Whose Power is It?

Introduction

Trade unions exist to do one thing: organize the collective power of their members. That makes them different from every other organization in the civil society landscape. An advocacy NGO can function with a passive donor base. A charitable foundation can function with staff alone. A union that cannot organize a credible strike threat has no power — and without power, everything else it does (lobbying, consulting, negotiating) rests on institutional position rather than on the capacity to halt production, or otherwise impose costs, on an opponent who refuses to concede. And within the union itself, the question of who steers its course and action — the members or the staff — determines whether it has real power or only a seat at the table.

What that power is for should be said plainly, because the Dutch institutional vocabulary is built to soften it. The employment relation is not an exchange between equals who occasionally disagree. Workers sell the only thing they have to sell and need the wage this month; the employer buys, and can wait. What is bought is not a fixed quantity of work but the capacity to work, which still has to be made to yield — so pace, hours, effort and conditions are settled after the contract is signed, on the employer's ground. And the wage and the profit are shares of one product: what the firm keeps is what workers produce over and above what they are paid. None of this requires bad employers. It is built into the relation and survives any amount of goodwill on either side, which is why individual workers have no remedy within it, and why combination is the only remedy there is: a union is what workers have instead of a bargaining position, the conversion of numbers into a capacity to halt production and so to make conceding cheaper than refusing. The polder framework's founding premise — that capital and labor are "social partners" whose interests can be reconciled around a table — does not abolish that antagonism. It administers it, and, as this essay argues, on terms that have moved steadily in capital's favor.

A companion article examines how professionalization produces class relations within NGOs and social movements more broadly.^[1] But the union is where this development causes the most trouble. An environmental lobby group can set the bar for its own 'success' wherever it likes. A union has no such freedom, and when it fails, members feel it in their wages, working conditions, pension terms, and their colleagues' willingness to take part in collective action. The failure is experienced, and it is more than a matter of arithmetic — especially on complex questions like what to make of discount rates and indexation ratios and their reasonableness.

And if we look at how the membership of Dutch unions has developed over recent decades, it seems clear that members have lost confidence in the union. The sectors with the highest union density are those that have much less to fear from international competition, while several economically important sectors (IT, the temp-agency sector, business services, agriculture and horticulture, hospitality, retail — and, to a lesser extent, financial services) are barely organized or not organized at all.^[2] The low density of many of those sectors is a problem in many countries — partly because workers change employers often and many work for a given employer only briefly, or can work for them only through a temp agency, often a foreign one, which has its own reasons for ensuring that workers accrue neither ties nor rights — but it is striking that organizations whose influence depends on members and organization appear to let this run its course.

That "letting it run its course" is the heart of the puzzle. Two readings present themselves. The first is that it cannot be otherwise: in individualized branches with high turnover and fragmented employers — IT and parts of business services foremost — no bargaining structure forms at all, and the labor movement's own research (De Beer, De Burcht) shows that CAO coverage in such sectors depends more on the density of employers than of the union. The second is that it could be otherwise but does not happen: where the FNV did commit resources — the "Schoon Genoeg" cleaners' campaign, using organizing techniques grafted from the SEIU — it recruited some two thousand new members in a low-paid, migrant-dominated sector. The capacity to reach such sectors therefore exists; whether what is built there is genuine organizing or merely mobilizing is not yet settled. In other unorganized sectors that presence has taken the form of advocacy for precarious migrant workers (horticulture, meatpacking) or of losing CAO rounds (retail) — present as advocate or negotiator, not as organizer of member power.

Both readings point, by different routes, to the same underlying fact: an organization that collects its dues by automatic direct debit and whose CAOs are declared generally binding for non-members^[3] can reproduce itself financially and institutionally without organizing anyone at all. Nor does that self-reproduction rest on dues alone: a substantial share of the union's income comes from employers directly, and the yellow unions — bodies employers fund into existence to sign CAOs of their own — are only the limit case of a financing that binds the apparatus to employers rather than to the workers whose power it is supposed to be. The union's survival is decoupled from whether it gets its members moving. In the sectors that remain comparatively organized — public administration, education, care, construction, largely sheltered from international competition — the union coasts on an inherited density it does little to renew: higher than elsewhere, but eroding all the same, and in public administration and construction faster than the national average;^[4] in the exposed, adversarial sectors where organizing is expensive, slow, and conflictual, the union lets it go. That outward pattern — whole economic sectors left untouched — is the mirror image of the inward one: the shift from organizing to servicing, from members to staff. It is the same decoupling, visible once in whom the union does and does not reach, and once in who sets its course from within. That is why, of all the NGOs, the union is the sharpest case: nowhere else does the question "whose power is it?" read so directly off what the organization does and fails to do.

This article compares the history of the FNV with that of the Chicago Teachers Union (CTU) — in part because parts of the FNV deliberately looked across the Atlantic for inspiration over the past fifteen years, among other things in organizing sectors that require no formal training and in which many migrants and children of migrants work. Dutch unions operate within a corporatist framework designed to make unions 'social partners.' Over the past four decades, the staff have largely displaced the members, through an interplay of professionalization, acquiescence, and exit. In the US the decline of unions has been more dramatic in general, the differences between unions are larger and more numerous, and the analytical vocabulary is more developed. The CTU's history provides both the diagnostic framework and the evidence that the shift in power is locally reversible. The institutional contexts differ sharply as a result (corporatism vs. adversarial bargaining, proportional representation vs. a two-party system, an eroded vs. an always-minimal welfare state), but the parallels are revealing precisely because of that difference: the decline dynamic — professionalization, then the replacement of members by staff — recurs in both.

First, a few general points about organizations. In class societies, positions of power supply both agency and material security, so those who hold organizational power have reason to use it to keep that position. Walther Müller-Jentsch gives this its trade-union-specific form: the official's interest thus also lies in maintaining the intermediary position between capital and labor. This means it becomes in his interest to contain member militancy rather than mobilize it. The FNV's archives show that this is a real problem.

In No Shortcuts, Jane McAlevey developed a vocabulary for the different ways of involving members in action and in achieving organizational goals — advocacy, mobilizing, organizing. In that book she also shows how the replacement of members by staff is reversible at the level of a single union. Both the American history and that of the FNV suggest that attempts to make fundamental changes in method at the federation level provoke resistance along several axes at once. From this it follows that the FNV faces not one problem but two: an organizational problem (the shift from organizing to advocacy) and a political problem (structural subordination to the polder model and its associated party relationships). Neither can be solved without the other.

The founding bargain and what it cost

De Beer and Berntsen's overview of Dutch unionism identifies the structural origin of the problem: when the Stichting van de Arbeid (Labour Foundation) was established in 1945, "the unions accepted that they would play no role whatsoever at the workplace in exchange for a prominent role at the national and industry levels."^[5] What this meant concretely: no recognized union representation at firm level, no workplace bargaining, none of the shop-steward institution that British or American unions rely on. The unions operated at the national level (SER, Stichting van de Arbeid) and at the sector level (CAO negotiations). At the workplace itself, the legally mandated works council (ondernemingsraad) became the primary institution — elected by all employees, not just union members, with consultation and consent rights but no power to bargain over wages or to call a strike. This is what "workplace absence" means in what follows: not that the union had no members inside firms — it had members, and in places workplace cadre and company member groups (bedrijfsledengroepen) — but that it had no bargaining structure of its own at the point of production. The workplace cadre that did form were given no bargaining power to match their numbers, as the next section shows — and becoming an active kaderlid has in most places stayed harder than the union's formal openness implies.

This is the founding bargain of the polder model. The unions that made it — the socialist NVV, the Catholic NKV, the Protestant CNV — were each embedded in their own pillar. They were not the only union federation in the field. Out of the wartime resistance strikes had come the communist-led Eenheidsvakcentrale (EVC), which by the end of 1945 counted roughly 170,000 members — rivaling the NVV — and organized by industrial sector around exactly the shop-floor militancy the bargain traded away: the wildcat Rotterdam dock strikes of 1945, the actions at Hoogovens and Werkspoor. The EVC was kept out. The Stichting van de Arbeid admitted only unions that accepted its rules, which the EVC refused, and the recognized federations used their monopoly of the national table to block it from collective bargaining before it fractured in the Cold War years. Part of what the founding bargain bought its signatories, in other words, was the standing to marginalize a militant rival that was building precisely the workplace organization they had agreed to forgo — the essay's "it could be otherwise" made concrete at the origin. Militant workplace unionism was not impossible in the Dutch context; it was defeated.^[6] For three decades the bargain appeared to work — but it worked on borrowed conditions rather than on its own soundness. Postwar reconstruction and product markets still largely sheltered from international competition kept the demand for labor high; the resulting scarcity of workers gave them structural power regardless of shop-floor organization; the still-intact pillarized infrastructure embedded unions in a broader community; and an expanding economy and welfare state meant that national-level bargaining delivered real gains. The professionals at the national table could point to results.

