Unite, the Union Bureaucracy & the Ballot
Reading Ernest Mandel through Unite, the Teamsters and Comisiones Obreras
Same Criteria, Opposite Conclusions
The ballot is running through Unite this summer, and the socialist left cannot agree on what to do with it. Sharon Graham, elected general secretary in 2021 as the insurgent, is the incumbent now. Simon Dubbins, the union’s long-standing international director, is challenging her as the candidate of United Left, the broad-left faction. And three currents on the revolutionary left have split along lines that will look, by the end of this piece, almost predictable.
One current gives Graham critical support, resting the case on her industrial record: very large numbers of disputes called, a high success rate, substantial sums won for members. It says plainly that it is not an uncritical supporter and goes on fighting for a socialist programme she does not hold. A second backs Dubbins, on the ground that the decisive questions sit on the political and internationalist terrain: the relationship to Labour, the survival of an organised collective left, and the charge that a leadership built around one person’s organising model is a concentration of power by another route. A third declares a plague on both houses. Two bureaucrats, it says; the incumbent first took the post on the active support of only a few per cent of eligible members; build a rank-and-file pole rather than vote for either.
Same tradition. Same criteria. Opposite conclusions. That three-way split is what happens whenever a particular set of criteria are forced to compete, in Britain or anywhere else, and the criteria belong to Ernest Mandel.
When an independent socialist stands in front of a union ballot, the practical question, support this candidate or oppose them or abstain, looks like a matter of weighing personalities and manifestos. It is not. Underneath it sits a theory of what a trade union is, what its officialdom is, and what a vote can and cannot do to either. The richest source for that theory in our tradition is Mandel; the most useful way to make his abstractions bite is to watch them operate in two contests International Viewpoint documented in full, the reform insurgency in the American Teamsters and the fight for the ‘soul’ of Spain’s Comisiones Obreras, and then carry the result back to the ballot papers now going out across Unite.
What Mandel means by ‘bureaucracy’
Mandel’s polemic ‘Sectarianism vs Revolutionary Marxism’, the expanded text of his November 1994 debate with the Spartacist League published in the Bulletin in Defense of Marxism, is ostensibly about the state bureaucracies of the former Soviet bloc, not about union officials. But the apparatus he builds there is the one the tradition turns on the union, and he says so in passing: the Stalinist bureaucrats, he writes, strangled democracy ‘within the Communist Party and the trade unions’, and did it to defend their ‘exorbitant material privileges’. The whole theory sits in that sentence.
The move that governs everything after, following Trotsky, is to treat the bureaucracy as a specific social layer with material interests to defend, not a collection of bad or corrupt individuals. This is the break from moralism. If the bureaucracy is just bad apples, the remedy is honest people. If it is a layer thrown up by a structural position, then honest people who occupy that position will, over time, be shaped by its interests. The problem is sociological, not ethical, and so it recurs.
And the bureaucracy plays a genuinely contradictory role, which is the point that matters most at a ballot box. Mandel’s own example is the Stalinist parties that suppressed capitalism, abolishing it across Eastern Europe after the war and in China, and did so, he writes, ‘in order to extend its power and privileges’, not from any commitment to socialism. ‘Suppressing capitalism,’ he goes on, ‘is not identical to maintaining it’; whoever denies it, the state-capitalist sect of the British SWP included, ends up claiming there was no structural difference between Mao’s China and Chiang Kai-shek’s. Carry the point to the union and it becomes the one thing to hold onto: a bureaucracy can deliver real gains, win strikes, raise wages, and remain a bureaucracy, because the test is never whether it delivers. The test is whether it develops or suppresses the members’ own capacity to act for themselves.
That test rests on workers’ self-emancipation, the principle Mandel goes back to again and again: Marx’s formula from the Inaugural Address of the First International, that the emancipation of the workers will be the work of the workers themselves, and the corollary from the Theses on Feuerbach that the educators themselves need educating. No party, no leadership, no apparatus, however wise, substitutes for the self-activity of the class. The opposite error he calls ‘substitutionism’, the apparatus acting in place of the members rather than through them, and against the Spartacists’ charge that opposing it is somehow anti-Leninist he roots the struggle against it in Marx himself. Democratic self-activity of the wage earners, on a public and large scale, is for Mandel a condition of socialism, not an optional nicety.
Which is why the programme matters as much as the personnel. Against substitutionism Mandel sets the democratic demands of the Transitional Program: bodies elected and recallable, workers’ control and veto at plant level, a ban on the sacking of union officials and activists, free election of all representative bodies. These are not trimmings on the real economic demands. They are the mechanism by which the ranks keep their own officials under them. A union programme silent on democracy has already conceded the ground to the apparatus.
