The question whether Marxism has a theory of democracy is usually asked in a form guaranteed to produce a useless answer. Asked as a loyalty test, it invites a verdict on a whole tradition, delivered in a sentence, and the sentence tells the reader more about the speaker than about the tradition. Asked precisely, it divides into two questions with opposite answers, and the division is the finding. The first question is whether Marxism developed a serious account of how democracy operates, and fails to operate, in societies where productive wealth is privately owned and heavily concentrated. The answer to that question is yes, in unusual depth and with arguments that political science has repeatedly absorbed without always acknowledging the source. The second question is whether Marxism developed a comparably serious account of the political institutions of the society it proposed: how such a society would make binding decisions, protect dissenters, organise disagreement, correct its own rulers, and remove them. The answer to that question is no, and the no is not a hostile invention. It was stated most precisely by a socialist political theorist writing inside the European left, debated openly by Italian Marxists in the 1970s, and partly conceded by the strongest Marxist respondent of that decade.

This article works through the second question as carefully as the first, because the imbalance between them is the subject. The pattern has a name in what follows: the asymmetry of the tradition. On one side stands a critique of liberal democracy of genuine analytical power, built from the distinction between political and human emancipation, from the analysis of formal equality coexisting with material inequality, and from a body of work on how ownership constrains what elected governments can do. On the other side stands a short inventory of institutional proposals, headed by the Paris Commune model of 1871, followed by long silences on the subjects where a theory of democracy is normally expected to speak: political rights held against the majority, the legitimacy of organised opposition, the separation and mutual checking of powers, and the procedure by which a post-capitalist society would settle the disagreements that would certainly arise inside it. The charge that these silences constitute a gap is examined component by component, the strongest replies are given their full force, and each component receives a verdict. Nothing in the inquiry requires a judgement about which political system a reader should prefer, and none is offered. The question is narrower and more answerable: does the gap exist, how large is it, and what have the tradition’s own best minds done about it.

A note on sources and fairness before the argument begins. The charge examined here is not the newspaper version, in which Marxism and democracy are simply opposites and the discussion ends. The serious charge was made by Norberto Bobbio, an Italian theorist of democracy who identified as a socialist, addressed it to Marxists as an internal problem, and framed it as a series of specific, answerable questions about institutions. The strongest reply was made by Nicos Poulantzas, a Marxist theorist who, in his final book, accepted a large part of the charge and rebuilt his position around representative institutions combined with direct democracy. An article that treats the exchange between these two as its centre of gravity is not conducting a prosecution. It is following the most honest conversation the tradition has had with itself on the subject, and reporting where that conversation arrived and where it stopped.

What is actually being asked about democracy

The word democracy carries at least three distinct loads in this debate, and much of the confusion surrounding the charge comes from speakers sliding between them mid-argument. In the first sense, democracy names a set of procedures: universal suffrage, competitive elections, freedom of association and expression, and rules by which the losers of one election can become the winners of the next without violence. In the second sense, democracy names a social condition: a rough equality of effective power among citizens, such that formal rights translate into comparable influence over collective decisions. In the third sense, democracy names an ideal of self-government, in which the people subject to decisions are, in some meaningful way, their authors. Liberal political theory concentrates on the first sense and treats the second and third as aspirations the procedures serve. The Marxist critique concentrates on the second sense and argues that the first, under concentrated ownership, delivers far less of it than its own self-image claims. The charge examined in this article concerns a fourth matter that neither sense covers: the institutional design question. Whatever a society’s procedures and whatever its distribution of wealth, somebody must specify the offices, the rules for filling and emptying them, the rights that limit what majorities may do, and the channels through which disagreement is expressed and processed. A theory of democracy, in the sense the charge demands, is an answer to that design question.

Does Marxism have a theory of democracy?

It has a well-developed critique of liberal democracy’s limits under concentrated ownership, and a thin account of the institutions that would replace it. The Commune model of elected recallable delegates is the main proposal. Critics argue the absence of a theory of opposition and rights is a serious gap.

That short answer is the whole article in miniature, and its two halves deserve equal patience. The critique half is sometimes dismissed by the tradition’s opponents as mere rhetoric, the word bourgeois attached to democracy as a spell to dissolve it. That dismissal will not survive contact with the arguments themselves, which are specific, largely empirical, and in their modern form shared by scholars with no Marxist commitments. The design half is sometimes dismissed by the tradition’s defenders as a pedantic demand, as though asking what institutions a socialist society would have were a category error, a request for blueprints that serious theory does not supply. That dismissal will not survive contact with the tradition’s own history either, because the demand was pressed hardest from within, by theorists who wanted the design work done and were frustrated that it had not been. The sections that follow take the halves in turn: first the critique at full strength, then the charge in its five components, then the replies, then the ledger.

One preliminary distinction disciplines everything after it. The theory of the state in general, the question of what the capitalist state is and whose interests it serves, is a separate subject with its own literature, and it is treated in the companion study of the marxist theory of the state guide. The democracy question overlaps that literature but is not identical to it. A theory can explain, even persuasively, why the existing state behaves as it does, and still say almost nothing about how a different state should be constituted. Indeed the sharpest form of the charge is precisely that the explanatory theory was allowed to substitute for the constitutional one: because the tradition believed it understood what the state was, it neglected to design what its successor would be. Keeping the two questions separate is the first act of fairness the subject requires, in both directions. The strength of the explanatory theory does not answer the design charge, and the design charge does not diminish the explanatory theory. Each must be weighed on its own evidence.

The critique of liberal democracy, stated at full strength

The critique begins with an essay of 1843, On the Jewish Question, in which the young Marx drew the distinction that organises the entire tradition’s thinking about rights and citizenship. Political emancipation, in his analysis, is the achievement of the modern state: the granting of equal citizenship, equal suffrage in principle, and equal standing before the law, regardless of religion, property, or birth. Marx was emphatic that this was a real achievement, a great progress, and the final form of emancipation possible within the existing order. His argument was about its limits. The state, he observed, can abolish the political significance of private property, of religion, of occupation, while leaving all of them fully operative in civil society. The citizen is declared equal in the heaven of the political sphere and remains unequal on the earth of economic life, and the split between the abstract citizen and the concrete property owner is not an oversight of the modern state but its constitutive structure. Political emancipation liberates the state from property and religion without liberating human beings from the power that property and, in his framing, confessional division exercise over their actual lives.

Human emancipation, the contrast term, named the condition in which the split itself would be overcome: when the individual no longer separates his social power from himself in the form of a political power standing over him, and when his actual relationships, of work and need and cooperation, are organised as his own collective life rather than mediated by abstractions. Whatever one makes of that destination, and later sections will press hard on its institutional emptiness, the analytical point embedded in the distinction retains force independently. Rights and votes are exercised by people who return each evening to workplaces they do not govern, to incomes set by markets they do not control, and to information supplied by media they do not own. A theory that counts only the moments of citizenship, the ballot and the courtroom appearance, as politically real is measuring an intermittent ritual and calling it self-government. The critique asks what happens between elections, in the economy, and it insists that the answer is a political fact of the first order, not a private misfortune outside politics altogether.

Did Marx oppose democracy itself?

Marx did not oppose democracy as such, and his earliest journalism defended press freedom and representative institutions against Prussian censorship. His target was the claim that political equality completes emancipation while economic subordination continues untouched. The complaint was that democracy in its liberal form stops at the threshold of the workplace and the market.

The textual record supports that reading without strain. The young Marx of the Rheinische Zeitung wrote against censorship with a liberal’s vocabulary and a radical’s energy. The older Marx praised the Paris Commune in the language of political form, as the political form at last discovered for working out economic emancipation. His polemics against fellow socialists were frequently polemics against their impatience with democratic means: against conspiratorial minorities who would substitute themselves for a majority movement, and against insurrectionary traditions that treated universal suffrage as a trap rather than an instrument. What he opposed, consistently, was the doctrine that the political sphere is self-sufficient, that a society has finished the work of freedom when it has equalised the vote. That doctrine he regarded as the ideology of the existing order in the strict sense his theory gives the word: a true description of one sphere, the political, generalised into a false description of the whole.

The mature version of the critique shifted from the essay’s philosophical register to an institutional one, and the Communist Manifesto supplied its most quoted sentence, describing the executive of the modern state as a committee for managing the common affairs of the whole bourgeoisie. The sentence is usually read as a conspiracy claim, as though ministers take instructions from boardrooms. Read in context, and read against the whole of Marx’s later work on France, it is a structural claim of a different and more interesting kind. The state, on this account, does not need instructions, because its personnel, its fiscal dependence, its legal framework, and its own interest in order align it with the class that owns the productive wealth of the society it governs. Marx’s journalism on France complicated the picture further, showing a state apparatus acquiring autonomy from every class, balancing among them, and ruling through a bureaucracy and an army that became ends in themselves. The tradition inherited both versions, the instrumental and the structural, and the tension between them generated a rich internal literature. What matters for the democracy question is the shared premise: political institutions are not neutral instruments equally available to whoever wins office. They are shaped, staffed, financed, and constrained by the economic order within which they sit, and any theory of democracy that ignores that shaping is describing the rules of a game while declining to ask who owns the board.

The critique’s second layer is empirical, and it concerns the mechanisms by which material inequality converts itself into political inequality without any law being broken and without any official acting corruptly. The first mechanism is the unequal distribution of the resources that political participation consumes. Voting is cheap; almost everything else in politics is expensive. Standing for office, funding a campaign, sustaining an organisation between elections, litigating, lobbying, commissioning research, and buying the attention of the media all require money, time, education, and confidence, and all of these are distributed along the same gradient as wealth. A political science that measures only the act of voting will report equality. A political science that measures who sets the agenda from which voters choose will report something else. The Marxist critique insists that the second measurement is the politically significant one, and on this point the mainstream literature has substantially come to agree, whatever vocabulary it prefers.

The second mechanism is the filtering of candidates and programmes before the electorate is consulted. In a system where campaigns are privately financed and media access is privately sold, the range of candidates with a realistic path to office is narrowed long before polling day, narrowed not by conspiracy but by the arithmetic of fundraising and coverage. Programmes that threaten the fundamental interests of large property rarely survive the filter with organisational backing intact, and those that do face a campaign environment in which the cost of alarming investors, advertisers, and proprietors is immediate and concrete, while the cost of disappointing unorganised voters is diffuse and deferred. The third mechanism operates after elections, on governments in office. A government depends for employment, tax revenue, and its own re-election prospects on the investment decisions of private owners, and those decisions are made continuously, without any vote, by actors responding to their assessment of the government’s policies. The result is a standing veto exercised not through parliament but through the bond market, the investment strike, and the flight of capital. None of these mechanisms requires secret coordination. All of them follow from the institutional separation between a democratic state and a privately owned economy, which is exactly the separation the young Marx identified as the structure of political emancipation.

What does bourgeois democracy mean in Marxist usage?

In Marxist usage the phrase names a democratic form whose content is limited by the bourgeoisie’s economic dominance, not a pretence or a ritual. Elections are real, rights are real, and governments do change. The claim is that the range of outcomes open to those procedures is bounded by ownership, and that the boundary, not the procedure, is the argument.

The phrase matters because its misreading powers the stalemate in public discussion. If bourgeois democracy meant fake democracy, the Marxist position would be easy to state and easy to refute, since the reality of electoral competition in liberal states is not in serious doubt. The tradition’s serious writers meant something more precise and harder to dismiss: that the class character of a society is located primarily in its relations of production, that the political form grows within limits those relations set, and that describing a state as democratic answers the question of its procedures while leaving the question of its social content untouched. Critics of the tradition are entitled to reject that relocation of the decisive question, and many do, arguing that procedures are not a shell but the substance of political freedom. What they cannot fairly do is refute the relocation by pointing at the procedures, since the procedures were never denied. A large share of the polemical literature on both sides consists of exactly that equivocation, and the ledger method adopted later in this article is designed to prevent it.

The fourth layer of the critique is the one the tradition itself considered decisive, and it concerns what happens to a democratic government that attempts, in earnest, to use its majority against the interests of concentrated ownership. The twentieth century supplied cases that every participant in the debate interprets and none can ignore. Elected governments pursuing expropriation or deep redistribution have faced capital flight, investment collapse, currency crises, hostile media campaigns, bureaucratic obstruction, and in several documented instances military intervention prepared or encouraged by domestic and foreign opponents of the programme. The Chilean experience of the early 1970s, in which an elected government pursuing a constitutional road to socialism was destroyed by a military coup after a period of economic disruption and institutional conflict, stands in the literature as the limiting case: a government that observed the procedures and was removed anyway. Defenders of the liberal order reply that such cases involve governments exceeding their mandates, wrecking economies, or provoking the very crises that destroyed them, and that other elected governments have enacted deep reforms, built welfare states, and nationalised industries without coups. Both readings are argued from the same record, and this article will not pretend the record speaks with one voice. What the record does establish, at minimum, is that the constraint is real enough to be the central strategic problem of any democratic socialism, which is what the tradition’s own internal debates, from Bernstein onward, always recognised it to be.

What is structural dependence, and why does it matter here?

Structural dependence names the constraint on governments that arises from their reliance on private investment for employment and revenue, modelled formally by economists in the decades of the twentieth century. It matters because it converts the Marxist critique of democracy from a slogan into a mechanism that scholars outside the tradition have stated in their terms and tested against evidence.

The formal literature deserves a paragraph of its own, because it marks the point where the critique of liberal democracy achieved something its institutional counterpart never did: a specification precise enough to be wrong. Writers such as Fred Block argued that state managers, whatever their own class origins or sympathies, are structurally compelled to maintain business confidence, because the performance of the economy on which their tenure depends is in private hands. Adam Przeworski, writing from within the socialist tradition and against its consolations, modelled the dilemma of democratic socialist parties as a set of trade-offs imposed by that dependence: a party that frightens capital before it has transformed ownership will preside over the crisis its opponents predict, and a party that reassures capital may find it has governed as capital’s manager and transformed nothing. These arguments were contested, refined, and in part absorbed into the ordinary equipment of comparative politics. Note the pattern, because it is the asymmetry in embryo. Where the tradition’s claim concerned the constraints on existing democracy, its best minds produced mechanisms, models, and testable propositions. The corresponding pages, the ones that should specify the institutions of the society beyond those constraints, were not written with anything like the same energy. The charge is about those missing pages.

