, is what the fork’s second prong meant, and the misery of the nineteen nineties is the measure of what the old system, for all its stagnation, had provided in security, equality of a grey kind, and the absence of mass unemployment. The tradition’s critics read the same decade as the old system’s delayed invoice: an economy so distorted by seventy years of command that its liberalisation could only be a collapse, its capital stock obsolete, its prices meaningless, its managers trained for a vanished game, so that the transition’s depth measures the inheritance’s rottenness rather than the reform’s cruelty. Both readings contain evidence and both commit the same fallacy in mirror image, treating a decade produced by the interaction of inheritance and policy as a pure sample of either. The grading discipline this article has used applies here too: the sequel discriminates among explanations only where the explanations made different predictions about it, and on that test the conversion mechanism scores again, since the fork’s account of a caste racing to convert position into property describes the privatisation decade’s personnel and methods with a fidelity that neither the triumphalist nor the purely economic accounts match. What the sequel cannot do is retrospectively grade the old system’s performance, which must be judged on its own record, in its own decades, by the directional evidence this article has used throughout.

There is a subtler use of the sequel that the grading should acknowledge, because it cuts against the article’s own emphasis and honesty requires it. The internal initiation explains the collapse by the stratum’s calculation that its future lay outside the system. The sequel tests that calculation, and for a large part of the stratum the calculation failed: the middle ranks of the apparatus, the industrial managers of the provinces, the officers, the teachers and doctors of the state sector, were not converted into owners but ruined, their savings destroyed by inflation, their salaries unpaid, their status annihilated. The men who dissolved the system in expectation of conversion were, in the majority, wrong about their own prospects, and an explanation that rests on the rationality of their calculation must absorb the fact that the calculation was, for most of them, a catastrophic error. The absorption is possible but must be stated: actors in cascades calculate with the information and the fears of the moment, each man’s exit rational given the others’, the aggregate outcome intended by none, which is why the article has consistently described the initiation as a cascade of conversions rather than as a conspiracy of beneficiaries. The fork survives the correction, since it predicted the attempt at conversion, not its universal success. But the correction disciplines the tone in which the conversion is described: not a caste coolly exchanging its perquisites for a richer order, but a stratum stampeding for exits of very unequal value, the swiftest and best placed seizing the assets, the rest discovering that the system’s end ended them too. That is a bleaker and a truer picture, and it brings the stratum’s story closer to the population’s, whose members also calculated, also stampeded, and also mostly lost.

A note on words: collapse, dissolution, fall

The vocabulary of the ending is not neutral, and a short note on it belongs in an article about explanations, because each camp’s preferred noun smuggles its explanation in before the argument begins. Collapse suggests a structure failing under its own weight, an engineer’s verdict, and it is the systemic account’s word, implying that the interesting questions are about loads and materials. Dissolution suggests a legal and political act, a body unmade by decision, and it is the internal initiation’s word: its questions concern the deciders. Fall suggests a height, a rival, and a push, and it is the triumphalist word, pointing at the contest as the thing to be explained. Decline, the mildest word, suggests a process without an event, and it is the word of those who think the event has been overrated. This article has used collapse in its title because the series assigns the title, and dissolution in its argument because the argument is about an act, and it flags the difference so that readers can watch their own vocabulary as they read the literature. The same discipline applies to the system’s name across the schools: workers’ state, state capitalism, bureaucratic collectivism, and the neutral actually existing socialism are not synonyms but verdicts, and a writer who adopts one has graded the table before presenting it. The article has therefore named the system by its own name, the Soviet Union, and its rulers by the neutral word stratum, wherever the argument was not explicitly reporting a school’s usage. Neutrality of vocabulary is not neutrality of judgement; the grading in these pages is sharp. It is the precondition of the grading being trusted, since a verdict delivered in the prosecutor’s vocabulary convinces only the prosecution.

The sector that worked: priority, secrecy, and the diffusion wall

A fact that every explanation must digest, and that the cruder versions of the systemic account digest badly, is that the Soviet system was not uniformly incapable. In its priority sectors it achieved results at the frontier of world practice: the weapons programmes that matched a richer rival across two technological generations, the space programme that led the world in its first decade and remained formidable in its fourth, the mathematics and theoretical physics whose schools were the equal of any. A system that can do these things cannot be described, without remainder, as an engine of incompetence, and the descriptions that survive contact with the fact are the ones that locate the failure precisely. The location the evidence supports is the boundary between the priority enclave and the civilian economy. Inside the enclave, the system’s methods worked as designed: resources were concentrated without limit, the customer was single and demanding, quality was enforced by a user who tested the product in earnest, and failure carried consequences that concentrated minds. Outside the enclave, the same methods produced the shortage economy’s familiar texture. The wall between the two was secrecy and priority itself: the enclave’s innovations were classified, its suppliers sequestered, its personnel forbidden the lateral movement by which, in market economies, military technology spins off into civilian industry. The system could innovate where it pointed its whole apparatus at a target. It could not diffuse, because diffusion requires the horizontal traffic, of people, components, and ideas between firms and sectors, that the vertical apparatus existed to prevent. The distinction refines the systemic account at its centre: the failure was not of science, nor of effort, nor even of investment, but of architecture, a vertical order that could storm a summit and could not irrigate a plain.

The schools divide over the enclave in ways that expose their premises once more. For state capitalism, the enclave is the system’s truth revealed: accumulation under military compulsion was the economy’s real purpose, the civilian sector its residual, and the frontier performance of the weapons programmes shows the system working at its essence rather than despite it. For the systemic account, the enclave is the exception that locates the rule: mobilisation succeeds at points and fails across surfaces, which is the extensive and intensive distinction wearing a uniform. For the degenerated workers’ state tradition, the enclave testified to the possibilities of planned production freed from profit, and its writers cited the space programme as the system’s promissory note; the collapse’s grading of that citation is unkind, since the note was never presented for payment in the civilian economy where the population lived. The consumer’s experience, which is where legitimacy was finally decided, was of the plain and not the summit: of goods that were shoddy, scarce, and unchanging across decades in which the rival’s goods transformed themselves repeatedly. A citizenry asked to take pride in orbital stations while queueing for sausage drew its own conclusions about the system’s priorities, and those conclusions, multiplied across a population and a generation, are the social form of the diffusion failure. The explanations that treat the enclave as a detail miss the political arithmetic: the system’s most expensive successes were invisible in daily life, and its daily failures were the most visible things in it.

The generation of the reform: where the initiators came from

The internal initiation has a biography, and tracing it answers a question the structural accounts leave hanging: why did the stratum produce, at the decisive moment, a leading faction committed to transforming the system that had made it? The faction’s personnel were formed in a specific passage of the system’s own history, the thaw that followed the dictator’s death, when the party’s new leadership denounced the terror in a secret speech, released the camps’ survivors, and licensed a limited honesty in literature and scholarship. The young men and women formed in those years, the generation of the sixties as Russian memory calls them, entered the institutes, the editorial offices, the party’s research departments, and the foreign service carrying a double inheritance: loyalty to the system’s founding promise and knowledge of its founding crimes, held together by the belief that the promise could be recovered from the crimes by reform. They spent the stagnation decades in the system’s internal exile of the possible: drafting the memoranda that were filed, editing the journals that pushed at the permitted boundary, staffing the institutes whose internal series told the truth the published ones concealed. When the succession finally opened, after the deaths of three general secretaries in rapid succession had discredited the gerontocracy even to itself, the reforming leader reached for exactly this population, promoting its members to the ideology secretariat, the foreign ministry, the editorial chairs of the glasnost press, and the commissions that drafted the reform. The internal initiation was thus not the stratum acting against its own formation. It was one formation of the stratum, the thaw’s children, finally holding the offices from which their lifelong project could be attempted, opposed by another formation, the stagnation’s custodians, whose counterstroke in the final August failed for the reasons this article has given. The point refines the caste analysis in a direction its founder would have recognised: the caste was never a monolith, its factions were formed by the system’s own history, and the faction that ended the system did so in the name of the system’s own promises, which is why the ending took the form of reform escaping its authors rather than of reaction overthrowing them.

