The appeal for the international dockworkers’ day of action reached me in early September through a forum for Fourth Internationalists doing trade union work.
Seven federations, forty-plus ports, one date: 30 October 2026. The flyer is headed Ports For Peace, and two of the Istanbul Declaration’s six bullet points are about war. The sixth is about work. Technological progress, mechanisation, automation and artificial intelligence, it says, must not be used to eliminate jobs or undermine workers’ rights, and then: “Dockworkers and port workers say NO to automation.” There’s more at work here than meets the eye. To be upfront, I support the day of action, even if that’s a pretty blunt soundbite. I’m against automation that makes things worse for workers, but I support automation that works in our favour.
RedMole.org is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Last week I put the placards against AI data centres through a test drawn from Lenin’s What Is to Be Done?: a slogan spreads one idea to many people; a demand is extracted from somebody and names who should concede it; a measure belongs in a government programme; a plank sits in a platform. “Shut down” failed it, because it borrowed a demand’s grammar without naming anyone to concede anything.
Run the same test on the dockers’ NO, but revise it first. Opening a negotiation with an outright ban isn’t naive; it’s standard practice, stating the maximum and letting the negotiating table pull it down toward whatever’s actually winnable, but in the workers’ interests. The International Longshoremen’s Association, which organises the US East and Gulf Coast ports, did exactly that in its 2024–25 contract fight with the employers’ body, USMX. It went from its president Harold Daggett’s flat demand for “no automation or semi-automation” to a final text reading no automation without prior union agreement and protection terms, and the distance between those two positions was three days on strike and a phone call from Mar-a-Lago. Read that way, the Istanbul Declaration’s NO isn’t a category error. It could be page one of a negotiation nobody has convened yet.
It’s also worth being unsentimental about what deals like this actually deliver. No negotiation of this kind ends clean. The ILA’s own contract leaves plenty unwon: disputed definitions of which new roles count as protected, an arbitration process weighted toward management’s calendar, a leadership with its own institutional stake in calling the outcome a triumph regardless of what the small print says. Trade union bureaucracies defend their own position as readily as they defend the membership’s, and a 99 per cent ratification vote says as much about a tired membership wanting the strike over as it does about the deal’s quality.
The organisations themselves are not the uniform bloc the flyer’s shared letterhead suggests, and this isn’t one unified international speaking with a single voice, the way the ITF might. It’s several autonomous, militant port unions making common cause after Istanbul. CGT is one of France’s oldest and largest confederations, Communist-led but unambiguously mainstream. Liman-İş is affiliated to HAK-İş, one of Turkey’s four recognised national confederations, representing roughly 7,000 dockers across the country’s major ports. LAB holds around a fifth of workplace delegates across the Basque Country, part of a nationalist-left bloc that, with the larger ELA, commands a majority there, though both sit outside Spain’s state-wide unions. USB is deliberately not mainstream: one of Italy’s rank-and-file “base unions,” built as an alternative to CGIL, CISL and UIL. ODT is a minor, recognised Moroccan centrale, some way behind UMT, CDT, UGTM and UNTM. ENEDEP organises container handlers at Piraeus’s Piers II and III, on terminals run by a COSCO-linked operator, and is the more militant of two rival unions there; SYNEDEP, the other, has at times been recognised instead as the majority bargaining representative. OrSA Porti, based at Gioia Tauro, belongs to the same autonomous, grassroots union tradition as USB, and has spent recent years fighting for recognition in company-level bargaining against unions the employer already deals with. Whether that fight has been won, lost or is still running, the reporting available doesn’t say.
Which raises the question the demand/slogan test skates past: addressed to whom? Not one of the six bullet points names an employer, a port operator or a government minister. No PMA, no USMX. Read the signatory list against that gap and a different picture forms: a coalition weighted toward smaller, more combative unions, sitting alongside a couple of genuinely mainstream ones, calling a day of action in ports where the largest local confederation hasn’t moved. The ENEDEP–SYNEDEP split at Piraeus is that pattern in miniature: a recognised, more accommodating union sits inside the official bargaining structure, and a more militant rival builds support by acting where the recognised union won’t. That’s not only, or even mainly, a demand aimed at capital. It’s a recruitment pitch aimed at the rest of the dock workforce: come to the union that’s acting, not the one negotiating quietly on your behalf.
And it has one piece of leverage no automation clause can match. Refusing to load arms shipments bound for a war zone doesn’t require an employer’s agreement, a minister’s signature or six months at a table. ENEDEP has already done exactly that, refusing to unload cargo its members believed was military-grade steel bound for Israel, action a union bound by an existing recognition agreement would find far harder to call. That’s presumably why Ports For Peace sits in the headline and automation gets one bullet among six: military cargo is the lever a minority union can actually pull, on its own initiative, right now, in a way an automation ban isn’t. Lenin’s point about the newspaper, made in Where to Begin? and quoted at length in What Is to Be Done?, the pamphlet this series keeps returning to, was that a paper is not only a collective propagandist but a collective organiser, that producing it builds the network its content merely describes. An international day of action does the same double work. Opposition to war and opposition to job-cutting automation are doing organisational labour here: building a network of unions willing to act together, independent of whether any single employer concedes a thing on 30 October.
The transatlantic comparison usually offered runs the other way: American dockers accept automation and get paid for it, European dockers refuse and get left behind. Neither half of that holds up cleanly.
The 2008 ILWU–PMA Pacific Coast agreement did explicitly authorise “fully mechanized and robotic-operated marine terminals,” in exchange for expanded union jurisdiction over maintenance work and pay guarantees when automated berths cut hours. PMA’s own account presents this as an unambiguous win for both sides. What PMA’s own fact sheet doesn’t mention is that the ILWU’s national leadership criticised the deal at the time, or that the union has resisted further automation at nearly every contract since, using the leverage it kept to block much of what the 2008 deal technically permitted, as reporting on the 2022 contract talks noted.
The 2024–25 ILA–USMX contract runs the other way entirely. Daggett went into bargaining demanding “no automation or semi-automation,” full stop, walked away from the table once management raised it, struck for three days, and came out with a 62 per cent wage rise and contract language barring any fully automated terminal and barring new semi-automated equipment until the union agrees staffing and protection terms first. What the ILA won is not the conditional yes the standard story assigns to American dockers. It is closer to the Istanbul Declaration’s flat no than anything the ILWU signed.
The easy conclusion is that the American record proves leverage beats language, that what matters is the strike and not the wording of the demand. That’s not quite right either. The wording did work of its own on both continents before any leverage existed to use it: the ILA’s flat NO built rank-and-file resolve behind a strike its leadership needed members to actually walk out for, just as this declaration’s flat NO is meant to build a coalition across seven federations before any of them has a strike ready to call.
None of which makes the flyer’s NO wrong, or premature. It changes what winning would look like. The American contracts were fought against a single named employer body, at ports where each union already held the field, with a strike either side believed would be used. The Istanbul Declaration doesn’t have that yet, because its immediate task is different: building the coalition and the willingness to act, not closing a deal with anyone in particular. Thirty October is worth marking on those terms. Whether the organisational gains (a wider network, new recruits, a demonstrated capacity to stop a ship loaded with arms) get converted into the kind of table ILWU and ILA each eventually sat down at is not something the flyer itself can answer.