Britain first legislated for young people to stay in learning until eighteen in 1918. It is still announcing it.

There is a step between a law being passed and a law doing anything, and it is called commencement. Parliament writes the power. A minister then decides, separately and often much later, whether to bring it into force. Sometimes that takes weeks. Sometimes it never happens at all, and the provision sits in the statute book looking exactly like law while doing none of the work of law.

Once you know to look for that gap, this week's announcement stops being news and starts being a pattern.

Take the oldest one. The Education Act 1918 raised the school leaving age from twelve to fourteen and, in the same Act, made provision for part-time continuation classes for those in work aged fourteen to eighteen. Parliament's own account of the period is blunt about what followed: many of those changes could be implemented only in part, or not at all, because of the public spending cuts forced by the economic depression of the 1920s. The idea that young people should stay in some form of learning until eighteen is not a modern one. It is a hundred and eight years old, and it was defeated by a Treasury.

The 1944 Act went further than re-stating the idea. Read its list of sections and the machinery is all there: county colleges, a duty to attend them in accordance with college attendance notices, enforcement of that attendance, a section on what those notices do to a young person's working hours, and a ban on charging fees at them. Parliament did not sketch compulsory part-time education to eighteen in 1944. It built it, clause by clause, with the enforcement attached.

The county colleges were never established. The accounts of why are consistent and they are the same as 1918: post-war money. By 1959 a Minister of Education was reviewing education for fifteen to eighteen year olds and describing colleges for required part-time attendance to eighteen as an unfulfilled promise of the 1944 Act. That the scheme existed in the Act is beyond doubt and can be read on the statute. That it was abandoned for cost is the settled account rather than something I can point to in a single official document, and it is stated here as such.

Then 1973, when work experience in the final year of compulsory schooling was lifted out of child employment law. Then 1998, when that was extended to the last two years, which is to say down to age fourteen. Then 2008, when Parliament created a duty on sixteen and seventeen year olds in England to stay in education or training until eighteen.

That 2008 duty is where the machinery is most visible, because it was commenced in two halves and only one half arrived. The duty itself came into force in stages in 2013 and 2015. The chapter behind it, providing for attendance notices, panels, penalty notices and ultimately a criminal offence, was left uncommenced. The Department for Education has said plainly that there is no penalty for a young person who does not comply and no current plan to switch those powers on.

So England has a legal obligation on every sixteen-year-old with nothing whatsoever behind it. That is not a scandal. It may well be the right call, since prosecuting a disengaged seventeen-year-old is a poor way to engage them. But it means the duty functions as an announcement wearing the clothes of a law, and the country has been treating it as settled for eleven years.

Now the structural point, which is the one I actually want to make.

In none of this is there anybody whose job it is to go back. Commencement has a paper trail. Implementation does not. A power can be passed and never switched on, a duty can be switched on without its enforcement, and a programme can launch without a published measure of whether it did anything, and no part of the machinery is obliged to notice any of it. Select committees and the audit bodies can look, and sometimes do, but they choose their subjects. Nothing compels the question.

That is a design fault, not a conspiracy, and it is shared across every administration that has held this brief since 1918. It also explains something that would otherwise look like carelessness: why a government can announce technical routes for fourteen-year-olds using a power it has held since 1998 without anyone in the room saying so. If nobody is charged with remembering, the institution does not remember.

The remedy is embarrassingly cheap. When a programme launches, publish the number it expects and the date it will report. Foundation apprenticeships launched last August. Their first published figure, 160 starts in eight months, arrived because it happened to sit in a routine statistical release, not because anyone had promised it.

And before anyone reaches for the excuse that the numbers are hard to get at, consider what the House of Commons Library does in its briefing on youth unemployment. It prints the European Union average. On the same page, in the same chart, it prints every individual country as well. Everything a reader needs in order to see through the aggregate is set out directly beside the aggregate.

The aggregate is what gets quoted. Not because anyone hid the rest, but because a single number travels and a spread does not. That is worth sitting with, because it means the problem is not always secrecy or spin. Sometimes the truth is published, in full, next to the misleading summary of itself, and the summary wins anyway. Which is a harder problem than dishonesty, and it will not be fixed by asking people to be more honest.

A hundred and eight years is long enough to notice the pattern. The fix is a column in a table.