Is the minimum wage to blame for falling youth employment?

This blog was originaly published on the Resolution Foundation’s Substack. 

Once a year the Low Pay Commission (LPC) assemble in their Canary Wharf office to decide the following year’s minimum wage rates. I went along to share my thoughts on falling youth employment. I said that we can’t confidently blame the minimum wage. But the LPC should still tread carefully and refrain from shrinking the gap between the youth rate and the adult rate next year.

Young people are having a tough time

Youth employment rates are ringing alarm bells, having fallen further than for older workers in recent years, as shown below.1

Employment rates among 18-24 and 25-64-year-olds fully recovered after Covid, but the trend for young people then diverged. The youth employment rate fell from 2022 onwards, now standing almost 6 per cent below 2019 levels. By contrast, the rate among 25-64-year-olds held up, reached a new high in 2023, and only then drifted down to 1 per cent below 2019 levels.

It’s not about studying or migration

Before we get to the minimum wage, it’s worth heading off some challenges to this negative picture.

Firstly – it’s not about studying. Student numbers are flat. This means that the share of young people not in employment or education (the NEET rate) has almost certainly been rising – which is also what the LFS tells us.

What about population growth? If the employment rate is jobs divided by people, then maybe it’s more people, not fewer jobs? If we focus on the numerator for a moment – the number of jobs – it looks like there might be some truth to that. The number of 18-24-year-olds in PAYE employment recovered post Covid and then stayed pretty flat.

Flat jobs and a rising population might be less worrying if we thought the population increase was coming from groups we’d expect to work at lower rates. We know there’s been high migration recently – so maybe it’s just that lots of people arrived with low employment rates?

But that’s not what’s been happening. Firstly, the number of jobs held by young UK nationals has been falling since 2022 – it’s not that population growth from overseas is making a flat picture on jobs look bad.2

Second, the number of UK nationals has probably been growing since 2022, due to a mini baby boom twenty years ago.3 This means jobs among UK nationals would have had to have been growing since 2022 for the employment picture to be stable among that group.

So, the picture is pretty much the same as where we started. There has been a serious fall in youth employment since 2022, which isn’t accounted for by rising education participation, or by population or migration changes.

Ok, so it’s genuinely bad. Why?

Honestly, we don’t have great answers. Lots of the obvious candidates don’t line up neatly with a deterioration that began in 2022.

Earlier in the year we dug into rising NEET rates and found that half of the post-2019 increase can be put down to a general labour market slowdown. The rest seems to be to worsening health, but that’s a slower moving trend, not something that accelerated in 2022 specifically. It’s tempting to reach for AI as an explanation, but Claude and ChatGPT weren’t very widely used until fairly recently, so they don’t really help with the 2022 question.

What about the minimum wage? There have certainly been big increases. Between April 2022 and April 2026, average weekly earnings rose 24 per cent, the adult minimum wage rose 34 per cent, and the older youth rate rose 59 per cent.

Lots of people have put two and two together: the minimum wage went up, youth employment went down, what more do you need to see? Well plenty else changed over that period, and it’s hard to strip that out. The IFS did some great analysis on the recent youth rate changes and couldn’t rule out a zero effect. And you find zero effects in some other studies.

What if this is all structural change, not policy? Perhaps the shape of the economy shifted against the young. Retail and wholesale jobs never recovered post-Covid and have been falling since 2022, while hospitality jobs have fallen since 2023. Together those sectors account for 2-in-5 youth jobs, and they stopped adding jobs just as more young people needed work. You can tell a story of big forces here: online shopping; automated warehouses; and a preference for screens to seeing our mates.

But maybe the minimum wage is why they shrank – it makes what they sell more expensive and causes firms to save on low-wage (disproportionately young) labour. That said, it’s not true that recent job shedding is all happening in low-paid sectors. This implies that either recent sectoral changes aren’t being driven by policies affecting low-paid labour costs (like the minimum wage, and employer NICs), or that those policies are having an effect, but there just happen to be other drivers affecting higher-paying sectors at the same time.

When you’re not sure what’s going on, be cautious

So, what to do if you are the Low Pay Commission? They have to make an actual pounds and pence decision for next year’s minimum wage rates – they can’t get away with the hand waving I’ve done here.

I think in the face of big uncertainty about policy effects, and when the labour market is weak, the right choice is caution. We can’t say that the minimum wage is causing the youth employment downturn. But absence of evidence is not evidence of absence. And with the UK’s high minimum wage rates, low hourly pay isn’t the problem it once was. So, I think the LPC should avoid another convergence hike to the youth rates this year.

One LPC Commissioner looked me in the eye and said give me a number. I’d eaten several of their biscuits so I could hardly say no. So my specific recommendation for 2027 was to uprate the whole framework – the adult rate and the youth rates – in line with average wages.

2 Data aside #2: These charts are counting PAYE jobs not people – because that’s what HMRC published – and there’s a probable discontinuity in 2021 from IR35 changes boosting the number of PAYE jobs per person in the PAYE system. But it’s safe to look at trends before and after this point.