Why school ICT spend looks like it doubled

We found this the boring way, by checking our own finance import hadn't broken. It hadn't. Across the 8,455 schools that filed a return in both 2023-24 and 2024-25, the average ICT figure really did go from £13,102 to £31,740.

Nobody's technology budget behaves like that. The obvious check is the neighbouring lines, same schools, same two years: premises came in at 1.06 times the year before, and energy actually fell, to 0.79. Spending surges don't confine themselves to one category and leave everything else flat.

What moved was the category. For 2024-25 the DfE broke E20 into seven sub-codes: E20A connectivity, E20B onsite servers, E20C IT learning resources, E20D administration software and systems, E20E laptops, desktops and tablets, E20F other hardware, and E20G IT support. Only E20C is the old E20. The other six hoover up spending that used to sit somewhere else on the return, or wasn't collected separately at all.

So the school that put £13,000 of ICT learning resources on its 2023-24 return is now putting its MIS licence, its broadband, its device refresh and its support contract in the same box. Same school. Same money.

The category grew. The spending didn't.

There's a second half that catches people out. ICT above the capitalisation threshold now goes to CE04 in the capital section instead of revenue, so a large one-off device purchase can vanish from the revenue line completely. A school can look like it slashed technology spending in the year it re-equipped every classroom.

We'd rather say this plainly than hedge it: don't benchmark ICT across the 2024-25 boundary at all. If your deck says schools grew technology spend by 145 per cent, you're quoting a paperwork change to the business manager who filed the return. They'll know. That meeting doesn't come back.

Compare inside the year instead. Rank a school against its peers in 2024-25, which is what the DfE's own benchmarking tool is built for, or wait for the 2025-26 returns to give you two clean years on the same definitions. Staff, premises and energy are still fine year on year, because nobody moved the goalposts there.

One thing we can't do, and won't pretend to: separate the genuine increase from the reclassification. Device refresh cycles didn't stop in 2024, so some of that 2.42 is real spending. We don't know how much. Neither does anyone else selling you a number.

Method. Counted from the CFR and AAR returns we import from the DfE's Financial Benchmarking and Insights Tool, limited to schools that filed in both years so the comparison isn't skewed by schools joining or leaving the dataset. Figures are the mean ICT line per school, not per pupil.