None of this licenses sectarian purism, and Mandel is at pains to say so. His running complaint against the Spartacists, and in a milder key against Lutte Ouvrière, is that they abandon the bridge between revolutionary conclusions and workers’ actual consciousness. He faults Lutte Ouvrière for refusing clear critical support to the Socialist Jospin against the conservative Chirac in 1995, reading that refusal as a failure to grasp the united front. It generalises straight to union elections. You may back a limited, imperfect, even bureaucratic candidate against a worse one, but critically, telling the truth about their limits, never sowing illusions in the apparatus.
The union bureaucracy in particular
Translate all this to the union and a familiar figure walks in: the full-time official as a mediating layer between labour and capital, with a material stake in the stability of the relationship it mediates. The official’s salary, security, pension and status do not rise and fall with the shop floor the way a member’s do. The official has something the member does not, the apparatus itself, and so acquires a structural preference for whatever preserves it: negotiation over confrontation, routine over mobilisation, the deal at the top over the assembly at the base.
Which is why the contradictory role matters most here. The union bureaucracy is not the employer. It organises workers, leads real fights, often wins them. But it fights within limits set by its own interest in survival and recognition, and it grows nervous the moment the members’ self-activity threatens to outrun its control. The dangerous figure, from a socialist standpoint, is not the openly right-wing, employer-friendly official; that one is easy to oppose. It is the left bureaucrat who talks militancy while doing deals in the background and keeping the ranks demobilised between set-pieces.
Two cases show the layer in its two forms. The Teamsters give it naked, corrupt, right-wing. CC.OO give it soft, left-talking, modernising. In both, the decisive question was Mandel’s, self-activity or substitution, and in both the answer decided how the revolutionary left voted.
The Teamsters: the bureaucracy stripped bare
The International Brotherhood of Teamsters is the textbook illustration of the apparatus with nothing to hide behind. By the time the reform movement came to a head, three of the union’s recent presidents had been sent to prison, Dave Beck, Jimmy Hoffa and Roy Williams, and a fourth, Jackie Presser, died under federal indictment; the top officials paid themselves two or three salaries running to hundreds of thousands of dollars a year while members’ pay and conditions fell; and the leadership had engineered an ‘election’ system that returned incumbents or their hand-picked successors term after term. International Viewpoint reported one reformer’s summary of how business was done: decisions taken ‘behind closed doors by a handful of men’.
Mandel’s sociological point is vivid here. The corruption was not incidental to the apparatus, it was the apparatus’s mode of self-perpetuation. Multiple salaries, backroom succession, suppressed member voting: these were the concrete shapes taken by a specific social layer defending its material interests.
Against it stood Teamsters for a Democratic Union, a rank-and-file reform caucus already sixteen years old, organised and patient and ready when the opening came. When the 1989 anti-racketeering settlement forced the union into a one-member-one-vote election, TDU had the machine to fight it. It endorsed Ron Carey, not a TDU member but a local officer with a fighting record and open contempt for the top leadership, and the reform slate around him was half working Teamsters and included the first woman and the first Latino ever to sit on the union’s executive board, along with an African-American member. Carey won in 1991 against an old guard tied to organised crime.
One detail shows the tradition’s anti-sectarianism in practice. The reformers accepted the government’s intervention, the court-supervised ballot, as the condition of a fair election, though much of the American labour left opposed state interference in unions on principle. A concrete-conditions judgement overriding a general rule; the bridge Mandel insists on. Purism would have refused the only mechanism that made the members’ own victory possible.
Then the second act, which is where it turns instructive. Carey was narrowly re-elected in 1996 and the result annulled: an illegal scheme had run some $800,000 of the union’s own political money, through outside consultants, back into his campaign. He was barred from the re-run. The reform movement fractured, TDU backing the clean candidate Tom Leedham while a more conservative wing floated a wavering-centre ticket, and James P. Hoffa, son of the union’s most famous boss and candidate of the old guard’s revival, moved into position to take the union, which he duly won in the 1998 re-run.
The question was hard and International Viewpoint did not flinch from it. Carey was still, plainly, preferable to Hoffa. Keep defending him, then? The ‘Teamster Trouble’ coverage answered no: diverting the members’ own dues money is an offence against their democratic rights, and radicals could defend it only by looking away from the evidence. Here the framework does its sharpest work. The very principle that justified backing the insurgency in 1991, the members’ democratic control of their own organisation, now forbade covering for the reform leader in 1998. Support had always been conditional on programme and conduct, never on the person. The reformer, become the incumbent, inherited exactly the scrutiny reformers had once aimed at the old guard. The fixed point did not move.