There is a fifth layer, quieter than the others, that concerns the formation of the citizens themselves. Democratic procedures aggregate preferences; they do not ask where preferences come from. If the institutions that form opinion, the press, the school, the workplace hierarchy, the inherited common sense of a society, are themselves shaped by unequal wealth and unequal authority, then the equality of the ballot aggregates an inequality manufactured upstream. Gramsci’s concept of hegemony gave this layer its classical statement: the dominance of a class is reproduced not only by coercion and not only by law but by the organisation of consent, by a whole civilisation of assumptions about what is natural, possible, and respectable. The concept has been stretched to cover almost everything, and in careless hands it becomes unfalsifiable, a universal solvent for electoral defeats. In careful hands it poses a genuine question that procedural theories of democracy have never fully answered, namely what the procedures are worth when the agenda, the information, and the range of the thinkable are set before the procedures begin. A theory of democracy confined to counting votes has no reply to that question. A theory that begins from it still owes an account of what institutions would do better, and it is at exactly this point, again, that the Marxist account falls silent. The critique is a searchlight. The charge is that the tradition never built the house the searchlight was supposed to illuminate.

The charge in its five components

With the critique at full strength, the counter-charge can be stated with the precision it deserves. It is not one accusation but five, and they differ in kind. Some allege an absence in the theory, some allege a positive commitment hostile to democratic institutions, and the fifth alleges that history has already run the experiment. Conflating them produces the familiar shouting match, in which a defender answers the easiest component and declares the charge refuted. The ledger method forbids that move. Each component is stated here in its strongest form, and each will receive its own verdict at the end.

The first component is the substitution charge. Marxism, the critics observe, developed an elaborate critique of the institutions it inherited and an elaborate theory of the transition it expected, the conquest of state power by the organised working class, yet between the transitional moment and the final horizon of a stateless society there is an institutional void where a constitutional theory should be. The texts specify what must be destroyed in unusual detail: the standing army, the appointed bureaucracy, the parliamentary talking shop, in Lenin’s phrase of 1917, the whole bureaucratic and military machinery of the bourgeois state. They specify what must replace it in a few celebrated paragraphs drawn from the Paris Commune, examined closely in a later section. And they specify almost nothing about the enduring questions of government in a complex society: how legislation is drafted and scrutinised, how administrators are supervised, how disputes between regions, industries, and generations are adjudicated, how a plan is corrected when it fails, how minorities within the working class itself are protected from majorities of it. The tradition’s answer, that the details would be worked out in practice, is treated by the critics as the confession the charge requires: every other political tradition, liberalism included, did its constitutional thinking in advance, on paper, in argument, and the refusal to do so was itself a theoretical position with consequences.

The second component concerns rights. Liberal democracy rests, at its foundation, on rights held by individuals against the community: rights of speech, association, conscience, and due process that no majority may vote away and no emergency may suspend without limit. The Marxist tradition’s canonical treatment of rights is the 1843 essay already discussed, and its thrust is diagnostic and deflationary: rights are analysed as the legal form of egoistic man, of the isolated property owner, and as destined to be surpassed rather than entrenched in a society beyond class division. The critics ask a question the texts never squarely answer. In the society the tradition proposes, what protects the individual dissenter, the unpopular minority, the citizen accused of an offence, against the collective in whose name the state acts? If rights are bourgeois in form, what is their successor form, and who enforces it against the majority’s own organs? It is not enough, the critics argue, to predict that in a classless society such protection will be unnecessary, because the prediction concerns the end of a long transition during which the protection is needed most, and because the prediction itself cannot be enforced by anyone. A theory that cannot say what a citizen may claim against its own state has, on this component, not a gap at the margin but a hole at the centre.

The third component concerns opposition. In a liberal democracy, organised opposition is not a tolerated nuisance but a constitutional institution: the alternative government, with recognised status, resources, and a legitimate path to power. Its legitimacy rests on a simple principle, that disagreement about policy and about the direction of society is permanent, reasonable, and entitled to organisation. The Marxist tradition, the critics argue, has no place for that principle. Its theory locates political disagreement in class antagonism, from which it follows that after the abolition of classes the ground of fundamental disagreement disappears, and organised opposition to the society’s direction can thereafter be interpreted only as the residue of the old order, as error, or as sabotage. The inference is rarely stated so baldly by the tradition’s serious writers, and individual Marxists have affirmed pluralism in terms. But a principle that must be imported from outside the theory, and that the theory’s own logic tends to dissolve, is not a principle the theory possesses. The critics’ question is exact: in the proposed society, may citizens organise to replace not merely the government but the system itself, campaigning, publishing, and standing for office on that programme? If yes, the society contains the permanent possibility of its own reversal and must be designed for it. If no, the society has answered the democracy question in the negative. The classical texts, the critics conclude, never face the question in that form.

The fourth component concerns the division of powers. Constitutional theory since Montesquieu has treated the separation of legislative, executive, and judicial functions, and their mutual restraint, as the elementary precaution against tyranny. Marx, following his reading of the Paris Commune, explicitly rejected the precaution. The Commune, he wrote, was to be a working body, executive and legislative at the same time, and the tradition took the fusion of powers as a positive principle, the institutional expression of the people’s undivided sovereignty against the fragmentation it associated with bourgeois constitutionalism. Lenin’s State and Revolution of 1917 carried the principle into the programme of a party that would soon govern. The critics’ case is that the precaution exists because of a fact about power that class analysis does not dissolve: those who make the rules, apply the rules, and judge disputes under the rules, when they are the same people, face no institutional resistance when they err or when they abuse. A judiciary appointed, paid, and removable by the executive it is asked to restrain is a decoration. The tradition’s counter-claim, that accountability flows upward from electors through recall rather than sideways through rival institutions, is a real alternative theory of control, and it will be weighed as such. The critics reply that recall by a mobilised majority is no protection for the minority the majority is mobilised against, and that the whole history of constitutionalism is the discovery that control must be institutional and impersonal precisely because majorities are passionate and temporary. That discovery, they argue, the tradition unlearned on purpose.

The fifth component is the appeal to the record. Between 1917 and the end of the twentieth century, a large number of states were governed by parties claiming the Marxist tradition, across several continents and at very different levels of economic development, and in none of them did competitive elections, an independent judiciary, a free press, or a legal opposition survive for long. The critics argue that a pattern so uniform across such different conditions cannot be explained by circumstance alone; that when a theory’s institutional silences are filled in practice by the same arrangements every time, the arrangements are evidence about the theory; and that the tradition’s defenders, by attributing every case to encirclement, backwardness, or betrayal, protect the theory from the only test history offers. The defenders answer with the strongest reply examined later in this article: that no theory is fairly judged by the worst regimes claiming its name, and that liberalism’s own centuries include colonial conquest, slavery’s long protection, and the disenfranchisement of women and the propertyless by states calling themselves democracies. That reply is serious and receives its section. But note what it establishes even if it succeeds entirely. It defends the critique of liberal democracy and the moral standing of the aspiration. It does not supply the missing institutions, which is what the first four components ask for. The record component therefore functions, in the ledger, less as a verdict on the theory than as a measure of the risk the silences carry: the risk that a movement armed with a magnificent critique and no constitutional design will improvise its institutions under pressure, and that improvisation under pressure has a known direction.

Bobbio’s version: the hardest form of the charge

The charge reached its precise form in Italy in the middle of the 1970s, and its author matters to its force. Norberto Bobbio was a professor of political philosophy at Turin, a theorist of democracy and of law, a man of the socialist left who had spent a career arguing that democracy’s rules were not a bourgeois trick but a precious and fragile achievement. In 1975 and 1976, in essays published in the journal Mondoperaio and in exchanges with Communist interlocutors, he posed a set of questions that became famous under a rubric the debate itself supplied: the Marxist theory of the state did not exist, he argued, in the form the tradition needed. What existed was a theory of the capitalist state, of its origins, its class character, and its extinction. What did not exist was a theory of the socialist state as a set of institutions: of the forms through which power would be exercised, limited, and transferred in the society after capitalism. Bobbio’s point was not that Marxists had failed to win arguments about democracy. It was that there was, strictly speaking, no Marxist doctrine of government to argue about, only a doctrine of the enemy’s government and a horizon beyond government altogether.

What makes Bobbio’s version the hardest form of the charge?

Bobbio was a socialist addressing Marxists, not an opponent collecting points, and his charge was narrow: not that Marxism is wicked, but that it lacks something specific, a theory of political institutions and of the rules of the game. A narrow charge from a sympathetic quarter cannot be answered by defending the tradition’s intentions, only by producing the missing theory.

Three features of Bobbio’s intervention gave it a weight the polemics lacked. The first is the identity of the questioner. A charge of theoretical incompleteness from a liberal opponent can be filed as hostility; the same charge from a theorist of the left, who shared the tradition’s ends and dissented only about its equipment, had to be answered on the merits or not at all. The Italian Communist Party, then pursuing its historic compromise and its road of democratic advance, understood this at once, and the exchange became a genuine debate conducted in journals and party schools rather than a ritual denunciation. The second feature is the charge’s modesty. Bobbio did not claim that democracy requires capitalism, or that socialist aims were illegitimate, or that the critique of formal equality was wrong. He granted the critique, in large measure, and asked about the morning after: by what procedures would a socialist society decide, and what in the theory guaranteed those procedures against the society’s own rulers? A charge that grants the opponent’s strongest ground and attacks only the empty ground is very difficult to evade, because the usual manoeuvres, defending the critique, citing the record of the other side, invoking the future, are all beside the point it actually makes. The third feature is Bobbio’s account of democracy itself, which he called, in a phrase his interlocutors could not dismiss, the rules of the game. Democracy, for Bobbio, is primarily a set of procedures for reaching collective decisions without violence, whose value lies precisely in their formality: they bind winners as well as losers, they are knowable in advance, and they survive the defeat of any cause, including the cause of socialism itself. To a tradition inclined to treat formality as a veil for content, Bobbio replied that the form is the content where the peaceful transfer of power is concerned, and that a movement unable to say which forms it would keep had not finished thinking about power.

Bobbio pressed a further question that sharpened the first four components into a single edge. Where, in the Marxist account, do rights live? He observed that the tradition treated rights as historically specific to bourgeois society, as the juridical dress of private property, and predicted their withering along with the state. Against this he set the actual career of rights in the liberal democracies: imperfect, uneven, often hypocritical, yet functioning as real constraints that real dissidents had used, in courtrooms and in newspapers, against real governments. A theory that dissolves rights into their class origin, he argued, leaves the citizen of the future society with nothing to hold in her hand. The point was not sentimental. Bobbio had written extensively on the theory of law, and his claim was technical: a legal order consists of norms that bind the powerful as well as the powerless, and a political theory that regards such norms as transitional scaffolding to be discarded has not explained what replaces their binding force. Promises of a higher harmony do not bind anyone. That, in one sentence, is the charge in its hardest form, and every reply examined in this article is, in the end, a reply to it.

The Italian debate of those years also produced the charge’s most interesting partial answers, and it is important for the ledger that some of the answering was done in good faith by Marxists who recognised the gap as real. Within the Italian Communist tradition, theorists argued that the party’s own practice, its defence of the republican constitution of 1948, its acceptance of alternation in government, its insistence on pluralism, amounted to an implicit theory that needed only to be made explicit and owned. Others, drawing on Gramsci, argued that hegemony as a concept already implied a politics of consent, alliance, and cultural leadership incompatible with a one-party monopoly, so that the missing theory was present in embryo in the tradition’s own most original thinker. These were not evasions. They conceded the canonical texts’ silence and relocated the theory to the tradition’s practice and its heterodox margins. But a concession that the theory must be assembled from practice and margins, rather than read off the founding texts, confirms the charge’s central contention even as it answers its spirit. The ledger will record the Italian exchange accordingly: the charge stands against the classical theory, and the repairs begin, significantly, by admitting it.

The textual evidence: Lenin and the theory of smashing

If Bobbio supplied the charge’s logic, A. J. Polan supplied its documentary evidence. In Lenin and the End of Politics, published in 1984, Polan read the tradition’s key transitional text, Lenin’s State and Revolution of 1917, with a lawyer’s attention to what its institutional proposals actually entailed. The book was written on the eve of the October Revolution, and it is the most sustained attempt in the canonical literature to describe the state that would follow the smashing of the old one. Its positive content can be summarised without caricature. The standing army and the appointed officialdom are abolished. Officials are elected, recallable at any time, and paid workmen’s wages. The functions of administration are simplified until they fall, in Lenin’s repeated phrase, within the competence of any literate person, a matter of accounting and control. Representation is reorganised through soviets, councils of delegates from workplaces and districts, which unite legislative and executive functions. And the whole apparatus is expected to wither as administration of things replaces government of men and as the habit of observing the elementary rules of social life makes coercion progressively unnecessary.

Polan’s reading fastened on what this design omits, and the omissions track Bobbio’s components exactly. The text contains no theory of politics as a continuing activity. Disagreement appears in it only as the resistance of the dispossessed bourgeoisie during a bounded transition, after which politics itself, understood as the contest of organised interests and opinions, is scheduled to disappear along with the state. There is no provision for parties other than the one carrying out the transition, no account of a press independent of the organs of power, no judiciary standing apart from the councils, no rights enforceable against the soviet majority, and no procedure for the situation the text does not imagine: a majority of the working class itself divided over what to do. Administration, in the text’s vision, is the application of a known general interest to particular cases, and a known general interest does not generate legitimate opposition, only error and obstruction. Polan’s conclusion was severe. The book does not merely neglect democratic institutions; its guiding idea, the end of politics, excludes the space in which such institutions would be needed. A reader may dispute the conclusion, and defenders have argued that a wartime pamphlet written in hiding should not be read as a constitution. The dispute itself is instructive, however, because the defence concedes the documentary point: the constitution was never written, in 1917 or afterward, and the pamphlet is the closest the canonical tradition came.

It is worth pausing on the phrase that does the deepest work in Lenin’s text and in Marx’s Commune writings behind it: the withering away of the state. The idea descends from Engels, who proposed that the state is not abolished but dies out as class antagonisms disappear. As a prophecy about a remote condition of society it is untestable in the present. As a piece of institutional reasoning it has a precise and damaging implication, which Bobbio identified. A theory whose horizon is the disappearance of the state has no motive to perfect the state’s forms. Why design checks on a power that is scheduled to evaporate? Why entrench rights against a coercion that history is about to retire? The eschatology does the damage: it converts every institutional question into a temporary expedient, and temporary expedients are not designed with the care that permanence forces on a designer. Liberal constitutionalism, whatever its hypocrisies, was written by people who expected their states to last and their enemies to inherit them, and that expectation is the mother of every safeguard in the liberal canon. The Marxist tradition expected its state to be a brief bridge to a condition in which safeguards would be obsolete. The charge is that the bridge lasted, in every historical case, far longer than the theory allowed, and that people had to live on it without the railings the theory had declined to build. That is the strongest statement of the fifth component, and it is a statement about the first four: about what a theory must contain if its predictions fail, since theories of society, unlike theories of tides, are acted upon by people who cannot wait for the prediction to be corrected.

