The policy didn't change. The optics did.

The story

Prime minister Andy Burnham has paused Labour's plan to release around 6,000 prisoners early from September, freeing up prison space, after fury that the scheme could see the killers of PC Andrew Harper released. Burnham said: "No prisoners will be let out under this policy until we have conducted an urgent review and done everything we can to minimise risk to the public. Where changes are needed, we will not hesitate to make them." The scheme could have meant early release for Jessie Cole and Albert Bowers, who were in the car that dragged PC Harper to his death in 2019; the driver, Henry Long, is not eligible. Harper's widow Lissie said it had "taken me years to rebuild a life that was torn apart in a matter of moments," adding: "It is deplorable to me that we have reached a point where releasing prisoners early is even considered acceptable."

The reframe, with the manuscript

The prison overcrowding problem the scheme was designed to solve has not gone away in the days since this backlash. Whatever case existed for releasing 6,000 prisoners early on Tuesday exists, in overcrowding terms, just as much today. What changed was not the underlying policy logic but its visibility: one case, attached to one widow's public grief and one officer's name the country already knew, became impossible for the government to defend in the same terms it had used for the other 5,999 or so releases the scheme would have covered.

Chapter 2's account of how a rule's force depends on when it is activated rather than what it says is worth setting alongside Burnham's own words here:

Under these conditions, rules accumulate power precisely because they are seldom used. A regulation may sit dormant for years; minor breaches overlooked, paperwork adjusted, informal approvals granted. Then a conflict arises. The same rule, unchanged on paper, is applied in full. What was previously tolerated becomes actionable. The rule did not change; its activation did. The Performance of Obedience, Ch. 2

The early-release scheme runs the same mechanism in reverse. A policy announced as general, covering a defined category of 6,000, was left to operate quietly until one case made its operation politically costly. Burnham's formulation, "where changes are needed, we will not hesitate to make them," concedes that the review's outcome will be shaped by which cases generate the most pressure, not by which prisoners pose the most risk. It does not promise a better rule. It promises a more defensible one.

That distinction, between a rule that is better and a rule that is merely defensible, runs through Chapter 13 directly:

Enforcement gravitates toward what is manageable rather than what is meaningful. The system optimises for defensibility, not resolution. The Performance of Obedience, Ch. 13

A blanket early-release scheme, applied uniformly to a defined category of prisoners, was manageable in exactly the administrative sense the book describes: a clean criterion, easy to explain, easy to implement. It became unmanageable the moment it produced a single case the public would not accept, and the government's response was not to defend the criterion but to carve out a pause around the case that made the criterion look indefensible. Every other prisoner swept up in this scheme, the ones without a widow able to speak to Metro, without a name the public already recognises, does not get the same review, the same urgency, or the same ministerial promise. The 6,000 were always going to be treated as a single category on paper. Whether that category actually functions as one in practice now depends on how loudly any individual case can be made to matter, and that is a test the policy document itself never mentioned.

The chapter is explicit that this is how punishment comes to signal reach rather than principle:

Penalties are applied not to uphold principles but to signal that authority still has reach. Highly visible punishment compensates for widespread tolerance elsewhere. The message is not that the rule matters, but that surveillance is intermittent. The Performance of Obedience, Ch. 1

This entry sits alongside the account of Gemma Barnes's confiscated body camera, covered here on 23 July: two different institutions, in two different weeks, reaching for the same phrase, "appropriate processes," "urgent review," a form of words that announces accountability is underway without disclosing what it will actually change or for whom. Norfolk Constabulary's process has run for two and a half years without a stated outcome. Burnham's review has existed for days. The pattern the chapter keeps naming is the same in both cases: the promise of process is offered as though it were itself the resolution, and only the passage of time reveals whether anything follows it.