Britain Doesn’t Have a Cost-of-Living Crisis. It Has a Cost-of-Policy Crisis.

We talk endlessly about Britain’s cost-of-living crisis, but I’m not sure that’s quite right. It feels increasingly like we’re describing the symptom rather than the cause. Britain has a cost-of-policy crisis.

The problem isn’t simply that government taxes too much, spends too much, regulates too much or gets individual decisions wrong. It is that government increasingly seems incapable of seeing the economy as a system. Policies are designed vertically, in silos. One department solves its problem, another department solves another, the Treasury finds some revenue and another minister announces an intervention. Each can be defended independently. Then they all collide horizontally in the real economy, and nobody appears responsible for the collision.

Government wants higher wages, so it raises wage floors. It wants more revenue, so it increases employer National Insurance. It wants greater security at work, so it increases the obligations and risks attached to employing people. It wants fewer people economically inactive, so it spends billions trying to get them into work. It wants businesses investing and growing, so it announces growth strategies. It wants Britain to lead the AI revolution, while simultaneously making human labour more expensive relative to automation.

You can agree with the objective behind every one of those policies and still ask the obvious question: does anybody ever put them all on the same piece of paper? Because businesses must.

The cost of the next job

Forget the macro models for a second and look at the actual decision a business makes: do I hire the next person? The employer doesn’t experience National Insurance policy separately from wage policy, employment law, pension costs, energy prices, interest rates or weak demand. They experience the cumulative cost and risk of employing somebody.

From 6 April 2025, employer National Insurance increased from 13.8% to 15%, while the threshold at which employers start paying it fell from £9,100 to £5,000.^[1] From 1 April 2026, the National Living Wage increased to £12.71, while the rate for 18–20-year-olds rose 8.5% to £10.85.^[2] Again, the argument isn’t that either policy is inherently wrong. The question is whether anyone adequately considered their cumulative effect alongside everything else.

Businesses rarely respond to those changes by immediately announcing mass redundancies. They adjust at the margin. A vacancy isn’t replaced, a junior role disappears, an extra shift isn’t added, automation gets pulled forward, expansion is postponed or three people absorb the work previously done by four. Individually, none of these decisions makes headlines. Collectively, they change the labour market.

The jobs haven’t necessarily been abolished; they simply never exist.

That matters when more than one million 16–24-year-olds were estimated to be not in education, employment or training in the first quarter of 2026 — 13.5% of the age group.^[3] It would be wrong to blame that on one policy. The IFS finds no clear evidence that recent minimum-wage increases have been a major driver, although it stresses that the estimates are too imprecise to rule out a meaningful effect; under-21s are largely insulated from employer NICs because a much higher secondary threshold applies to them.^[4] But complexity doesn’t mean incentives cease to exist. The same IFS research found the proportion of 16–24-year-olds in payrolled employment fell 4.3 percentage points in the three years to December 2025, equivalent to around 330,000 workers.^[4]

Something is clearly happening at the entry point to Britain’s labour market. One part of government worries about youth employment, another increases employment costs, another funds skills programmes, another funds welfare, another tries to increase tax receipts and another worries about productivity. Each part of government can be doing something individually rational while government collectively produces an irrational outcome.

The first job also matters disproportionately. It creates experience, skills, references, confidence and eventually the second job. Losing the first rung doesn’t simply mean losing today’s employment; it damages tomorrow’s productivity.

Paying for our own incoherence

This is where silo policymaking becomes particularly expensive. Somebody in productive employment generates output, pays tax and spends money in the economy while generally requiring less support from the state. Prevent that job from existing and the equation begins to reverse: less earned income, lower tax receipts, less consumption and potentially more state support.

Universal Credit then quite reasonably catches people whose incomes are insufficient. Once the taper applies, UC falls by 55p for each additional £1 of relevant earnings.^[5] Again, each individual component has a rationale. But stand back and look at the system. We increase the cost of creating employment, spend money dealing with economic inactivity, lose tax revenue that employment would have generated and increase welfare expenditure. The Treasury then needs additional revenue to fund that expenditure, so productive activity is taxed again.

The loop becomes remarkably easy to see: higher costs lead to weaker hiring and investment, which contribute to weaker growth and revenues, which increase pressure on spending and taxation, which raises costs again. Government isn’t merely struggling to solve the problem. Its different arms can end up financing the consequences of one another’s policies.

None of this is an argument against minimum wages, welfare or employment protections. It is an argument about sequencing, interaction and coherence. If wages are going to rise sustainably, productivity needs to rise with them. If we want businesses to create entry-level jobs, we need to think about the cumulative cost and risk of creating them. If welfare is supposed to support people into employment, the transition into work has to be worthwhile. If investment is the goal, government needs to understand what makes investment attractive in the round rather than simply announcing another investment strategy.

AI makes the contradiction harder to ignore

AI makes this incoherence more consequential. Government wants Britain to become an AI superpower because AI can increase productivity, create new products and allow businesses to scale. That is entirely sensible. But technology responds to incentives too.

In a confident, expanding economy, a business asks how AI can help it grow. In a weak, high-cost economy, it is much more likely to ask how many people AI means it doesn’t need to hire. It is the same technology producing a completely different economic outcome because the incentives surrounding its deployment are different.

If government simultaneously encourages adoption of AI, increases the cost and risk of human employment, worries about youth unemployment and then spends money trying to get young people into work, we should at least acknowledge the contradiction. AI doesn’t automatically destroy jobs, but if human labour becomes progressively more expensive while automation becomes progressively cheaper, businesses will do the maths. Once again, the first consequence may not be redundancies. It may be jobs that are simply never created.

The economy is not organised like Whitehall

This is the fundamental mistake. Government is organised into departments, budgets, policy teams and ministerial responsibilities. The economy isn’t.

The Treasury can model a tax rise. DWP can model welfare spending. DBT can produce a growth strategy. DfE can fund skills. Another part of government can increase employment regulation. Every spreadsheet can work independently, but there is only one economy underneath them all.

Businesses experience tax, wages, regulation, energy, skills and financing costs simultaneously. People experience wages, tax, housing, benefits, transport and employment simultaneously. Capital considers returns, taxation, regulation, political risk and opportunity simultaneously. Reality integrates what government separates.

That is why the problem isn’t simply bad policy. It is incongruent policy. A government simultaneously trying to make employment more expensive, increase employment, reduce welfare dependency, increase tax receipts, accelerate automation and stimulate investment has to explain how those objectives fit together. It isn’t enough for each policy to have a PowerPoint explaining why it makes sense. The combined system has to make sense too.

Britain doesn’t need artificially cheap labour; it needs productive labour. It doesn’t need low wages; it needs wages rising because productivity rises. And it doesn’t need an ever-larger welfare system compensating for an economy unable to create enough productive work. It needs policy designed horizontally rather than merely administered vertically.

Before introducing another individually “reasonable” intervention, government should be forced to answer a much harder question: what does this policy do when it interacts with everything else we’ve already done?

Because Britain can no longer afford a state where every department can claim its policy is working while the country produced by all of them isn’t.

That is the real cost of incoherent government.