The biography also explains the reform’s characteristic blindness, which the sequence section recorded without explaining. The thaw’s generation had formed its project against the terror and the stagnation, and its enemies list contained the bureaucracy’s obstruction and the doctrine’s dogmatism. It did not contain the federation, because the generation’s own formation was union wide and Russian speaking, its members drawn from every republic into Moscow’s institutes, and the national question belonged in its mental world to the solved problems of the founding era. Nor did its list contain the conversion of its own stratum, since the generation thought of itself as the apparatus’s conscience rather than as its advance guard into property. The reform therefore opened the two doors its authors were not watching: the republics walked out of one, and the apparatus walked out of the other, carrying the assets. An explanation of the collapse that begins from the stratum’s interests alone cannot say why the initiators misread their own situation so completely; an explanation that begins from the initiators’ ideas alone cannot say why their project, of all the projects on offer, commanded the offices. The complete account needs the biography, the interests, and the structure in one frame, and the schools’ partiality, graded throughout this article, is nowhere more visible than in each school’s decision about which of the three to leave out.

Misreadings to correct

The brief for this article lists five misreadings that the popular literature repeats, and each has been corrected in place as the argument proceeded; they are gathered here because a reader who carries away only the corrections has still carried away something worth having. The first misreading is that the system was defeated militarily. No army invaded, no front collapsed, and the state’s armed forces, the largest in the world, ended the story in their barracks, their commanders declining orders from a committee that claimed to speak for the state. Military pressure contributed through budgets and calculations, as this article weighed, but a defeat requires a battle, and the record contains none. The second misreading is that mass revolt brought the system down. Crowds mattered in the Baltics and at the barricades of the final August, but the initiative at each successive turning point, the reform, the elections, the monopoly’s surrender, the dissolution itself, belonged to officeholders, and the largest popular intervention in the sequence, the referendum, voted for the union’s preservation and was overridden by its elites. The third misreading is that economic collapse preceded political liberalisation and caused it. The chronology runs the other way: a stagnating but functioning economy was politically opened first, and the disintegration of supply, finance, and output followed the political dissolution link by link, as the section on the final years documented. The fourth misreading is that some Marxist account predicted the event. None predicted its timing, manner, speed, or geography; one analysis anticipated, structurally and decades early, the agent, the motive, and the destination of an ending from within, and the distance between that anticipation and a prediction is the distance this article’s ledger exists to measure. The fifth misreading is that the national question was secondary, a decorative complication on an economic story. The union dissolved into its republics, along borders drawn by its own federal constitution, through institutions its own nationalities policy had built, and the schools that treated that dimension as decoration were surprised by the shape of the ending even when they had diagnosed its approach. A reader who finds any of the five misreadings in a new account of the collapse can grade that account accordingly: it has not yet reckoned with the sequence, and the sequence is the one witness none of the schools can cross examine into silence.

To the five, the article adds a sixth misreading of its own finding, because it is the one the grading most depends on refusing. It is the misreading that the collapse, being a single event, must have a single explanation, such that the schools are contestants for a crown. The record supports a division of labour instead: the systemic account owns the premise, the caste analysis owns the initiation’s anatomy, the state capitalist owns the continuity, the collectivists own the desertion’s vocabulary, the legitimacy account owns the silence, and the national analysis owns the map. The crown model of explanation is a habit imported from polemic, where a tradition must win or lose entire, and it produces the literature’s characteristic waste, in which each school defends its weakest cell with the energy it should spend on its strongest. The table with which this article began is offered as the alternative habit: explanations as instruments with specified ranges, graded cell by cell against a record that none of them commands whole. Readers who adopt the habit will find it works on the next collapse too, whichever system supplies it.

The untranslated shelf: a note on sources and standing

A note on sources is owed, because the evidentiary base of this subject is divided by language in ways that silently shape the literature. The Western debate was conducted for decades on estimates and emigré testimony, and its standard works, Nove’s economic histories, Berliner’s studies of management, the reconstruction agencies’ series, are available in English and form the shelf an English reader can reach entire. The post-Soviet documentary literature is partly translated: Gaidar’s anatomy appeared in English, the Chernyaev diaries were published in translation by an American archive project, and the principal memoirs of the union’s last leadership exist in English versions of varying completeness. The alternative Soviet growth reconstructions, by contrast, remain substantially a Russian literature. Khanin’s recalculations, the statistical controversies of the glasnost years in which the official series were publicly indicted by Soviet statisticians themselves, and the reform era’s internal economic debates, conducted in the journals and institute papers of Moscow and Novosibirsk, are available in English only in fragments and summaries, usually at second hand through the Western scholars who used them. The same is true of the republican dimensions: the literatures of the dissolution in Ukrainian, in the Baltic languages, and in the languages of the Caucasus and Central Asia, where the ending is remembered as liberation, as catastrophe, or as both in sequence, reach the English reader filtered through a small number of translators and area specialists. The consequence for grading is practical. An English language account of the collapse is built disproportionately on the deliberations of the Moscow summit and the estimates of Western agencies, because those are the records its language opens, and it will systematically underweight the statisticians’ internal indictment and the republics’ own archives unless it corrects for the tilt deliberately. This article has tried to correct by resting its economic claims on direction rather than on any series, and its political claims on the sequence of public acts, which no language barrier conceals, rather than on the interior monologues of any capital. Readers pursuing the subject should treat any account’s bibliography as a map of its blind spots: the languages it cannot read are the rooms it has not entered.

What the explanations leave open

The article closes its grading with the questions the record, as it stands, does not answer, because an explanation’s honesty is measured by its residue. The first open question is the weight of the individual. The structural accounts agree that a crisis was probable and a dissolution chosen; they divide, and the archives have not settled, on how much of the choosing belonged to the specific men, above all to the reforming general secretary, whose combination of determination and restraint, pressing the opening and refusing the bloodshed that might have closed it, appears in every documentary account as a variable no structure supplied. A different man in that office plausibly conserves the system for another decade or drowns the opening in repression, and the explanations that cannot price that plausibility have a hole where the biography should be. The second open question is the working class’s counterfactual: whether any organisation, programme, or leadership could have converted the population’s grievances into an independent politics in the years of opening, or whether seventy years of atomisation had foreclosed the possibility in advance, as the fork’s custodians came bitterly to conclude. The record shows what happened; it cannot show what was possible, and the schools’ confidence on the point exceeds their evidence in both directions. The third open question is comparative and belongs to the future: whether the mechanisms this case has refined, the cascade of performed unanimity, the turning of office into assets, the federation that arms its own dissolution, are general instruments for reading the endings of party states and other brittle orders, or whether they are the biography of one system wearing the dress of a theory. The article’s wager, stated once and plainly, is that they are instruments, and that the reader who has followed the grading now owns them: ask of any threatened order who would have to initiate its ending, what those initiators hold, what they could carry across, and whether the forms that proclaim its permanence are believed or only performed. The Soviet case will not answer for other cases. It has taught the questions, and the questions, unlike the prophecies, have kept their value.

Everyday life as evidence: queues, jokes, and the second economy

The explanations disagree about the summit and the statistics, but they all claim the same ground floor, and a grading that never descends to it leaves the evidence half read. Everyday life in the late Soviet Union had a texture that contemporaries described with a consistency rare in social evidence: the queue as an institution, consuming hours of every citizen’s week; the deficit good, available by luck, connection, or the back door of the shop; the blat network, the web of reciprocal favours through which the official economy’s failures were privately repaired; the second economy of moonlighting repairmen, traded services, and garden plots whose produce, grown on a sliver of the land, supplied a share of the nation’s food that every school cites and none can measure within the protocol this article observes. The political joke flourished as the era’s characteristic literature, its targets the shortages, the aged leadership, and the gap between the slogans and the shops, and the jokes are evidence of a specific kind: a population that had stopped believing the forms and had not stopped performing them, which is Yurchak’s condition rendered as folklore. Each school reads the ground floor through its own lens, and the readings are worth grading like the rest. The systemic account reads the queue as the shortage economy’s signature, the visible form of prices that do not clear and plans that allocate by quantity, and the reading is exact as far as it goes. The incentive account reads blat as the market’s revenge, exchange reconstituting itself illegally inside the plan’s interstices, proof that the abolished signals regenerated wherever two people had something to trade. The legitimacy account reads the jokes as the belief system’s autopsy performed by its own subjects. The caste analysis reads the deficit good and the closed shop, the special distribution networks through which the stratum fed itself outside the queue, as the caste’s material constitution, the daily experience that taught the population what the stratum was. All four readings are true of the same queue, which is why the ground floor, like the summit, refuses single ownership.

