A school can take years to build.
A child can tak but a moment.
That mismatch is becoming one of the quieter challenges facing England's education system.
For years, the question was how to find enough school places.
Now the question is what happens when some of those places are no longer needed.
England created a net increase of around 1.3 million school places between 2009/10 and 2024/25 as pupil numbers grew. But the direction has changed. The National Audit Office reported this year that primary-school demand fell by 3 per cent between 2018/19 and 2024/25, with a further 7 per cent decline projected between 2025 and 2030. The proportion of unfilled primary places rose from 10 per cent to 14 per cent over the same period.
On paper, that sounds like a problem of numbers.
Inside a school, it looks rather different.
A classroom does not become cheaper because fewer children sit inside it.
The heating still needs to come on. The roof still needs repairing. Teachers still need to be paid. A school cannot simply lose ten pupils and remove ten per cent of its costs.
And then there is the question nobody can answer from a spreadsheet:
What will the children who do arrive need?
A population forecast can estimate how many five-year-olds may live in an area.
It cannot tell us which child will need a wheelchair-accessible classroom.
It cannot tell us who will need a specialist teacher, additional support or a quieter space.
It cannot tell us who will flourish because somebody happened to notice that they were struggling.
That is why the changing number of school places is not simply an estates problem.
It is a planning problem.
And planning is becoming harder because the change is not happening evenly.
The National Audit Office found that primary pupil numbers actually increased in 16 per cent of local authorities between 2018/19 and 2024/25. Even in authorities expecting an overall decline, many smaller areas are still forecasting growth.
One part of a council can be looking at empty classrooms while another is asking where its next school will go.
Housing can change the picture overnight.
Families move.
Birth rates change.
A development that exists on a planning document today can become hundreds of households tomorrow.
A school that looks too large in a national forecast can suddenly become exactly the right size for its neighbourhood.
This is why forecasts are useful — and why they should never be mistaken for certainty.
The Department for Education's latest pupil projections, published in July 2026, use population estimates, birth data and school-census information to estimate future demand. They are designed to help governments and schools plan.
They cannot tell the future.
Nobody can.
And that matters because school buildings have long lives.
A decision made today can shape a community for decades.
Close a school too quickly and a growing neighbourhood may find itself short of places a few years later.
Keep every school open indefinitely and money can become tied up maintaining buildings that are increasingly empty.
There is no painless option.
The financial pressure is already visible.
School funding largely follows pupil numbers. The National Audit Office estimates that a projected fall of 56,300 primary pupils in 2027 could reduce schools' per-pupil funding by around £288 million, although the impact will vary significantly between schools and areas.
That creates a temptation to look at an empty classroom and see waste.
But an empty classroom is also capacity.
It could become a nursery.
Specialist provision.
A community space.
A room for smaller classes.
Or simply somewhere to put the children who arrive after the forecast was wrong.
That last possibility may be more important than it sounds.
Because the children being planned for today are not yet statistics.
Some have not been born.
Some have not moved into the area.
Some have not yet discovered what they are good at.
And some will need things that nobody thought to put into the original plan.
This is particularly important as England attempts to reform support for children with special educational needs and disabilities. Local authorities have warned that delivering those changes will require additional specialist provision, staff and capacity.
The challenge, then, is not simply building enough places.
It is making sure the places we build can change.
A school should not have to be designed around one forecast and then remain trapped by it.
The best planning is not perfect prediction.
It is resilience.
It is leaving enough room for the future to surprise us.
Because somewhere, right now, there are children we haven't met yet.
We do not know their names.
We do not know what they will need.
We do not know which school they will attend.
But one day, they will walk through a school gate and inherit the decisions being made now.
We cannot give them certainty.
We can give them something better.
A system flexible enough to listen when they arrive.