25th August 2026
The latest figures suggest that the UK is moving into a "low-hire, low-fire" labour market. Businesses are not necessarily making large numbers of redundancies, but they are becoming much less willing to take on new employees.
Vacancies have fallen to 707,000, the lowest level outside the pandemic since 2014, while payrolled employment is also below a year ago.
That raises a bigger question than simply whether unemployment is rising. What happens if employers decide that the marginal cost of employing another person is simply too high?
Employer National Insurance, the National Living Wage, pension contributions, holiday entitlement, employment regulation and other costs all form part of the calculation a business makes when deciding whether to employ someone. Individually, each policy can have a reasonable justification. Collectively, however, they can change the economics of employing people.
The latest ONS figures provide some evidence that something is already happening. Payrolled employee numbers fell by 78,000 between June 2025 and June 2026, while the early estimate for July was 94,000 below the previous year. Unemployment was 4.9% in April-June 2026.
At the same time, vacancies have fallen 19,000 over the year to May-July, and the decline is particularly noticeable among the smallest businesses. Vacancies in businesses employing between one and nine people fell by 8,000 in the latest quarter.
That last figure is particularly important for Britain because small businesses are often the entry point into the labour market for young people, people returning to work and those without extensive qualifications.
The danger is that businesses change behaviour rather than simply cut jobs
This is where I think the Government needs to be careful.
It is relatively easy to see the effect of a policy if a company announces 500 redundancies. It is much harder to see the effect when a company simply decides not to employ the next five people it would otherwise have taken on.
The business may instead ask existing employees to work longer hours, invest in software, automate part of the process, outsource work, use temporary labour or simply accept a lower level of service.
None of those decisions necessarily appears as a redundancy in the statistics.
But collectively they can have a substantial effect on employment.
The recent collapse in graduate vacancies illustrates the problem particularly starkly. Adzuna's latest figures show graduate vacancies falling by about 45% year-on-year to only 8,383 advertised positions in July, the lowest level since its records began in 2016.
That is not simply a graduate problem.
It potentially represents the beginning of a much wider change in how businesses recruit.
Could higher wages actually reduce employment?
This is one of those subjects where the answer cannot simply be "yes" or "no".
Increasing the minimum wage puts more money into the pockets of lower-paid workers. Those workers are likely to spend a substantial proportion of their additional income, which can support other businesses.
But there is a point at which the cost of employing someone becomes important to the employer's decision.
If an employee becomes significantly more expensive, the business has to generate enough additional revenue or productivity to justify that cost.
For a large company, there may be scope to absorb it.
For a small café, shop, care provider, garage or local manufacturer working with relatively narrow margins, the calculation can be very different.
That business might decide that instead of employing another person, it will reduce opening hours, introduce technology, increase prices or ask existing staff to do more.
This is why the Government should not look at National Insurance, minimum wages and employment regulation individually.
Businesses experience the combined cost.
National Insurance is particularly interesting
The increase in employer National Insurance was intended to raise substantial additional revenue for the Government.
That is understandable given the pressure on public finances.
But employer National Insurance is effectively a tax on employing people.
It increases the cost of taking somebody onto the payroll without increasing the value of the employee's work to the business.
A business considering whether to employ someone at, say, £30,000 a year doesn't simply consider the £30,000 salary. It considers the employer's National Insurance, pension contributions, holiday costs, equipment, premises, training and administration.
The more expensive employment becomes, the greater the incentive to find alternatives.
That doesn't mean that every increase in employment costs destroys jobs.
It means that the cumulative effect eventually matters.
And the latest labour-market figures suggest that businesses are becoming increasingly cautious.
The Government may therefore face an awkward choice
If unemployment begins rising substantially, the Government may eventually have to reconsider whether the additional tax revenue generated from employment is worth the possible reduction in employment.
That doesn't necessarily mean abolishing employer National Insurance or freezing the minimum wage.
It could mean designing the system more carefully.