The ground floor also supplies the collapse’s most quietly damning social fact, the one that connects the everyday sections to the internal initiation. When the ending came, the population’s dominant response was not grief, not triumph, and not resistance, but a turning inward to the private survival strategies the second economy had rehearsed: the plot, the network, the moonlight skill, the relative abroad. A citizenry that had spent decades learning to live around its state did not experience the state’s disappearance as the loss of a provider so much as the removal of an obstruction it had long routed around, and the energy that in other collapses flows into the streets flowed instead into the shuttle trade, the market stall, and the ransacking of the workplace for anything convertible. The schools that expected the population to act as a class, in defence or in revolt, had mistaken the second economy’s lesson. Decades of shortage had not collectivised the citizenry’s survival; they had privatised it, household by household, and a privatised citizenry meets the end of its state as a change in the weather, to be dressed for, not as a summons to history. That sentence, if it is true, is the deepest single explanation of the silence with which the union ended, and it was available to any observer of the queue in any year of the final two decades. The tradition’s failure to draw it belongs in the ledger beside the failures of timing and geography, as the failure of attention to the society directly in front of it.

The cleanest comparison: the system against itself

The comparison discipline stated at this article’s opening forbids loose cross-system contrasts, and it permits one comparison without reservation: the Soviet system against itself across time, measured within its own accounting traditions and judged by its own proclaimed standards. That comparison is the systemic account’s home ground and it is devastating in its simplicity. The system of the nineteen thirties, at a far lower level of development, with a peasant workforce and a primitive capital stock, could mobilise resources for priorities with an effectiveness its successor could not match in the nineteen seventies with an educated population and the accumulated plant of half a century. The system of the nineteen fifties could put the first satellite and the first man in orbit and terrify a richer rival into reorganising its science; the system of the nineteen eighties could not put a reliable personal computer, a decent shoe, or a queue free supermarket into its own capital. The system of the war years could evacuate an industrial base across a continent under invasion and out produce the invader in the weapons that decided the war; the system of the final decade could not evacuate its own authority from a television studio. Nothing in these contrasts requires a Western yardstick, a contested price index, or a figure this article’s protocol forbids. They are the system’s own biography, and they establish the one proposition every explanation needs and none disputes: something in the formation consumed its own capacities over time, converting a mechanism formidable at mobilisation into one incapable of adaptation, and the conversion happened inside the formation’s normal operation rather than through any external wound. The schools divide, as ever, over naming the something: the exhaustion of extensive reserves, the caste’s consolidation, the new class’s sclerosis, the accumulation regime’s maturity. The self-comparison cannot choose among the names. It can and does refute the accounts, still encountered, that treat the collapse as an accident of leadership or a bolt from the world market, since a formation whose capacities declined across four decades of peace, under leaders of every temperament, was travelling toward its crisis on its own rails. What the leadership of the nineteen eighties supplied was not the direction but the decision to stop travelling and dismantle the train.

The succession: gerontocracy and the queue for office

A final structural feature deserves its place, because it connects the system’s biology to the stratum’s psychology in a way the schools underuse. The Soviet order never solved the problem of succession. Its founder’s death produced a terror; his successor’s death produced a thaw and a palace removal; thereafter the system settled into a pattern in which leaders died in office at advanced ages and were replaced by the next oldest guardian of the same settlement. In the reform’s immediate prehistory, three general secretaries died within a span of under three years, and the spectacle of state funerals becoming a genre of public life, of a leadership whose average age climbed as the society’s patience fell, did specific work in the stratum’s interior. It discredited conservation from within: the custodians’ claim to embody stability curdled into the image of a queue for office in which the prize was a brief tenancy of a mausoleum. It empowered the reform faction by giving it the one argument no conservative could answer, that the alternative to transformation was not permanence but a succession of funerals over a stagnating country. And it shaped the reform’s tempo, since the new leadership, itself no longer young and conscious of the queue behind it, chose speed in political opening over the patience its economic advisers counselled, fearing that a slow reform would be interred with its author like the thaw’s project a generation earlier. The schools have generally treated the succession as biographical colour. It is better treated as a mechanism: a system that cannot renew its leadership cannot renew its policies, and a stratum that watches its summit become a hospice draws conclusions about the future that no statistic could teach it faster. The internal initiation was launched by men who had counted the funerals.

A compressed genealogy: how the schools were born

The three classifications graded in this article were not invented to explain the collapse; they were forged, most of them, within a few years of each other, in arguments among people who had expected to make the revolution together, and a compressed genealogy explains both their durability and their blind spots. The argument began inside the Soviet party itself in the nineteen twenties, when the Left Opposition, analysing the bureaucracy that was strangling it, produced the first anatomy of the stratum as a social phenomenon with interests of its own: Christian Rakovsky’s letter on the professional dangers of power, written from exile, described how office transforms the officeholder’s psychology and material situation until a stratum of functionaries stands between the class and its state. Trotsky’s exile writings systematised the anatomy into the caste analysis and the fork. The first great schism came in the closing years of the nineteen thirties, when the Soviet state’s invasion of Finland and its partition of Poland with the Third Reich forced the classification question on the tradition’s Western sections as a practical matter: if the bureaucracy could act as an imperialist power in its own right, was the state it ruled still, in any sense the tradition could honour, a workers’ state? The American organisation split over the question in nineteen forty, and the minority that left, carrying Shachtman and Burnham’s arguments toward bureaucratic collectivism, founded the third row of this article’s table in the act of walking out. The state capitalist row was codified in the same years, by Dunayevskaya from within the American debate and by Cliff, systematising it after the war, from the study of the Soviet economy’s war driven accumulation. The three rows are thus siblings, born of one crisis of classification between nineteen thirty-six and nineteen forty-eight, and their subsequent seventy years of polemic have the character of a family argument: each school defined itself against the other two, refined its categories on their objections, and preserved, in its founding texts, a set of expectations about the stratum that the collapse finally tested all at once.

The genealogy explains the blind spots as faithfully as the insights. All three schools were founded to answer a question about the stratum’s relation to property and production, because that was the question the founding crisis posed, and all three therefore organised their attention around the factory, the plan, and the surplus. The national question, which the founding generation had treated in a different and earlier debate, sat in a separate drawer of the tradition’s cabinet, and the schools of the classification dispute rarely opened it; the federation’s architecture was, for them, an inheritance from that earlier debate, assumed settled. Legitimacy sat in another drawer, labelled superstructure. The collapse, arriving half a century later, drew its decisive facts from the unopened drawers as much as from the contested one, which is why the grading in this article has repeatedly found the schools strongest where their founding question pointed and weakest where it did not. A reader who understands the genealogy will not be surprised, and will also understand why the tradition’s internal debate, for all its ferocity, functioned as a single research programme with a shared perimeter: the schools policed each other’s answers and conspired, without intending to, in each other’s questions. The non-Marxist explanations entered the subject from outside the perimeter, from price theory, demography, and nationalities scholarship, and their comparative advantage on the unopened drawers is the advantage of the outsider who never agreed to the family’s terms. That is the deepest reason the convergence and the complementarity run as they do in this article’s balance sheet, and it is a lesson about traditions generally that the case teaches better than any methodological treatise: a school’s founding question is its destiny, and the facts that refute it are usually filed, by its founders, under somebody else’s subject.

