A forecast made in 2015 said driverless technology would create 320,000 British jobs by 2030. We are close enough now to check

In March 2015, the Society of Motor Manufacturers and Traders published research it had commissioned from KPMG. It was billed as the first comprehensive analysis of what connected and autonomous vehicles would mean for Britain.

The findings were specific. A fifty-one billion pound boost to the economy. More than twenty-five thousand fewer serious road casualties a year. And three hundred and twenty thousand jobs, by 2030.

A transport minister welcomed it, noting the government was investing nineteen million pounds in real-world trials of driverless cars and a hundred million in research.

That was eleven years ago. The deadline is four years away.

I raise it because forecasts about work almost never get marked. The usual reason is that they are written to expire beyond anyone's attention span. A prediction for 2050 costs nothing to make and nothing to get wrong, because by the time it is due, the person who made it has moved on, the organisation has rebranded, and the number has been superseded by a newer number with a further deadline.

This one is different only in its timing. It was made for 2030, in 2015, which means it will fall due while the people who made it are still working.

I am not going to tell you it has failed. That would be a forecast of my own, and I have no better claim to know what the next four years hold than KPMG had in 2015. Employment in a new industry can arrive suddenly. The technology has taken longer than expected and could still take less than expected from here.

What I can tell you is what the record shows so far, and what it does not.

It shows around fifteen hundred jobs generated in the UK self-driving sector between 2018 and 2022, on roughly four hundred and seventy-five million pounds of direct investment. That is the only published measurement of delivery I can find.

It does not show anybody tracking the 2015 forecast. No annual update, no revision, no statement that conditions changed. The number was made, reported, and left.

Britain has done this before, and the pattern is consistent enough to be worth naming.

In 1965 the National Plan set a target of twenty-five per cent economic growth over six years. It was abandoned within a year in a sterling crisis, and the department that produced it was wound up in 1969. The plan was not marked wrong. The apparatus that would have marked it stopped existing.

In 2008 Parliament created a duty on sixteen and seventeen year olds to remain in education or training. The duty commenced. The enforcement provisions never did, and the department confirms there is no penalty and no plan to introduce one.

In 2009 Parliament created a duty to offer an apprenticeship to every qualified sixteen to nineteen year old who wanted one. It never came into force at all, and was removed by another Act two years later, before it commenced.

None of that is dishonesty. It is the ordinary behaviour of a system where making a commitment is somebody's job and checking it later is nobody's. The making has an author, a date, a press release and a minister. The checking has none of those things, so it does not happen.

Which brings me to the only recommendation I have, and it is small.

The 2030 forecast is four years out. The 2035 government projection of thirty-eight thousand is nine. Both are close enough that somebody could commit now to publishing, on those dates, what actually happened, alongside what was said would happen.

That commitment would cost almost nothing. It requires no new survey and no new body. It requires one organisation to write down, in advance, that it will mark its own homework in public on a stated date.

The reason it will not happen is not cost. It is that nobody has ever been asked to, and a forecast whose author expects to be marked is a different kind of forecast from the ones we have been getting.