Why TVs Get Cheaper And Nurses Don't

Start With Televisions

It is a fair question, and a hundred thousand people have seen it.

She is right that government has something to do with it. Britain's energy costs are substantially a political choice, built out of decisions about market design, levies, and what the country chose to build and what it chose not to. That is a real argument, and we have made the case ourselves.

But look at the first half of her own sentence. Televisions get cheaper every year, and no government did that either. They get cheaper because making a television is the kind of work that compounds: better machines, better processes, more output per worker per hour, without any obvious limit.

Once you have noticed why televisions get cheaper, you have to ask what happens to the things that cannot work that way. A hip operation. A reception class. An hour of a district nurse's time.

Britain is being squeezed by two forces at once, and almost every argument you will hear picks one and denies the other.

Policy comes first. Choices about energy, planning, tax and regulation that make things cost more than they need to. Contestable, fixable, and the proper business of politics.

Cost disease comes second. A permanent tendency for hands-on human work to get dearer relative to anything a machine can make. Not contestable, not fixable, and running under every government in every rich country.

Fix the policy and you still have the cost disease. Ignore the policy and you get both. Anyone offering you one without the other is offering half an answer.

Four Musicians, One HourWhy the violinist never got faster

In 1826 it took four musicians one hour to play a Schubert string quartet. In 2026 it takes four musicians one hour to play a Schubert string quartet. Everything else got faster.

That is the whole of Baumol's cost disease in two sentences. William Baumol and William Bowen were not studying hospitals or schools. They were studying why live classical music kept needing more money every year. The answer was not that the musicians had got worse. It was that the rest of the economy had got better. Four hours of human labour now produce vastly more wheat than they did before industrialisation, and unimaginably more manufactured goods. They still produce exactly one Schubert quartet.

Wages, though, are not set quartet by quartet. They are set across the whole economy. If a car plant triples what each worker produces and pays accordingly, the orchestra has to match that pay rise to keep its violinists, without any matching gain in output. So the price of an hour of skilled human attention rises forever, relative to the price of everything a machine can make.

Now substitute a nurse for the violinist. Or a midwife, a care worker, a reception teacher, a health visitor. Baumol's disease stops being a curiosity about orchestras and becomes the central fiscal fact of every developed country on earth.

Government Is Unusually ExposedWhy a string quartet prices your hip operation

So much of what government does is the kind of work that cannot be sped up without turning into something else. You can automate the manufacture of a hip replacement. You cannot automate the surgeon, the anaesthetist, the recovery nurse, or the physiotherapist who gets the patient walking again.

So the bill climbs. UK health spending rose from 2.8% of GDP in 1955-56 to 8.4% in 2022-23. Health is now around 39% of all day-to-day departmental spending, with a Department of Health and Social Care revenue budget of £202bn in 2025-26. Education, the second largest, ran at about 4.1% of national income in 2024-25. Between them, those two absorb roughly a third of everything the tax system collects.

Every pound of that growth crowds something else out. That is not a scandal. It is arithmetic, and it is happening in Tokyo and Paris and Ottawa too.

Read those three numbers together. More tax. More staff. Less output per pound than seven years ago. Anyone who tells you that is simple has not looked at it.

Where This Becomes An AlibiThree things Baumol does not excuse

Baumol's disease is real, and it has become the most convenient explanation in British politics, because it is the only one that blames nobody. That is what makes it useful to a government, and what makes it worth checking.

First: Baumol predicts rising cost, not falling quality. If Britain were only paying the Baumol tax, we would pay more each year and get roughly the same service. The theory says an hour of nursing gets dearer. It says nothing about a waiting list of 7.27 million cases and a median wait of 11.9 weeks against 7.5 weeks in June 2019. Rising prices and deteriorating service are two different diagnoses, and only one of them is a law of economics.

Second: not all of the state is a string quartet. Payroll, procurement, scheduling, records, estates and back office are not hands-on human work. Those are car plants. They should have got cheaper. Government spends over £26bn a year on digital technology, with 28% of central government systems classed as legacy and nearly 80 rated at critical risk, of which 21 had no remediation funding at all. Nobody in Whitehall can tell you what running the old estate costs per year, because nobody has ever totalled it.

Third: Baumol's own conclusion was optimistic, and it was conditional. He argued that a society can afford the rising price of human care precisely because productivity elsewhere is rising fast enough to pay for it. The share of your income spent on the nurse goes up, but your income goes up faster. Remove that condition and the whole argument inverts.

Britain removed it. Productivity growth ran at around 2.2% a year before 2008 and around 0.5% since. The flat-lining since 2007 is unprecedented in almost 160 years, and between 2008 and 2017 the UK ranked 31st out of 35 OECD countries for growth in output per hour.

