Here are two numbers from the same government table, for the same year, about the same thing. One says 12,774. The other says 5. Both are labelled 2023. Both come from the Ministry of Justice's family court statistics (Source: Ministry of Justice, Family Court Statistics -- Family_Court_Tables__Jul-Sep_2024_ -- Table_3).

Run the two against each other and you get a change of 255,380%. That figure is real arithmetic and complete nonsense. It is not a trend. It is not a surge. It is what happens when two cells that measure different things get lined up next to each other and treated as a before and after.

So let's deal with the number that actually matters. A prohibited steps order is the family court telling a parent what they may not do: no taking the child abroad, no moving them to a different school, no changing their name, no contact with a named person. It is the court stepping into a private decision because the adults cannot be trusted to make it between them. In 2023 the courts recorded 12,774 of them. That is roughly 35 a day, every day, including Christmas.

Set that against the 5. Whatever that cell represents, a subcategory, a rounded or suppressed count, a single quarter, a specific order type, it is not the national picture. And nothing in a headline percentage tells you which of the two you are looking at.

This is where I get annoyed, and it is my own view rather than the data's: a table that can be misread this easily is a publication problem, not a reader problem. The family court tables are free, quarterly, and genuinely rich. They are also dense enough that an automated tool, a councillor, or a journalist on deadline can pull two adjacent figures and produce a story about an explosion in court intervention that never happened. The Ministry of Justice knows its own footnotes. Most people downloading the spreadsheet do not.

The strongest alternative reading of my own finding is worth stating plainly. The 12,774 may count applications rather than orders made, or orders across multiple linked cases involving the same family, which would inflate it relative to the number of children affected. The 5 may be entirely legitimate in context: a narrow category, a partial period, or a figure held back because it is small enough to identify individuals. Neither possibility rescues the 255,380% comparison, but both mean the honest sentence is "12,774 recorded in 2023, definition depending" rather than anything grander.

One more thing about age. The tables in question are the July to September 2024 release, and the annual figure here is for 2023. That is the most recent full year available in this cut, not a snapshot of today's courts. If you see this number quoted as current, it isn't.

The useful takeaway is small and unglamorous. Thirty-five times a day, a court in England and Wales formally restricted what a parent could do with their own child in 2023. That is a big enough fact on its own. It does not need a fake percentage bolted to the front of it.