By Sophie Lindqvist

The RMT has called eight days of strike action between November and February, three of them landing in the Christmas shopping window — including the final Saturday before Christmas Day, the date retailers call "Super Saturday" or "Panic Saturday" because of its outsized sales. The announcement came this week from Eddie Dempsey, the union's General Secretary.

The union's leverage is real, but so is the political arithmetic it is testing. London has seen more than 160 days of strike action under Mayor Sadiq Khan — more than three times the combined total under both of his predecessors at City Hall. That record sits awkwardly against Khan's election pledge that there would be zero strikes if he won the mayoralty. A promise made; a promise measured against 160-plus days of disruption.

How automated is the Tube already?

More than most passengers assume. Drivers actually drive trains on only three lines — the Piccadilly, the Bakerloo and the Waterloo & City. On the remaining eight, acceleration and braking are computerised under Automatic Train Operation (ATO), known in industry jargon as "Grade of Automation 2" (GoA2); the person in the cab largely opens and closes the doors.

The Docklands Light Railway sits at GoA3: trains run without drivers, but a Passenger Service Agent remains onboard. The endpoint is GoA4 — fully unstaffed operation — and it is hardly experimental. The Riyadh metro, the Dubai Metro, Singapore's Mass Rapid Transit, the Vancouver SkyTrain and the Copenhagen Metro all run at GoA4. The USA has 33 such systems; Paris runs three fully automated Metro lines. In the UK, fully automated transit is confined to airport links such as Heathrow's Terminal 5 Transit and Birmingham's AirRail.

Public anxiety has never been the main obstacle, the argument runs — the DLR has been automated since 1987 and carried 90.8 million passenger journeys last financial year. Cost is the barrier that actually bites.

What does £20bn buy?

Transport for London (TfL) has estimated that fully automating the three oldest lines — the Central, Bakerloo and Piccadilly — would cost £20bn, with the Central Line alone absorbing £10bn. New signalling, platform edge doors, more modern trains and other upgrades are prerequisites, not extras. Any claim that automation is imminent should therefore be read against a two-decade-scale capital programme, not a budget line.

What changed last month is the funding backdrop. Budget plans to fully devolve business rates to local mayors by the end of the decade could put an extra £9bn a year on the table. Set against Greater London Authority spending of £22.9bn in 2026/27, that is a material uplift — before counting the £10bn of additional borrowing the Budget unlocked. TfL had earmarked much of that windfall for line extensions and improved stations; redirecting a significant proportion toward driverless conversion is now at least plausible. Plausible, however, is not funded, and no reallocation has been announced.

What does this change for Londoners?

In the short term, very little: the strikes proceed, and three of them fall on the most commercially sensitive days of the retail calendar. The longer-term question is whether repeated disruption erodes the public's resistance to driverless conversion. The RMT can currently hold the capital to ransom with impunity — a phrase advocates use deliberately — because anti-strike legislation is off the table: the Labour government overturned the law that would have limited walkouts. Automation is thus cast as the structural fix rather than the legislative one: a strike-proof network, more resilient and reliable, that breaks the union's hold on the capital's economy.

The counterweights remain unspoken in this framing. A £20bn estimate covering only three lines implies a network-wide figure nobody has published. Devolved business rates arrive "by the end of the decade," not next year. And the DLR's 1987 precedent involved a new-build railway, not retrofitting Victorian tunnels. What the strikes change, concretely, is the politics: every walkout strengthens the case that the cheapest union to negotiate with is the one that no longer exists.