Was there a third thing? The transition that was supposed to happen

Running beneath the schools’ debate, and surfacing in the reform years as the initiators’ own hope, was an idea that the collapse refuted more completely than any classification: the idea that the system could evolve, by reform from within, into a third thing, neither command nor market capitalism, in which planning’s coordination and social ownership would be married to democracy, legality, and a market’s information. The idea had a long pedigree inside the tradition. It animated the reform communisms that the system’s armies had suppressed in Budapest and Prague, whose programmes, drafted by the parties’ own economists and philosophers, proposed exactly the marriage the Soviet reformers later attempted. It animated the inside critics whose memoranda this article has cited, Zaslavskaya’s renewal and Medvedev’s reformed socialism alike. And it animated the reform leadership’s self understanding through most of its tenure: the general secretary’s published programme spoke of more socialism, of a humane and democratic form of the founding promise, and the men around him believed, on the evidence of their diaries, that they were building the third thing until the building’s collapse instructed them otherwise. The grading must record the idea’s fate precisely. The third thing was not refuted as a logical possibility; no single case can do that. It was refuted as a path from this starting point, by the mechanisms this article has traced: a reform that must democratise the party that is the state’s skeleton, marketise the economy whose prices are fictions, and federalise a union whose republics are sovereignties in waiting, cannot proceed in the order the third thing requires, since each element’s reform dissolves a precondition of the others. The reformers discovered the interlock by walking into it. Whether some other starting point, an earlier decade, a different succession, a slower hand, could have threaded the interlock is the counterfactual the alternatives section audited and left, as the record requires, unproven in either direction. What the record does establish is narrower and harder: from the position of the decade’s middle years, with the oil income gone, the stratum’s conversion options open, the republics built and waiting, and the belief already thinned to performance, the third thing was attempted by the most favourable leadership the system ever produced for the purpose, and it failed in six years. Explanations of the collapse that ignore the attempt miss the event’s tragedy and its evidence alike, since the attempt is the experiment that tested the system’s reformability under near ideal conditions of intention, and the result of the experiment is the strongest argument the inevitability side possesses.

The security organs: the institution that watched

One institution’s conduct in the ending is so contrary to its reputation that it deserves a section, because the explanations divide over it revealingly. The security police, the KGB, was the system’s most informed organ: its reports on the economy’s condition, the population’s mood, and the nationalities’ temper were, on the archival evidence, the most accurate the leadership received, and its chairman in the early nineteen eighties, Yuri Andropov, carried its diagnosis into the general secretaryship itself, becoming the reform impulse’s first, authoritarian embodiment. The institution that knew most about the crisis thus stood at the origin of the attempt to answer it. In the ending, the same institution split along the system’s fault lines. Its chairman in the final years joined the emergency committee and supplied its claim to authority; its special units, ordered into the capital in the crisis, declined the decisive assaults; its republican branches transferred their allegiance to the republican governments that employed them; and its archives, the memory of the whole apparatus, passed substantially intact to successors who have used them selectively ever since. The coercive core of the state behaved, at the end, exactly as the rest of the stratum behaved: it calculated, it hedged, it converted its local positions into futures under the successor arrangements, and it declined to die for the union. The point is worth more than its irony. Apparatuses of coercion are supposed, in the cruder theories of the state, to be the last defenders of any order, the men whose corporate existence is bound to the system’s survival beyond all calculation. The Soviet record shows a security organ whose members’ calculations ran the other way, because their information told them the system was failing, their careers offered republican and commercial exits, and their chairman’s adventure in August demonstrated that even the committee could not make the machine obey. A theory of the stratum that cannot accommodate its policemen is incomplete, and the caste analysis accommodates them without strain: officeholders all, holding revocable warrants in a system whose centre had stopped warranting.

The security organs also supply the grading with a control case for the legitimacy argument. If any population inside the system was selected, trained, and rewarded for belief, it was the officer corps of the security police, whose entry required demonstrated loyalty and whose work required performing it. Their conduct in the final crisis is therefore the hardest test the belief hypothesis can be given, and the hypothesis survives it in its refined form: what the officers displayed was not conviction betrayed but the performance’s end, the same transition from ritual to calculation that Yurchak described in the society at large, executed by the ritual’s own priests. When the priests stop believing the liturgy saves, the temple’s fall is a matter of scheduling. The schools for which ideology was a mask the stratum wore for others must explain why the mask’s keepers dropped it in unison; the accounts that treat the performance as the system’s actual substance, sustained by mutual expectation and collapsing when expectation turned, predicted the manner of the dropping if not its date. It is one more cell in the table where the grading favours the analysts of expectation over the analysts of force.

The economists’ war: the programmes that fought on paper

In the union’s penultimate year the collapse acquired a paper trail of a peculiarly instructive kind, as the reform camp split over the economy’s future and fought its argument in rival programmes whose fates map the ending’s logic. The government’s economists proposed a regulated transition: prices freed in stages, the plan’s commanding heights retained while the market grew around them, the union’s economic space preserved under a renewed treaty. A rival team, assembled by the Russian republic’s new leadership and the union president in uneasy collaboration, produced the Five Hundred Days programme: a rapid, sequenced leap to market relations, republican control of property and taxation, and a union reduced to a coordinating shell among sovereign economies. The union president commissioned the radical programme, recoiled from its political price, which was the union’s own demotion, and attempted a synthesis that satisfied neither camp; the government’s programme was adopted in diluted form and implemented in none. The episode is sometimes told as a failure of economic nerve. Its analytical value is greater. The programmes were not really rival techniques; they were rival constitutions wearing economic dress. The gradual programme presupposed a union centre strong enough to sequence a transition, and no such centre any longer existed, since the party that had been the centre’s skeleton was dissolving and the republics were confiscating the revenues. The radical programme presupposed republics willing to pool sovereignty in a new economic union of their own design, and the republics’ leaders, having acquired sovereignty as a weapon, valued it as a possession. Each programme therefore required, as its silent premise, the political order whose absence was the problem, and the war of the programmes is the record of intelligent men discovering that economics cannot be sequenced ahead of sovereignty. The systemic account predicts the discovery, since it locates the crisis in institutions rather than in ideas. The internal initiation frames it: the programmes’ authors were the stratum arguing with itself about the terms of its own conversion, gradual and compensated or rapid and competitive, and the dissolution settled the argument by abolishing the shared treasury over which it was fought.

The episode also corrects a sentimental misreading that the five canonical misreadings do not cover. It is often said that the union died because no one had a plan. The record shows the opposite embarrassment: plans in profusion, drafted by the system’s best economists, debated in its press with a freedom that would have been unimaginable five years earlier, and voted, leaked, and counter drafted through a whole autumn. What the system lacked was not plans but a sovereign capable of choosing among them and surviving the choice. Every plan’s implementation required an authority, union or republican, whose writ ran across the whole economic space, and by the time the plans were ready the writ had been divided among fifteen treasuries in the making. The lesson generalises beyond the case, and the article notes it as the economists’ war’s residue for the reader’s own use: in the collapse of a composite state, the binding constraint is never the quality of the transition programme. It is the prior question of whose signature the programme carries, and that question is settled by the politics of the stratum, not by the economics of the institutes.

The other solidarities: faith and memory beneath the official creed

A system that claims the whole of life leaves, when its claim thins, a landscape of suppressed solidarities waiting to be reactivated, and the collapse’s speed owed something to how many were waiting. The official creed had spent seventy years attempting to replace the older loyalties, of nation, of faith, of locality and kin, with a single Soviet identity administered from the centre. The attempt had real achievements to its name: generations were educated in its vocabulary, its rituals marked the life cycle from the October children’s organisation to the pensioner’s medal, and its identity was genuinely felt by many of its citizens, as the referendum’s result later testified. But beneath the administered identity the older solidarities persisted in the forms the system itself had licensed or failed to eradicate. The republics preserved nationhood in institutional form, as argued in the national discussion earlier in this article. The faiths preserved it in personal form: the Orthodox Church, compromised by its accommodations and persecuted in waves, retained the allegiance of grandmothers and the architecture of a thousand reopened parishes the moment toleration arrived; Islam in Central Asia and the Caucasus had survived as custom, kinship law, and the unregistered mullahs of the parallel society; Catholicism in Lithuania had fused with the national cause until the two were indistinguishable, and its Chronicle of current events, typed and retyped across two decades, was the dissident press’s most durable achievement. When the centre’s creed thinned to performance, these were the solidarities with living roots, and the politics of the ending flowed into them because they were there: dug channels, waiting for water. The Marxist schools, whose method taught them to look for the solidarities a society actually organises around, might have been expected to map the channels in advance. With honourable exceptions in the tradition’s national writings of an earlier generation, they did not, because their classification debate had trained them to see the union’s population as classes and strata, categories the official statistics themselves supplied, rather than as nations and believers, categories the official statistics suppressed. The collapse punished the omission in its geography and its speed alike, and the grading records it as the tradition’s most consequential failure of attention, the more instructive because the tradition’s own classical texts, on the national question above all, contained the instruments that its later schools declined to pick up.