Comisiones Obreras: the bureaucracy that speaks left
If the Teamsters show the crude form, the Sixth Congress of Spain’s Comisiones Obreras, January 1996, shows the subtle one. CC.OO had been forged in clandestine struggle under Franco as a self-consciously new kind of union: alternative, anti-capitalist, pluralist, autonomous, participative, socio-political. No corruption scandal, no jailed presidents, no funnelled cash. And yet International Viewpoint, printing both sides, ran the congress under the headline of a fight for the union’s ‘soul’.
The dividing lines fell on Mandel’s two axes, programme and democracy. On programme, the majority around General Secretary Gutiérrez was charged by the minority ‘Critical Sector’ with accommodation: treating the neoliberal architecture of the Maastricht Treaty as inescapable, reducing the fight against Spain’s labour ‘counter-reform’ to piecemeal bargaining, putting negotiation before mobilisation everywhere. Against which the Critical Sector set concrete fighting demands: a referendum on Maastricht, a 35-hour week with no loss of pay, reversal of the counter-reform, defence of the public sector.
On democracy, the majority’s statute changes would have limited members’ rights, increased the powers of the apparatus, stripped dissident voices from the executive. The congress ran on designated rather than directly elected delegates. Afterwards Gutiérrez governed through hand-picked ‘counsellors’ and shut the minority representatives out of any real responsibility. The union’s historic founder, Marcelino Camacho, had already resigned the presidency in 1995 over the leadership’s drift from the Communist Party and its rightward turn; when the Critical Sector proposed him for the post again at the 1996 congress, the officialist majority voted the candidacy down, 366 to 571. The Critical Sector campaigned for proportional representation, regularly renewed and recallable posts, rank-and-file assemblies, and took more than a third of the delegates and seven of the nineteen executive seats, with outright majorities in many of the base assemblies where ordinary members actually took part.
This is the left bureaucrat in Mandel’s sense. No theft, no thuggery: a steady concentration of power in the apparatus, a preference for the deal over the mobilisation, and, the giveaway, a set of rule changes whose whole purpose was to insulate the leadership from the ranks. Substitutionism wearing a modernising, professional face, which makes it harder to name and so more instructive.
The revealing thing about CC.OO is that the revolutionary left did not agree on how to vote. Some comrades gave critical support to the Gutiérrez majority; others were in the Critical Sector; International Viewpoint published both. Joaquín Nieto’s case for critical support weighted the congress’s confirmation of the union’s autonomy from the Communist Party, and its membership and workplace-election gains, above the democratic complaints. Jesús Albarracín, Pedro Montes and Agustín Moreno weighted the programmatic accommodation and the bureaucratic power-grab as decisive. Same tradition, same criteria, opposite conclusions, because the criteria genuinely pull apart when a concrete situation forces them into competition.
Back to Unite
Set beside Mandel, the two cases yield a method, not a checklist, and the quickest way to see the method work is to carry it back to the ballot now running through Unite, which has the useful property of letting one person occupy both structural positions in turn, insurgent and then incumbent: the Carey arc replayed.
The papers went out in the middle of July and the disagreement started the same week. Inside several socialist WhatsApp this summer, comrades in the Your Party diaspora who share traditions, lessons and almost a framework could not converge: one voting for the challenger while his partner, at the same kitchen table, voted for the incumbent; one crossing both names off the ballot and writing Free Palestine across it; another writing something about the climate and trans rights that neither candidate would say aloud; several settling on a spoiled paper as the only honest mark left to make. Not confusion. The evil of two lessers, as one of them put it, and no way to rank the two evils that did not first require choosing which question the union is for. The criteria pulling apart in a single room, before they ever reach a printed page.
Recall the 2021 contest. Three left candidates, Steve Turner, Howard Beckett and Sharon Graham, faced one centre-right candidate, Gerard Coyne. United Left balloted internally and chose Turner as its single official candidate; Beckett ran anyway before withdrawing to back Turner; Graham, then head of the union’s organising and leverage department, ran without seeking the faction’s endorsement and refused to stand aside so the left could unite behind Turner. She won, 37.7% to Turner’s 33.7% and Coyne’s 28.5%, on a turnout of roughly 10%. ‘Back to the workplace’; no ‘blank cheque’ to the Labour Party: the workplace-organising, self-activity register. The criticism from her left was that industrial militancy without an explicit political project collapses into economism.
2026 inverts the positions, and the three currents now line up on the axes the two cases taught us to look for. Read the three cases for endorsement side by side and the strange thing is how little they disagree about the facts.