But the bargain's viability depended on that whole conjuncture — and every prop under it was giving way. When depillarization eroded the confessional unions' distinct rationale, the NVV and NKV merged in 1976 to form the FNV with CNV remaining separate. The merger created a larger organization with a larger professional apparatus, but it coincided with the loss of the social world that had given the founding bargain its sustainability. As depillarization severed the social infrastructure, and as the labor scarcity that had substituted for shop-floor organization was undone from both sides — the end of full employment slackening demand, the entry of women into paid work and the recruitment of 'guest' workers expanding supply — the absence of workplace organization was exposed as a trap the union had no capacity to escape. Internal attempts to re-establish a workplace presence in the 1980s — which the next section examines — failed precisely because the necessary organizational infrastructure had never been built.^[7]

Several mechanisms reinforced the shift. The founding bargain itself removed the union from the workplace, so there was no organizational infrastructure through which members develop agency. The polder model meant that union officials had a place at the table irrespective of member activity. And automated dues collection removed the last material incentive for regular contact between members and staff, so the attention shifts to providing services and managing the membership at the lowest possible cost so as to minimize employee time needed to perform these tasks.^[8]

The Wassenaar Accord of 1982 sealed this by institutionalizing the polder model as the primary framework for labor relations. From that point, the werkorganisatie — the parallel structure of paid staff that holds the real operational power alongside the union's formal democratic bodies — dominated the membership. Those yellow unions operated freely because the polder framework treats any organization that signs a CAO as a legitimate social partner, and Dutch law imposes virtually no independence requirements on organizations claiming to negotiate on behalf of workers.^[9] And the FNV leadership, having committed to "poldering," found itself confronting faits accomplis that it had helped create — agreements reached in consultative bodies where employer and state interests dominated, presented to members as inevitable — and lacked the organizational capacity to challenge them.

The turn to servicing then became explicit strategy. Facing membership it was already losing — among women, migrants, and flexworkers, and in the IT and commercial-service sectors where it had never gained a foothold — the FNV's 1987 report FNV 2000 set out to make the union "recognisable" through personal services to members: the FNV, it proposed, should become a "sociale ANWB," a social version of the motorists' association. Bert Breij, a sympathetic historian of the movement, reads the move as the run-up to "a union that dismantles itself, makes itself narrower," preparing for zaakwaarnemerschap — acting as its members' agent rather than as the vehicle of their power.^[10]

The results, measured over four decades, are stark: real contractual wages have been "virtually stable since 1979." The union has, in aggregate, managed to adjust wages to inflation and nothing more — while the wage share in GDP declined from 58 percent in 1980 to 48 percent in 2018. De Beer and Berntsen's time-series analysis confirms what you would expect: members join when they see the union winning. There is a measurable positive relationship between wage gains and membership growth — which means that structural membership decline is the predictable consequence of a union delivering no real wage growth for forty years.

And while the union under-delivered on wages, the material base of the bargain kept shifting out from under it. The labor scarcity and the shelter from competition that had once done the work of organization were never recovered — they were undone further, and on a larger scale, by European integration: the single market, the eastward enlargements of 2004 and 2007, and the posted-workers arrangements that let firms staff jobs with workers hired on cheaper cross-border terms, each reopening the competition in labor and product markets that the postwar boom had suspended. The union organized almost none of it — not the women and guest workers who reshaped the postwar workforce, nor the EU-mobile and posted workers who reshaped it again — responding to these shifts, when at all, as advocate rather than organizer. That is why the sectors most exposed to them are the ones this essay opened with: barely organized, or not organized at all.

One issue is the exception that proves the rule. The entry of women into paid work did push the union to organize around something — workplace sexual harassment — and for a time it organized in the fullest sense. From around 1980, union women (the FNV's Vrouwensecretariaat, under Elske ter Veld, and members on the shop floor) forced ongewenste intimiteiten onto the agenda as a labor-conditions question rather than a private matter, over the open resistance of a male leadership that thought it beneath the class struggle — the Industriebond's Arie Groenevelt called it an insult to the male half of the working class. The methods were organizing methods: a complaints mailbox that "nearly exploded," a "black book" of assaults compiled by women at Hoogovens that dragged a reluctant Industriebond into action, activist courses under the banner "union women in revolt against sexism." And it delivered — a complaints bureau in 1985, confidential advisors, and eventually the 1994 Arbowet amendment that made every employer legally responsible for preventing harassment, aggression, and bullying.^[11] But look at what it became. Four decades on, the union's presence on workplace safety is a trust hotline open four evenings a week, a reporting point, downloadable advice leaflets, research reports that name bad employers, and letters to parliament — the servicing-and-advocacy register, staffed by professionals, with the member militancy that won the original gains nowhere in sight. The same demographic shift that eroded the bargain's material base also produced, briefly, exactly the organizing the union otherwise lacked — and the apparatus and the state between them absorbed it expertly into the model this essay is about, by giving workers to understand that the matter was settled now that it stood 'in the law.'

The guest workers are the other half of that sentence, and their case runs the opposite way. Where union women forced their issue from inside, migrant workers mostly had to act outside it: of the forty-odd wildcat strikes by Spanish, Turkish, Yugoslav, and Italian workers between 1961 and 1974 — more often over housing, food, and medical care than over wages — most erupted beyond the union structures, and the union's most consistent intervention was defensive advocacy against the state, opposing the deportation of workers who struck. The unions had welcomed guest workers to fill jobs the employers could not staff, but feared them as wage-competition and potential strikebreakers, and stayed largely passive as Turkish and Moroccan families settled in the 1970s. Migrants organized themselves where the union would not — the Komitee Marokkaanse Arbeiders Nederland (KMAN), the Turkish HTIB — and the FNV was at first fiercely opposed to migrant self-organization of any kind. Its own minority policy, formulated across the 1980s, largely called the government to account while staying thin on what the union itself would do; it began advocating in earnest only late in the decade, under government pressure, through CAO clauses and hiring programs that achieved little — defeated, as Roosblad's study documents, not only by employers but by discrimination and competition-anxiety on the union's own side.^[12] And the daughters of those workers, who entered paid work in far greater numbers than their mothers, landed in exactly the exposed, low-density sectors — cleaning, care, hospitality, logistics — where the union organizes least. The one serious push toward them, the migrant-women cleaners of "Schoon Genoeg," is the campaign the rest of this essay reads as mobilizing rather than organizing. Women's entry produced organizing the apparatus absorbed; the guest workers' produced organizing it mostly never attempted — self-help outside the union, and belated advocacy within it. The same decoupling, seen from its other side.

What the archives show

The IISG's Precaire polder report (2018), based on previously unopened FNV archival material, documents how the shift from members to staff played out inside the union.^[13] As early as 1992, the Dienstenbond's own internal project council diagnosed the problem: the professionalization of union work had created a "gulf" between the world of active cadre members and the world of professional union staff. Paid officials had "such an advantage... that cadre members always feel inferior. This inhibits cadre members from becoming active, since 'the official knows better anyway and has an answer to everything.'" The report's authors note that this was not a recent development but traced back to the origins of the modern union movement under Henri Polak: the decision to employ professional staff "already created a certain distinction between the world of the active cadre members and the world of the professional union people. This distinction has in a sense become a gulf."

The archival evidence shows that every attempt to close this gap reproduced it at a higher level. When the Industriebond tried to involve cadre members more in collective bargaining in the late 1980s, participation was high (4,500 members came to preparatory meetings) but "the influence of their participation was negligible." When the Dienstenbond decided to make "the workplace central" in its union work, the solution took the form of "professionalization of cadre activities" — training specialized cadre members to take over tasks from paid officials, which simply shifted the gap from ordinary members vs. werkorganisatie to ordinary members vs. specialized cadre. A districtshoofd wrote to the reform working group in 1988: "Let's not pretend we're really doing something, because that could backfire badly. You'd hate to be a cadre member and figure it out."

The report's conclusion is damning: the "efficiency drive" of professionalization and specialization "seems to have come at the cost of democracy in the union, and may have actually strengthened the problems that (cadre) members had identified in the 1980s." The shift from workplace organizing (bedrijfsledengroepen) to participation in works councils (ondernemingsraden) — which accelerated through the 1990s — "seems to have contributed in the longer term to undermining the union movement's strike capacity."

Why does the pattern reproduce? Not because unions are unusual in this — every organization that delegates develops a stratum with an interest in its own position, which is why the companion essay on NGOs finds the same dynamic across foundations and advocacy groups. The interest is close to universal; the question is not whether it exists but how to combat it — and combatting it grows harder as the organization scales, because scale concentrates the apparatus worth defending and raises the stakes of holding a position at all. What is specific to unions is the direction the interest points. Müller-Jentsch's analysis of unions as "intermediary organizations" locates it: the position that individual officials and the leadership of the werkorganisatie have an interest in maintaining is the mediating one, between capital and labor — which tilts the interest toward delivering a governable membership to the bargaining relationship, not toward the member militancy that could disrupt it, and tilts it harder the more institutionally embedded the union becomes.^[14] Consider the typical official: salary not tied to outcomes in the sector they negotiate for, often not from that sector and unlikely to stay long, working on a permanent contract insulated from the consequences of bad deals. Such an official has no material incentive to win — and every incentive to preserve the institutional relationships on which their position depends. When the Industriebond redesigned its participation process, the officials who designed it had a structural interest in a form of participation that preserved their own relevance — and that is what they produced. The result is a self-reinforcing loop: officials who don't organize have no basis for assessing their own power, which makes them risk-averse, which means they don't organize — and that absence means there is no countervailing pressure from below to break the cycle.

Bob Hancké's 1986 study of a Belgian public-sector union local provides the complementary mechanism — what happens on the members' side. He found that members with high class consciousness participated more at every level of union activity, were more willing to formulate and organize internal opposition, and had systematically higher democratic aspirations. His central conclusion was that the shift toward "client-oriented" unionism — treating members as consumers of union services rather than as agents of collective action — would undermine democracy, because it was precisely the members with a broader political definition of unionism who drove democratization. The relationship is mutually reinforcing: active member engagement produces political consciousness, which drives democratic participation, which sustains the engagement. The staff-run model reverses the cycle — depoliticizing members, reducing participation, entrenching the werkorganisatie.^[15]

The two mechanisms compose into a spiral. Professionalization means the union delivers less; members who recognize this leave; those who stay increasingly treat the union as a service provider. Both responses reduce the internal pressure that might contest the werkorganisatie — which enables further professionalization, further exit, further acquiescence. The membership-decline data above is the exit side of this spiral quantified: a union that delivers nothing loses the members who would have demanded that it deliver something.

