Unemployment ticked back down to 4.9%. The rest of the labour market still looks softer underneath.

A lot has been going on in markets over the last couple of weeks, so it is worth pausing to remind ourselves exactly where UK unemployment is currently standing — and what the underlying labour market data is actually telling us.

The latest figures show that the UK unemployment rate fell to 4.9% in the three months to February 2026. That is down from 5.1% in September to November 2025, and 5.2% in October to December 2025. So the headline number has improved. But it still remains 0.5 percentage points above the level it was a year earlier, and this is what matters more than the latest tick lower.

At the same time, the employment rate slipped to 75.0%, payrolled employees fell by 74,000 year on year, and vacancies fell to 711,000. There are now 2.5 unemployed people per vacancy, up from 2.0 a year before. So while there is nothing major to be concerned about at this stage, this is the picture of a labour market that is starting to loosen — not one that is accelerating or re-accelerating.

On a historical basis, these figures remain pretty good. And we mustn't forget that we have just been through a period of significant inflation, especially wage inflation, which should have a dampening effect on the employment figures all on their own.

The fall in unemployment did not come with a clearly stronger employment backdrop. ONS reports unemployment down by 60,000 on the quarter — but economic inactivity up by 95,000, while the employment rate was slightly lower. The Office for National Statistics says the rise in inactivity was largely because of students not looking for work, and those inactive for "other" reasons.

That nuance matters. Part of the unemployment drop reflects people moving out of the labour force entirely rather than a broad-based pickup in hiring. The ONS itself advises caution with short-term Labour Force Survey moves at the moment, given ongoing data-collection improvements and elevated volatility, and recommends reading the unemployment rate alongside payroll, claimant-count and vacancy data. When you do that, the cross-checks all point to a softer labour market than the 4.9% headline alone implies.

Vacancies are clearly weaker than a year ago. Total vacancies fell by 29,000 on the quarter to 711,000. That is the lowest level since February to April 2021 — and outside the pandemic period, the lowest since November 2014 to January 2015. That might be a cause for a little concern.

The unemployment-to-vacancy ratio is one of the better high-level measures of labour-market tightness, because it captures both sides of the balance — how many people are looking for work, and how many openings are out there to fill. That ratio sat at 2.5 in December 2025 to February 2026. Unchanged on the previous quarter, but up materially from 2.0 a year earlier. There are now noticeably more unemployed people competing for each open job than there were a year ago.

Wage growth is also easing. Regular earnings growth slowed to 3.6%, and real regular pay growth — that is, adjusting for inflation on the CPIH measure — was only 0.2%. That is essentially flat in real terms.

That might be some cause for concern as well. It is another sign that labour-market tightness is easing rather than intensifying. Workers are no longer in a position to bid up wages aggressively, which historically tends to coincide with rising slack. And with my view on inflation being that we will see headline CPI bouncing toward 3.5 to 4.5% through the summer, real pay is likely to actually go negative again before this cycle is over.

So where is being worst hit by a lack of jobs? A few sectors stand out clearly.

Construction vacancies were down 38.7% year on year, to only 26,000. Arts, entertainment and recreation was down 25.7% year on year to 14,000. Education had 47,000 vacancies, while administrative and support services stood at 48,000. Those are the sectors where hiring appetite has visibly retreated.

It is worth pairing total vacancy numbers with vacancies per 100 employee jobs, because that adjusts for sector size. On that intensity basis, the weakest sectors are construction (1.6), education (1.7), arts and recreation (1.7), and administrative services (1.8). All sit below the wider economy average and well below the parts of the market that are still recruiting actively.

Where demand is still relatively firm

By contrast, vacancy intensity remains comparatively higher in financial and insurance activities (3.0 per 100 jobs), accommodation and food services (2.9), human health and social work (2.7), information and communication (2.6), and professional, scientific and technical activities (2.5). Those are the parts of the economy where recruitment pressure has not yet broken — though the absolute trend in vacancies is lower across the board.

Of the worst hit, it appears to be the relatively small businesses. Firms with one to nine employees saw 32,000 fewer vacancies year on year, with the biggest quarterly fall coming from this same group. By contrast, companies with 2,500 or more staff showed a slight increase on the quarter.

So labour demand is weakest among smaller employers. That fits with the broader picture of higher financing costs, tight margins after several years of cost increases, and the recent rise in employer National Insurance contributions hitting small payrolls disproportionately. Smaller firms are the ones that show stress first when the economic backdrop tightens, and that is what we are seeing in the vacancies data.

Over the last 24 months, what we are seeing is a gradual rise in unemployment rather than anything dramatic. Nothing like a sharp break — just a steady drift higher.

The pattern looks like this: in early 2024 the unemployment rate was in the low-4% range, around 4.2 to 4.3%. Through mid-2024 it was broadly stable. Then 2025 brought a steady deterioration, with the rate climbing through 4.5% in February to 4.7% in spring, and on to 5.0% by August. Late 2025 saw a local peak around 5.2%, and only in early 2026 has there been a partial improvement — the latest 4.9% reading.

That pattern fits well with the vacancies data, which flattened for much of 2025 and then fell again into early 2026. It also fits with the rise in the unemployment-to-vacancy ratio over the past year. None of these data series are conflicting with each other; they are all telling the same story of a market past peak tightness.

Where the trend looks like it is heading is soft-to-sideways, not strong. The reasons:

• Vacancies are falling, and at a cycle low.
• The unemployment-to-vacancy ratio is much less tight than a year ago.
• Payroll employment is lower year on year.
• Wage growth has eased and real pay is essentially flat.
• Small-business hiring demand is notably weak.

Against that, the latest unemployment rate did fall, and some sectors still show decent vacancy intensity — particularly hospitality, health, finance and parts of information-rich services. So the most balanced conclusion is that the UK labour market is cooling, but not collapsing.

In line with my expectations on rising inflation, I suspect we will continue to see this softness for the next 12 to 24 months. I don't think it is going to be disastrous, but I don't think it is going to improve much from here for some period of time. The conditions that would be needed for a clear re-tightening of the labour market — stronger growth, falling inflation, easier financial conditions, more confidence among small employers — are mostly not in place. If anything, the energy shock and the fiscal drift discussed in last week's inflation piece point in the opposite direction.

The most evidence-based framing is that the UK unemployment rate has improved at the margin, but the wider labour market still looks cooler and less tight than a year ago. The drop to 4.9% is real, but it sits alongside weaker vacancies, weaker payroll employment, higher inactivity and a larger pool of unemployed people per vacancy.

This is not yet a convincing recovery story. It is more a story of a labour market that has stopped worsening as quickly, while remaining materially softer than it was in early 2024.