The provinces: how the ending looked from outside the capitals

The literature of the collapse is a literature of capitals, of Moscow above all, and the bias is worth correcting in a section, because most of the stratum whose behaviour this article explains did not live in Moscow and did not experience the ending as a drama of summits. In the provincial city and the industrial town, the ending arrived as a sequence of quieter substitutions. The party committee that had governed the town in fact became, by a change of letterhead, the administration that governed it in law, and its first secretary became, by election or appointment, the mayor or the governor. The enterprise director who had bargained with ministries bargained now with banks and buyers, and was better placed than any newcomer to end as the owner. The collective farm chairman became the head of a joint stock company whose shares the workforce held and whose meetings he chaired. Nothing in the provinces resembled a revolution, and that was the point: the internal initiation, viewed from the oblasts, was a relabelling conducted by the incumbents, ratified by populations whose attention was fixed on wages, supplies, and the terrifying novelty of prices that moved. The conversion mechanism this article traced in the cooperatives and leasing arrangements of the capital had its provincial form in this seamless incumbency, and the provincial evidence is in one respect stronger than the capital’s, because it cannot be attributed to a conspiracy of reformers. No committee in Moscow designed the provincial relabelling. It happened everywhere at once because the incentives were everywhere the same, which is what a structural explanation claims and what a conspiratorial one cannot supply. The sequel in the provinces was correspondingly bleaker than the capital’s: the single factory towns, whose enterprise had been employer, landlord, clinic, and pension fund in one, discovered that the conversion of their director into an owner did not convert their town into a viable economy, and their long decline across the following decades is the transition’s true social history, written in depopulation rather than in the privatisation scandals of the capital. A grading of the explanations that rests only on Moscow will overstate the role of intention and understate the role of structure, since the provinces show the structure operating without intentions anyone recorded. The article’s balance, between the stratum’s choices and the situations that chose for it, was struck with the provinces in view.

The Baltic exception: where the crowds did lead

The article’s central claim has an exception, and the exception deserves examination rather than a footnote, because it tests the claim’s edges. In the Baltic republics, the ending was driven substantially from below: popular fronts mobilised hundreds of thousands, the human chain that linked the three capitals in the summer of the union’s antepenultimate year remains the era’s most reproduced image of peaceful mass action, and the republican leaderships were pushed by their movements at least as much as they led them. Three conditions made the Baltics different, and each condition confirms the general argument by specifying where it does not apply. The claim of interrupted statehood gave the movements a legal and historical frame no other republic possessed: the incorporation of the nineteen forties could be represented, with the secret protocols finally published, as an occupation to be ended rather than a union to be reformed. The republics’ small scale and cultural cohesion made mobilisation nearly total; a movement that can plausibly assemble a third of the nation has a physics unavailable to a movement in a vast federation. And the republican party leaderships, calculating as officeholders do, concluded earlier than their counterparts elsewhere that their futures lay in leading the national cause rather than in resisting it, so that even in the exception the stratum’s conversion, into national leadership, was the hinge on which mass pressure acted. The Baltic case therefore refines rather than refutes the internal initiation: where a ready made alternative sovereignty existed with a living memory of statehood, the initiative could be shared between the street and the office, and even there the office’s defection was the event that made the street’s victory irreversible. The grading consequence is a caution against the claim’s imperial form. The internal initiation describes the union’s centre and its Slavic core exactly, the Baltics partially, and Central Asia, where the stratum conserved itself by relabelling, in a third register again. An explanation is not weakened by knowing its own range. It is weakened by claiming a range the map refuses.

The dissident mirror: the regime’s enemies audited

Between the schools and the inside critics stands a population the grading has touched only in passing: the dissidents, the regime’s declared enemies inside the country, whose anticipations form a mirror in which the tradition’s record looks different and no better. The dissident spectrum ran from Sakharov’s liberal convergence, through the neo Slavophile and nationalist currents, to Aleksandr Solzhenitsyn, whose writings from exile prophesied the empire’s end from the right: in his Letter to the Soviet Leaders and his later essays he urged the regime to abandon Central Asia and the Caucasus as burdens, predicted that the union’s non Russian peoples would depart, and located the system’s fatal flaw in its ideology’s war on the nation’s moral and religious substance. The dissidents’ collective record on the event is the mirror image of the schools’. They foresaw the ending’s moral meaning with a clarity the schools never approached, having never doubted that a system built on lies would be abandoned the moment lying became optional; Sakharov’s essays read, in retrospect, like the legitimacy account written twenty years early by a participant. They foresaw its geography partially, the nationalists among them expecting the empire to shed its borderlands while keeping a Russian core, an expectation the Slavic core’s own defection refuted. And they foresaw its manner hardly at all, having imagined the end as a moral awakening, a repentance, or a national resurrection, rather than as a cascade of officeholders converting their warrants. The mirror’s lesson is the ledger’s lesson in another dress. Every camp that studied the system saw the facet its founding question illuminated: the schools saw the stratum’s economics, the dissidents saw the creed’s falsity, the demographers saw the cradles, the nationalities scholars saw the seams. The event, when it came, was the sum of the facets and the property of none, and the article’s method, grading each account on the cells where its question gave it sight, is the only response to that record that does not repeat the error of claiming the whole.

The absent Thermidor: why no saviour emerged

The tradition’s historical imagination, formed on the French Revolution’s script, expected the Soviet drama to produce its Thermidor and its Bonaparte: the reaction that stabilises a revolution by betraying it, and the general who crowns the reaction. Trotsky himself had mapped the bureaucracy onto the Thermidorian role, and generations of the tradition’s readers watched the Soviet succession for a man on horseback. The ending refused the script completely, and the refusal is diagnostic. No saviour emerged in the final crisis, though the emergency committee had the forms of one, because a Thermidor requires a settlement that a decisive minority is willing to impose and a majority is willing to accept as the price of order, and by the union’s final year no settlement commanded that combination. The committee offered the past without its belief; the republican presidents offered the future without a union; the reforming president offered a treaty without an enforcer. A Bonaparte is a solution to the problem of a revolution that has exhausted its believers and not yet found its form, and the Soviet case inverted the problem: the believers were exhausted, the forms were all present, and every candidate form had already been tried by the men now abandoning it. The general who might have played the role commanded an army whose officers were calculating their futures in the successor states, and the role itself, the saviour of the union, was the one part in the drama for which no actor auditioned, because the union was the one asset none of the stratum’s factions any longer wished to hold. The script’s failure is the internal initiation’s final confirmation. Revolutions produce saviours when a class or a cause still needs the state that the revolution built. The Soviet state ended unsaved because, by the end, nobody’s future required it: not the owners in waiting, not the national presidents, not the population with its private survival strategies, and not the army with its divided command. A state dies when its defence becomes nobody’s interest, and all the explanations in this article, whatever their other quarrels, are footnotes to that sentence.

