Like Groundhog Day, every month sees a new headline showing apprenticeships down compared with last year. Yesterday was no exception. The only thing that varies is the percentage fall. So is it time to panic, or to keep calm and carry on?

1) Things are going to get worse before they get better

There was a big spike in apprenticeship starts in April 2017, the month before the levy’s introduction. So even if apprenticeships continue at their current levels, this will be a bigger percentage fall compared with the spike last year. Conversely, the sharp fall in May 2017 when the levy started will make May 2018 look like a great month even if it isn’t.

2) The government is increasingly unlikely to hit its 3 million target

Even if numbers pick up (which they probably will in the second year of the levy), it would take what looks like an unrealistic spike to catch up lost ground by 2020. Governments should generally try to meet their manifesto commitments, but starts were always the wrong thing to target and ministers are already preparing the ground by saying they won’t sacrifice quality for quantity.

3) It is too early to tell if quality is rising

The government argues that quality is rising as a higher proportion of apprenticeship starts are at level 3 and above, and in standards that require more training. You can see why officials have put this in minister’s lines to take, but there is not enough evidence to conclude it yet.

We need to look at apprenticeship completions, and employment and earnings outcomes across sectors, before we say quality has risen. And, of course, Ofsted told the Commons Education Select Committee earlier this week that half of the apprenticeship providers it has inspected were inadequate or needed improvement.

4) There should not be another 3 million-type target

Targets drive focus and perhaps without the 3 million target, there may not have been the levy. But targets also drive behaviour and the target has created pressure to shoehorn everything into being an apprenticeship and an emphasis on quantity rather than quality.

When you only have a hammer, everything looks like a nail. Beyond 2020 we need to look at other measures of success, such as the proportion of young people taking apprenticeships and their impact on jobs, pay and productivity.

5) We need to build on the current system, not scrap it

With each passing month, pressure will grow for a fundamental rethink on the levy and other reforms. But this would be to throw the baby out with the bath water. Instead, we need to develop new measures of success instead of starts alone.

We need to widen access to apprenticeships, perhaps as the Learning and Work Institute has proposed through a new apprentice premium. We also need to underpin quality, including through making sure apprenticeship standards match the best in the world, and prepare people for careers not just their current jobs.

The fact that apprenticeship numbers are headline news is a good thing. It shows a bigger focus on them and technical education than for many years. And the current reforms are overall a good thing, too. We’ll be exploring the next steps we need in an essay collection bringing together leading thinkers next month.

Stephen Evans is CEO of the Learning and Work Institute