Accommodation is not the same as clear sight, and neither exhausts what is going on. Take even the members who are under no illusion about the werkorganisatie: what looks like their consent is accommodation, a rational adjustment to an organization that delivers little and offers no low-cost channel for changing it. Their acquiescence is not a matter of being deceived — so focusing there would change nothing. And for most members the question of illusion barely arises: they have no embodied experience of a union that runs on their own activity rather than on staff, and little vocabulary for imagining one, so its dominance registers less as an injustice to be contested than as simply what a union is — or as something mildly positive that scarce resources keep from going further. What is scarce is not awareness that the union is failing — that circulates freely — but two harder things: the lived reference from which a member-run union could appear as a real possibility rather than a utopian one, and an in-depth grasp of why it fails, of the causes and the structure behind them. The second is not merely absent but actively blocked — by a technical framing that makes the causes look like a matter for officials and experts rather than something members can judge from their own working lives, by a staff that has internalized the very division of roles it would have to contest and pushes back when members reach past it, and by the erosion of the settings where such understanding was once built collectively. And because the disposition the first absence leaves is affective as much as cognitive, even showing that such unions have worked does not by itself restore the wish for one.

The exit side of the spiral compounds this. The members with the experience and standing to generate enthusiasm — the ones who could turn discontent into action — are disproportionately those it has already driven out or worn into passivity; those who stay tend to grasp the failure without being able to move it, so that even lucid grievance sits inert. So the remedy the spiral seems to invite — educate the members, raise their consciousness — misdiagnoses the problem: it broadcasts a failure they already see, while there is no perspective for action. What has to be built is harder and more deliberate: first-hand experience of the alternative, short-term goals concrete enough to seem attainable and worth fighting for, yet that ladder up toward a larger objective, the structural literacy that can name the causes and hold under pressure, and a layer of members able to turn that understanding into collective motion. None of this is delivered by information alone or arrives spontaneously; it is built through organizing together with sustained cadre formation — the political-economic education that turns a grasp of the failure into the capacity to analyze, argue, and lead.^[16]

The pension fight

The pension fight makes the consequences concrete. The cuts worked through a technical change most members never saw. The 2007 Pensioenwet required pension funds to value their future liabilities at the risk-free market interest rate; when the 2008 crash and the euro crisis drove that rate toward zero, the funds' liabilities ballooned on paper and their coverage ratios (dekkingsgraden) collapsed — even though their assets, some €1,300 billion, were intact. Under the financial-assessment framework (financieel toetsingskader), a low coverage ratio legally compels a fund to stop indexing pensions to inflation, and below a floor to cut them outright. So in 2013 several of the largest funds imposed nominal cuts on every entitlement, and indexation was frozen for years — not because the money was gone but because the accounting rule said so. What determines a payout had been quietly shifted from the promise written into a pension agreement to a single financial-market figure, the coverage ratio.^[17] For more than a decade the FNV mounted no union-wide resistance to what amounted to reneging on the pension promise, and the resistance that did form came from below and against the leadership rather than through it: FNV cadre committees — "Red het Pensioenstelsel," "Breed Protest tegen de Pensioenroof" — that named the collapsed rekenrente an accounting trick and demanded it be raised, surfacing only around 2018–2019 and remaining, on their own account, "largely unknown even within the FNV."^[18] Part of why the challenge came so late is that the apparatus had gone along with the market-valuation regime and staffs the boards that administer it: the frame it would have had to contest was one it had already internalized, and when members did contest it, the union's own campaign team pushed back. The union itself brought behind-the-scenes negotiation (SER consultation, backroom deals) and occasional demonstrations to a fight where real concessions could only have been won by an organization capable of credible collective action — which the FNV, having replaced member agency with staff agency, could not deliver. The outcome was largely inevitable long before the board agreed to it — not because the fight was unwinnable, but because the parts of the organization that would have had to wage it did not dare or want to.

Behind the missing power lay a missing literacy. The reneging on the pension promise worked because its form was technical: recasting that promise as a coverage ratio turned a political decision into what looked like a neutral accounting fact, legible only to specialists and defensible as mere prudence. Contesting it required not awareness that something had been lost — that was never scarce — but an economic literacy that could name the risk-free rekenrente as a political choice serving capital rather than a fact of nature, and hold that argument under pressure. That literacy was thin at every level, the officials often included. Where it is absent, willing resistance folds the moment the reneging is restated as responsibility.^[19]

That the FNV mobilized exactly this kind of countervailing power a few years earlier makes the failure sharper. In 2004, when the government moved to abolish early-retirement arrangements (VUT and prepensioen), the union put more than 300,000 people on the Museumplein — the largest union demonstration in Dutch history — and the cabinet and employers backed down. Pension policy was not immovable. The difference was that the 2004 threat was immediate and legible — a benefit every member could see being taken away — whereas the 2010s cuts arrived through a technical channel most members never registered as an attack at all. The fight that could have been mobilized was never triggered, and neither the leadership nor the werkorganisatie did anything to make the threat visible.

In the pension case, the werkorganisatie's institutional interests were not abstract. The FNV oversees pension funds and owns real estate through vehicles that intertwine with capital funds, giving the organization as an institution financial interests that can diverge from its members' interests as workers. Cees Grimbergen's investigative series Zwarte Zwanen documents the consequences: in the case of Pensioenfonds Vervoer, the transport workers' pension fund, Goldman Sachs speculated against the very mortgage- and credit-backed products it had placed hundreds of millions of the fund's money into — a concealed conflict of interest the fund valued at €250 million — and the FNV and CNV leadership who sat on the fund's board went along with the secret settlement that buried it. Members of the fund's accountability body were silenced through NDAs and threats of legal action.^[20] The pension deal that the FNV board eventually agreed to — and the expulsion of the sector chair who publicly withdrew his sector's support for it, both discussed below — follows the same pattern Grimbergen documents across funds: union officials with board seats in the pension-finance complex develop institutional loyalties that override their accountability to members.

The same dynamic has a longer and starker precedent one sector over, and it turns on a restructuring the unions themselves helped design. The dock workers' fund — the Pensioenfonds voor de Vervoer- en Havenbedrijven — sat on reserves that unions and employers had long drawn on for early-retirement schemes; when, in 1996, the FNV and CNV proposed with the employers to take €136 million out to lower the retirement age, the fund's board agreed only on condition that the fund be restructured so such "grabs from the kitty" could never recur. At the end of 1997 it was folded into an independent insurer, Optas, at a safe distance from workers and employers, its billions henceforth run by professionals rather than the social partners. What the money was protected from, in short, was the members.

The unions did not have that structure imposed on them by an adversary, and were not talked into it by an expert. Until then the unions and the employers' association SVZ had governed the fund between them, which is precisely why it could be drawn on. When containerization meant the port needed fewer hands, the union kept its promise that no man would be forced out of the gate — but paid for that promise out of the pension fund's free reserves: ƒ385 million between 1989 and 1991, and the 1996 withdrawal was not the first. The costs of automation were met, in other words, from what the members had accumulated. And the plan meant to end this came from an FNV officer, Joop Verroen, together with the PVH's director Peter Rietbergen: what was called protection was protection against the trustees themselves, and Verroen sold that outcome to the dock workers whose say it removed.

The reserves this produced were beklemd vermogen — legally spendable only on pensions, but carrying no enforceable obligation to index. Optas could therefore stop indexing and hoard the profits perfectly lawfully; that made it a takeover target, Aegon acquired it in 2007, and the 2018 Optas-Aegon merger let Aegon turn the earmarked reserves — by then €2.5 billion, some €58,000 per insured worker — to its own ends. De Nederlandsche Bank, the regulator charged with protecting members, approved the merger in secret; when it had to release the merger documents, it asked Aegon's own lawyer to mark what was "commercially sensitive," and he designated sixteen without any substantiation, whereupon the bank declared them secret; a court annulled that consent in 2023, finding the bank had shown "insufficient regard" for the dock workers — after which nothing happened, leaving pensioners, most now past eighty, still in court. The union's own hand in building the depoliticized structure is the whole point: the restructuring sold as protection was the instrument of dispossession.^[21]

But the polder framework also constrained what was structurally available, and the constraint was not one the FNV could have overturned by itself, however militant: the rekenrente-and-coverage-ratio regime was law, backed by the central bank and the EU's fiscal architecture. That exogeneity is real but narrower than it looks. Valuation rules bind the fund; they do not extinguish the employer's promise. What a pension agreement recorded was a commitment employers — the state among them — entered voluntarily and were paid for in wage restraint, and while a firm is solvent and distributing profits, who covers a shortfall is a bargaining question before it is an accounting one. David Hollanders puts the point sharply: the union could in principle have litigated against ABP and PFZW holding up the pension regulations of the 1970s, 1980s and 1990s, in which cuts do not appear — except that, having agreed to the rekenrente when it was introduced and to the financialization around it since, it had no standing of the kind that counts, which is moral rather than legal. Some of what presents itself as exogenous constraint is the residue of a fight never had.^[22] The two mechanisms then produce identical observable behavior — leadership restraining action to preserve the organization — from opposite causes: endogenous capture and exogenous constraint. In the pension case they are not even cleanly separable, because the FNV helped build the regime it is now trapped by — it went along with the financialization of the pension system and staffs the boards that administer it, which makes it, in Hollanders' phrase, victim and carrier at once. The corrective implications still differ: endogenous capture calls for redistributing authority within the union; exogenous constraint calls for changing the organizational form's vulnerability to the polder framework. The FNV pension case involves both, which is part of why it resists the standard prescriptions.