The intelligentsia’s defection: the clerks of the creed changed sides

Every order keeps a clerisy, the stratum paid to produce its justifications, and the Soviet order’s clerisy, its writers, journalists, academics, and cultural officials, defected earlier and more completely than any other part of the establishment, a fact the explanations centred on economics tend to treat as decoration and the legitimacy account treats as the main event. The defection had a structure worth specifying. The cultural apparatus was the first zone the reform liberalised, deliberately: glasnost was aimed at the bureaucracy through the press, and the editors appointed to conduct it were the thaw generation’s own, men and women who had waited thirty years for the licence and used it past its intended limits within months. The thick journals, the system’s licensed literary monthlies, became the arena in which the buried past was exhumed, and their circulation figures, which this article’s protocol forbids it to quote, moved in a direction no one in the apparatus had anticipated, as a reading public formed around revelations the regime itself was publishing. The academics followed: the institutes whose internal memoranda had told the truth in camera now told it in print, and the economists and sociologists of the reform became public figures, converting their filed diagnoses into programmes. By the time the political struggle proper began, the creed had lost its producers. The party’s own ideology departments could still print the formulas, but the men and women who had once made the formulas plausible were writing for the other side, or for the market in revelation that glasnost had created. The grading consequence falls on the legitimacy account’s side of the ledger. Belief systems do not evaporate in the abstract; they are produced, daily, by identifiable people in identifiable offices, and they collapse when the producers stop producing. The schools that treated ideology as the reflection of deeper processes were watching the depths while the surface, where the reflection was manufactured, changed hands. A complete explanation of the collapse needs the clerisy’s defection in its sequence, between the leadership’s licence and the population’s indifference, since it is the mechanism that converted the summit’s decision to open the archives into the society’s conclusion that the creed had been a lie. The stratum initiated the ending in the ministries. Its clerks initiated it in the mind, and the two initiations were performed, characteristically, by the same generation for the same reasons.

The treaty that never was: the negotiation as evidence

The union’s final year contained a negotiation whose failure is among the most informative documents in the whole record, and the explanations should be graded against it explicitly. Through the spring and summer of that year, the union president and the leaders of nine republics conducted the process named for the presidential residence where they met, drafting a treaty for a renewed federation: sovereignty conceded to the republics, defence and a single economic space retained at the centre, the union re founded by consent rather than by the old constitution’s fiction. Drafts were initialled; a signing date was set for the late summer. The emergency committee struck on the eve of the signing, in the treaty’s name as much as in the system’s, and its failure destroyed the treaty’s remaining premise, which was a centre credible enough to be worth contracting with. Thereafter the republics that had negotiated in good faith drew the conclusion the committee had intended to prevent: that the centre was either too weak to guarantee a treaty or too dangerous to be trusted with one, and the final dissolution, signed in December by three republican presidents, was the negotiation’s corpse. The episode discriminates finely among the explanations. For the inevitability side, the treaty’s near completion proves too much contingency for their comfort: a re founded union was, months before the end, a document awaiting a signature, and its destruction required a coup, a counter mobilisation, and a defection, three events none of which the structure dictated. For the pure contingency side, the treaty’s terms are the embarrassment, since the draft conceded so much sovereignty that the union it would have preserved was already a confederation in substance, and the distance between the treaty and the dissolution is shorter than the drama suggests. The article’s grading takes the episode as the sequence’s clearest single exhibit for its whole method: structure set the treaty’s terms, making them confederal; decisions destroyed the treaty, making the outcome dissolution; and the stratum’s members, on every side of the table, behaved as men pricing futures rather than as men defending a faith. Whoever explains the collapse must explain why a negotiation that had produced a draft ended in a hunting lodge, and the explanations that route the answer through the initiators’ calculations are, once again, the ones the record rewards.

The core’s grievance: when the largest nation defected

The national analysis of the collapse usually begins with the borderlands, and the habit conceals the decisive defection, which was the core’s. The Russian republic was the union’s foundation and its anomaly: the largest republic had the fewest institutions of its own, possessing, across most of the union’s existence, no separate party organisation, no separate academy of sciences, no separate capital apparatus, because Russia and the union were treated as interchangeable and the union’s institutions were, in effect, Russia’s. The arrangement generated, across the postwar decades, a grievance literature of a distinctive kind, in which Russian writers, the village prose school above all, mourned a Russian countryside depopulated and a Russian culture subordinated to an internationalist project that drew on Russia’s resources for the whole. The grievance was available in two political colours. In its conservative form it indicted the union as Russia’s exploiter and dreamed of a national restoration inside a reformed state. In its oppositional form it fused with the democratic movement: the Russian republic’s new leadership, elected in the reform’s opening, built its platform on the claim that Russia was subsidising the union and being ruled by it simultaneously, and declared Russian sovereignty as the remedy for both. The declaration was the dissolution’s single most consequential act, because a union can survive the defection of its borderlands, as many empires have, and cannot survive the defection of its core, whose territory, population, and resources are the state. When the Russian government began withholding taxes, claiming the union’s assets on its territory, and negotiating with foreign powers as a sovereign, the union centre became a tenant in its own capital. The grading consequence extends the national section’s verdict. The schools underweighted not merely the borderlands’ nationalisms, which they might dismiss as survivals, but the core’s, which their frameworks had no category for at all, since a ruling nation’s nationalism against its own state contradicts the picture of the union as that nation’s instrument. The record shows the instrument’s handle turning in the hand: the stratum’s Russian majority found, in the republic’s sovereignty, the most valuable conversion vehicle of all, a state of its own to inherit. The internal initiation and the national question, the article’s two master facts, meet in that discovery, and an explanation that holds both is the one the balance sheet has been approaching throughout.

The party’s last congresses: an institution voting itself away

The party’s end, when it came, was procedural, and the procedure is worth a short section because it is the internal initiation photographed at the level of the institution. The party’s final congress, held in the reform’s fifth year, was broadcast to a country that watched its general secretary fight for a programme his own delegates half supported, watched the Russian republic’s leader announce his resignation from the party on the congress floor and walk out to greater authority outside it, and watched the organisation that had monopolised truth for seventy years discover that its procedures could not manufacture a majority for anything. Within a year the party’s leading bodies were dissolving their own structures faster than any opponent could have dissolved them: republican parties seceding to become national parties, regional committees converting themselves into administrations, the central apparatus negotiating the transfer of its buildings, its newspapers, and its funds. After the emergency committee’s failure, the party was suspended and then banned on Russian territory by decree of the Russian president, a former candidate member of its own Politburo, and the ban was accepted by the membership with a quiet that remains, for many historians, the single most astonishing fact of the entire collapse. Millions of members, an apparatus of full time officials in every town, an ideology department in every institution, and the organisation evaporated in weeks without a single local defence action of consequence. The evaporation is the legitimacy argument’s closing exhibit and the caste argument’s as well: the members did not defend the party because the party had long since become, for most of them, a career instrument rather than a faith, and a career instrument is discarded when the career moves elsewhere. The schools that had described the party as the caste’s skeleton described its end best. The skeleton did not break. It was unbolted, joint by joint, by the men who had been its bones, each carrying his joint into the successor arrangements where it might still bear weight.

The memory of the ending: nostalgia, indictment, and the uses of the past

The collapse did not end when the flag came down, because its meaning remained in production, and a short account of the memory belongs in a complete grading, since the explanations compete in the successor societies as living politics and not only as scholarship. The memory divided along the line the sequel’s evidence suggested. For the populations that bore the transition’s costs, the union was retrospectively revalued: surveys across the post Soviet decades, whose figures the protocol of this article forbids quoting, recorded persistent majorities in several successor states regretting the dissolution, and the regret attached less to the ideology than to the securities the old order had supplied, the predictable wage, the pension that arrived, the status of a superpower’s citizen, the absence of the new order’s humiliations. For the Baltic societies the same years consolidated the opposite memory, of an occupation ended and a statehood restored, institutionalised in museums and in law. For the successor elites, the memory was an instrument to be tuned: the Russian leadership of the union’s final crisis was retrospectively cast, by the governments that followed, first as the liberator from a totalitarian past and later, as the successor state’s own ideology hardened, as the naif or the accomplice of a geopolitical catastrophe, a phrase applied to the dissolution by a later Russian president in a formulation that entered the world’s press. The schools’ explanations circulate inside these memories as ammunition. The caste analysis is quoted by those who hold the conversion to have been a theft organised in advance; the systemic account by those who hold the ending to have been a necessity no leadership could have escaped; the national analysis by every successor state’s founding myth. A researcher entering the field should understand that the sources are still warm: the participants give interviews as combatants, the archives open and close with the political weather of the states that hold them, and the historiography’s third generation, described earlier in this article, writes under pressures the first generation would recognise. The grading offered here has tried to rest on the traces least subject to the weather, the dated texts, the public sequence, the internal series, and it offers its balance sheet in the knowledge that the memory war will go on grading the graders.

