The freedom nobody uses

Posted on: 11 August 2026

Since the Deregulation Act 2015, every state school in England has been free to set its own term dates, a flexibility academies had already enjoyed for several years before that. The reform was presented at the time as a genuine devolution of judgement to the people closest to the pupils, and the Department for Education made a point of naming schools that had already broken the mould, among them the Boulevard Academy in Hull, which cut its summer break from six weeks to four. A decade on, the six week summer holiday is still the six week summer holiday, taken at very nearly the same time by very nearly everybody. The freedom was granted, it was never withdrawn, and it has gone almost entirely unused.

The usual explanation is institutional inertia, which is the explanation people reach for when they have not looked at the incentives. A closer reading of what schools actually do suggests something more interesting: the freedom is unused not because nobody wants it but because the first school to exercise it bears the entire cost of doing so. A primary that moves its summer break two weeks later immediately creates a problem for every family with a child at the secondary down the road, for every parent whose childcare arrangement depends on a sibling being off at the same time, for the shared transport contracts, and for its own staff, whose children are at other schools on other calendars. The gain from a more evenly distributed year, whatever it might be, accrues to everybody. The loss from deviating falls on the deviator. That is not inertia. It is a coordination problem with a stable and thoroughly unloved solution.

The evidence for this reading is not hidden in academic literature, it sits in plain sight in local authority consultation documents. Cambridgeshire County Council explains that it sets its proposed term dates in conjunction with neighbouring authorities under what is known as the Eastern Region Model Calendar, and gives its reason openly: a coordinated approach minimises the impact on families and on school transport providers. The same document concedes that alignment cannot be guaranteed because there is no formal mechanism for the regional setting of term dates. Read that twice. Authorities coordinate voluntarily, at some administrative effort, in order to converge. The absence of a formal mechanism is treated as a problem to be worked around rather than as a space in which to experiment. Nobody in that consultation is asking whether Cambridgeshire might deliberately sit a fortnight away from Norfolk.

Germany asked exactly that question, and asked it in 1964. The Hamburg Agreement established that the summer holidays would be regionally staggered, and the wording of the Standing Conference of Ministers of Education has a bluntness that would struggle to survive an English consultation: the purpose is to prevent the population seeking rest from beginning and ending its holidays at the same moment, given the consequent damage to traffic and to accommodation demand in holiday regions. From that came a rotating system built on five groups of Länder, each group assembled to hold a roughly comparable share of the national population, with the order rotating annually so that no region is permanently disadvantaged. In 2026 Hesse, Rhineland-Palatinate and the Saarland begin on 29 June while Bavaria waits until 3 August, a spread of thirty five days.

Honesty requires the rest of the German picture, because it is routinely told as a fable about continental competence. The rotation is not universal. Bavaria and Baden-Württemberg sit outside it on fixed late dates, historically justified by the agricultural calendar, and the consequence is that on 1 August 2026 fifteen of the sixteen Länder have their schools closed simultaneously anyway. Staggering flattens the peak, it does not abolish it. What it does is move the equilibrium by a few weeks and take the pressure off the edges, which is considerably more than is achieved by a system that never attempts the exercise at all.

Meanwhile England has spent the past two years addressing the same peak from the opposite direction. The National Framework for Penalty Notices came into force on 19 August 2024, standardising what had been a patchwork of local practice: eighty pounds per parent per child for unauthorised absence, rising to a hundred and sixty if unpaid, with a second penalty within three years charged at the higher rate from the outset. Councils issued more than four hundred and fifty nine thousand fines for unauthorised term time holidays in the 2024/25 academic year, and such absences now account for the majority of attendance penalties in England. The public argument about this has settled into a familiar shape, parents pleading the arithmetic of holiday pricing against a state insisting on the value of an uninterrupted education, and both sides are arguing about behaviour that neither of them designed.

Look at the sequence rather than the merits. The synchronised peak came first and was produced by the coordination structure described above. The price premium came second and is the entirely rational response of an industry facing a demand curve with a cliff in it. The fines came third, and their function, whatever their stated purpose, is to make deviation from the peak more expensive at precisely the moment when the market has made conformity more expensive. A family that cannot afford August is now offered the choice between paying the premium and paying the penalty. The policy assumes the problem is individual opportunism, that parents are defecting from a norm for private gain. The structure suggests the opposite: the parents taking children out in June are the ones least able to sustain the equilibrium everybody else is trapped in, and the penalty falls hardest on them.

There is a natural control on this reading available inside the United Kingdom, which is that the peak is not uniform across it. Scottish schools finish in late June and return in the middle of August, Northern Irish dates are set centrally by the department rather than locally, and English and Welsh schools run to late July and early September. Scotland's divergence is a historical inheritance rather than a designed instrument, but it demonstrates that the calendar is not fixed by climate, by daylight or by any external constraint. It is fixed by what everybody else is doing. Move the reference group and the date moves with it.

I have spent enough time between Ticino and London to notice the one structural feature that neither England nor Italy possesses, and it has nothing to do with the length of the summer. The Swiss school year is broken into five separate holiday windows across autumn, Christmas, carnival, spring and summer, which means the summer is not required to absorb a family's entire annual demand for rest. The February sports holidays are staggered between cantons quite deliberately, so as not to overwhelm the ski resorts, which is the Hamburg Agreement logic applied to winter. England has Christmas, Easter, three half terms and then six weeks. The question of how much load any single window has to carry is prior to the question of when that window opens, and it is asked far less often.

What follows from all this is not a policy proposal, because the mechanism suggests that exhortation and penalty are both aimed at the wrong level. Campaigns urging families to travel outside the peak assume an information deficit. There is none. Every parent in England can recite the price difference between the third week of August and the second week of September, and books the third week of August regardless, because the alternative is a fortnight in which the office expects a reply, the childminder is unavailable and the return brings no shared backlog to rejoin. Telling somebody constrained by the behaviour of others what their constraint is costing them does not loosen it, it merely adds the discomfort of clarity.

Which leaves the observation I find most useful. The date on which somebody takes their holiday says very little about their income and a great deal about their position in a network of dependencies. The self employed consultant with an entirely domestic client list is more tightly bound than the salaried employee of a firm whose counterparties sit in another time zone. Financial independence and calendar independence turn out to be different things, and the second is written by other people. Anyone taking a fortnight in late September, and not yet retired, is telling you something quite specific about the shape of their working life.