A vocabulary for what went wrong

The FNV's trajectory — from workplace absence to professional dominance to organizational paralysis — is that tilt running its course where nothing counteracts it. What the Dutch archives register as a "gulf" between officials and cadre, McAlevey's fieldwork lets us name more precisely. Drawing on decades as a union organizer and chief negotiator, she distinguishes in No Shortcuts (2016) three modes that the single word "professionalization" collapses together.^[23]

Advocacy means professionals acting on behalf of a constituency. Lawyers, lobbyists, researchers, and communications specialists wage the battle. Ordinary people are not involved in any real way. This model can produce small gains — seatbelt regulations, product safety rules — but it cannot challenge power, because it never deploys the only concrete advantage ordinary people have over elites: large numbers, organized to act collectively.

Mobilizing means a professional staff directs existing supporters to actions: rallies, petitions, media events, get-out-the-vote drives. This is a substantial improvement over advocacy, because bodies appear. But — and this is the critical point — the bodies are stage-managed. The staff sees itself as the key agent of change; the participants are there to be photographed and counted, not to shape strategy. Mobilizing can look radical from the outside — mass marches, social media campaigns, civil disobedience — while actually demobilizing the base, because ordinary people learn that their role is to follow instructions rather than exercise agency.

Organizing means the agency for change rests with a continually expanding base of ordinary people who are not activists, who did not show up because they were already sympathetic, and who help develop the power analysis, design the strategy, and achieve the outcome themselves. The organizing approach relies not on staff-directed actions but on building majority participation among a bounded constituency — a workplace, a congregation, a neighborhood — and testing that participation through escalating collective actions that the members themselves lead.

McAlevey's core finding is that only organizing can generate the power needed for high-stakes fights. Whether the demand is a pension plan, production control, school funding, or the right to unionize in a hostile-law state, the power required to win is proportional to what it costs the opponent to concede. Low-cost demands (a marginal wage increase for janitors, where wages are a tiny fraction of building costs) can be won through mobilizing. High-cost demands (a defined-benefit pension, control over working conditions, the right to strike) require the credible threat of majority collective action — which requires organizing.

The distinction matters because much of what passes for "organizing" in contemporary movements is actually mobilizing. If a staff member writes the petition and asks supporters to sign it, that is mobilizing. If organic leaders on the shop floor get a supermajority of their coworkers to sign a public petition demanding recognition — knowing the employer will see their names — that is organizing. The petition may look the same in a photograph. The power behind it is fundamentally different.

What organizing looks like

The US case shows what the alternative looks like — both when it works and when it is destroyed.

The Congress of Industrial Organizations (CIO), built in the 1930s through mass strikes, sit-downs, and rank-and-file militancy, was grounded in what McAlevey calls deep organizing. Left-wing organizers — Socialists, Communists, IWW veterans — were the CIO's best, precisely because they were committed not just to winning campaigns but to building the capacity of workers to act collectively across racial and skill divisions — bringing together workers of different ethnic backgrounds and skill levels whom the craft unions of the American Federation of Labor (AFL) had largely written off as unskilled. They developed specific methods: identifying organic leaders (workers to whom their coworkers naturally turned, who were often not the first to volunteer for the union), recruiting and developing those leaders, and using escalating "structure tests" to assess whether the organization was strong enough to strike. The resulting unions were, as Stepan-Norris and Zeitlin showed, not only the most effective but also the most democratic — equipped with the internal caucuses, open elections, and alternative newsletters that Michels's "iron law of oligarchy" claimed were impossible. Oligarchy did not destroy these unions. McCarthyism destroyed them: the Taft-Hartley Act of 1947 banned solidarity strikes and forced anti-Communist affidavits; the subsequent purges drove out the left organizers who were the backbone of the CIO's method — often with the active collaboration of the unions' own right wing.^[24]

What replaced them was business unionism — a professional staff managing the decline. The AFL had practiced this for decades: its leaders, as Piven and Cloward noted, "came to function more as labor contractors than as labor leaders, depending more on collusive arrangements with employers." After McCarthyism destroyed the CIO's organizing cadre, the AFL model won by default. Sweeney hired the American Management Association to train his growing SEIU staff, with the Harvard Business Review at the top of the curriculum (Moody). Automatic dues checkoff, as with the FNV, decoupled revenue from engagement. The servicing model became the path of least resistance.^[25]

Then came what McAlevey calls "New Labor." In 1995, a new generation of union leaders led by John Sweeney won the first contested election in AFL-CIO history by promising revitalization through aggressive organizing; Andy Stern succeeded him atop their flagship, the SEIU, the following year. Twenty years and hundreds of millions of dollars later, union density had declined further: from 10.3% to 6.7% in the private sector. What went wrong?

Part of the answer is the relocation of production to low-wage countries and regions. But what certainly did not help is that New Labor replaced organizing with the corporate campaign: a strategy of tactical warfare waged not by workers but by a professional staff of researchers, lawyers, and communicators who target the employer through brand damage, shareholder actions, political leverage, and neutrality deals. In the corporate campaign model, workers are one of a dozen "leverage points" — equivalent to a city council member, a supply-chain partner, or a journalist. Communications staff select a few telegenic workers as "authentic messengers," coach them, and present them to the media. The rest are irrelevant. As Peter Olney, organizing director of the International Longshore and Warehouse Union, described the internal discussions: "The conversations were about how workers really got in the way of organizing."

The clearest illustration of the difference between the two models is the contrast between two locals of the same union — SEIU — both organizing nursing home workers in the new millennium. In Connecticut, 1199 New England maintained the CIO tradition: organic leader identification, open negotiations where all workers were welcome, and routine majority strikes. The result was the highest nursing home contract standards in the country — including defined-benefit pensions enabling health service workers to retire when the physical toll of caring for patients became too much. In Washington state, the SEIU local followed the Stern model: corporate campaigns, neutrality deals negotiated by staff, contracts that stripped workers of shop-floor rights and the right to strike. What workers got in exchange was marginal wage increases. What the union got was organizational growth — more dues-paying members, not better conditions for the workers. Same union, same sector, same country, opposite approaches — and the results were as different as the methods predicted.^[26]

Whole-worker organizing: the Smithfield case

The contrast between 1199NE and Washington state SEIU demonstrates the difference between organizing and mobilizing within the workplace. The Smithfield case demonstrates why the workplace alone is not enough.

Smithfield Foods operated the world's largest pork production facility in rural Tar Heel, North Carolina — a right-to-work state^[27] with the lowest unionization rate in the country. The workforce was mostly male, racially and ethnically diverse (Black, Latino, white, and Native American — mostly Lumbee — workers), and subject to conditions that made the meatpacking plants Upton Sinclair described in The Jungle look modern: missing fingers, lost limbs, line speeds that treated workers as extensions of the machinery. The company exploited racial and ethnic tensions systematically — using immigration enforcement to terrify Latino workers and racial division to prevent solidarity between Black and white workers.

The UFCW tried to unionize the plant three times. The first two attempts (1994, 1997) failed. Both treated the campaign as a workplace fight: get enough cards signed, win the NLRB election.^[28] Employer intimidation, racial division, and the absence of any community infrastructure defeated both.

The third attempt succeeded — and succeeded decisively — when the campaign was reframed as a moral fight in which race and class were given equal weight, and the workers' own community networks were systematically engaged. Reverend Nelson Johnson, a civil rights veteran, was central to this: he helped the campaign reach workers through their churches, their families, their sports teams, their neighborhood relationships. Workers themselves educated their communities about conditions in the plant. Wildcat strikes created a crisis that made the community take notice. The national consumer campaign helped, but without the agency of the workers and their community, there would have been no victory.

McAlevey calls this "whole worker organizing": the recognition that workers are not just employees but people embedded in families, congregations, and neighborhoods, and that their power in the workplace depends on — and can be multiplied by — their power in those other settings. The CIO's left organizers understood this in the 1930s; William Z. Foster's Organizing Methods in the Steel Industry devotes an entire chapter to engaging churches, fraternal organizations, and unemployed workers' groups. What McAlevey adds is the argument that most contemporary unions have lost this capacity — not because they don't talk about "community allies" but because they outsource community engagement to staff or to weak allied organizations rather than building it through the workers' own organic ties.

Could it work here?

The cleaners' strikes of 2010–2014 are the case most often cited as evidence that organizing could work in the Dutch context. The "Schoon Genoeg" campaign — the slogan means both "clean enough" and "sick and tired" — launched by the FNV in 2007 using SEIU-inspired organizing techniques, produced the longest strike actions in the Netherlands since 1933: nine weeks in 2010, fifteen weeks in 2012. The cleaners — mostly women, mostly from migrant or ethnic minority backgrounds — won wage increases, better training, workload assessments, and improved sickness absence terms. The campaign's first phase brought in some two thousand new members; by the 2010 strike the union had around fifteen thousand in the sector. It also built up a Schoonmaakparlement — a body of elected cleaning workers, seventy-five strong by the end of 2009, which served as the campaign's decision-making structure. Within the FNV's history, this was unprecedented: workers exercising collective agency through a body they controlled rather than through officials negotiating on their behalf.^[29]

The question is what kind of power was built. As Connolly, Marino, and Martinez Lucio documented through extensive fieldwork (2008–2013), the answer is ambiguous. On one hand, the campaign produced genuine worker agency. One organizer captured the shift from institutional legitimacy to membership power: the goal was a deal signed "from a position of power... Not from the institution, not from the fact that they say, 'I accept that you're sitting at the table.' No! They have to say, 'Well, I'm not able to get rid of you at the table because you're in power.'" On the other hand, the top initiated the campaign and the specialist organizers it hired built the infrastructure, more than organic leaders identified and developed among the cleaners themselves did.