The disputed term at the centre of the transitional theory deserves its own clarification, because public discussion of it is almost pure noise. The phrase dictatorship of the proletariat, examined in full in the companion article on the dictatorship of the proletariat, was used by Marx rarely and by the later tradition constantly, and it carried two very different meanings that the debate perpetually confuses. In its sociological sense it meant the class character of state power, so that a parliamentary republic was, in that vocabulary, a dictatorship of the bourgeoisie however free its elections. In its constitutional sense it came to mean the suspension of democratic forms during a revolutionary transition. Marx’s own usage, sparse and tied to the Commune and to the Gotha Programme critique, sits closer to the first sense with the Commune as its institutional illustration: the working class in power, ruling through radically democratic forms, against a dispossessed minority expected to resist. The slide from the first sense to the second, from a description of whose interests a state serves to a licence for ruling without opposition, is one of the consequential equivocations in modern political history. It matters for the ledger because the charge’s defenders and attackers both exploit the slide: attackers read the constitutional sense back into every text, defenders retreat to the sociological sense whenever the record is raised. An honest accounting fixes the term at each use. What the classical theory indisputably lacks is not goodwill toward democracy but a worked-out constitutional sense at all, in either meaning of the phrase.

Fusion of powers: the precaution refused

Among the five components, the fourth is the one where the tradition’s position is a positive doctrine rather than a silence, and it therefore repays separate treatment. The doctrine has a genealogy. The French Revolution’s assemblies had claimed undivided sovereignty against a crown; Rousseau had theorised the general will as indivisible; and the Paris Commune of 1871, governing a besieged city for two months, fused its functions by necessity and by principle. Marx, in The Civil War in France, elevated the necessity into a model. The Commune was to be a working, not a parliamentary, body. His objection to parliamentarism was not to elections, which the Commune held, but to the separation between a legislature that talks and an executive that acts, with the electorate permitted, as he wrote elsewhere, to decide once every few years which member of the ruling class would misrepresent it. Fusion promised accountability of a different kind: the delegates who decide are the delegates who execute, they can be watched doing both, and they can be recalled for either failure.

The critics’ answer begins by granting the diagnosis and refusing the remedy. Parliamentary irresponsibility, the gap between electoral promises and governing conduct, is a real pathology, and the liberal tradition’s own reformers have complained of it for as long as Marxists have. The question is what follows. The separation of powers was never defended as a device for making government efficient or even, in the first instance, for making it representative. It was defended as a device for making government resistible: for ensuring that the citizen who challenges a decision faces an institution that did not make it, staffed by people whose careers do not depend on the decision’s authors, armed with procedures the authors do not control. Courts, audit offices, ombudsmen, and a legislature with its own mandate are friction, deliberately installed. Fusion removes the friction in the name of the people’s unity, and the critics ask the obvious question: when the unified body errs, or is captured, or simply hardens into a self-perpetuating administration, where does the citizen stand? Recall, the tradition’s answer, is a weapon of majorities. The whole point of separated powers is to give minorities, including the minority of one dissenter, a lever that is not a majority. A theory of democracy that cannot explain the dissenter’s lever has failed a test that the liberal tradition, to its credit, set itself after its own century of religious war and executive terror.

The tradition’s strongest internal voices recognised the problem in practice even when the doctrine denied it in theory. The soviet constitutions of the revolutionary period formally proclaimed the unity of power while the actual state developed, within a few years, a judiciary, a procuracy, administrative courts in all but name, and the rest of the differentiated machinery that governing a large society forces into existence. The differentiation happened, but because the doctrine forbade acknowledging it as a good, it happened without the independence that alone makes differentiation protective. Judges existed; judicial independence, as a principled claim against the ruling party, did not, and could not, because the theory of fused sovereignty classified such independence as a bourgeois fragmentation of the people’s will. Here the ledger’s logic bites hardest. The issue is not that the tradition failed to imagine courts. It imagined them and subordinated them in principle. A silence can be filled by later work. A principle must be revised, and revising the fusion principle means conceding that the Commune model, the tradition’s one institutional exemplar, is defective at its centre, which is a concession the tradition’s guardians have historically been reluctant to make in words even when their own state practice made it in fact.

There is a further, subtler cost of the fusion doctrine, which concerns legislation itself. A body that executes its own decisions at the speed of a campaign has no structural reason to draft its measures for generality, clarity, and durability, the virtues that separated legislatures are forced, however imperfectly, to respect because other hands will apply their words to cases they did not foresee. Revolutionary legality, in the historical record, tended toward the decree: situational, retroactive, addressed to persons and campaigns rather than to continuing states of affairs. The critics connect this tendency directly to the doctrine. Where the lawmaker, the enforcer, and the judge coincide, law as a stable framework within which citizens plan their lives gives way to command as an instrument of the rulers’ current purposes. Bobbio, whose life’s work was the theory of legal norms, felt this component of the charge in his bones: a legal order, for him, was precisely the machinery by which power announces in advance what it will do, and thereby limits itself. A tradition that treats such machinery as a veil for class rule will, when it governs, rule by announcement of a different kind: the campaign, the directive, the exemplary punishment. The historical chapters of this article will show that tendency appearing too consistently to be dismissed as slander, while also showing the pressures, of war and encirclement and inherited backwardness, that pushed every revolutionary government in the same direction. Both things are true, and the ledger keeps them in separate columns on purpose.

The Commune inventory: what the one model actually specifies

Every accounting of this question must sooner or later open the tradition’s single institutional exemplar and count its contents. The Paris Commune governed from March to May 1871, and Marx’s defence of it, The Civil War in France, extracted from its practice the principles he offered as the discovered form of working-class government. The full analysis of that text and its occasion belongs to the companion study of the paris commune marx analysis; here the task is the narrower one of an inventory, because the charge turns on how much institutional specification the model contains and how much it leaves out.

What does the Commune model actually contain?

The Commune model contains election of officials by universal suffrage, recall of delegates at any time, officials paid at workmen’s wages, abolition of the standing army in favour of the armed population, and the fusion of legislative and executive work in one elected body. It is a genuine institutional sketch, aimed at preventing a separate bureaucratic caste.

The inventory, stated completely, runs as follows. Municipal councillors are elected by universal suffrage in the districts of the city, responsible to their electors, and revocable at short notice. The Commune is both legislature and executive, working through commissions drawn from its own members. Public functions are performed by elected and recallable agents, from judges in Marx’s account of the model to the police, who are stripped of their political functions and made responsible instruments. No official receives more than a workman’s wage, a rule aimed at the material root of bureaucracy, the conversion of office into a career and a property. The standing army is suppressed and replaced by the National Guard, the armed people. The clergy are disestablished and church property secularised. Education is opened, freed from church and state interference in Marx’s description, and made accessible. Beyond the city, the model projects a federation: communes across the country sending recallable delegates to a national assembly, so that the structure scales by delegation under imperative mandate rather than by the creation of a separate sovereign parliament. And the whole is summed in Marx’s verdict that the Commune was the political form at last discovered under which to work out the economic emancipation of labour.

Measured against the charge of total absence, the inventory acquits itself better than polemics suggest. Election, recall, wage equality, the destruction of a permanent coercive caste, federal delegation: these are institutional choices, they are specific, and several of them address real pathologies of representative government that liberal theory also recognises, above all the professionalisation of politics and the autonomy of the administrative machine. A defender of the tradition is entitled to insist that the Commune is a sketch of accountable government, not a blank page. The charge, in its careful Bobbian form, never claimed a blank page. It claimed a thin one, and the inventory itself, once complete, shows the thinness. The sketch occupies a few pages of a pamphlet written in the middle of a war. Its author never developed it into a constitutional theory, never subjected it to the questions a second week of reflection would have raised, and treated it explicitly as a set of emergency measures of a besieged city generalised by hope. What the sketch does not contain is as definite as what it does. It contains no theory of parties, no provision for an opposition press, no rights charter, no judiciary independent of the assembly it serves, no rules for the competition of programmes, no treatment of the rights of minorities within the majority class, no procedure for resolving conflicts between the federation’s levels, and no account of how economic planning, which Marx expected to replace the market’s anarchy, would be directed, audited, and corrected by the political structure the sketch describes. Two months of improvised municipal government, however heroic, are being asked to bear the weight of a constitutional order for industrial societies of tens of millions. The tradition’s reverence for the model spared it the criticism that might have thickened it, and that reverence is itself part of the charge.

There is also a question of scale that the model’s admirers rarely face squarely. The Commune administered a city. Its mechanisms assume an electorate that can know its delegates, observe their work, and operate recall as a living practice rather than a formal possibility. Delegation upward to a national body already strains the assumption, and the strain grows with every level of the federation and every technical function the state absorbs. Modern government involves central banking, pharmaceutical regulation, epidemic control, the management of power grids and pension funds, competences that cannot rotate among any literate persons and cannot be supervised by assemblies meeting after working hours. Lenin’s accounting and control formula, the reduction of administration to bookkeeping within everyone’s reach, was already an evasion of this problem in 1917, when states were far simpler than they have since become. A serious institutional theory for a socialist society would have to reconcile the Commune principles with the division of expert labour, designing the boundary between what delegates decide and what specialists execute, and the instruments by which lay sovereignty audits expert power without collapsing into either technocracy or amateurism. That reconciliation is exactly the kind of unglamorous design work the tradition did not do, preferring to repeat the inventory of 1871 as though recitation were development. The replies section will show that when Marxists finally undertook the design work, in the Eurocommunist years and in later democratic theory, they did so by quietly retiring the Commune from a constitution to an inspiration, which is where, the ledger suggests, it always belonged.

Rights and opposition: the two silences

The second and third components of the charge, rights and opposition, are best examined together, because a single theoretical root feeds both silences. The root is the tradition’s account of where disagreement comes from. In the classical theory, fundamental political conflict is the expression of class antagonism: interests collide because the relations of production set classes against one another, and ideologies are the forms in which those collisions are thought. The account explains a great deal of the conflict in capitalist societies, arguably more than any rival. Its consequence for institutional design is the problem. If fundamental conflict is class conflict, then the abolition of classes removes the foundation of fundamental conflict, and the remaining disagreements, over technique, timing, and administration, are in principle soluble by knowledge. A society whose disagreements are in principle soluble by knowledge does not need institutions for processing insoluble disagreement, which is to say, it does not need opposition parties, adversarial courts, or rights entrenched against majorities. It needs administration, expertise, and correction of errors. The silences of the theory are not careless. They follow from its most characteristic premise, and that is why they proved so durable: each generation of the tradition’s theorists inherited not just an omission but an argument that the omission was justified.

The premise deserves challenge on its own terms, and the strongest challengers have again been internal. Human disagreement does not reduce to class. People who share a class position, and who would share it in a society without classes, disagree about risk and the future, about the claims of the present generation against its descendants, about the environment’s weight against consumption, about the proper place of religion, art, and national culture, about the upbringing of children and the organisation of care, about the thousand matters on which interests are not opposed but judgements simply differ. Scarcity, moreover, does not retire with capitalism: time, attention, positional goods, and the carrying capacity of ecosystems remain scarce under any mode of production, and scarce goods generate conflicts that no growth abolishes. Even the tradition’s own practice refuted the premise. The socialist movement of Marx’s lifetime was a garden of factions, and Marx spent a substantial fraction of his working life in ferocious dispute with other socialists whose class position was identical to his own. A theory of institutions built on the coming disappearance of fundamental disagreement was contradicted by the biography of its founder. The critics conclude that the theory needs a doctrine of permanent, legitimate, organised disagreement, and that supplying one is not a decorative addition but a revision of the theory’s anthropology.

Rights present the same structure in miniature. The tradition’s critique unmasked the formal character of rights under inequality, and the unmasking was largely just: a right to property protects the factory owner and the homeless man in beautifully equal language, and a right to publish means one thing to a proprietor and another to a compositor. But from the proposition that rights are insufficient without material equality, the tradition slid to the proposition that rights are dispensable once material equality arrives, and the slide is a fallacy with a body count in the historical record. For rights do two kinds of work, and the critique addressed only one. They distribute advantage, certainly, and there their formality can mock the poor. But they also constrain power, and in that function their formality is not a defect but the entire point: the rule that no one may be imprisoned without trial protects the unpopular precisely because it is formal, indifferent to the merits of the prisoner’s opinions and to the size of the majority against him. A society that has equalised wealth and abolished classes still contains power, still contains officials, majorities, and the temptation to silence the awkward. Against those, only the formal constraint works, and a theory that has taught its adherents to hear formality as fraud will disarm them at the moment they most need the constraint. Bobbio’s whole career in legal theory was a brief for this distinction, and it is the part of his charge that the subsequent debate has done least to answer.

The opposition question has a sharp practical form that concentrates the theoretical issue wonderfully. Imagine the society the tradition proposes, its founding generation in power, its programme enacted, and now a group of citizens concludes that the programme itself is mistaken and organises to reverse it: to restore markets in some sector, to reopen a question the founders considered settled, to campaign for the system’s replacement by something the founders would have called regression. Does the constitution of that society protect their organising, their press, their candidacies? The classical theory offers no answer, because its categories classify such citizens in advance: their position expresses an interest that the new order was built to abolish, and their activity is therefore not politics but residue. Every historical regime of the tradition answered the question in practice by refusal, some brutally, some bureaucratically, all consistently. The defenders’ strongest counter, that liberal societies have also refused this protection at their margins, to colonial subjects, to proscribed parties, to movements surveilled and infiltrated by the state, is historically true and will be weighed in the replies. But the liberal refusal contradicts the liberal theory, and dissidents could and did appeal to the theory against the practice, in the practice’s own courts, sometimes successfully. The Marxist refusal expresses the classical theory, and the dissident appealing to it finds in the founding texts no ground to stand on. That asymmetry, between a theory betrayed by its regimes and a theory implemented by them, is the fifth component’s strongest evidence, and no reply examined in this article fully dissolves it. What the replies can do, and the best of them does, is argue that the theory can be revised from within until the asymmetry disappears. Whether the revision leaves the theory recognisably itself is the question the ledger leaves open, deliberately.