The critical-support-for-Graham current rests on the delivery. Her campaign points to a record over five years of well over a thousand disputes and, on its own accounting, more than £700 million put into members’ pockets from successful disputes alone; the Socialist Party’s people in Unite back her re-election on those grounds of industrial delivery, counting more than 1,500 disputes at a high success rate and reserving the right to criticise, which is critical support by the book. That is the Nieto position from CC.OO transposed. The obvious reading is that delivery vindicates her. But that is not quite right either: the record is real, and the question it steps around is also real, because Mandel’s warning is precisely that gains never settle the matter, a bureaucracy being fully capable of delivering to the members it also keeps as spectators. Delivery answers the delivery test. It does not touch the bureaucracy test, which asks whether the model builds the members’ own capacity to act or runs their disputes for them.
The back-Dubbins current rests on programme and on democracy: the political and internationalist terrain, the relationship to Labour, the survival of an organised collective left, and the charge that a personalised leadership built around one organising model is a concentration of power dressed as dynamism, with the Labourist and caucus-patronage baggage named as its cost. That is the Albarracín–Montes–Moreno position transposed, treating the political battlefield and the rule of the ranks as the axes that decide.
The plague-on-both-houses current rests on the last axis, rank-and-file democracy unmet by either, noting that the incumbent first took the post on the active support of a few per cent of eligible members, and concludes that neither candidate clears the bar: principled abstention combined with base-building, which is exactly where part of the revolutionary left landed at CC.OO. Legitimate on Mandel’s terms, that one, but only where it is genuinely tied to organising and does not lapse into the fastidious purism he condemns, the abstention that mistakes clean hands for a position.
And, in the Carey pattern, some of Graham’s own 2021 backers have publicly switched to the challenger, citing governance and internal-culture concerns. Those allegations come from her opponents and from a hostile press, and should be weighed accordingly. But the structure is the one the Teamsters teach. The insurgent who becomes the incumbent inherits the scrutiny once aimed at the old guard, and the fixed point, the members’ democratic control of their own union, does not move to accommodate her.
The reason there are strong arguments on every side is not indecision. It is that the criteria pull apart under real conditions, and where a socialist lands depends on which axis they treat as primary: industrial militancy as the material base of any politics, and you give critical support to the incumbent; the political and Labour battlefield and the survival of an organised left, and you back the challenger; rank-and-file democracy above all, unmet by either, and you abstain and organise. The choice of candidate is downstream of the choice of axis, which is precisely the disagreement the revolutionary left had, in good faith and in the open, over Comisiones Obreras thirty years ago.
Which is the point at which the framework stops handing out answers. It can tell you what each current is weighting and why each is defensible; it cannot, on its own, say which axis ought to be primary in Unite in the summer of 2026, because that is a judgement about the concrete balance of forces in this union at this moment, and a method was never going to make it for you. Perhaps that is where it should be left, or nearly.
The measure that outlasts the count
Mandel’s lasting contribution is to refuse the two easy answers. The bureaucracy is not a gang of crooks to be swapped for honest people, the moralist error, nor an irredeemable enemy no different from the boss, the sectarian error. It is a social layer with its own interests, capable of leading real fights and of strangling its members’ initiative, sometimes in the same breath. A union election is one moment in the longer struggle over which tendency prevails, and a vote, however cast, is worth far less than the rank-and-file capacity built around it.
Which is why, on Mandel’s account, the decisive activity is never the ballot but the assembly: the caucus, the shop-floor network that can hold any leadership, reform or old guard, right or left, to account after the count. The Teamsters won a reform president and then had to watch the fixed point of democratic control judge him too. Comisiones Obreras kept its founder’s memory and lost his presidency to a modernising apparatus. In both, the members’ self-activity, not the winning candidate, was the real measure of the outcome. Support, opposition and abstention are three ways of serving that measure, and the whole art is knowing, in the concrete case, which one does; which is why Unite this summer can put the same question to three currents of one tradition and get back three answers, none of them yet disproved by a count that has not happened.
Sources
The Mandel quotations are from ‘Sectarianism vs Revolutionary Marxism’, the expanded text of his 11 November 1994 debate with the Spartacist League, published as ‘World Socialist Revolution Today’, Bulletin in Defense of Marxism, no. 125 (May/June 1995). The Teamsters material draws on International Viewpoint no. 221 (February 1992), on the 1991 reform victory and Teamsters for a Democratic Union, and nos. 296, 297 and 303 (1998), on the annulled 1996 election, the financing scandal and the movement’s fracture. The Comisiones Obreras material draws on International Viewpoint no. 273 (January 1996), ‘Spain: fight for the soul of the union’, and no. 276 (April 1996), which carried the Critical Sector and majority perspectives side by side; the 1996 congress vote figures are from the union’s own record. The Unite delivery figures are as claimed by Graham’s 2026 campaign and by the Socialist Party in support of her re-election; the 2021 result is from the union’s declaration and contemporary reporting; the 2026 contest is drawn from coverage across the British left.