The post-campaign trajectory reveals the deeper problem. After the strikes, the organizers effectively became what Connolly et al. call "managers of discontent" — containing and channeling worker expectations that had been raised but could not be sustained within the existing union structure. The cleaning sector outcome was, in the end, a tripartite commission — employers, unions, and client companies monitoring fair pricing — which is social partnership by another name. The leadership used the organizing tactics to get a stronger seat at the table, not to replace it. And when the campaign ended, much of the mobilization infrastructure dissipated — the Schoonmaakparlement did not survive as an ongoing democratic body. The democratic structure was a campaign tool, not a permanent transformation of the union's operating model — which is what makes it mobilizing in McAlevey's terms, despite being more participatory than anything the FNV ordinarily does.

The organizing experiment remains marginal within the FNV's structure: roughly 100 organizers out of approximately 2,000 total paid staff.^[30] Campaigns in low-wage sectors have been cross-subsidized by more affluent sectors, generating resentment. Transfers to other sectors were limited: officials in older sectors like manufacturing — sectors with higher density — were, as one FNV-Bondgenoten official put it, "not really all that interested in organizing." The fundamental question Connolly et al. cite from Simms and Holgate — "what are we organizing for?" — remains unresolved. If the answer is "to strengthen our position within regulated social partnership," then organizing is a tactical innovation within the advocacy model, not a transformation of it.

The sharpest form of the objection comes from inside the federation. Han Noten — a former FNV Dienstenbond official, and one of the two mediators in the pension conflict that nearly split the FNV in 2011 — takes organizing at the bottom of the labor market to be self-defeating: the union built its power base by organizing the middle, the most vulnerable "will also be the first to take their leave of you," and so "if you want power, you have to address yourself to those who have it."^[31] On the ledger he is not obviously wrong. The 106-day strike of 2012 cost the FNV around ten million euros; most of the members it recruited did not stay; the scarce Dutch research on what organizing yields found higher negotiated wage increases in organized sectors but an otherwise roughly equal bargaining end-state; federation membership went on falling. What the ledger does not establish is his claim that addressing those who already have power yields more. Set against what the apparatus did with the rest of its resources over the same period — the decades of polder routine whose wage returns this essay has already traced — the campaign is not the outlier — a comparison Ron Meyer, who was involved in the Schoon Genoeg campaigns, made at the time. The churn is at least as well explained by an organization that had no use for the people it recruited once the agreement was signed. Organizing was made a department rather than a way of working; the Schoonmaakparlement was not kept; the campaign was never extended to the sectors where the union was already strong, as Agnes Jongerius afterwards said it should have been. Which returns the weight to the question left open above, and makes it a practical one: unanswered, there was nothing for the new members to be retained into. The campaign's most durable product was the one nobody planned — the occupations of 2012–2013 produced a Supreme Court ruling that widened the right to strike.

There is a sharper sign still of what was and wasn't imported. The campaign took its name and methods from the SEIU's "Justice for Janitors" — but not its agenda. In the United States that same campaign made sexual violence against the overwhelmingly migrant, overwhelmingly female cleaning workforce one of its central organizing fights: the 2015 Frontline documentary Rape on the Night Shift broke the story, janitors then won California laws (AB1978 in 2016, AB2079 in 2018) mandating anti-harassment training, and women built survivor-led structures — the Ya Basta Center, a promotora peer-education program — to organize other women around the dangers of isolated work. Schoon Genoeg organized a demographically near-identical workforce and carried none of it: neither its demands (naming wages, travel costs, in-hours language training, recognition), nor the fieldwork account of it, nor the Schoonmaakparlement's own later inventory of misstanden (managerial intimidation, workload, contracts, sick pay) registers workplace harassment or gendered safety at all — after the FNV's own women's wing had forced exactly that issue onto the union's agenda three decades earlier.^[32] Here the difference between mobilizing and organizing shows in what gets fought for rather than in who does the fighting: a staff-set agenda of winnable, table-friendly demands, rather than surfacing and fighting over what most threatens the workers themselves. Deep organizing takes its agenda from what the workers most need to change; import the tactics without that generative core and the issues a membership would raise itself stay invisible. And if we cannot say how far the problem reaches in the Dutch case, that is in no small part the FNV's own doing: it neither investigates nor reports on it, even as TNO and European bodies flag precisely these occupations and populations as high-risk.

The Smithfield case suggests what a deeper approach might look like. The FNV's membership includes precisely the kind of workers for whom whole-worker organizing would be most natural and most powerful: migrant workers in logistics, agriculture, meatpacking, and cleaning, embedded in ethnic, religious, and neighborhood communities that the werkorganisatie does not reach. Who actually makes decisions inside the union? If the pattern holds, the werkorganisatie should be whiter, more male, and more credentialed than the membership it serves. If this is a class problem rather than a diversity problem, the standard response of "more representation" misses the point. The problem is not the demographic composition of the professional caste but its existence.

Internal reform and its limits

The FNV has had its own rank-and-file reform moment — and the comparison with the Chicago Teachers Union is instructive precisely because of what went differently.

In 2010, the CTU was — in McAlevey's words — "a fairly typical weak, unimaginative organization." A rank-and-file caucus called CORE won the union presidency by running on a platform of deep organizing. Within two years, the CTU defied a 75% strike-authorization threshold designed to make strikes impossible and conducted a nine-day strike that defeated Mayor Rahm Emanuel — a Democrat with close ties to the corporate class. The strike's most profound result was not the contract but the durable power it built beyond the workplace: it shifted opinion across Chicago and made CTU president Karen Lewis a credible challenger to the very mayor the strike had beaten.^[33]

In 2009, a group of cadre members within Abvakabo FNV calling themselves the "Kloofdichters" (gap-closers) organized an internal opposition. The name was the diagnosis: in Lot van Baaren's words, "er gaapte een enorme kloof tussen het bestuur en de meeste actieve kaderleden" — an enormous gulf had opened between the board and most of the active cadre members. Initiated by van Baaren and Ger Geldhof, both bondsraad members, the Kloofdichters wrote a program for a democratic and militant union and ran eleven candidates for the unpaid board at the May 2010 congress, seven of whom were elected; with two counter-candidates also taking salaried posts, at least nine of the new fifteen-member board belonged to the reformers. Their slogan: a "strijdbare bond van, voor en door de leden."^[34]

The congress adopted sharpened positions across the board — strengthening the "action leg" of the union, increasing member control over board elections, rejecting the technocratic line of the outgoing leadership.

When the 2011 pension crisis hit, the Kloofdichter-influenced Abvakabo board refused to accept the deal negotiated between the FNV confederation, employers, and government. FNV chair Agnes Jongerius had signed the pensioenakkoord — raising the retirement age from 65 to 67 and shifting financial risk to employees — despite opposition from the two largest unions. Ninety percent of surveyed members opposed the agreement; FNV Bondgenoten rejected it with 96% voting against. But the federatieraad operated on a one-vote-per-union basis regardless of membership size: Bondgenoten and Abvakabo together represented 60 percent of FNV membership and were involved in 90 percent of collective bargaining, yet held only 2 of 19 votes. A coalition of smaller unions passed the deal over the explicit opposition of the two largest. The federation's architecture produced an outcome against the membership's expressed preferences — which became the primary justification for the restructuring that followed. Bondgenoten and Abvakabo withdrew confidence; Jongerius resigned in June 2011. The confederation nearly split.

What followed was not a CTU-style transformation but a four-year restructuring process controlled from the top. In December 2011, the 19 FNV union chairs were gathered at a crisis meeting in Dalfsen, convened by two external mediators: Herman Wijffels (CDA, former Rabobank director, former SER chair) and Han Noten (PvdA, former NS executive). Phones were confiscated. The result was a mandate for structural reorganization toward "De Nieuwe Vakbeweging" ("the new union"). The kwartiermakers (quartermasters) appointed to design the new structure were polder-establishment figures, not organizers: the lead went to Jetta Klijnsma (PvdA MP, former state secretary), and the advisory team she assembled ran from a CNV consultant and an SP official to a care employers' representative (Frank Bluiminck of ActiZ) — with no active rank-and-file organizers among them, despite Abvakabo's requests. The class composition of the design process shaped the design: what emerged was a service-provider union with a consultative parliament, not a fighting union with democratic control.

By mid-2012, Ton Heerts — former AFMP military union chief, former FNV vice-chairman, former PvdA parliamentarian — became temporary chairman. In May 2013, the transitional entity "FNV in Beweging" was formally established at a founding congress, with Heerts elected chair by direct member vote (62%) over Corrie van Brenk (38%). The new structure introduced the ledenparlement — 108 members elected by sector — as the highest decision-making body, replacing the old federatieraad.

The ledenparlement was, from the beginning, a weak democratic instrument. It barely communicated with the membership or externally; meeting minutes were effectively inaccessible; votes were secret. The body existed as a formal structure without the transparency or communicative practices that would have made it a real vehicle for member agency. Its authority was tested and found hollow within weeks of its establishment. In April 2013 — before the founding congress had even taken place — Heerts signed the Sociaal Akkoord with employers and cabinet, accepting reduced unemployment benefits duration and weakened dismissal protections in exchange for vague employment promises. The ledenparlement had imposed a strict motion: no negotiations on WW or dismissal law reforms. Heerts objected — "I won't go to The Hague bound hand and foot" — and the federation board softened the mandate. The membership was presented with the fait accompli with barely any time to consult constituents. The government subsequently violated its commitment to suspend EUR 4.3 billion in cuts. NRC Handelsblad's deputy editor called it "the miraculous pacification of the FNV." The pattern was the same as the pension deal that had precipitated the crisis: leadership acting, membership informed afterward. The new governance structures had changed the org chart without changing the power relations.