One further concept, routine in the constitutional tradition and missing from the classical canon, crystallises both silences into a single institution: the loyal opposition. The phrase names a settlement so familiar in the democracies that its strangeness goes unnoticed. A body of politicians is paid by the state to oppose the state, given office, staff, broadcast time, and the formal title of the alternative government, on the shared understanding that their opposition is a service to the polity rather than an attack on it, and that the government they attack will, on losing an election, hand them the keys and retire to the benches they vacated. The settlement presupposes everything the classical theory lacks: that disagreement is permanent, that it is legitimate in organised form, that the system itself survives the defeat of every programme advanced within it, and that the majority of the moment holds power on loan from a procedure it does not own. Building such an institution into a theory requires the theory to imagine its own adherents out of power, governing their conduct in defeat and their tolerance in victory, and to value the imagination. The canon never performs it. Its transitional writings imagine the old ruling class out of power, permanently, as a class abolished rather than as an opposition accommodated. Its horizon writings imagine disagreement dissolved. Between the abolition and the dissolution there is no scene in which yesterday’s rulers sit across an aisle, funded and protected, criticising the new order in its own chamber and preparing to inherit it by vote. The absence of that scene, mundane as it sounds, is as exact a measure of the gap as any in this article, because a theory of democracy can be recognised, in the end, by whether it has made imaginative room for its own defeat.

The historical record, reported plainly

The fifth component requires the record to be stated, and this article states it in the same register as everything else here: as evidence, without decoration in either direction. The facts that bear on the democracy charge are not, in their outline, matters of serious historical dispute. In Russia, the revolution of October 1917 was followed within months by the dispersal of the Constituent Assembly, in which the Bolsheviks were a minority, and within a few years by the prohibition of other socialist parties, the suppression of independent trade unions as autonomous centres of power, and the criminalisation of organised dissent inside the ruling party itself, formalised in the ban on factions of 1921. The press was brought under state control. The political police, founded weeks after the revolution, operated outside effective judicial restraint, and under Stalin the machinery of repression expanded into mass arrests, executions, and a camp system whose victims historians count in the millions, with scholarly estimates of excess deaths under his rule varying by method and by category but converging on a scale that no honest summary may soften. Political pluralism did not exist in the Soviet Union at any point in its seven decades; the constitution of 1936 proclaimed extensive rights on paper while the terror proceeded around them, a fact that belongs in any discussion of what paper rights are worth without independent institutions to enforce them.

The pattern repeated, with local variation and at different levels of violence, in the states established in the tradition’s name after 1945 in Eastern Europe and in Asia. Parties other than the ruling one were dissolved, absorbed, or retained as licensed decoration without independent power. Elections were held with single lists or with outcomes settled in advance. Courts were subordinated to the ruling party through appointment, through the doctrine of socialist legality, and through the understanding, universal among judges, of which cases were political. The press was owned by the state or by organisations the party controlled. Independent unions, churches, and associations were suppressed or converted into transmission mechanisms of the party. In China, the campaigns of the 1950s and the Cultural Revolution of the following decade destroyed, among much else, the brief flowerings of criticism that the regime itself had invited, and the death tolls of the famine years after the Great Leap Forward are estimated by historians in the tens of millions, again with ranges varying by source and method. In Cambodia, a regime claiming the tradition carried the logic of a society remade without opposition to an extreme that killed a large fraction of the population in under four years. These are the extreme points of the distribution, and fairness requires noting that they are points on a distribution: other states of the tradition governed with less terror, particularly in the later decades, and some permitted limited cultural and economic space while never conceding political pluralism. The constant, across the whole distribution and across every level of development and every national culture involved, is the one the charge predicts: nowhere did an organised opposition with a legal path to power, a press independent of the ruling party, or a judiciary capable of ruling against the state in a political case survive. Variance in brutality was enormous. Variance on pluralism was nil.

Three interpretive cautions belong with the record, and an honest article owes all three. The first is that the regimes in question inherited, in most cases, societies without prior traditions of constitutional government, and several faced invasion, civil war, or blockade that would have strained any institutional order; the record does not come with a control group, and the counterfactual, what these societies would have become under governments of another doctrine, is unknowable. The second caution is that the tradition also produced movements and governments outside the state-socialist pattern: communist and Marxist parties that contested elections, accepted defeats, governed municipalities and regions under liberal constitutions, and in the Italian case developed, over decades, an explicitly pluralist self-understanding. The record of the tradition is therefore not exhausted by the record of its one-party states, although those states are where its theory of government was implemented without a liberal constitution standing in the way, which is why the critics weight them as they do. The third caution cuts the other way. The explanation that each regime betrayed the theory must eventually confront the number of the betrayals and their uniformity on the specific point of pluralism. A theory betrayed once tells against the betrayer. A theory betrayed identically on the same clause by every government that implemented it, across forty years and four continents, tells a investigator to look at the clause. The clause, the critics say, is the one examined throughout this article: the theory contains no principled ground on which pluralism, once power is held, must be preserved. The regimes did not violate a safeguard. They encountered an absence, and filled it with the apparatus of their survival. That is the record’s testimony in the ledger, stated as strongly as the critics state it, because the replies that follow are worthless unless the charge they answer is present at full weight.

One more feature of the record is regularly overlooked and bears directly on the rights component. The constitutions of the one-party states were not silent about rights; several were voluminous, guaranteeing speech, press, assembly, and conscience in language as generous as any liberal text. The guarantees failed not for want of words but for want of machinery: there was no court to which a citizen could carry them against the state, no opposition to publicise their breach, no press to report it, and no election at which to punish it. The lesson the record teaches about rights is therefore the opposite of the lesson drawn by the tradition’s early deflation of them. Rights without institutions are decoration, certainly; the Soviet constitutions prove it. But institutions without rights are blind, and the liberal coupling of the two, parchment plus machinery, each checking the abuse of the other, is the discovery the tradition’s theory arrived too late and too reluctantly to make its own. Where later Marxists did make it, in the Eurocommunist doctrines discussed below, they made it by adopting the coupling wholesale from the liberal canon and declaring it permanent rather than transitional, which is either the tradition maturing or the tradition conceding, and the ledger will treat it as both.

The replies at full strength

The replies deserve the same discipline as the charge: stated as their best proponents state them, without softening, before any weighing. There are five, of increasing strength.

The first reply is that the demand for a constitutional theory is a demand no emancipatory theory can meet in advance, and that meeting it would be a defect rather than a virtue. Marx refused, famously, to write recipes for the cook-shops of the future, and the refusal was principled: institutions are the creation of the people who will live under them, formed in the struggle that brings the new society into being, and a theorist who drafts their constitution from his study commits the utopian error of substituting his imagination for their self-activity. On this reply, the absence Bobbio laments is the signature of the theory’s democratic character at the deepest level. The Commune was not designed by Marx; it was discovered by him, in the practice of Parisian workers, and the theory’s method is to generalise from such discoveries rather than to legislate ahead of them. The critics answer that the refusal proves too much: the same Marx wrote hundreds of pages of detailed economic analysis, projected the tendencies of a system he had not seen mature, and did not consider foresight a vice in economics. A movement is entitled to refuse blueprints. It is not entitled to refuse, in the name of that refusal, to think about courts and opposition at all, since the refusal does not prevent institutions from arising; it merely ensures they arise undesigned, under the pressures least favourable to liberty. The reply establishes a genuine dilemma, the blueprint against the blank, but the space between the two, the articulation of principles and safeguards without floor plans, is exactly where constitutional theory lives, and the tradition did not occupy that space either.

The second reply defends the Commune model as more than a sketch: as a different theory of control, whose logic the critics misread because they measure it against the liberal instruments it deliberately replaces. Control, on this theory, does not flow sideways among separated powers, each checking the others while all stand at a remove from the people. It flows in a circle: the electors delegate, the delegates act in the open, the electors watch and recall. Short mandates, imperative instructions, rotation, publicity, and the wage rule are the working parts of a machinery of accountability at least as considered as Montesquieu’s, and aimed at a pathology Montesquieu’s machinery barely touches, the crystallisation of a political class distinct from the governed. The reply has real force, and the modern revival of interest in recall, rotation, sortition, and deliberative assemblies among democratic theorists of no Marxist affiliation shows that the Commune’s questions are alive. Its weakness is the one already identified: the circle of control protects the electors against the delegates, but nothing in it protects a minority of electors against the majority whose delegates govern, and the history of majoritarian passion is long enough that a theory of democracy may not treat that protection as optional. The reply answers the first component partially, the substitution charge, by producing a genuine design. Against the second and third components it is silent, because the design is.

The third reply returns to the 1843 distinction and reads it more carefully than the critics do. Political emancipation, on this reading, is not a fraud Marx exposed but a stage he honoured, and human emancipation is its completion rather than its repudiation. The rights of man are criticised for their insufficiency, for securing the individual in his separation while leaving the material conditions of his unfreedom intact; and a completion that adds material equality to formal right would preserve the right while redeeming it. Supporters of this reading point to Marx’s own practice, his defence of press freedom, his insistence that the working class required universal suffrage and the freedoms of combination and publication as the very conditions of its self-organisation, and to the programmatic documents in which democratic demands appear not as camouflage but as the first chapter of the movement’s own programme. On this reply, the theory of rights is present in the tradition as a floor rather than a ceiling: the freedoms are assumed as the movement’s element, and the argument begins from their inadequacy, not their worthlessness. There is textual justice in this reading, and it softens the second component considerably. What it cannot supply is the account of rights as constraints on the post-revolutionary state itself, held against the community in the strong sense, enforceable when the majority of the community finds them inconvenient. Marx defended the freedoms as weapons of the movement against the old order. The critics ask about their status as shields of the individual against the new one, and the texts pass from the first question to the second only in silence.

The fourth reply is the tu quoque, and it is stronger than its name suggests. No political theory, it argues, is entitled to be judged by the worst regimes that have claimed it. Liberalism’s own centuries were written in colonies governed without consent, in the protection of slavery by constitutional democracies, in the exclusion of women and the propertyless from franchises that coexisted with the language of universal rights, and in emergency regimes and counterinsurgencies conducted by states that never stopped calling themselves free. If the gap between liberal theory and liberal practice does not refute liberalism, because the theory contains the standards by which the practice is condemned, then judgment of Marxism should proceed identically: locate the theory, and ask whether the practice follows or betrays it. This reply is just, and this article has applied its principle throughout, which is why the record was presented as evidence about a gap in the theory rather than as a simple indictment of an aspiration. But the reply, at full strength, also reveals its own limit, and the limit is the asymmetry identified in the earlier sections. The liberal dissident appealing against liberal practice finds his standard inside the theory: the rights are written down, the procedures are specified, the courts exist in the design even when they fail in the event. The dissident appealing against the practice of a Marxist state finds, in the classical canon, a critique of the other side’s institutions and a two-month municipal sketch for his own. The tu quoque defends the tradition’s honour. It does not fill its library. Bobbio, who knew the liberal record as well as anyone and had no illusions about it, put the point by asking not whose history was cleaner but whose theory contained the rules by which rulers could be peacefully removed. On that question only one side of the exchange had done the work.

The fifth reply is the strongest, because it does not defend the canon but extends it. It is the reply of Nicos Poulantzas, and its setting matters. Poulantzas was the leading Marxist theorist of the state of his generation, formed in the structuralist school, a man whose earlier work had treated parliamentary democracy with the tradition’s customary suspicion. In State, Power, Socialism, published in 1978, the year before his death, he executed a turn that his readers experienced as a rupture. The problem of a democratic road to socialism, he argued, had to be posed without the tradition’s founding evasion. A democratic socialism worthy of the name required the maintenance and deepening of representative institutions, universal suffrage, political pluralism, and the freedoms, not as a tactical concession to be withdrawn after victory but as permanent and constitutive features of the society being built, combined with, and only with, an expansion of direct democracy at the base: self-management, popular initiatives, the Commune inheritance reworked as a complement to representation rather than its replacement. And he drew the conclusion the canon had avoided: a strategy that treats the state apparatus as an instrument to be smashed and refounded in a moment mistakes both the state and the moment, because the apparatuses condense social relations that a decree does not dissolve, and because the attempt at a refounding from zero has a historical tendency, which he named, toward the authoritarian outcome the theory disowns. The neighbouring debate over the authoritarian outcome as such is treated in the companion article on marxism and authoritarianism, but Poulantzas’s place in the present ledger is distinct: he is the tradition’s strongest witness for the defence precisely because he begins as its strongest witness for the prosecution, conceding the absence and then supplying, in outline, the missing combination of institutions.

Eurocommunism and the democratic road

Poulantzas was theorising a movement that already existed. Across the 1970s the large communist parties of Western Europe, above all the Italian, and for a time the Spanish and French, developed the doctrines gathered under the name Eurocommunism, whose political core was an answer to the democracy charge delivered not in journals but in party programmes. Its elements were consistent across the national variants. Political pluralism was affirmed as a permanent value rather than a transitional tactic. Alternation in government was accepted: a communist party voted out of office would go, and said so in advance. The freedoms of press, association, and conscience were declared irreversible conquests of civilisation rather than bourgeois forms to be surpassed. The road to socialism was defined as democratic in means as well as in end, proceeding by consent assembled through elections and sustained by a plurality of parties, unions, and associations that the party did not control. Enrico Berlinguer, the Italian party’s leader, gave the doctrine its most considered statements, insisting that the values of liberty realised, however imperfectly, in the liberal democracies were historically universal acquisitions that socialism must preserve and extend rather than transcend. Santiago Carrillo’s Eurocommunism and the State of 1977 carried the argument to the point of redefining the transition itself as a democratisation of the state apparatus carried out within legality.

How did Eurocommunism alter the tradition’s position on democracy?

Eurocommunism altered the position by conceding the charge’s substance: its parties affirmed pluralism, alternation in office, and the political freedoms as permanent values rather than transitional tactics, and redefined the transition itself as the democratisation of the state conducted within legality, by consent assembled and renewed through elections.