Run Your Own NumbersWhat the stall cost you

Whitehall sees this as a line in a spreadsheet. You see it on a payslip. Enter what you earn.

Half Of It Nobody CountsWhere the disease actually bites

Every public argument about this cites the NHS budget. That number is too small, and the ways in which it is too small are not accidental.

Four places carry that bill. The DHSC revenue budget of £202bn, of which around 90% goes to NHS England. £28.7bn of adult social care in council budgets in England, funded by council tax and the social care precept, which means the most Baumol-exposed service in the country is paid for by the most regressive tax in the country. A cross-subsidy paid privately by care home self-funders, who the Competition and Markets Authority found were paying on average 41% more than council-funded residents for the same care. And £162bn a year of unpaid family care in England and Wales, which appears on no balance sheet anywhere.

So everything councils are blamed for cutting, the libraries and the bus routes and the leisure centres and the state of the roads, was crowded out by care. Baumol did not arrive at the town hall as a rising price. It arrived as a closed swimming pool.

Written Into LawWhere the cost disease stops being a metaphor

Median care worker pay in the independent sector was £12.00 an hour in March 2025. Care wages are anchored to the National Living Wage, which is explicitly set to track median earnings across the economy. The NLW rose 32.7% in real terms between 2015-16 and 2025-26. In March 2024, 59% of independent sector staff were earning below the level the NLW was about to be raised to, meaning the statutory floor lifts most of the sector directly.

Follow that through. When wages rise in the wider economy, the cost of care rises by law, whether or not a single care worker becomes more productive. Baumol described a tendency. Britain has written it into statute, and then declined to fund it.

Look at what that has already produced. Social care has been through everything commentators propose doing to the NHS. It is means-tested. It is charged for. It is rationed. It is topped up by a 41% private cross-subsidy and £162bn of unpaid family labour, most of it done by women. If you want to know what "make people pay for more of it" looks like in practice, you do not need a think tank white paper. You need to visit a care home in Wigan.

What Can Actually Be DoneSix families, honestly costed

There are only six things any country can do about this. Every policy you will ever hear proposed is a variant of one of them.

Paying more and promising less are opposite ends of one lever, and the entire political system behaves as though there is a position between them. There is not. You either raise the money or you narrow the promise, and refusing to choose is itself a choice, made silently, in favour of the waiting list.

Inflation gets used far more often than it gets proposed. Hold budgets in cash terms, let inflation do the rest, and the real value of both the debt and the public sector paybill falls without anybody legislating anything. Between early 2022 and spring 2023, average real pay in the public sector fell by about £2,000, or 6%, twice the fall in the private sector.

Be precise about what that achieves. Inflation moves the price level. Baumol is a statement about relative prices. Double every price in the country overnight and the nurse still costs what she did relative to the television, because the ratio between them is untouched. Nothing about the cost disease has changed.

What inflation does change is the real burden of a fixed nominal debt, and even there Britain is a poor candidate. The OBR's own assessment is that a period of higher inflation may no longer be an effective way of reducing the debt-to-GDP ratio, partly because a large share of UK borrowing is short-dated or index-linked, and on index-linked gilts both the coupon and the capital repayment rise with inflation. The country has arranged its own finances so that the oldest trick in the book works less well on it than on most.

So debasement belongs in the table and not in the answer. It buys some time against the debt, less in Britain than you would think. It buys nothing at all against the disease.

Efficiency is what gets announced, because it promises money without pain. Add up every efficiency saving in British government that carries a defensible independent number and you get somewhere between £1.5bn and £2.5bn a year, against public sector staff costs of £260bn. Where the National Audit Office has gone looking, historically about a quarter to a third of headline efficiency savings survives contact with the ledger.

Technology is what gets promised. The best evidence for it comes from Sweden, where AI cut radiologist screen-reading workload by 44% in a properly randomised trial. But note what else that trial did. It increased cancer detection by 29%, which generates more diagnosis, more surgery, more oncology and more nursing, all of it hands-on. Automating the reading of the scan creates work for the people who cannot be automated. Baumol reasserts itself one step downstream.

HonestyThree true things at once

Back To The Television

Nobody needs to blame nurses, or teachers, or care workers, or their own parents. The rising relative price of human attention is a fact of every rich country, under every government. What is distinctively British is the missing half of Baumol's equation: the growth that was meant to pay the bill.

Eighteen years of politics has been spent arguing about how to slice a pie that stopped rising. Five of the six families above are arguments about slicing. Only one is about the pie.

Give Britain the cheapest energy in Europe tomorrow and the nurse would still cost more each year than she did the year before, because that is what happens to work done by people in an economy that keeps getting better at everything else. Policy decides how much worse the bill is than it needed to be. The cost disease decides that there is a bill at all. Growth decides whether you can pay it.