The formal merger on 1 January 2015 dissolved most of the individual bonds into an "undivided FNV." The dissolution was incomplete. Several bonds — the Politiebond, the Horecabond, the AOB (Onderwijsbond), Druk en Papier — maintained their independence within the federation, retaining their own union councils and boards. One member organization left the federation outright. ANBO (seniors, some 200,000 members) had joined the FNV only in 2009 and withdrew again in January 2013, before the merger; because the federation pooled its member unions' dues and reserves into a single central treasury, a departing union could not readily reclaim its funds — a disincentive to exit built into the structure.^[35] The new structure absorbed the Kloofdichters' energy. It made organizing a self-contained department — a specialist unit that could be walled off and left marginal — rather than changing how the union as a whole operated. The pension question, which had precipitated the entire restructuring, was eventually settled along lines the membership had originally rejected: in 2019, under FNV chair Han Busker, the FNV signed a pensioenakkoord that Tuur Elzinga — then vice-president with responsibility for pensions, later chair — had negotiated and championed. When Jan de Jong, chair of the Seniors' sector, issued a press release withdrawing the sector's support for the pension reform in April 2022, Elzinga's board expelled him. The justification: publishing internal division was "explicitly giving external publicity to internal division within the FNV" and could only be intended to "damage the FNV." The mechanism is recognizable — dissent reframed as disloyalty, organizational unity invoked to silence the constituency most directly affected.

The structural unevenness of the 2015 merger matters for what happened next. In early 2025, a governance crisis escalated rapidly: the werkorganisatie unilaterally established a code of conduct; the ledenparlement rejected it and passed a motion of no confidence against the board; the general director and general secretary reported sick; staff went on strike on 17 February; the board resigned in March. FNV Personeel — the staff organization — together with 480 members petitioned the Ondernemingskamer, supported by sector councils, the ondernemingsraad, and the sitting raad van toezicht. On 13 June the court appointed two temporary toezichthouders — Lodewijk Asscher (former PvdA minister) and Ton Heerts (the same former FNV chair who had signed the Sociaal Akkoord) — with authority over new elections and organizational reform. Instead of facilitating the board elections the ledenparlement had been preparing, Asscher and Heerts proposed new statutes ("Voor de toekomst: een sterke FNV!") that would gut the ledenparlement's role in favor of sector-based governance. The ledenparlement rejected the plan by a narrow margin and passed motions with large majorities demanding preservation of elected boards and the ledenparlement; the toezichthouders returned to the Ondernemingskamer for expanded authority. The ledenparlement and organized rank and file could not mount effective resistance: the process was conducted under tight scheduling, with minimal information reaching ordinary members, and the structural implications of the proposed statutes were not widely understood until too late. The pattern is recognizable — a structural reorganization that changes the governance architecture without changing the werkorganisatie's dominance over the membership — but with a further turn: where the 2013 Sociaal Akkoord circumvented the ledenparlement, the 2025 proposal aims to abolish its effective power permanently.

The bonds that maintained their independence in 2015 have institutional reasons to go along with the restructuring: they already have their own governance structures and fit naturally into a sector-based model. The sectors within the "undivided" FNV, meanwhile, have few active members — partly because vakbondsbestuurders discourage cadre activity — which makes them easier to control from above. Shifting power from the ledenparlement to the sectors is, in practice, shifting power from the one body that — however weakly — provided a formal channel for organized member voice to bodies where the werkorganisatie already dominates. The ledenparlement's democratic practice was attenuated from the start, but even a nominal democratic structure can become a site of contestation if enough members organize within it. The 2025 restructuring forecloses that possibility.

Both CORE and the Kloofdichters were member-led reform factions that won internal elections against entrenched professional leadership, and both achieved temporary changes. CORE sustained the change longer — the CTU struck, won, and the organizational infrastructure survived because it was rooted in organic leaders at every school. The Kloofdichters' reforms were absorbed by a structural reorganization that left the werkorganisatie's dominance intact. The reform energy was channeled into a ledenparlement that barely communicated with the membership — a body that could do little to enlarge member agency, and whose first major test (the Sociaal Akkoord) showed that leadership could circumvent it at will.

CORE's organizational transformation was tested and proven through collective action that showed every teacher in Chicago that the new model worked. The Kloofdichters never had that test. Their reforms had to be carried out within the existing polder model and through the werkorganisatie, which meant they were always vulnerable to being outmaneuvered by the institutional machinery they were trying to transform. A reform faction that wins internal elections but cannot demonstrate power through collective action, and is not broadly rooted, stands little chance.

What a strike does that an election cannot is anchor power below the level of the leadership. The name for the layer it works on is the militant minority — the old syndicalist term Kim Moody revives: the politically conscious core — here, the organic leaders at every school — that carries an organizing culture between upheavals and can act without waiting for staff direction.^[36] The 2012 strike proved that capacity was anchored there and not only in the officers a reorganization could sideline — which is why CORE's gains were robust. The Kloofdichters had won a board election but lacked a broad base: their majority rested on an undermobilized membership, and so depended on the werkorganisatie for implementation. The polder model thereby obstructs not only strikes but internal reform.

But the comparison now cuts both ways. A decade after the original capture, the same pattern had re-emerged: withheld financial audits, the House of Delegates banning publication of its own minutes, leadership acting against its own bylaws. A reform caucus called REAL — itself an outgrowth of CORE — ran on transparency in May 2025. Because of its structure, that fight was possible: an internal opposition could form and run. The Kloofdichters' gains were shallower and their absorption faster and more complete than CORE's, which suggests the strike did structural work — but the work was permissive, not preventive. The question for both is not whether the pattern re-emerges but whether the earlier reforms left members enough room to contest power again before the organization neutralizes that resistance.

The bureaucratic gap does not only breed reform caucuses inside the union; it pushes energy clean outside it. In 2017, primary-school teachers frustrated with the distance between the education union and the classroom founded PO in Actie and drove the primary-education strike wave of 2017–2019 — the gulf the Kloofdichters had named, now acted on from the outside. It did not so much fade as get absorbed: its co-founder Thijs Roovers joined the AOb, entered its executive, and became its chair, the reform energy institutionalized the way the Kloofdichters' was, only from the outside in. University staff founded WOinActie the same year over workload, with demonstrations, "white" strikes, and complaints to the labour inspectorate; years on it is still issuing calls, but it never built the rank-and-file structure that would turn attention into power — awareness and lobbying, led by established academics, mobilizing in McAlevey's exact sense. Outside the union the ceilings are the same as inside it: absorption, or the mobilizing plateau.^[37]

Organizing as vocabulary

The union's own most recent statement of intent confirms the diagnosis. In June 2026 the FNV congress adopted De Kracht van Beweging, the federation's meerjarenbeleid for 2026–2030, submitted by the board under Hans Spekman and settled through 489 amendments across 54 tekstvoorstellen, each carrying the board's pre-advies to adopt, adapt, redirect, or reject.^[38] The document reads as the werkorganisatie's self-portrait, and what it concedes is as telling as what it withholds.

It concedes the vocabulary. The half of the plan headed "what is needed" now opens not with visibility or lobbying but with organization: "Our power comes first of all from organization on the shop floor. From kadergroepen that build the bond with colleagues every day. A strongly organized FNV is the foundation." Organizing, once the marginal experiment of a hundred staff among two thousand, is named a proven and effective method; the conversation between colleagues is called the basis of durable membership; a "strongly organized FNV" is listed as the first organizational priority rather than the last. On the surface, organizing appears to have won.

It is at the level of mechanism that the concession stops. Every amendment that would have lodged power below the werkorganisatie — where a reorganization cannot reach it — was rejected or hollowed. Metaal's demand for financial transparency between the sectors and the centre: not adopted. The proposal to seat kaderleden at the cao table rather than merely survey them: not adopted, on the ground that the household rules already provide for it. The first institutional check the document ever entertains on the apparatus itself — an independent FNV ombudsperson with authority over the werkorganisatie as well as the membership — was rejected as "too detailed." The broadly supported demand for board elections from below survived only as a promise to "discuss the statutes with the bondsraad." What passed was wording, declared priorities, and a research agenda — never a binding mechanism: the Ghent model of union-administered social security, a statutory minimum-CAO, and a legally anchored automatic price compensation — the levers that would tie membership directly to provision — appear only under the heading "continuing to learn."

The strike undergoes the same operation in miniature. It survives in the text — "striking counts as our most powerful means of strengthening our bargaining position" — but under a bullet retitled Actievoeren, folded into an undefined wider repertoire, and subordinated to the negotiating table rather than to the members' own capacity. The expanded right to political and solidarity strikes across sectors is a real gain; but a right is something to be defended in law, and the plan commits to nothing that would build the capacity to use one — none of the escalating collective actions, a supermajority petition, a sticker day, a practice walkout, by which a union tests and builds whether it can strike at all. It names the members' strongest weapon and undertakes nothing to forge it.

The same gap runs through what the plan says about its own decline. It is candid that the numbers have stalled — the union "stabilizes" but "large membership growth fails to materialize" — and in places it does commit to organizing: cadre groups are "the basis of union power," new cadre are to be recruited and trained, members are "structurally trained" to build power with colleagues, and action together with strong cadre is said to produce "better results and membership growth." What it does not do is supply the two links that would make that chain deliver. On growth, the concrete commitment is marketing — an integrated strategy fusing workplace recruitment with "data-driven marketing," member-gets-member drives, and the positioning of individual services — with the growth question itself demoted to a research priority. On capacity, as the strike showed, it commits to none of the escalating collective actions that would turn trained members into a membership able to strike. So in a period of decline the plan names the chain from recruitment through cadre to action and staffs its two load-bearing links with marketing and study — the same vocabulary without mechanism as the rest.