The alternative tradition against which Eurocommunism defined itself is instructive here, since the comparison clarifies what its concessions cost and what they bought. The social democratic parties of the same century had made peace with the liberal constitution generations earlier, treating its procedures as the permanent framework within which social reform is pursued, a settlement whose own history and costs are traced in the comparison at marxism vs social democracy. Eurocommunism arrived at a similar destination from the opposite bank of the river, retaining an objective, the transformation of property relations, that social democracy had progressively abandoned, while adopting the constitutional framework social democracy had never questioned. Critics from the revolutionary left charged that the destination determined the journey: that a party which guarantees in advance the rules that protect its opponents’ property will find those rules invoked, lawfully and effectively, at the exact moment its programme touches ownership, so that the democratic road leads everywhere except to its stated end. Critics from the liberal side charged the reverse, that the commitment to pluralism was a mask worn for respectability, to be dropped when power allowed. The Eurocommunists answered both charges with the same move, and the move is the theoretically interesting thing about them: they argued that means and ends cannot be separated in politics, that a socialism built by pluralist means would be a different and better society than one built against them, and that a movement unable to win durable consent for the transformation of ownership has not been cheated of its victory but has not yet earned it. Whether that answer is a profound revision of the tradition or an abandonment of it depends on premises about how societies change that this article does not adjudicate. For the ledger, what matters is narrower. The democratic road is a genuine institutional position, the first in the tradition’s history to be held by mass parties with a realistic prospect of office, and it was constructed by conceding the substance of Bobbio’s charge and building on the concession.

Its fate belongs to the record too. The Eurocommunist parties never governed a major state alone. The Italian party reached the threshold of national power in the years of the historic compromise and drew back, or was held back, at the threshold, supporting governments it did not lead during the national solidarity period of the late 1970s. The French experiment ended earlier, in the rupture of the common programme’s alliance. The Spanish party shrank. By the end of the following decade the current had dispersed, and after 1989 the parties themselves transformed, renamed, or dissolved amid the general collapse of the state-socialist world they had criticised. Defenders of the democratic road attribute the failure to circumstance: to the Cold War’s discipline, to the opposition of both superpowers’ camps, to electoral arithmetic, to the onset of the economic turn of the 1980s that shrank every left project’s room. Critics answer that the road was never tested because it could not be: that its theory of gradual transformation through a state whose structural dependence on private investment it left intact was refuted by the same logic the tradition had aimed at everyone else. The ledger records the episode as the replies’ most substantial entry and also as their clearest limit. The democratic road answered the charge at the level of the party’s intentions and guarantees. It did not live long enough, or govern long enough, to answer it at the level of institutional design in operation, and the design work it required, on the courts, the media, the administration, and the economy’s commanding heights under a pluralist constitution, remained programmatic rather than tested. Partial answers are the currency of this entire subject, and the ledger that follows will be found to trade in nothing else.

Later design work: the tradition begins to draw plans

A further body of work, produced mostly after the Eurocommunist moment and largely outside the parties, took up the design question in the spirit the charge demanded: not as apologetics but as construction. Its most cited exemplar is the project associated with the sociologist Erik Olin Wright, whose programme of real utopias, examined in the companion study of real utopias explained, proposed that emancipatory theory should identify institutional designs that are simultaneously desirable, viable, and achievable, and should test existing experiments, participatory budgeting, cooperative firms, unconditional basic income, against those standards. The significance for the present question is methodological. Wright’s programme institutionalises the concession this article has been tracking: it treats the absence of worked-out alternatives as a research failure to be repaired by design work, case study, and the candid grading of proposals, rather than by appeals to a future that will improvise them. Participatory budgeting in Porto Alegre, where a city’s investment budget was allocated through layered popular assemblies over three decades, became the canonical case because it is a Commune principle, direct popular decision over real resources, translated into a durable institution inside a representative system, with rules, delegates, recall in practice, and published accounts. Whatever its limits of scale and its dependence on a sponsoring party’s tenure, it demonstrates the form an answer to Bobbio takes when an answer is seriously attempted: specified procedures, operating in the open, evaluated against results.

Around such cases a wider literature accumulated. Theorists of deliberative and associative democracy, some working within the Marxist inheritance and some adjacent to it, developed designs for the democratisation of economic life that the classical theory had left as a slogan: workers’ councils with defined competences alongside parliaments, social ownership administered through plural funds and competing cooperatives rather than a single ministry, constitutional entrenchment of social and political rights against future majorities including socialist ones, media ownership dispersed by law among public, cooperative, and private holders under rules designed to prevent reconcentration. Market socialist models of the period attempted the hardest reconciliation of all, specifying how an economy could retain prices and competition among firms while socialising the ownership of capital, and subjecting their own blueprints to the objections, on information, incentives, and political capture, that the tradition had once reserved for its opponents. None of this work commands the tradition as a whole, and none of it has been enacted as the constitution of a state. Its existence nevertheless changes the ledger’s arithmetic for the substitution component. In 1975 Bobbio could say that the theory of the socialist state did not exist. A fair statement at a later date is that fragments of such a theory exist, produced by the tradition’s heirs under the pressure of his question, scattered across journals and experiments, uncanonised, untested at the scale of a nation, and carrying no guarantee that a movement in power would consult them. The gap between a literature and a doctrine is precisely the distance the charge measures. It has narrowed. It has not closed, and the writers doing the narrowing have generally been the first to say so.

Can a critique be true while the replacement theory is missing?

Yes, and the two judgements are independent. A diagnosis does not stand or fall with the therapy proposed beside it. The critique of formal democracy survives the absence of a successor design, just as the absence is not excused by the critique’s power. Keeping the judgements separate is what allows each to be graded honestly in the ledger.

That separation is the intellectual hinge of the whole subject, and it is worth stating why both camps resist it. The tradition’s opponents resist it because the missing replacement is their most effective rhetorical weapon, and they prefer to wield it as a refutation of the critique itself: if you cannot say what comes next, your account of what is wrong must be worthless. The inference is a fallacy, and its exposure is a service the tradition’s defenders perform reliably. The tradition’s defenders resist the separation for the mirror reason: they prefer the critique’s strength to stand as the theory’s vindication entire, so that conceding the gap feels like surrendering the whole case. It is not. A theory of society can be simultaneously the best available account of how an existing order constrains democracy and an unfinished account of how a freer order would govern itself, and the mature position, reached by Bobbio from one side and by Poulantzas from the other, is to hold both halves without allowing either to anaesthetise the other. The critique without the design leaves a movement powerful in opposition and unready in power, which is a fair summary of a century. The design without the critique leaves a constitutionalism naive about the economic forces that bend its procedures, which is a fair summary of the liberal complacency the young Marx attacked. The ledger’s purpose is to grade the tradition as it stands, not as either camp wishes it were.

Why the silence lasted: four causes inside the theory

A gap this consequential invites a causal question, and the literature offers four causes that operate together. The first is the sequence of Marx’s own project. His life’s work was the critique of political economy, a project that consumed the decades in which a political theory might have been written, and the political writings he did produce were overwhelmingly occasional: journalism, polemics, defences of movements under fire. The Communist Manifesto is a pamphlet of agitation. The Civil War in France is a defence of an insurrection already drowned in blood. The Critique of the Gotha Programme, the text that comes closest to describing a post-capitalist order, is a private memorandum on a party programme, unpublished in his lifetime, and even it devotes its institutional attention to distribution in two phases rather than to government. The founder’s bibliography bequeathed a method and a critique at monumental scale and a politics at pamphlet scale, and a tradition that venerates its founder’s every paragraph will reproduce the proportions of his desk.

The second cause is the intellectual prestige of the economic within the theory’s self-image. Marxism understood itself as a science of society’s material foundations, against the idealism it attributed to rivals who began from constitutions, rights, and ideas. Within that self-image, constitutional questions looked superficial almost by definition: forms of the superstructure, to be explained by the base rather than lingered over as causes in their own right. A theorist ambitious within the tradition wrote about production, crisis, and class. A theorist writing about electoral systems, judicial review, and the design of second chambers marked himself as a legalist, in the vocabulary of the schools a half step from a liberal. Doctrines allocate prestige, and prestige allocates careers, and the allocation across a century is a material fact about why the library has the shape it has. Bobbio, whose formation was in the supposedly superficial disciplines of law and political philosophy, was placed by that allocation outside the tradition’s high road, which is part of why his questions arrived as an external shock although they were asked in friendship.

The third cause is strategic, and it operated in every period. A movement in opposition, contemplating an enemy state armed and hostile, does not advertise constitutional scruples it expects to need only after victory, and it positively fears some of them in advance: guarantees to the opposition read, to militants facing prison and exile, as guarantees to the counter-revolution. Under conditions of illegality, conspiracy, and war, which were the conditions of the tradition’s most formative decades in Russia and much of Europe and Asia, pluralism inside the movement itself appeared as a luxury of peacetime parliamentarism. The habits formed under those conditions then travelled, codified as theory, into conditions where they no longer applied, because the parties that carried them owed their identities to the generations that had formed them. Path dependence is not an excuse the critics are obliged to accept, but it is an explanation they are obliged to include, since the alternative explanation, that millions of adherents across a century were simply indifferent to freedom, explains nothing and insults the evidence of the sacrifices many of them made in freedom’s name under regimes of the right.

The fourth cause is the subtlest, and it returns to the premise examined earlier: the theory’s expectation that the state, and with it coercive politics as such, was a transient phenomenon of class society. Every other cause could have been overcome by a sufficiently determined theorist. This one operated inside the theory’s logic, devaluing the design task at its root. One does not write the constitutional law of a bridge. The prediction embedded in the premise was, moreover, self-insulating. Each failure of the state to wither could be attributed to the survival of the conditions that explained the state, encirclement, scarcity, the habits of the old society, so that the premise survived every disappointment by consuming it as evidence of the transition’s length. A prediction that digests its own falsifications is a liability in any science, and the tradition’s critics in the philosophy of science have made that point about other parts of the doctrine. Here its cost was specific and institutional: it licensed a hundred years of postponement, during which the design questions were always about to become urgent and never quite were, until the states built on the postponement collapsed with the questions still unanswered and the populations they governed left to import their constitutions, in the end, from the liberal canon the tradition had spent a century criticising. That final irony is a fact of the record, not a flourish, and it belongs in any honest account of what the gap cost.

What would count as answering the charge?

Answering the charge would require a worked-out doctrine, held by the tradition as its own, specifying how a post-capitalist society constitutes and removes governments, entrenches rights against its own majorities, protects organised opposition, and separates the functions of rule-making, execution, and judgement. Partial answers exist. A canonical one has not been produced.

The charge and reply ledger

The ledger below compresses the article’s argument into its auditable form. Each row states one component of the charge, names who presses it in its strongest version, gives the strongest reply the record affords, records whether the reply lands, and names what a defender of the tradition must concede even where the reply succeeds. The verdicts are this article’s, argued in the sections above; a reader who disputes a verdict has the full text of both sides in those sections with which to dispute it.

Component of the charge	Who presses it	Strongest reply	Verdict	What a defender must concede
No theory of replacement institutions: critique and transition are specified, the successor constitution is not	Bobbio; Polan on the textual evidence	The Commune model supplies a genuine design of accountable, recallable government	Partially answered	The Commune is a municipal sketch of a few pages, never developed into a constitutional theory for a complex society
No theory of rights held against the community	Bobbio, from legal theory	Rights are honoured as a floor in Marx’s practice and completed, not abolished, by material equality	Partially answered	The canon contains no account of rights enforceable against the post-revolutionary state and its own majorities
No theory of legitimate opposition	Bobbio; the critics of the end-of-politics premise	Eurocommunism and Poulantzas affirm pluralism and alternation as permanent	Partially answered	The classical theory grounds disagreement in class and therefore has no principled place for opposition after classes; pluralism enters by revision
Fusion of powers affirmed against their separation	The constitutionalist critics; Bobbio on legality	Recall and publicity form an alternative machinery of control aimed at the political caste	Standing	The alternative machinery protects majorities against delegates and offers the dissenter no lever that is not a majority
The record of one-party states reflects the theory’s absences	The critics of the record, in the strong form	No theory should be judged by its worst claimants; the liberal record is also stained	Partially answered	The uniformity of the outcome on pluralism, across every implementing regime, is evidence about the theory and cannot be wholly discharged onto circumstance
The critique of liberal democracy is mere rhetoric	The tradition’s dismissive opponents	Structural dependence and the filtering mechanisms state the critique as testable social science	Answered	Nothing on this row; the concession runs the other way, and opponents must meet the mechanisms rather than the slogan

Read as a whole, the ledger shows a pattern that neither camp’s slogan captures. The row that is fully answered is the row defending the critique. The rows constituting the charge are answered partially or not at all, and the partial answers share a signature: they were produced late, by the tradition’s revisers, against the resistance or the silence of its guardians, and they proceed by importing into the theory precisely the elements, rights, pluralism, separated and independent institutions, alternation in power, whose absence the charge alleged. This is not a dishonourable pattern. Traditions that survive are traditions that revise, and the history of liberalism itself is a history of such imports, of suffrage extended under pressure, of social rights grafted onto a doctrine of property, of emergencies survived by rules bent and then restored. The difference the charge finally identifies is one of timing and acknowledgment. The imports into Marxism arrived after the regimes built on the unamended theory had already demonstrated, at scale, what the unamended theory permits in practice, and the acknowledgment that imports were needed at all is still, in parts of the tradition, resisted as betrayal. A gap confessed is a research programme. A gap denied is a hazard carried into the next attempt.

The asymmetry of the tradition

The ledger’s pattern is the asymmetry of the tradition, stated now in its full form, because the article’s central claim is not simply that a gap exists but that the gap has a shape, and the shape explains the tradition’s career. Marxism is, by the standards of modern political thought, an unusually successful critical theory and an unusually unsuccessful constitutional one, and the two facts are connected by more than accident. The critique of liberal democracy was built by the tradition’s characteristic method at its best: take the order’s own professed values, freedom, equality, self-government, and measure the order’s institutions against them under the actual distribution of wealth and power. That method is a solvent. Applied to parliamentarism it dissolves the pretence that equal votes mean equal influence. Applied to rights it dissolves the pretence that formal protection is protection in fact. Applied to the state it dissolves the pretence of neutrality. A tradition armed with the solvent sees through the existing order with a clarity its rivals have repeatedly had to borrow. But the solvent dissolves the tradition’s own constructions with equal indifference, and a movement that has internalised the habit of asking, of every institution, whose interest it serves and what material inequality it conceals, will struggle to build institutions of its own with the innocence construction requires. Every court it designs it suspects of becoming a caste. Every right it entrenches it hears in the accent of the property owner. Every separation of powers it drafts looks to it like the fragmentation of the people’s will. The critique, turned inward, becomes a reason the building never starts, and the theorists who did start building, from the Italian debaters to Poulantzas to the designers of real utopias, did so by consciously suspending the solvent at the point of construction, an act of theoretical self-discipline the canon nowhere teaches.

