The last plan to send prisoners abroad was scrapped. The announcement survived anyway.

The story

Reform UK says it will look to send foreign criminals to jails in countries including El Salvador, home to a prison where dozens of inmates per cell are crammed onto four-tier bunk beds. Party chairman Lee Anderson would visit El Salvador for talks, following a similar deal struck by President Trump to deport migrants from the US. Reform has vowed to deport more than 10,000 foreign nationals from British jails to solve overcrowding if it wins power, and also plans pop-up "Nightingale prisons," named after the temporary hospitals used during the pandemic, to keep inmates in jail for longer. The previous government's scheme to send criminals and asylum seekers to Rwanda was scrapped after years of legal challenges and no removals.

The reframe

The Rwanda scheme is a useful marker precisely because of how completely it failed to do the thing it was announced to do. Years of legal challenge, hundreds of millions of pounds committed, and by the time it was scrapped, a number of actual removals that rounded, in practical terms, to nothing. As a policy for reducing prison overcrowding or deterring crossings, it did not work. As an announcement, it worked for years, generating headlines, dividing opposition, and giving successive ministers something concrete to defend at the despatch box.

This sits alongside the entry of 29 July on Fifa's Forward Enterprise proposal, another case where a plan's political value arrived immediately while its practical consequences, if they arrived at all, arrived on a different and much slower timeline. The El Salvador proposal, and the "Nightingale prisons" plan alongside it, are being announced against exactly that backdrop. The previous version of this idea did not deliver. It was expensive, litigated, and eventually abandoned. None of that appears to be a precondition for announcing a new version aimed at the same underlying problem.

The manuscript's account of what an announcement is actually for, separate from whether it works, describes this pattern precisely: language ceasing to merely describe action and beginning instead to substitute for it.

Whether jails in El Salvador will ever hold a meaningful number of British-sentenced foreign offenders is, at this stage, unknowable, in the same way it was unknowable in 2022 whether flights to Kigali would ever leave the ground. What is knowable now is that overcrowded prisons are a genuine, immediate, and politically inconvenient problem, and that a policy addressing it abroad is available to announce well before any of the practical questions, cost, capacity, legal challenge, diplomatic negotiation, have been worked through. The announcement does not need to survive contact with those questions to do its job. It only needs to exist long enough to be reported.

None of this means overcrowding is not a real problem, or that every version of offshore capacity is doomed to repeat Rwanda's outcome. It means that the interval between announcing a solution and delivering one has, on the most recent precedent, been long enough for the entire scheme to be conceived, litigated, and abandoned without a single prisoner's circumstances changing as a result. A new announcement, arriving before that lesson has finished being absorbed, is not obviously evidence that the lesson has been learned.