This is why the plan and the 2025 restructuring belong together: they are the programmatic and the governance faces of a single decoupling. The restructuring rearranged the bodies through which members might have spoken; the plan then tells those members what the apparatus will do on their behalf — servicing put in place of the agency the restructuring removed. The pensioenakkoord the membership once rejected returns as the plan's stated point of departure, its history unmentioned. The document does not describe an organization preparing to hand its members the capacity to act. It describes an organization that has learned to say it will.

The political problem

McAlevey deliberately brackets the question of how unions should relate to political parties. The bracketing is a real gap, not a safe simplification — her own US cases show that a union cannot ignore the political context. And the FNV analysis cannot leave it aside at all: where the American party-union tie is informal, the polder model institutionalizes the union-party-employer relationship. The Dutch case is not a different problem but a more entrenched, more visible version of the same one.

Kim Moody's work illustrates this for the US context. His analysis of the Democratic Party in On New Terrain (2017) shows that every attempt by unions to realign the party from within — the Rainbow Coalition, the Progressive Alliance, DSOC — was absorbed by a party that simultaneously moved rightward under the pressure of corporate PACs, the Democratic Leadership Council, and structural reforms that concentrated power in elite hands.^[39]

McAlevey's own cases show why this matters even within her framework. Thirty-eight percent of union households in Wisconsin voted to retain Governor Scott Walker — who had just stripped public-sector unions of collective bargaining rights. In Michigan, voters rejected a ballot measure to enshrine collective bargaining protections in the state constitution, after which the governor swiftly passed right-to-work legislation. These are failures of political consciousness that workplace organizing alone can neither explain nor prevent. The cause is political: having tied their members' politics to the Democratic Party — which delivered nothing while demanding loyalty — the unions never built an independent working-class politics of their own, and left members with party messaging in place of class analysis. So when the attack came, union households had little to fall back on, and many broke toward the very forces dismantling them.

The Dutch variant is less dramatic but structurally parallel. The FNV invests in party relationships — historically with the PvdA, more broadly with GroenLinks and the left-of-center bloc — rather than in member organization. De Beer and Berntsen document the continuing informal ties: former FNV chair Agnes Jongerius went into the European Parliament for the PvdA; former CNV vice-chair Aart Jan de Geus became Minister of Social Affairs for the CDA; former Young-CNV chair Jesse Klaver went on to lead GroenLinks. The SP, which claims to be the exception, follows the same pattern: Ron Meyer, involved in the Schoon Genoeg cleaners' campaigns, became SP party leader — but brought the mobilizing model, trying to translate the Our Revolution template of staff-directed mass campaigning into Dutch. The next party leader, Lilian Marijnissen, also came from the FNV werkorganisatie, with no visible organizing track record or interest in it. What is telling is not the direction of the circulation between union and party but who is admitted to it. One stratum moves freely among union leadership, party politics, and the bodies convened to manage the union's own crises: werkorganisatie officials and union chairs pass into parliament, and the movement runs back the other way when the apparatus needs managing — the PvdA and SER-connected figures who led the 2011–2013 restructuring, the former PvdA minister and former FNV chair the court installed as toezichthouders in 2025. The cadre member is admitted nowhere in it: no rank-and-file organizer rises into union or party leadership, and none was allowed near the redesign of the union. Admission turns on already holding a professional or political position — and the party-union pipeline it forms produces the very people who manage the crises its own logic generates. The result is that workers' political consciousness is shaped by professional officials rather than by their own experience of collective struggle.

De Beer and Berntsen, themselves no radicals, conclude that "marginalization seems the most likely scenario for the Dutch unions." The union's formal position persists only as long as employers and the state find it useful — and as membership continues to erode, "ultimately a point may be reached at which the employers and the government no longer need the unions to secure societal and political support for their aims." This is the advocacy model's endgame: a union whose legitimacy rests on institutional position rather than membership power can be bypassed the moment the institution no longer needs it.

The subordination is not only political but financial. The FNV's own published figures show that 24 percent of its income comes from employer sources — 7 percent from direct werkgeversbijdragen and 17 percent from employer-financed social funds. A 2023 Follow the Money investigation into CAO financing documents the system through which this operates: 122 of 180 sectors with a CAO have paritaire social funds, jointly governed by employers and unions, through which employer contributions flow. The FNV dominates these funds — it holds the most board seats and in some cases receives the entire payout first, with other unions required to submit accountant declarations before receiving their share.^[40] This gives the FNV institutional power over smaller unions, but it also ties the FNV's organizational survival to the system it would need to challenge.

Employers are open about the logic. In 2008 the ING Bank began offering new hires, at their option, a union membership paid for by the bank — at a firm where union density was low. The point, as Breij observed, is that employers above all do not want to have to do without the unions as a negotiating partner, and will pay to keep them — through CAO provisions, and now by funding membership directly. "Thanks to employer contributions," he noted, "quite a few unions now seem able to keep themselves afloat" — which is precisely why he asked whether their independence was thereby "at stake."^[41]

When Follow the Money asked for financial details, the FNV refused, explaining that it doesn't want employers to know how long it could sustain a strike — a response that reveals both the dependency and the awareness that it is one.

As Paul de Beer observes in the same investigation, declining membership makes unions ever more reliant on employer contributions, undermining the autonomy on which the system's legitimacy rests. And because the fund that pays a union for a CAO pays only the unions that sign it, that financing is a lever as much as a subsidy: "a union that doesn't sign will soon get no money either," as one FNV negotiator put it — "a means of pressure with which employers can influence the outcome of CAO negotiations." Employers and yellow unions are accordingly concluding an increasing number of agreements without the FNV — at the limit through bonds that exist to be paid, like the LBV, which signed a CAO covering 200,000 temp workers over the objections of the FNV, CNV, and De Unie, who called it a "family business" that is "willing, for a fee, to help employers," and which the employers' own federation VNO-NCW had long before nicknamed "the Antilles route of labor relations."^[42] The FNV is thus simultaneously undermined by and invested in the same system: yellow unions threaten its position, but the structural reform that would address it — decoupling union financing from CAO negotiations, as labor-law scholars have proposed — would also eliminate the FNV's own dominance of the fund infrastructure. The FNV's stance of refusing to sign "bad" agreements is principled, but it directly reduces income, creating a structural tension between organizational integrity and organizational survival that the advocacy model cannot resolve.

Pension fund governance adds a further layer. FNV-nominated board members — often specialists — sit on the boards of pension funds that manage billions in workers' deferred wages. Both the legal framework and the institutional culture treat this management as a technical question about returns and risk parameters — not as a political question about what happens to workers' money. The fiduciary duty framework defines responsible management as financial optimization, which depoliticizes the exercise of power at the point where it is largest: the investment decisions that determine where workers' capital goes and whose interests it serves. Ordinary members have no visibility into these decisions. The ledenparlement may be informed, but the opacity means the membership has no basis on which to form a view, let alone hold officials accountable. The cases documented in the pension section above are the extreme — captured reserves, cover-ups, NDAs — but they occur within a broader routine: delegates and officials embedded in a financial governance system whose norms they absorb, exercising power over resources whose owners cannot see what is being done with them. That this is not generally understood as a political problem is itself part of the problem.

The seat is also weaker than it looks, because of how the funds themselves are built. Around the turn of the century the large sector funds were split in two: ABP and PFZW became small rump funds — a board with a secretariat — while the work of running them moved into separate executors, APG and PGGM, employing thousands. The executor invests, runs the ICT, handles the communication, writes the reporting; ABP and APG share a building. The split was justified by competition — executors would compete for the funds' custom and so become better or cheaper — and the competition never arrived, which leaves the salaries it was meant to justify resting on nothing. What it did produce is a board dependent on the organization it nominally commissions for staff, information and expertise, and which in some cases does not know what the fund is invested in. The shape is this essay's own: a lay body formally in charge, a professional apparatus holding the competence, and the asymmetry between them settling the outcome — except that here the lay body is the one the union's nominees sit on, so an FNV representative reaches the investment decisions across two such gaps rather than one. The accountability bodies are further out still: they have no say over the three things that matter most — what the fund invests in, how it votes its shares, and what asset management may cost. Nor does the regulator supply what the governance lacks. DNB caps what a fund may pay out and sets no ceiling on what it may spend, so the rule that forced the 2013 cuts has no counterpart on the cost side — where, on Hollanders' figures, roughly a quarter of the €30 billion or so paid in premiums each year was going, most of it to asset management.^[43]

This sword cuts both ways. A seat on the boards that govern these funds is a real gain, not a trap the union should have refused: better that organized labor sits where the investment decisions are made than that the money is left wholly to the finance industry — which is exactly why they sit there. But what they administer is, of course, not simply workers' deferred wages, and calling it that is part of the problem. What the funds hold is finance capital — well over a trillion euros that has to be invested to grow, on returns extracted from labor elsewhere — so a funded pension ties workers' security to the very accumulation that squeezes them, and makes them mediated rentiers whose stake in high yields pulls against their stake as workers, and against solidarity with workers elsewhere. That is why the seat encases whoever sits in it: the fund's own logic is capital's. To sit on the board is to administer that logic — to absorb its fiduciary norms and become, once again, victim and carrier at once; the gain and the cooptation come through the same door. The pressure to be coopted is structural, not a failure of individual resolve, and only countervailing member power — members who hold whoever sits to account — can keep that insight from being lost, though even that means contending with the fund's own imperatives, not merely with a captured staff.