The asymmetry also has a sociology, continuous with the causes traced earlier. Critique is the work of opposition, and the tradition spent most of its life in opposition, where the solvent is the right tool and its absence would be the scandal. Constitutional design is the work of founders, and the tradition’s founders, in the few places they held power, arrived there through collapse and war and governed through emergency, delegating the constitutional question to a future that their emergency kept postponing. Between the opposition’s habit and the emergency’s pressure, the design fell through, generation after generation, each inheriting the critique in finer condition and the design in none. By the time the tradition’s Western parties were positioned to need the design in earnest, in the Italy of the 1970s, the need arrived as Bobbio’s questions, and the most honest of them answered as Poulantzas did: by treating the absence as the central problem of a democratic socialism and the combination of representative and direct institutions as the beginning of its repair. That the repair came from the tradition’s own greatest theorist of the state is the strongest evidence that the charge was never an enemy’s caricature. It was the tradition’s unfinished business, recognised at last from inside, and the measure of the asymmetry is that recognising it took a century.

Nothing in this analysis licenses the two verdicts the shouting match prefers. It does not license the conclusion that the critique was a fraud, since the critique’s mechanisms have been absorbed into the common equipment of political science and its questions about ownership and influence remain open and empirical. It does not license the conclusion that the gap was a detail, since the gap governed what happened wherever the theory was implemented without external constitutional constraint, and the populations concerned paid for it in the currency the record section itemised. And it does not license any conclusion, in either direction, about which political system the reader should support, a question that turns on values and on judgements about alternatives actually on offer, on which a scholarly article on a theoretical gap has no commission to rule. What the analysis establishes is a finding about a body of thought: that it is two theories of unequal completeness wearing one name, and that anyone who takes it up, in sympathy or in opposition, does better to know which of the two they are holding at any given moment.

Five misreadings that keep the argument stuck

Public discussion of this subject recycles a small set of errors, and naming them is a service to readers who will meet the argument elsewhere in worse condition. The first misreading is that Marx opposed democracy, full stop. The record of his journalism, his suffrage demands, and his praise of the Commune’s elected and revocable government contradicts it. His objection was to the doctrine that political equality completes emancipation in a society of economic subordination, which is an objection to a stopping point, not to the road. The second misreading is that bourgeois democracy means sham democracy, a stage set behind which a dictatorship hides. The serious usage, established earlier in this article, is a claim about the limits ownership sets to democratic outcomes, and it is compatible with elections being genuine, hard fought, and consequential within those limits. The third misreading is that the charge examined here is an attack from the tradition’s enemies. Its precise author was a socialist theorist of democracy, its venue was a journal of the socialist left, its addressees were communists, and its most serious respondent was the leading Marxist theorist of the state. Arguments about the tradition’s completeness have been an internal affair at their best, and readers should suspect any presentation that casts them as a war between camps. The fourth misreading is the mirror image, that the Commune model is no answer at all, a riot mistaken for a constitution. It is an answer of a specific and limited kind, a machinery of accountability against bureaucratic caste, and the ledger graded it as partial rather than worthless because that is what an inventory supports. The fifth misreading is that the historical record settles the theoretical question by itself, in either direction. Records do not settle questions of theory without an account of the connection between the two, which is why this article built the connection explicitly: the record is evidence about the theory where the theory contains an absence that practice must fill, and it is weak evidence where the practice contradicts the theory’s explicit commitments. Readers equipped with that distinction can weigh cases the polemics merely hurl.

Why is the Commune treated as the tradition’s main institutional text?

Because it is the one place where Marx endorsed a concrete set of governing arrangements rather than a critique or a prophecy. Later writers returned to it whenever the institutional question was pressed, and that repeated return, across a century, is itself the measure of how little was added to the inventory afterward.

Marx’s own democratic commitments: the evidence assembled

Any verdict on the tradition should begin by getting its founder’s position right, and on democracy the founder’s record is more substantial than either his detractors or his more impatient followers have usually allowed. Begin with the young journalist of the Rheinische Zeitung in 1842 and 1843, writing under Prussian censorship. His articles on the press laws defend freedom of the press not as a concession to be granted by authority but as the natural medium of a people’s reason, and his commentary on the proceedings of the Rhenish assembly attacks the representation of estates and property in the language of a democrat: the assembly’s deputies, he argues, represent wood and land and privilege, and the scandal is precisely that human beings go unrepresented. The essays of the following years radicalise the argument rather than abandoning it. Democracy, in the manuscript critique of Hegel’s philosophy of the state written in 1843, is described as the solved riddle of all constitutions, the form in which the constitution is brought back to its actual basis, the human being. The vocabulary is philosophical and the destination of the argument, the abolition of the separation between state and society, leads toward the silences this article has documented. But the starting point is unambiguous: self-government is the standard against which every existing form is measured and found wanting, and the measurement begins from the most democratic premise available in the philosophy of the time.

The mature record continues the pattern in a practical register. Marx supported the Chartists, whose programme was a democratic programme of suffrage and representation, and treated the winning of universal suffrage in England as a measure that would place political power in the hands of the class that constituted the immense majority. The rules he drafted for the International Working Men’s Association in 1864 open by declaring that the emancipation of the working classes must be conquered by the working classes themselves, a sentence whose anti-vanguardist plain sense later generations of his followers had to labour to qualify. His addresses for the International treat the ballot, the union, and the public meeting as the movement’s natural instruments in the countries where they exist. The most quoted evidence is the speech at Amsterdam in 1872, after the Hague Congress, in which he stated that in countries such as America, England, and possibly Holland, the workers might attain their goal by peaceful means, while in most continental countries force would be the lever, a distinction drawn not from doctrine but from an inspection of institutions: where democratic channels existed and were real, they were to be used, and the statement is unintelligible unless the channels’ reality mattered to the speaker. The same man who wrote that the working class cannot simply lay hold of the ready-made state machinery wrote also, of the Commune, as an elected body answerable at short intervals to universal suffrage. The thread connecting the record is not hostility to democracy. It is the insistence that democracy is a beginning whose completion lies beyond the political sphere, an insistence whose strength and whose danger this article has been weighing from the start.

Why, then, did the tradition’s institutional theory not grow from this soil? Part of the answer is the fate of the founder’s own political writings within his Nachlass and his movement. The 1843 critique of Hegel remained unpublished until the twentieth century. The Gotha memorandum circulated in manuscript and reached print only in 1891, against the wishes of the party leadership it criticised. The Civil War in France was an address to a defeated insurrection, read thereafter as scripture rather than as the occasional piece it was. A tradition builds its canon from what circulates, and what circulated most powerfully in the founder’s name during the decisive decades of party formation were the economic works, the Manifesto, and the polemics, in which democratic demands appear as the movement’s presupposition rather than as a subject for design. Presuppositions are not theorised; they are inherited, and when the inheritance was spent, in the crises of the twentieth century, there was no developed theory to draw on. The founder had assumed democracy as the element of a victorious movement. His science was directed at the economy that deformed it. The assumption and the science together produced the asymmetry with which this article began, and they produced it innocently, which is why the strongest form of the charge is a finding about structure rather than about anyone’s intentions.

The Italian exchange: what was actually asked and answered

Because the Bobbio exchange is the article’s evidentiary centre, its particulars deserve a fuller record than summary allows. The setting was an Italian left in motion. The Communist Party under Berlinguer had proclaimed the historic compromise, a strategy of broad democratic alliance, and had drawn from the Chilean catastrophe of 1973 the lesson that a narrow majority attempting a rapid transformation invites its own destruction; the party was therefore publicly reconsidering the relation between socialism and the liberal constitution at the moment Bobbio wrote. His interventions of 1975 and 1976, essays in Mondoperaio and contributions to the debate the party’s own intellectuals joined, posed the question of the Marxist theory of the state as a question of existence: such a theory, in the sense of a doctrine of the institutions of a socialist polity, did not exist, and the belief that it existed, somewhere in the canon, awaiting application, was itself an obstacle to producing it. Around the central claim he arranged subordinate questions that map onto the ledger’s rows. What happens to the rules of the game when the game’s outcome is supposed to be the end of the game? Which rights survive the transition, and in whose custody? Is pluralism a value in itself or a concession to the enemy’s strength, to be withdrawn when the strength fails? Can a party that regards the state as an instrument of class domination describe, in advance, the state it would regard as legitimate when its opponents hold it?

The replies from within the party’s orbit were of three kinds, and distinguishing them clarifies what the exchange achieved. The first kind appealed to practice: the party’s thirty years of constitutional loyalty, its defence of the republic against coups and terrorism in the years to come, its municipal governments, were offered as a theory enacted, needing codification rather than invention. The second appealed to Gramsci, whose concepts of hegemony and the war of position were read as already containing a pluralist politics of consent, so that the missing theory existed in the tradition’s most original mind and required only development. The third kind, the rarest and the most consequential, accepted the charge and changed the doctrine: pluralism was affirmed as a permanent and universal value, the freedoms as irreversible, alternation as binding, in the party’s programmatic statements of those years. The exchange’s historical importance is that the third kind prevailed within the largest communist party of the West, so that the charge’s practical answer was delivered not by its author but by its addressees, in the form of a revision. Bobbio’s own later work, above all The Future of Democracy of 1984, generalised the questions into a theory of democracy’s promises and their partial non-fulfilment, in which the Marxist case appears as one instance of a universal problem: every modern doctrine, liberalism included, is measured by him against the rules it proclaims. The even-handedness of the later work confirms what the Italian debaters sensed in the earlier exchange. The questions were not a trap. They were an invitation to the tradition to finish a part of its own thinking, and the measure of the invitation’s seriousness is that where it was accepted, the accepting changed the tradition more than a decade of polemic from its enemies ever had.

A late echo of the exchange deserves recording, because it shows the charge outliving the world that produced it. After 1989, when the one-party states collapsed and their successor societies set about importing constitutions, the questions of 1975 returned in the mouths of the tradition’s surviving thinkers as a reckoning: the design work postponed for a century had to be done in a decade, by societies in ruins, and it was done, where it was done at all, with liberal instruments. Defenders of the tradition replied, with some justice, that the collapse proved nothing about the democratic road that was never taken, and that the Eurocommunist alternative had been strangled by the same Cold War that preserved the regimes whose fall was now treated as the theory’s verdict. Both statements can be true, and the ledger makes room for both. They do not alter the finding about the classical theory, which concerns what the texts contain. The texts contain a critique of permanent value and an institutional sketch, and between the sketch and a constitution there stretches the distance this article has surveyed. The distance is the answer to the title’s question, stated without animus and without consolation: the theory of democracy in Marxism is a theory of democracy’s limits, and the theory of its institutions remains, in the canon, the missing volume that its ablest readers have been writing, piecemeal and against resistance, ever since one of them first pointed at the empty shelf.

The strongest objection to the charge, given its full weight

Fairness to the replies requires one of them to be stated as an objection to the entire enterprise of this article, because its proponents regard it as dissolving the charge rather than answering it piecemeal. The objection runs as follows. The charge judges a theory by a standard no living political tradition could meet in its founding texts, and then confirms the judgement by a record compiled under conditions the theory never chose. Liberalism’s founding centuries contained no theory of universal suffrage, no place for women or the propertyless as citizens, no restraint on colonial government, and a doctrine of property that licensed dispossession on several continents, yet the liberal tradition is permitted to be judged by its best texts and its subsequent self-corrections, while Marxism is judged by the police files of its worst regimes. The asymmetry, on this objection, lies not in the Marxist tradition but in the bookkeeping of its critics. Grant the objection its full force before weighing it, because it is the strongest thing that can be said against the ledger method, and a ledger that cannot survive it deserves to fail.

The objection succeeds in two of its three movements. It is true that traditions must be judged by their theories and not only by their claimants, and this article has applied that rule to liberalism’s credit throughout, treating the liberal canon’s safeguards as real even where liberal practice traduced them. It is also true that the conditions under which Marxist parties took and held power, invasion, civil war, embargo, and inherited despotism, belong in any causal account of what followed, and that no counterfactual entitles a critic to assume a gentle outcome under other banners. The objection fails in its third movement, at the point where it must explain why the self-correction it celebrates in liberalism has an equivalent in the Marxist case. Liberalism’s corrections were executed with tools the canon itself supplied: the suffrage was extended by appeal to the canon’s own universalism, in its own legislatures and courts, against the resistance of its own beneficiaries, and the standards by which colonial rule stands condemned were drawn from the tradition’s own account of rights. The correction of the Marxist record, where it was attempted from within, required the importation of the instruments the canon lacked, and the importation is not a detail of the story. It is the story. The objection, pressed to its conclusion, does not dissolve the charge. It relocates it, from the morality of the aspiration, where the charge never belonged, to the equipment of the theory, which is where Bobbio put it in the first place.

There remains the deepest stratum of the objection, and it deserves a direct answer because capable defenders rest their whole case on it. The critique of ownership’s constraints, they observe, stands unanswered by the tradition’s critics in any form that accepts the evidence: the filtering of agendas, the structural veto, the manufacture of consent are mechanisms named by scholars across the spectrum and denied, in their specifics, by almost no one who has studied them. A political discourse that spends its energy on the Marxist theory’s missing chapters while treating those mechanisms as scenery is not neutral between the traditions. It is applying the solvent selectively. The observation is just, and it is the reason this article opened with the critique at full strength and graded its ledger row as answered. But the observation cuts at the charge’s misuse, not at the charge. That liberal democracies remain constrained by concentrated ownership is a reason to pursue the question of what institutions could do better. It is not a reason to excuse the tradition that posed the question most sharply from answering it, and the tradition’s own best minds, measured by Poulantzas, did not ask to be excused. They asked to be read as having begun the answer. The ledger honours them by taking the request literally.

Disagreement inside the movement: the internal evidence

The theory’s difficulty with opposition has a domestic history that predates any Marxist government, and the domestic history is in some ways the cleanest evidence, because it cannot be attributed to the pressures of ruling. The socialist movement of the nineteenth and early twentieth centuries was an ecology of factions, journals, and rival internationals, and its internal life was conducted with a ferocity that the later legend of a monolithic tradition conceals. Marx and Engels fought the anarchists to expulsion at the Hague Congress of 1872. The German party lived with its revisionist controversy as a permanent condition. The Russian movement split in 1903 over, among other things, a clause in a membership definition, and the split hardened into two traditions that regarded each other as the principal obstacle. Nothing in this record is discreditable in itself; intellectual movements argue. What matters for the institutional question is the theory each faction carried about the meaning of its opponents’ existence. Within the classical framework, an opponent inside the movement was not a citizen with a different judgement but a representative of a different class interest wearing a socialist vocabulary, and the classification licensed the conclusion that the dispute would end not by persuasion but by the disappearance of the social basis the opponent expressed. Debate conducted on that premise tends toward a particular terminus: the organisational elimination of the error, since errors rooted in class position cannot be argued out of their holders.