A revolution in reverse: the case that ran backward

The tradition graded in this article possessed, from its founding texts, a theory of how orders end: a ruling class that has outlived its function is overthrown by the class whose interests the new order embodies, in a crisis that concentrates decades of contradiction into weeks of decision. The Soviet ending ran the film backward, and the reversal is worth a final section because it explains the peculiar disorientation that marks even the best of the schools’ accounts. Every element of the classical script appeared in negative. The initiative belonged not to the rising class but to the incumbents. The ideology that collapsed was not the old order’s enemy but the old order’s own, discarded by its priests. The property transformation that followed was executed not by expropriating the rulers but by the rulers expropriating, in the legal sense of converting to their own ownership, the assets they had administered. The crowds, where they appeared, demanded not the future the script assigned them but cheaper goods, national sovereignty, or the punishment of the men in front of them. Even the violence was inverted: the classical script reserves its bloodshed for the old order’s defence, and this order declined to shed any. A tradition reading the event through its script kept misassigning the roles, expecting the bureaucracy to play the doomed aristocracy, complete with an emigration and a Vendée, when the bureaucracy was in fact playing the script’s other part, the class that carries out the transformation and founds the new order on the old one’s assets. The fork alone had glimpsed the possibility of the reversal, since a caste converting itself into owners is precisely a revolution conducted by the incumbents against their own system, and the glimpse is the deepest layer of the anticipation this article has credited. The lesson the case leaves for any theory of how orders end is austere and general: the agents of a transformation are whoever holds the assets and the information when the old legitimacy fails, and they are as likely to be the old order’s servants as its enemies. Scripts that assign the future to a class in advance will be surprised by every ending in which the servants get there first, and the Soviet ending, the largest of the century, was the servants’ work from the first memorandum to the last signature.

That is also why the case retains its hold on readers who have no stake in the tradition’s internal quarrels. The Soviet collapse is the best documented instance in the historical record of a great order ending without war, invasion, or insurrection, dissolved by the deliberations of the men appointed to preserve it, and every subsequent ending of a brittle order, in whatever region and under whatever banner, has been read against it by the participants themselves. The explanations graded in this article are therefore not antiquarian. They are the standing instruments for the next occasion on which a ruling stratum must decide whether its arrangements are worth defending, and the reader who has learned to grade them, cell by cell, against a known outcome is rehearsing a judgement that the future will examine under less forgiving conditions than a seminar. The table, the sequence, and the ledger are the rehearsal’s equipment, and they are now the reader’s.

For students and exam candidates

This article is built to be used under examination conditions, and its architecture is the revision strategy. Master the explanations tested table first: four rows, each with a mechanism, an expectation about who initiates an ending, an expectation about the class outcome, and a verdict from the record. A candidate who can reproduce that table from memory owns the skeleton of any answer on the collapse, whatever the question’s wording, because every stem on this topic is secretly asking for one row, or for the grading of two rows against each other. Hang the flesh on the skeleton in a fixed order: the sequence, in which political liberalisation preceded economic disintegration; the internal initiation, the named claim that grades the rows; the one anticipation, with its four specified elements and its four blanks; and the predictive ledger, stated honestly, since examiners reward the candidate who concedes the tradition’s failure as a prophecy and locates its success as an anatomy. Keep the explanations table as a set of cards in the VaultBook study notes, one card per row with its strain point on the reverse, and drill the real stems of past papers through the ReportMedic previous-year question explorer, because the stems recur in costume and the costume changes faster than the questions. Observe the figures discipline as an examination rule and not only as this article’s protocol: assert no growth rate, no price, and no spending share you cannot source, describe the deceleration directionally, and name the measurement dispute when you cite the slowdown. An answer that grades the schools against the initiation will outscore an answer that merely narrates the ending, in every room where this topic is set.

For teachers and seminar leaders

Teach this topic as an exercise in discriminating evidence, and the internal initiation is the instrument. Open the seminar by establishing the sequence from documents, reform, elections, the monopoly’s surrender, the failed emergency, the dissolution, and require the class to fix the order before any theory is admitted, since most errors on the topic are chronological before they are ideological. Then set the article’s central exercise: distribute the four explanations without their names, as pure claims about what the bureaucracy was, and have students derive, for each, an expectation about who would end such a system and what would happen to its rulers, before revealing the record and grading the derivations against it. The exercise teaches the difference between an explanation and a narrative more efficiently than any methodological lecture, because students discover in their own work that some claims about a stratum generate expectations and others merely decorate outcomes. Follow with the objection seminar, in which the Popper and Kołakowski charges are prosecuted by students assigned to make them unanswerable, and defended only with the one anticipation, stated with its blanks attached. Close with the national question as the seminar on blind spots: have the class identify which drawer of the tradition’s cabinet the federation was filed in, and what their own favoured framework is filing away unexamined. The topic’s gift to a teacher is that its outcome is known and its theories were in print beforehand, a combination almost no other subject offers; use the combination ruthlessly, and mark down any essay, however elegant, that grades a school by its intentions rather than by its expectations.

For researchers, historians, and journalists

The research frontier on this subject is no longer the sequence, which the archival generation settled in outline, but the mechanisms at republican and enterprise level, where office was turned into property and the sovereignty cascade was actually executed and where the records are least centralised and least translated. Three cautions from this article’s grading apply directly to new work. First, respect the measurement dispute in everything quantitative: the divergence between the official series, the Western reconstructions, and Khanin’s alternative accounts is not a problem to be averaged away but a finding about the system’s self knowledge, and any figure quoted without its accounting concept and its construction should be treated as unquoted; the companion study of the economic record of the socialist states assembles the rival series with their health warnings. Second, read the schools’ founding texts as dated commitments rather than as positions in a timeless debate; the fork’s evidentiary value depends entirely on its date, and the genealogy of the classification dispute, from the exile arguments of the nineteen twenties through the schisms around Trotskyism and Stalinism, is the indispensable context for grading any anticipation claim. Third, treat the memoir literature as the victors’ archive problem in its purest form: the stratum that might have defended the system wrote almost nothing, the liquidators and critics wrote shelves, and the asymmetry should be priced into every inference about motive. Journalists covering the anniversaries should note, finally, that the participants’ incentives have not retired with them; the conversion’s winners and losers still brief, and their briefings are data about the sequel, not about the event.

Frequently Asked Questions

Q: Why did the Soviet Union collapse?

Three causes interlocked, and removing any one of them breaks the explanation. A way of growing that depended on endlessly adding workers and capital ran out of things to add. The men who inherited that exhaustion opened the political system faster than they rebuilt the economy, so authority drained away before new coordination existed. And the officials who staffed the state discovered prospects beyond it that paid better than loyalty, so when the crisis peaked nobody with power chose to save the arrangements. Stagnation made reform unavoidable, sequencing made reform fatal, and self interest made the fatality unopposed. Single cause stories, whether they blame oil, armaments, or one leader, mistake a component for the machine.

Q: Who initiated the Soviet collapse?

The initiative sat at the top of the state from its first internal memorandum to its final dissolution decree. Reform was drafted at the summit, staffed by the summit’s nominees, and passed by the summit’s own parliament. The elections that humbled the apparatus happened because the apparatus permitted them. The party surrendered its guaranteed monopoly by a vote of the parliament it controlled. Republican presidents, every one of them formed by the system, signed the final dissolution. Popular movements supplied pressure in the Baltic republics and crowds at the barricades during the August emergency, but they were answering openings created above them. When citizens voted in a referendum to keep a renewed union, the elites dissolved it anyway. Direction of initiative is the first fact any answer should state.

Q: What is the difference between extensive and intensive growth?

One grows by quantity, the other by quality. Extensive expansion adds inputs: peasants moved into factories, women drawn into employment, fresh mines and fields opened, investment poured into plant. Intensive expansion squeezes more from unchanged inputs, as better methods, machines, and skills raise what each worker yields. Command administration is superb at the first, because mobilisation can be ordered, counted, and enforced. It is feeble at the second, because improvement must be adopted voluntarily in thousands of workplaces whose managers were punished for the disruption new methods cause. The Soviet economy consumed its reserves of labour and resources across five decades, then faced a transition its own institutions blocked, and the slowdown recorded by every statistical series, official or recalculated, was the sound of that blockage.