That the fund is capital is not only a trap; it is also a lever — and how unions have used it is itself telling. There are three ways to politicize the funds. The first is divestment as a naked political act: Dutch labor has, as far as I know, taken it only once, when in 2014 PGGM, which runs the healthcare fund PFZW, pulled out of five Israeli banks over their financing of settlements in the occupied territories. The BDS movement had called for exactly this divestment; it drew an Israeli protest and a summons of the Dutch ambassador, and it stands out precisely because it is so isolated. The second is divestment that can be re-described as prudence: the fossil-fuel exits — ABP's €15 billion withdrawal in 2021, then PFZW and PME — came only after years of campaigning, and were sold and depoliticized by being framed as long-term financial risk rather than political choice. That contrast is the lesson in miniature: the lever moves when the political can be dressed as "good policy," and stays fixed when it cannot. The third road went the other way, through ownership: Rudolf Meidner's wage-earner funds, proposed to the Swedish unions in the 1970s, would have made profitable firms issue new shares each year into union-controlled funds until labor held the controlling stakes — socializing capital through the very instrument that now encases the union — before the plan was gutted after fierce resistance and, in 1991, abolished. Each road asks the union to see the fund as capital to be wielded rather than deferred wages to be safeguarded, and broadly unions do almost nothing with it unless they can present it as optimization.^[44]

The question is whether the polder model itself is the mechanism through which the werkorganisatie's dominance over the membership is maintained — and whether any organizational reform can succeed without also breaking the union's dependence on consultative structures designed to produce consensus between capital and labor. Organizational reform without political independence is insufficient — a union that organizes its members brilliantly but channels their power into supporting parties that serve capital will reproduce the subordination at a higher level. Political independence without organizational transformation is empty — a union that declares independence from the polder model but cannot organize a credible strike threat has nothing to be independent with. Neither McAlevey nor Moody offers a framework that holds both. The argument about class relations within movements suggests why: what needs to be fought is not just a set of organizational practices but the material interests of the people who benefit from those practices — interests that will resist reorganization for the same reasons any form of class privilege is defended.

Conclusion

The question this essay's title asks has an honest answer: right now, it is the werkorganisatie's power. The FNV's institutional position, its consultative access, its pension fund board seats, its revenue stream — these belong to the professional apparatus, not to the members whose collective action is the only thing that would give the apparatus real leverage. The members are formally sovereign. In practice, they are an audience.

That is a claim about positions, not about the people holding them, and the difference matters for what follows from it. Nothing here rests on the motives or the character of the union's professionals: most work hard at genuinely difficult jobs, many see the problem described here more clearly than any outside critic does, and much of the sharpest resistance in these pages came from inside the organization — the Kloofdichters were cadre and bondsraad members, the pension committees were kaderleden, the sector chair expelled in 2022 was an elected officer. What the argument identifies is what a position makes rational: an apparatus whose income, access and continuity do not depend on whether members move will, over time and on average, behave as though it does not depend on them, whoever is hired into it. That is also why personalizing the diagnosis changes nothing — replace every official while leaving the structure of dependence intact and the same behavior reassembles within a few years, which is close to what the FNV's own reform episodes show. The question is not who is at fault. It is what a given position depends on, because that is the part that can be rebuilt.

The evidence reviewed here suggests this can be changed. The CIO's left organizers built unions that were both the most effective and the most democratic. The CTU's CORE caucus rebuilt one in two years. These are not utopian examples — they are documented cases of unions where the answer to "whose power is it?" was, for a time, "the members'." That the answer tends to drift back does not mean the question is unanswerable. It means the answer has to be fought for repeatedly, and that the organizational features which keep it fightable matter more than any permanent solution.

The question is not unique to unions. It is the question any organization claiming to represent a constituency must answer — and the companion article on NGOs traces the same dynamic across foundations, advocacy organizations, and social movements. But the union is where the answer matters most: for other organizations "the staff" is a shortcoming; for a union it is a self-contradiction — because a union whose power is not its members' has no power at all.

Where that fight resumes is a question of scale. The organizing-minded cadre that would carry it — the militant minority — is, for now, too small to win and hold a confrontation on the scale of the pension fight or the federation itself; a caucus that reaches for those before it has built a base is the Kloofdichters again. The nearer task is the choice of ground: the sectors and subsectors where the werkorganisatie is weakest or most dependent on its members — the exposed, low-density, adversarial pockets where its position cannot reproduce itself on inertia and checkoff alone, and where a union presence already runs on cadre activity rather than in spite of it. That is where a militant minority can accumulate the capacity that later distributes, and where a strike, when it comes, tests something real. The 2026 plan, for all it withholds, left the footholds to build with — shop-floor access, worker-to-worker recruitment, local self-organization, adopted as text the members can hold the apparatus to — and the sectors that named the mechanism, Metaal and Jong and the sharpest of the local networks, are where a base could first cohere. Whose power it is will not be settled at the federation. It will be built, if it is built, in the places the apparatus has the least reason to look.

Bibliography

Breed Protest tegen de Pensioenroof. breedprotesttegendepensioenroof.nl.

Breij, Bert. Twee miljoen leden: Over het verleden, de toekomst en het heden van de Nederlandse vakbeweging. Amsterdam: Vakbondshistorische Vereniging, 2008.

CBS (Centraal Bureau voor de Statistiek). "Percentage vakbondsleden onder werknemers verder gedaald." 2024. https://www.cbs.nl/nl-nl/nieuws/2024/37/percentage-vakbondsleden-onder-werknemers-verder-gedaald.

CBS (Centraal Bureau voor de Statistiek). "Wat kenmerkt vakbondsleden?" Statistische Trends, 2023. https://www.cbs.nl/nl-nl/longread/statistische-trends/2023/wat-kenmerkt-vakbondsleden-?onepage=true.

Connolly, Heather, Stefania Marino, and Miguel Martinez Lucio. "'Justice for Janitors' Goes Dutch: The Limits and Possibilities of Unions' Adoption of Organizing in a Context of Regulated Social Partnership." Work, Employment and Society 31, no. 2 (2017): 319–335.

De Beer, Paul, and Lisa Berntsen. "Trade Unions in the Netherlands: Erosion of Their Power Base in the Stable Polder Model." In Trade Unions in the European Union, edited by Jeremy Waddington, Torsten Müller, and Kurt Vandaele, 799–832. Brussels: Peter Lang/ETUI, 2023.

De Beer, Paul, et al., eds. Polderen en strijden. Amsterdam: Van Gennep, 2025.

Driessen, Chris, David Hollanders, et al. Positie en strategie van de vakbeweging: beschouwingen, analyses en voorstellen. Amsterdam: De Burcht (Wetenschappelijk Bureau voor de Vakbeweging), 2018.

Fantasia, Rick, and Kim Voss. Hard Work: Remaking the American Labor Movement. Berkeley: University of California Press, 2004.

FNV. De Kracht van Beweging: Meerjarenbeleid FNV 2026–2030. Amsterdam: FNV, 2026.

Foster, William Z. Organizing Methods in the Steel Industry. New York: Workers Library Publishers, 1936.

Grimbergen, Cees. Zwarte Zwanen. Documentary series. Omroep MAX, 2013–2025.

Guinan, Joe. "Socialising Capital: Looking Back on the Meidner Plan." International Journal of Public Policy 15, no. 1/2 (2019): 38–58.

Hancké, Bob. Sociaal bewustzijn en vakbondsdemocratie. Licentiaatsverhandeling, Vrije Universiteit Brussel, 1985.

Hancké, Bob. "Vakbondsleden en vakbondsdemocratie." Tijdschrift voor Arbeidsvraagstukken 2, no. 4 (1986): 30–43.

Hyman, Richard. Industrial Relations: A Marxist Introduction. London: Macmillan, 1975.

Knotter, Ad. "Justice for Janitors Goes Dutch." International Review of Social History 62, no. 1 (2017).

Massop, Sjarrel. "Interview met econoom David Hollanders over de pensioenen (1): De mythe van de dekkingsgraden" and "(2): Pensioenen, vermogensbeheer en rol vakbeweging." Solidariteit, 2019. solidariteit.nl.

McAlevey, Jane. No Shortcuts: Organizing for Power in the New Gilded Age. New York: Oxford University Press, 2016.

Moody, Kim. On New Terrain: How Capital Is Reshaping the Battleground of Class War. Chicago: Haymarket Books, 2017.

Müller-Jentsch, Walther. "Trade Unions as Intermediary Organizations." Economic and Industrial Democracy 6, no. 1 (1985): 3–33.

"Pensioencomités in debat met vicevoorzitter FNV Elzinga." Globalinfo.nl, 2020.

Piven, Frances Fox, and Richard A. Cloward. Poor People's Movements: Why They Succeed, How They Fail. New York: Vintage, 1977.

Red het Pensioenstelsel (Landelijk Actiecomité van FNV-kaderleden). redhetpensioenstelsel.nl.

Rouffaer, Casper. "Vakbond verkoos geld boven achterban — onder druk van oud-VVD-minister." Follow the Money, 2025.

Rouffaer, Casper, and Tom Claessens. "Vakbond De Unie morrelt aan de fundamenten van het poldermodel." Follow the Money, 2023.

Stepan-Norris, Judith, and Maurice Zeitlin. Left Out: Reds and America's Industrial Unions. Cambridge: Cambridge University Press, 2003.

Van den Berg, Ewout, Lisa Berntsen, and Saskia Boumans, eds. Hoger onderwijs in verzet: documentatie van de actiebeweging tegen de bezuinigingen 2024–2025. Amsterdam: De Burcht (Wetenschappelijk Bureau voor de Vakbeweging), 2025.

Van der Meer, Tom. "Het electorale fundament van de SP – en de dramatische campagne van 2012." StukRoodVlees, 13 juni 2014. https://stukroodvlees.nl/het-electorale-fundament-van-de-sp-en-de-dramatische-campagne-van-2012/.

Tamminga, Menno. De vuist van de vakbond. 2017.

Van Dijk, Jan Joost, Matthias van Rossum, Moira van Diepen, Rosa Kösters, and Arthur Scholte. Precaire polder. Amsterdam: IISG, 2018.