What happened to organised disagreement inside the revolutionary party?

At the Tenth Congress of the Russian party in 1921, factions within the party itself were formally banned, in a measure presented as temporary and emergency-bound. The ban outlived the emergency by the party’s whole remaining history, and it converted the theory’s premise about disagreement into an organisational rule whose effects every later institution inherited.

The episode repays attention because it is the hinge between the theory’s premise and the states’ practice, occurring before the terror, before the cult, at a congress of revolutionaries many of whom had spent their lives demanding free debate inside their movement. The same congress season saw the suppression of the Kronstadt rising, whose sailors, veterans of 1917, demanded soviet elections and freedoms for socialist parties and were put down by force. A reader need not romanticise the rebels or minimise the regime’s peril in that winter to see the theoretical shape of the events. A demand for pluralism arising inside the working class itself, from the revolution’s own celebrated constituency, had no category in the theory under which it could be legitimate. It was processed as the premise directed: as the voice of a social basis, peasant, petty bourgeois, exhausted, that history was removing, and therefore as a phenomenon to be administered rather than a position to be answered. The later regimes did not invent this processing. They inherited it, with the ban on factions as its organisational charter, and applied it at scales the men of 1921 did not foresee. When the ledger grades the opposition component as partially answered, the partial credit belongs to the Eurocommunist repudiation of exactly this inheritance, a repudiation that had to name the ban and the processing as errors of theory, not merely excesses of men, before it could claim to have repaired them.

Three test cases: press, courts, administration

Abstractions about institutions become testable when they are run against the specific machinery a modern society cannot do without. Three cases show the design gap in cross-section.

The press first. A socialist society requires information to flow upward against the gradient of power, since its planners and delegates must learn what their decisions are doing. The classical theory treats the press primarily as a weapon and an interest: bourgeois newspapers as the property of owners, party newspapers as the voice of a class. Both descriptions capture something real, and neither yields a doctrine of the press as an institution with claims against the society’s own government. Who owns the presses and the transmitters in the proposed order, and by what rule are they allocated when demand for them exceeds supply? If the state allocates them, the state decides which criticisms are printed. If some independent body allocates them, the theory owes an account of that body, its appointment, its own accountability, and its protection against capture by the power it polices. The one-party states answered the question by ownership, and the answer strangled the upward flow of information their own planning required, a cost their economists documented in the language of falsified reporting and invisible shortage. The later design literature answers with dispersal: mixed ownership, statutory independence for public media, rights of access enforceable in court. Each element of that answer is an institution of recognisably liberal pedigree, adopted because nothing in the canon generates an equivalent. The press, in short, is the case where the gap ceases to be philosophical. A society that cannot describe how its critics get printed has not finished designing its politics, whatever the completeness of its economics.

The courts second. The tradition’s jurisprudence, where it exists, treats law as superstructure and courts as its clerks, and the treatment generates the subordination described earlier. But adjudication is also where a society’s promises become specific: a plan target disputed between ministries, a citizen’s flat allocated twice, a manager accused of falsifying output, a citizen accused of anything at all. Each case requires a decider whose authority the loser accepts, and acceptance depends on the decider’s visible independence from the winner’s patrons. The regimes of the tradition staffed their courts and taught them their limits, and the resulting legal systems handled routine disputes competently while failing, by design, at the single class of case a constitutional court exists for: the case where the state is a party and the citizen is right. The design question for a socialist judiciary is therefore not whether judges should be elected, the Commune’s answer, since election by the majority reproduces the majority’s power in robes. It is how a judge acquires the standing to rule against the assembled will of the society’s governing institutions and survive the ruling. Liberalism’s answer, tenure, salary protection, appointment insulated from the executive of the day, precedent, a professional culture of independence, is imperfect and often hypocritical in application. It is nonetheless an answer, accumulated over centuries of exactly the failures the question describes. The canon offers no rival accumulation, and the later Marxist literature, Poulantzas included, gestures toward the problem rather than solving it.

Administration third, the least glamorous and in daily life the most important. Governing a complex society means delegating to permanent officials whose knowledge the delegates cannot replicate. The tradition identified the danger of this stratum with unmatched vividness: the bureaucracy as a caste, the official’s career as a property, the machine’s tendency to become the master. Its remedies were rotation, recall, wage ceilings, and the simplification of functions, remedies calibrated to the administration of a city in 1871 or a poor and largely illiterate country in 1917. Against the administration of a modern economy they are not so much wrong as unscaled, and the unscaledness is diagnostic. A serious answer would specify which functions rotate and which accumulate expertise, how specialists are audited by lay authority without the audit becoming either a rubber stamp or an inquisition, and how the administrator’s legitimate professional judgement is distinguished, in rules rather than in slogans, from the caste interest the theory rightly fears. Weber, the tradition’s great contemporary antagonist on this ground, predicted that the abolition of private capital would not diminish bureaucracy but fuse the economic and state machines into a single hierarchy harder to resist than either alone. The prediction’s accuracy in the one-party states is a matter of record. What the defenders of the tradition must show, and what the design literature has only begun to attempt, is the institutional arrangement under which Weber’s prediction fails. Until that arrangement is specified, the administration case stands where the ledger leaves the component it belongs to: the diagnosis brilliant, the prescription a sketch for a smaller world.

The workplace: the tradition’s counter-claim to extend democracy

Justice to the tradition requires a section for the direction in which its democratic imagination was expansive rather than silent, because the charge, fairly stated, concerns political institutions and not the whole of its thinking about self-government. Marxism’s founding complaint against liberal democracy was that it stops at the factory gate: that the citizen who votes as an equal on Sunday is governed as a subordinate from Monday, taking orders in a hierarchy he did not choose, under rules he cannot appeal, for most of his waking life. The extension of democracy into economic life, into the enterprise, the office, and the plan, is the tradition’s positive vision, and it is not a small one. A theory that asks why self-government should end where work begins has identified a genuine incoherence in the liberal settlement, and every subsequent movement for industrial democracy, codetermination, and cooperative ownership has drawn on the identification, usually without the accompanying theory of the state.

The experiments in workers’ self-management that the twentieth century actually ran, most extensively in Yugoslavia after its break with the Soviet model, supply the record for this counter-claim, and the record is mixed in instructive ways. Councils in enterprises exercised real functions, elected managers within limits, and disposed of a share of the surplus, demonstrating that the extension of formal democracy into the firm is administrable and not merely rhetorical. The same system showed the limits that appear when enterprise democracy is nested inside a one-party state and a soft budget constraint: the councils operated under the shadow of the League of Communists, investment was politicised through the banking system, and the federation’s richer and poorer republics reproduced, inside a socialist economy, the regional antagonisms the theory expected solidarity to dissolve. The lesson the ledger draws is not that economic democracy failed, since its local successes were real. It is that economic democracy cannot substitute for the missing political institutions, because the questions it cannot answer, who guards the federation’s rules, who protects the minority republic, the minority firm, the minority worker against the democratic organs themselves, are precisely the political questions. A movement that extends the vote into the workplace while leaving the constitution undesigned has widened democracy’s territory and left its capital unfortified. The counter-claim therefore survives the scrutiny as a genuine contribution and fails as an answer to the charge, which was never that the tradition imagined too little democracy. It was that it imagined democracy’s substance richly and its safeguards barely at all.

The transition problem: safeguards deferred are safeguards denied

Between the critique and the design stands a temporal structure that does much of the gap’s practical damage, and it deserves isolation because it operates even where a movement’s intentions are democratic. The classical theory organises history into a before, an after, and a between. The between, the transition, is the period in which the old order’s resistance is expected, the new order’s institutions are immature, and the extraordinary measures are justified by the emergency of being between. Every safeguard the charge demands is, in the theory’s own sequencing, an institution of the after: rights in full flower belong to the society beyond scarcity and antagonism, pluralism to the condition in which disagreement has lost its class root, legality to the settled state. The between is governed by necessity, and necessity, as the tradition’s own historians of other revolutions could have testified, is the oldest enemy of every safeguard ever promised. The design flaw is therefore not only that the after was undesigned. It is that the between was designed, by the theory’s silence, to be constitution-free, and the between has a property the theory systematically underestimated: it lasts. It lasted, in the founding case, for the entire life of the state that proclaimed it, because the conditions for ending it, the disappearance of antagonism, the withering of the state, were defined so that their non-arrival could always be explained and never be admitted.

Other political traditions have faced the identical structural temptation and built defences against it, which is what makes the comparison fair rather than rhetorical. Liberal constitutionalism was born in transitions, in settlements after civil wars and revolutions, and its characteristic inventions, the entrenched right, the independent court, the loyal opposition, are technologies for surviving the between: devices that bind the victorious precisely when they feel most entitled to be unbound, on the theory, learned in blood, that the emergency is when the binding matters. Even the liberal tradition fails its own technologies often enough, in internments, emergencies, and colonial exceptions, and the record of those failures is part of why the Marxist critique of its hypocrisy retains force. The difference the charge identifies is one of equipment and intention at the moment of victory. The liberal founder, whatever his practice, hands his dissident a text and a court. The classical Marxist founder hands his dissident a prediction, that the dissident’s grievance will shortly lose its social basis, and the prediction is cold comfort in the cell. A theory of the transition worthy of the tradition’s own standards would have specified the between’s constitution with more care than the after’s, since the between is where its adherents and its victims alike would actually live. That no such specification exists, in a literature that specified the after’s distribution rules down to the labour certificate and the deduction for the common fund, is the single most concrete demonstration the article can offer of where the tradition’s attention was and was not directed.

The Eurocommunist revision can be reread, profitably, as the moment the tradition grasped the transition problem and inverted the sequence. Its doctrine amounted to the claim that there is no between exempt from the rules: that the road is walked under the constitution from the first step, that the freedoms are not the after’s luxury but the between’s method, and that a movement which cannot advance under pluralism has discovered something about its support rather than something about pluralism. Read so, the revision is the deepest of the replies, deeper than its authors’ programmatic language sometimes suggests, because it abandons the temporal structure that generated the gap rather than decorating the gap with promises. Its abandonment is also the measure of what the classical theory would have to surrender to answer the charge completely: not a policy, not a tactic, but its philosophy of the interim, the licensed exception between the critique and the constitution. The ledger’s partial verdicts encode exactly this. The components are partially answered where the tradition’s revisers rebuilt the sequence, and standing where the sequence, and the silences it licenses, remain the doctrine of the canon itself.

The plan and the ballot: the economic constitution gap

One province of the design gap is so large that it deserves separate survey, because it concerns the daily substance of government in the society the tradition proposed. Abolish the market as the coordinator of production, as the classical programme intended, and every allocation the market performed silently must be performed by a decision someone makes and someone can contest. What is produced, in what quantities, at what quality, for which regions, with what sacrifices of present consumption to future investment: under capitalism these outcomes emerge from prices, badly and unequally, but without a decider who can be lobbied, captured, or voted out for them. Under a planned order they are choices, and the theory of democracy for such an order must specify the chooser, the procedure, and the appeal. The classical texts specify none of the three. Marx’s remarks on planning are confined to the image of freely associated producers regulating their interchange with nature rationally and to the distribution rules of the Gotha memorandum’s two phases. Between the image and a ministry there is nothing, and the nothing governed the actual construction of the planned economies, whose institutions, the ministry, the planning commission, the enterprise under instruction, were improvised by administrators and then theorised after the fact by a literature of apologetics the tradition’s serious minds mostly declined to write.

Who decides what a planned economy produces?

In the classical texts, nobody is specified: the associated producers decide, by a procedure never described. In the historical planned economies, ministries and planning commissions decided, through bargaining among officials. The absence of a designed procedure for aggregating the society’s preferences is the economic form of the article’s central gap.

The information problem that critics of planning raised from the nineteen twenties onward intersects the democracy problem at exactly this point, and the intersection is rarely drawn. The familiar objection, associated with Mises and Hayek, is epistemic: without prices formed in exchange, planners cannot know what anything costs or what anyone wants. The democratic objection is distinct and, for the present subject, more damaging. Even supposing the information gathered, by survey, by statistics, by the ingenious trial and error of the market socialists, someone must aggregate the wants of millions into a single plan, and aggregation is a political act wearing a technical costume. Consumers want more housing and more leisure and more investment in the future and lower sacrifices in the present, and no arithmetic reconciles the wants into a plan without a rule for whose want prevails when they conflict. Liberal economies evade the question by dispersing it into markets and budgets, where it is answered brutally by income and quietly by default. A planned economy must answer it in the open, by a constitution for the plan: by votes in assemblies on priorities, by delegates bargaining under mandate, by experts optimising a welfare function a politician hands them. Each answer creates a different politics, a different locus of power, a different opposition’s grievance. The tradition, which had thought so penetratingly about how the aggregation of votes conceals the inequality of the voters, did not think about how the aggregation of needs would conceal, or reveal, the inequality of the needy, and the planners who inherited the silence filled it with the only aggregation they knew: the hierarchy’s own preferences, laundered through the technical costume until the costume was mistaken for the thing itself.

The market socialist tradition, from Oskar Lange’s model of the nineteen thirties onward, is the honourable exception that confirms the pattern’s shape, because its authors saw the aggregation problem and engineered answers: a planning board setting prices by trial and error, managers instructed to follow rules, consumers sovereign in their purchases, and, in the democratic versions, an investment share of the social product fixed by parliament. Observe what the exception required. Lange’s model works by simulating the market inside a socialist constitution and by assigning the genuinely political residue, the rate of accumulation, to an elected body, which is to say it solves the economic constitution by importing both the market’s procedure and parliament’s. The model’s critics within the tradition despised it for exactly those imports, and its defenders outside the tradition prized it for them. Either way, the lesson for the ledger is constant. Where the tradition’s heirs produced a working design for the economic province, they produced it the way Poulantzas produced his politics and the Eurocommunists their pluralism: by taking the missing machinery from the storehouse of the order the canon had criticised, and fitting it, with whatever modifications, to socialist purposes. The storehouse has been generous. The canon, on its own, remains what Bobbio found on the shelf in 1975: the critique in many volumes, the constitution in a pamphlet, and the plan, the largest single object a modern state administers, in a silence that a century of governing never managed to fill with anything but improvisation. That is the gap measured at its widest, and it is why the asymmetry of the tradition is not a scholar’s quibble but the load-bearing fact of its history: the theory knew, better than any rival, what was wrong with the house it inherited, and it had not drawn the plans for its own.

