Metro, Thursday 20 August 2026
A month ago, prime minister Andy Burnham stood outside No.10 and said his "first instruction" as prime minister would be to end rough sleeping in the UK. According to the Chain database, which tracks rough sleeping in London, around 782 people have made their bed on the streets around Westminster in the hours since that pledge was made. The borough consistently records the highest levels of rough sleeping in the UK. While numbers have been dipping slightly, the capital's street population still hovers close to 4,000. Burnham and housing secretary Angela Rayner have now set out their first target: bringing that figure to zero by ensuring nobody in England sleeps rough this Christmas. Emma Haddad, chief executive of homeless charity St Mungo's, told Metro the ambition is "absolutely right" but that she does not know whether it can be delivered. She said most rough sleepers would need to be offered somewhere, not committed to permanently, "so that there are choices and options for people." Both Haddad and the prime minister invoked the Everybody In scheme from the pandemic, when empty hotels and lockdown restrictions provided a wealth of individual rooms. Haddad noted that scheme was always an emergency measure and its circumstances were unique. Writing in the Guardian, Burnham said he does not "pretend this is a permanent fix, far from it," calling it instead the first of three pillars: emergency measures, early intervention as people leave institutions like prison, and addressing the drivers, unaffordable homes, inadequate housing benefit, and unmet mental health and substance misuse needs, that cause homelessness in the first place.
Notice what Haddad is actually being careful about here. She does not doubt the intention. She does not doubt the resources exist, in principle, to bring people indoors this winter. What she cannot say is whether the offer, once made, resolves anything, or whether it simply produces another version of what has already been produced for over a decade: contact, a bed, a Christmas without visible failure, and then a return to the street once the emergency measure ends. The manuscript's account of how systems respond to homelessness starts from exactly this distinction.
A crisis demands resolution. A condition invites containment.
Burnham's own language, however honestly meant, already sorts into the second category. "First of three pillars" is a phrase that describes a sequence, not an outcome. The emergency offer is pillar one. Early intervention is pillar two. Addressing unaffordable housing, benefit levels, and mental health and substance provision, the actual causes Haddad and Burnham both name, is pillar three, and pillar three is the one with no Christmas deadline attached to it. The manuscript describes this exact administrative shape, activity that is measurable and immediate sitting in front of resolution that is neither.
Annual reports often show improvement. Contacts increase. Referral pathways expand. Bed placements become more efficient. Documentation grows more detailed.
The manuscript's point is not that this reporting is dishonest. It is that the reporting measures activity rather than resolution, and the population it is meant to be shrinking rarely shrinks by anything like the same margin.
The 782 figure matters here precisely because it did not wait politely for the pledge to take effect. It is not evidence that Burnham has failed within a month, nobody involved is claiming that. It is evidence of the gap the manuscript identifies between a system's stated priority and its operational tempo. The pledge was made instantly, in a doorway, as a first instruction. The 782 people who have since bedded down around one borough did so inside a machinery that moves at the pace of housing stock, hotel contracts, and benefit reform, not the pace of political announcement.
Haddad's "three pillars" framing is worth reading against the manuscript's account of what containment actually requires to sustain itself.
The system often justifies itself through scarcity, limited beds, limited funds, limited staff. These constraints are real, but they are unevenly applied. Vast resources exist elsewhere in the system, absorbed by inefficiency, duplication, and moral theatre.
Everybody In is the case study that makes this legible. During the pandemic, empty hotel rooms existed because lockdown had emptied them, not because anyone had built new capacity. Haddad's caution about repeating it this Christmas is not caution about resources. It is caution about the fact that emergency measures are, definitionally, temporary, and a temporary fix to a chronic condition produces exactly the cycle the manuscript describes: contact, temporary accommodation, short-term stabilisation, exit, return.
What gives this particular moment its weight is that the prime minister making the pledge is, by his own account, aware of the distinction he is at risk of falling into.
A society that can manage visible failure indefinitely without resolving it is signalling something precise: that the appearance of order matters more than the reduction of harm; that stability is preferred to justice; that containment is safer than change.
Burnham naming pillar three, the housing, benefit, and mental health provision that would actually shrink the number of people who become homeless in the first place, and naming it as the one without a deadline, is either the most honest thing said about this policy so far or the clearest signal of where it will eventually be quietly deprioritised. Both readings are available. What is not available, on the evidence in front of Metro today, is a mechanism by which pillar three happens on any timetable at all. The 782 people who found themselves outside Westminster this month did not experience three pillars. They experienced the first one, if they were lucky enough to be offered it, and the manuscript's account of managed failure is precisely about what happens to everyone the first pillar does not reach.
This entry reads Burnham's pledge, and Haddad's careful hedging around it, as the manuscript's account of homelessness policy operating as a managed condition rather than a resolved crisis, where activity, contact, referral, emergency accommodation, is produced continuously while the underlying population on the street barely moves.