For example, the Government might consider whether very small businesses should receive greater relief when taking on their first employees, whether the employment allowance should be expanded, or whether increases in employment costs should be phased differently for sectors where margins are particularly tight.
There is also a wider question about whether taxation should encourage businesses to employ people or encourage them to replace people with machines.
That question is becoming increasingly important because artificial intelligence and automation are advancing at exactly the same time as the cost of employing people is rising.
AI changes the calculation
This is perhaps the most important part of the story.
Imagine a business considering hiring an administrative worker for £30,000 or £35,000 a year.
If new software costing several thousand pounds can perform a substantial part of that person's work, the economics have changed.
The company may invest in the software rather than create the job.
That doesn't necessarily mean fewer jobs forever. New technology creates jobs as well as destroying them.
But it does mean that the Government should be very careful about making human labour progressively more expensive at precisely the moment when technology is making automation increasingly attractive.
The danger is particularly significant for entry-level jobs.
Young people traditionally acquire experience by doing relatively routine work.
If those jobs disappear, graduates and school-leavers can find themselves in a vicious circle: employers want experience, but young people cannot obtain experience because employers aren't hiring them.
The graduate vacancy figures suggest that this is already becoming a serious concern.
There is another problem hiding behind the unemployment rate
The unemployment rate of 4.9% doesn't look disastrous by historical standards.
But it may not tell us enough.
The more interesting question is how easy it is for somebody to move from unemployment into work.
When vacancies are plentiful, somebody losing their job may find another relatively quickly.
When vacancies collapse, the same person can remain unemployed for much longer.
That is when unemployment becomes economically and socially more damaging.
The latest figures show around 2.5 unemployed people for every vacancy, compared with 2.3 a year earlier.
The difference may look small, but multiplied across the whole economy it represents a substantial change in the balance between employers and jobseekers.
Government needs to watch the direction, not just today's numbers
I wouldn't argue that the latest figures prove that National Insurance and the minimum wage are causing the decline in recruitment.
There are other factors.
The economy has been growing only modestly. Businesses face high energy costs. Consumer demand remains uncertain. Interest rates are still relatively high. International trade is becoming more difficult. And companies are investing in automation and AI.
It would therefore be far too simplistic to blame one Government policy.
But that is not the same as saying the policies have no effect.
When a business is already uncertain about demand, another increase in employment costs can be the factor that changes the decision from "Let's hire someone" to "Let's manage without them."
That marginal decision is precisely what economic policy needs to understand.
Britain could end up with a strange labour market
We could eventually have a situation where businesses complain that they cannot find particular skills, while simultaneously reducing the number of vacancies they advertise.
At the same time, young people could struggle to get their first job, older workers may find it harder to return to employment and businesses may increasingly automate routine work.
That would be a very different problem from traditional mass unemployment.
It would be a labour market in which jobs still exist, but the doorway into employment becomes narrower.
And that could be particularly damaging to the next generation.
The Government may need to change course — but carefully
I don't think the answer is to abandon the minimum wage or reverse every increase in employment costs.
Low-paid workers need protection, and there is a strong economic argument for ensuring that people who work full-time can earn enough to live on.
Nor would it make sense to suggest that businesses should be allowed to ignore employment rights simply to create more jobs.
But the Government does need to recognise that every additional cost placed on employment changes the calculation made by employers.
At some point, policy has to ask whether it is encouraging businesses to employ people or encouraging them to find ways of operating with fewer people.
And the latest evidence suggests that the question is becoming more urgent.
The UK has not yet reached a jobs crisis.
But the warning lights are beginning to flash.
Vacancies are falling. Payrolled employment is below last year. Small businesses are reducing recruitment. Graduate vacancies have collapsed. Private-sector wage growth has slowed sharply to 2.8%, its weakest rate since late 2020.
That combination deserves attention.
The Government may discover that the most important employment statistic isn't how many people have lost their jobs. It is how many jobs employers have decided not to create.
And if that number continues to grow, ministers may eventually have to reconsider the whole balance between taxing employment, protecting workers and creating the conditions in which businesses actually want to employ more people.