What remains open

An audit that ends with verdicts owes its reader a statement of what the verdicts leave undecided, and this subject leaves a great deal undecided in both directions. It remains open whether the design fragments now scattered through the tradition’s later literature can be assembled into a doctrine with the coherence and authority the canon gives the critique; the fragments were written by scholars and movements working in liberal orders, under their protections, and a doctrine assembled from them would carry that parentage in its structure, which partisans of the canon would read as contamination and its authors as maturity. It remains open whether the critique itself requires revision in the light of the same century that exposed the design gap, since the democracies the critique described have since extended suffrage to universality, built welfare states under electoral pressure, and in several countries subjected concentrated wealth to forms of publicity and regulation the nineteenth century did not know, while also generating new concentrations, of media, of data, of platform power, that the critique’s categories fit with unsettling ease. A critique that fits every era equally risks the unfalsifiability the tradition’s critics in another debate charged against it, and its defenders owe the same discipline demanded throughout this article: the specification of what evidence would count against it. It remains open, finally, whether any theory of society can complete both halves of the democratic problem at once, the diagnosis of power’s deformation of procedures and the design of procedures proof against power, or whether the two tasks require a permanent tension between traditions, each auditing the other’s characteristic blindness. Bobbio, who spent a life in that tension, ended his own account of democracy not with a system vindicated but with a list of promises unkept by every regime he knew, and the list’s impartiality is the note on which a fair audit should close. The charge against the Marxist theory of democracy is sustained in its precise form and graded in the ledger. The question the charge was always in service of, how a society of equals would actually govern itself, remains where the tradition’s best minds left it: open, and waiting for its volume.

For the scholarly reader, the open question has a practical corollary worth stating plainly, because it governs how the whole subject should be studied from here. The materials for the missing volume exist, but they exist under other flags: in the Eurocommunist programmes and their critics, in the market socialist blueprints and their referees, in the case files of participatory institutions, in the constitutional settlements that successor societies negotiated after 1989. Studying the Marxist theory of democracy therefore can no longer mean studying the canon alone and reporting its silences, necessary as that report remains. It means following the tradition’s question into the literatures where its answers were attempted, and grading each attempt by the ledger’s standard: does it specify who decides, who dissents, who judges, and who removes, under rules that bind the theory’s own adherents when they hold power and protect them when they do not. By that standard the audit of the tradition is not finished, because the tradition, in its scattered and self-critical afterlife, is not finished either. The verdicts recorded here are verdicts on the canon and on the first century of its career. The book they audit is still being written by other hands.

For students and exam candidates

This topic is a standard evaluation prompt, and the marking logic is predictable: candidates are rewarded for structure, not for allegiance. The highest-value structure is the one used here. Separate the critique from the design question, state the charge in its five components, give Bobbio as its precise author and Poulantzas as the strongest internal reply, and grade each component separately rather than delivering a single verdict on a whole tradition. Two errors lose the most marks. The first is treating bourgeois democracy as a synonym for fake democracy, which signals that the candidate has not read the critique they are evaluating. The second is citing the historical record as a self-sufficient refutation, without stating the mechanism, the theory’s missing safeguards, by which the record is supposed to bear on the texts. Keep a paired file for the topic, the charge in one column and the replies in the other, and revise from cases: the Commune for the design sketch, the 1936 Soviet constitution for rights without machinery, the Italian debate for the internal repair. The study notes at VaultBook suit exactly this kind of two-column revision, with the ledger’s rows as headings and the verdicts left for the candidate to defend in their own words.

For teachers and seminar leaders

The seminar that teaches this topic best is run as a hearing. Assign the charge first, in Bobbio’s own framing, and require students to state it to the satisfaction of an imagined Marxist before any reply is permitted; assign Poulantzas second, and require the same discipline in reverse. The pedagogical centre is the distinction between the two questions, critique and design, and the moment a cohort grasps that a thinker can be right in the first and absent in the second, the topic stops being a loyalty exercise and becomes an education in how theories are built and left unfinished. Productive exercises include the Commune inventory, in which students list what the model specifies and then what a constitution for a large society would still require; the thought experiment of the organised opposition in the post-revolutionary society, whose constitution the students must draft a clause for; and the comparison of constitutions on paper, liberal and Soviet, as a lesson in the difference between rights declared and rights machined with enforcement. Guard against the two classroom failure modes: the session that becomes a referendum on the twentieth century, in which the texts vanish, and the session that becomes a textual sanctuary, in which the century is never mentioned. The topic exists because both the texts and the century are real, and the syllabus should force them into the same room.

For researchers, historians, and journalists

For researchers, the live frontier is the uncanonised design literature: the Eurocommunist archive, the market socialist models, the participatory institutions documented in Latin America and elsewhere, and the question of whether these fragments can be assembled into a doctrine a movement could actually carry. Historians have a parallel task in the record’s fine grain, distinguishing what each regime’s arrangements owed to the theory it claimed, what they owed to war, inheritance, and the logic of a besieged apparatus, and resisting both the prosecutorial summary and the apologetic one. Journalists meet this subject in its degraded form, as the exchange of slogans about whether a movement or a politician is democratic, and the accurate form of the charge is a professional asset precisely because it is narrower than the slogan. The charge is not that a tradition hates elections. It is that a tradition produced a powerful account of democracy’s constraints and an unfinished account of democracy’s institutions, that its own best members identified the gap fifty years ago, and that the repairs remain partial. Reported in that form, with Bobbio named and Poulantzas answering, the story is both true and more interesting than the version in which one side is simply lying.

Frequently Asked Questions

Q: Does Marxism have a theory of democracy?

It has a well-developed critique of liberal democracy’s limits under concentrated ownership, and a thin account of the institutions that would replace it. The Commune model of elected recallable delegates is the main proposal. Critics argue the absence of a theory of opposition and rights is a serious gap. Researchers generally grade the two halves separately: the critique is a substantial body of theory with mechanisms that political science has partly absorbed, while the institutional half consists of the Commune sketch, the transitional doctrine, and the later repairs by Eurocommunists and democratic theorists.

Q: Is liberal democracy just bourgeois democracy?

No, and the phrase does not claim what its sound suggests. In the tradition’s serious usage, bourgeois democracy names a system in which democratic procedures are genuine but operate within limits set by the private ownership of productive wealth: the agenda is filtered by money and media ownership, and governments are constrained by their dependence on private investment. Elections change governments and sometimes change a great deal else. The argument is about the boundary of what they can change, in particular whether they can change the ownership structure itself, and that argument is empirical, pursued by scholars inside and outside the tradition with evidence on both sides.

Q: What institutions would a socialist democracy have?

The classical answer is the Commune model generalised: elected and recallable delegates, officials on ordinary wages, no standing army or permanent bureaucracy, and federation from local communes upward. The revised answer, from Poulantzas and the Eurocommunists, keeps representative institutions, universal suffrage, plural parties, and entrenched freedoms, and adds direct democracy at the base through councils, self-management, and popular initiatives. Later design work adds participatory budgeting, plural forms of social ownership, and dispersed media ownership. No single version has been canonised by the tradition as a whole, and none has been operated as the constitution of a state, which is the factual core of the charge this article examines.

Q: What does Marxism say about individual rights?

The canonical texts treat rights analytically and deflationarily: the 1843 essay reads them as the legal form of the isolated property owner, valuable as far as political emancipation goes and insufficient beyond it. Marx’s practice was more protective than the theory’s tone; he defended press freedom and the freedoms of combination as conditions of the workers’ movement itself. What the canon lacks is an account of rights as permanent constraints on the post-revolutionary state, enforceable by independent courts against the majority’s own government. Later Marxists, above all the Eurocommunists, supplied the omission by declaring the freedoms irreversible, a repair that concedes the original gap while closing it prospectively.

Q: What did Bobbio argue about Marxism and politics?

Norberto Bobbio, the Italian theorist of democracy, argued in the nineteen seventies that Marxism possesses a theory of the capitalist state and of its extinction but no theory of the socialist state as a set of institutions. He asked by what rules a post-capitalist society would reach decisions, protect rights, and transfer power, and he held that democracy consists essentially in such rules, binding winners and losers alike. Because Bobbio was a socialist addressing Marxists in their own journals, the charge could not be dismissed as hostility, and it drew the Italian debate and Poulantzas’s late work in reply. His is the hardest form of the charge because it grants the critique and contests only the empty ground.

Q: Is there room for opposition parties in Marxist theory?

In the classical theory, effectively no, and the reason is structural rather than incidental: if political conflict expresses class antagonism, then after the abolition of classes organised opposition has no principled ground and appears as residue or obstruction. The transitional texts make no provision for parties other than the one carrying the transition, and the regimes of the tradition prohibited or licensed opposition accordingly. Marxists of the Eurocommunist generation revised the position explicitly, affirming pluralism and alternation as permanent features of a democratic socialism. The revision is widely regarded as necessary and is equally widely recognised as a revision, imported into the theory rather than derived from its classical premises.

Q: What is the democratic road to socialism?

It is the strategy developed by the Eurocommunist parties of Western Europe in the nineteen seventies and theorised most rigorously by Poulantzas: the transformation of society pursued through elections and consent, within a pluralist constitution whose freedoms are accepted as permanent, combining representative institutions with expanding direct democracy at the base. Its parties accepted alternation in office in advance, renounced the model of smashing the state in a single moment, and treated the democratisation of the state apparatus as the transition itself. The road was never travelled to its end by a governing party, so its institutional promises remain programmatic, but as a doctrine it is the tradition’s most complete answer to the democracy charge.

Q: Does the historical record settle the democracy question?

Not by itself, and the article’s argument depends on seeing why. A record settles a theoretical question only through a specified connection, and here the connection is the theory’s silences: where a doctrine provides no safeguard, practice fills the absence, and the uniformity of the filling, the disappearance of pluralism in every implementing regime across very different conditions, is legitimate evidence about the doctrine. By the same standard, the record also acquits parts of the theory: Marxist parties have contested elections and accepted defeats inside liberal constitutions, and the critique of formal democracy is not touched by the record of regimes that claimed it. The record is a witness, not a judge.

Q: What is the difference between political and human emancipation?

Political emancipation, in the 1843 essay, is the achievement of equal citizenship within the existing order: the state declares men equal regardless of property or religion while leaving both fully powerful in civil life. Human emancipation names the overcoming of that split, the condition in which individuals no longer confront their own collective power as an alien political sphere standing over them. The distinction is the seed of the whole critique, since it locates democracy’s limit inside its triumph. It is also the seed of the charge, since the condition beyond the split is described in the language of reconciliation rather than of institutions, and everything the ledger records missing is missing from that description.

Q: Have Marxists answered the democracy charge?

Partially, and the honest answer requires the ledger’s distinctions. The Commune model answers the substitution charge in sketch. The careful reading of the emancipation distinction answers part of the rights charge. The Eurocommunist doctrines and Poulantzas’s final book answer the opposition charge at the level of programme by affirming pluralism, alternation, and permanent freedoms, and the later design literature has begun the institutional work in earnest. What has not been produced is a canonical, worked-out constitutional theory owned by the tradition as a whole and tested in government. The strongest respondents were also the charge’s strongest witnesses, since each began by conceding that the gap Bobbio identified was real.

Q: What institutions does the Commune model actually specify?

Election of councillors by universal suffrage in districts, responsibility and recall at short notice, the fusion of legislative and executive functions in a working body, public functions performed by elected and revocable agents including judges and police, a ceiling on officials’ pay set at a workman’s wage, suppression of the standing army in favour of the armed population, disestablishment of the church, and a federation of communes delegating recallable representatives upward to a national body. The specification is real and aimed squarely at bureaucratic caste. It does not extend to parties, a rights charter, an independent judiciary, protection of minorities, intergovernmental disputes, or the direction and audit of a planned economy.

Q: How should a student answer the charge that Marxism lacks a politics?

Begin by refusing the slogan form and crediting the charge with its precise author and scope: Bobbio, a socialist, alleging a missing theory of institutions rather than a wicked intention. Then separate the critique from the design question and grade five components individually: replacement institutions, rights, opposition, division of powers, and the record. Deploy the Commune as the partial answer on design, the emancipation distinction on rights, Poulantzas and Eurocommunism on pluralism, and concede what the ledger concedes, above all the dissenter’s missing lever under fused powers. Examiners reward the candidate who states what would count as a complete answer and measures the tradition against that standard.

Q: Why does concentrated ownership limit formal political equality?

Because the resources politics consumes, money, time, education, organisational continuity, and access to publicity, are distributed along the same gradient as wealth, so that influence over agendas and candidacies is unequal before any vote is cast. Because campaigns and media are privately financed, the range of viable programmes is filtered by the tolerance of those who fund them. And because governments depend on private investment for employment and revenue, an elected majority faces a standing veto exercised through capital flight and investment strikes rather than through parliament. None of the mechanisms requires conspiracy; all follow from combining democratic procedures with a privately owned economy, which is why scholars without Marxist commitments have restated them as ordinary political science.

Q: What safeguards for dissent does the tradition propose or omit?

Proposed: recall of delegates, publicity of proceedings, short mandates, rotation, and the accountability of officials paid at ordinary wages, all directed at protecting the majority of electors against a crystallising political caste. Omitted: rights entrenched against majorities and enforceable by courts the rulers do not control; legal status and resources for an organised opposition; a press independent of the organs of power; and any procedure for the citizen who dissents not from a delegate’s conduct but from the system’s direction. The proposed safeguards are genuine, and their direction is the finding: every one of them guards against betrayal from above, and none guards against the majority itself, which is the case constitutionalism exists for.

Q: What makes Bobbio’s version of the charge the hardest to answer?

Three things together. Its author was a socialist and a theorist of democracy of the first rank, so the charge arrives as an internal audit rather than an enemy’s brief. Its scope is narrow and technical, a missing theory of institutions and of the rules by which power is limited and transferred, so the tradition’s strengths, its critique, its intentions, its enemies’ records, are all beside the point. And its standard is one the tradition’s own best minds accepted: Poulantzas answered by rebuilding his position around representative institutions and pluralism, which is a concession in the form of a construction project. A charge that the strongest respondent answers by changing his theory has been graded by the only jury that matters.