Q: How do state capitalist accounts explain the collapse?

By denying that anything fundamental ended. In this reading the bureaucracy had always been a capitalist class in collective form, piling up capital under pressure of military rivalry with the West, and the workers had always been wage labourers confronting owners they did not choose. What finished in the final years was merely one organisational shell of that class rule, swapped for open markets in which the same networks could operate as private proprietors. Continuity is the school’s exhibit: directors kept their factories, connections kept their value, subordination kept its shape. The embarrassment is equally plain. Classes that own the means of production are not supposed to vote their own ownership structure out of existence, and rescuing the account requires arguments that bring it close to the caste theory it was built to destroy.

Q: Did the arms race cause the collapse?

It loaded the dice without throwing them. Defence consumed the choicest engineers, materials, and machine tools for forty years, and the renewed competition of the final decade raised the cost of parity precisely when civilian stagnation could least afford it. Yet the burden had been heavier, relative to a smaller economy, in earlier decades that produced no collapse, and no Soviet army was beaten in the field. The race worked by persuasion: it taught the leadership that a model which could match missiles but not consumer goods had failed on its own terms. External pressure set the exam; it did not write the answers, choose the candidates, or sign the surrender. Explanations that stop at the arms race cannot say why the ending took the form of dissolution from within.

Q: Did oil prices matter to the collapse?

They decided the timing, which is not the same as deciding the outcome. Energy sales through the nineteen seventies brought in the foreign currency that purchased imported grain and Western machinery, quietly subsidising the system’s weaknesses and purchasing a decade of postponement. When world prices dropped steeply in the middle of the following decade, the subsidy vanished while the import habit remained, borrowing bridged the gap briefly, and the new leadership found the treasury bare at the moment it contemplated reform. A fiscal emergency concentrates minds wonderfully, and this one fixed the season in which desperation became policy. But commodity prices cannot draw maps. They cannot explain fifteen successor states on federal borders, a party evaporating undefended, or officials racing to become owners.

Q: Why do Marxist accounts underweight the national question?

Their founding quarrel trained them to look elsewhere. The schools were born arguing about what the bureaucracy was in relation to property: a caste squatting on nationalised wealth, a capitalist class in state dress, or a new ruling order altogether. Factories, plans, and surpluses filled their field of vision, while nationhood was filed under an older controversy the founders were assumed to have settled. Republican borders looked like administrative decoration on a unitary state, and the institutions inside those borders, parties, academies, cultural unions, were read as transmission belts rather than as states in waiting. When the centre faltered, the decoration turned out to be the architecture. The blindness was inherited, not stupid: traditions see what their originating questions illuminate, and stumble over the furniture those questions left in darkness.

Q: Did anyone predict the collapse?

The event itself, nobody: not the year, not the peaceful manner, not the six year speed, not the map of fifteen successors. Fragments of foresight were scattered across the camps. Trotsky had described, two generations early, a bureaucracy tempted to turn administrative privilege into hereditary property, and got the agent and the motive while missing the route entirely. A French scholar of nationalities foresaw the empire cracking along its ethnic seams; a dissident essayist foresaw disintegration and attached it to a war that never happened; a demographer read doom in infant mortality figures. Meanwhile the professional watchers, Western institutes and intelligence services alike, budgeted for permanence. Fragility was widely visible. The specific death, at that hour, in that shape, by those hands, surprised every observer on earth.

Q: Was the collapse inevitable?

Separate the layers before answering. That growth would falter was built into a model that lived by adding inputs and throttled the improvements that might have replaced them. That faltering would become a crisis was likely once easy energy money ended. That the crisis would end in dissolution, along republican borders, peacefully, within a single year, was chosen step by step: in the decision to free politics before markets, in conceding republican elections, in a coup attempted by men who could not shoot, in three presidents preferring dissolution to a renegotiated federation that lay drafted on the table. Elsewhere in the same years a comparable party state picked the opposite sequence and endured. Necessity governed the pressure. Contingency, working through the ruling stratum’s choices, governed the ending.

Q: Do Marxist and mainstream explanations converge?

On how the system died, largely; on what its death means, not at all. A Marxist economist of the systemic school and a Western economic historian will give you the same skeleton: mobilisation exhausted, productivity blocked by perverse incentives and guaranteed rescues for failing enterprises, energy income collapsing as trigger, political opening racing ahead of economic rebuilding, and a leadership whose decisions turned breakdown into dissolution. The most respected mainstream narrative even describes an establishment unmaking itself, which is this article’s central claim in different vocabulary. The quarrel resumes over the corpse’s identity: whether a distorted socialism, a disguised capitalism, or a novel formation expired, and whether planning itself stood condemned or only one historical specimen of it. Agreement about mechanisms, disagreement about meanings, is the honest summary.

Q: How does the internal initiation grade the rival explanations of the collapse?

It charges each explanation a different price for the same fact, the fact that the governors began the ending. Trotsky’s caste pays least: men holding revocable office rather than inheritable wealth were always liable, in his account, to trade the system for property, and the trade is what the record shows. The collectivist new class pays more, explaining its desertion by the formation’s competitive failure. The systemic account pays nothing and gains nothing, since in a centralised order only the centre can start anything. State capitalism pays most, because a class genuinely owning the state should have fought for its own arrangements. Grading by this price produces a ranking of comfort, not a knockout, and rankings of comfort are what honest comparison of theories usually yields.

Q: What should a student separate between a system that was overthrown and one dismantled by its own stratum?

The direction of the evidence. Overthrow points outward, toward the rebels: their complaints, their organisation, their victory over the soldiers. Self dismantling points inward, toward the officeholders: the nature of what they held, the insecurity of holding it, the exits available when holding stopped paying. Confuse the two and every subsequent judgement wobbles, because you will hunt for insurrectionary heroes in a story whose protagonists carried briefcases, and you will read a referendum for preservation, overridden by elites, as a people’s verdict. In answers, declare the direction early. The Soviet case had crowds, strikes, and barricades, but they decorated a procession led from the summit, and the student who keeps that order straight can deploy every other fact in the topic without contradiction.

Q: Why did no ruling stratum defend the arrangements that constituted its power?

Because what the arrangements gave its members could not be kept, inherited, or sold, and the reform years opened markets in which position could be swapped for things that could. Loyalty promised a lifetime of revocable privilege inside a system its own chief had pronounced failed. Defection promised the director his factory as property, the official his network as capital, the republican boss a presidency, and everyone an escape from reciting a creed that had become pure ceremony. When the emergency committee finally ordered a defence, it discovered that the army was split, the party’s command chain had dissolved, and its own constituency had one foot out of the door. Men defend what they own. This stratum, at the decisive hour, concluded that it owned nothing worth the risk.

Q: How does the explanations tested table order the mechanisms?

Chronologically, by the birth of each analysis, because an explanation’s claim on our respect depends on whether it committed itself before or after the outcome was known. The caste analysis leads, its founding argument dating from the nineteen thirties, when the system’s permanence seemed assured to almost everyone else. State capitalism follows from the nineteen forties schisms, bureaucratic collectivism beside it from the same quarrel, and the systemic economic account closes the table as the postwar decades’ joint product of scholars inside and outside the tradition. Every row answers identical questions: the mechanism proposed, the initiator the mechanism leads you to expect, the class result it leads you to expect, and the verdict the record returns. Disagreement with the grading is welcome at the level of the individual cell, which is where the table invites it.

Q: What separates predicting a collapse from explaining one after it happened?

Exposure to being wrong. A genuine forecast is written before the event, names who will act and why, and can be refuted by what occurs; the fork in Trotsky’s analysis earns its partial credit exactly because it was printed half a century early with its reasoning attached. A retrospective explanation faces no such discipline: its author can promote the facts that fit, demote the awkward ones, and redescribe every surprise as an implication, and the journals of the following decade show every school performing some of that tailoring. The practical test for readers is to ask what a theory’s holder would have written in the year before the ending. Accounts that survive the question deserve trust. Accounts that dissolve under it are history written backward, however elegant their prose.