Something's got to give
Demographically-adjusted public spending and tax
Many people are frustrated or puzzled at the way that increased resources don’t always seem to translate into improved public services.
For example: with little fanfare, the government recently hit its manifesto target to increase the number of GP appointments by 50 million each year. That’s equivalent to 44 more appointments per practice per working day.
GP output is up. But GPs are not being carried shoulder-high across the land by cheering crowds. Why? Because demand is growing insatiably. As the primary care recovery plan pointed out:
“The number of people in England aged 70 or over is up around a third on 2010, from 6.1 million to 8.1 million, and this group has on average five times more GP appointments than young people.”
I am not a pessimist about public service reform. Spending per head and demographics aren’t the only things that matter, and often more can be done with the same budget.
For example, schools in England have soared past those in Scotland and Wales in international rankings not because they got more money, but because of a consistent 20 year-programme of reform to push up standards, weed out weak management and spread techniques that work.
But… demographics do have a massive impact on public services. Changes in the age profile of the population have given politicians helpful tailwinds in the past, and are now starting to deliver powerful headwinds, of such force as to require a reconsideration of many things.
Tax and spend by age
What would have happened if we had combined today’s spending patterns with yesterday’s demographics? Or tomorrow’s?
The OBR provide a useful breakdown of public spending received and tax paid by age.
The red line shows spending by age. Children consume a bit more spending than working-age adults - the cost of school and a bit more health spending. Retired people require a lot more spending - much higher costs for pensions, health and social care (shown with dotted lines)
The blue line shows tax receipts by age. Children basically don’t pay any tax. People then earn more and spend more, leading to a peak in tax payments in middle age. People then start to work less and spend less and receipts fall away.
How the age structure has changed
The ONS produce data for the historic age profile of the population and also a forecast. They produce it by single year of age back to 1971, and five year bands before that.
I have spliced them together here. To simplify things, here I have created three broad bands: young people, working age people, and pensioners.
In 1953 the proportion of people who were pensioners was much lower. It increased until about 1980, then flatlined till about 2010, since when it has increased rapidly. The proportion of working age people declined from the 1950s to about 1980, then rose again, giving the public finances a nice bump, then started falling in about 2010. The proportion of young people has been falling since the mid-1970s.
And looking at the same data again with more detail we can see that within the pensioner category the group that’s growing fastest are the oldest groups over 85.
Back-casting and forecasting tax and spending
What would have happened if we’d had today’s patterns of public spending and tax with yesterday’s demographics? Obviously in practice the pattern of spending by age will probably have been a bit different, but the basic fiscal lifecycle of taking out a bit as a child, paying in as an adult, then taking out again as a pensioner is an enduring one.
The graph below holds spend by age constant and changes spending and tax receipts in line with the demographics of that year, with 2021 used as the base year.
You could have achieved the same service for each age group with quite a lot less cash in the past, mainly because of the smaller proportion of pensioners. Despite the larger number of kids, you could have spent about 9% less in 1953 compared to 2021, and still delivered the same service for each age group.
Demographics made it steadily harder to raise tax revenues from 1953 to 1970, but then then steadily easier from the mid 70s to the financial crisis. You could think of this as the period where the boomers had their career peaks.
From 1979 to 1990 demographics were helping Mrs Thatcher on both the tax and spending side.
In contrast, since 2010 we have been running up the down escalator. More older people has been pushing spending up, while the working age share falling has been bad for tax receipts. (The boomers retiring, if you will).
And if you think the period since 2010 has been hard going, you ain’t seen nothing yet.
Here’s the same chart rolled forward, showing how the coming decades will put upward pressure on spending and downward pressure on receipts.
This growing gap between spend and tax is the underlying cause creating the Graph of Doom which the OBR produce in their Fiscal Risks Report each year:
Obviously all of what I’ve written above is a very rough and amateur look at these issues - I’m sure someone could do a much better version of this. There is a stack of good stuff in each years’ Fiscal Risks Report for starters.
Conclusions
There are about a million things that flow from this, and you can take your pick, but for me:
Policy decisions aren’t the only things driving tax levels and service quality. The ageing society is leading to ever-higher levels of demand. This is the main cause of the high tax burden.
A whole bunch of sacred cows are probably going to be abandoned at some point. for example, will we really keep the aid spending target in the face of these pressures? The conversation about working age welfare spending reform needs to restart - particularly given the explosive growth of claims for mental health conditions.
In every public service the process of reform needs to be massively stepped up1.
While migration isn’t the answer to our demographic challenge, if we reform migration to make it more selective that will improve the public finances.
We need to accelerate the (now growing) conversation about demographics and supporting parents.
For example in the NHS this should involve; the full rollout of productivity-increasing digital tech like electronic bed management; far more accountability and risk of sacking for senior managers; mending the DHSC/NHSE divide with a single finance function; more drastic rebalancing towards primary care and prevention; and changes to the way we hospitalise the end of life. In General Practice I think the reforms now rolling out will make a difference: the new Pharmacy First service will free up about ten million GP appointments a year, the huge investment in technology will increase productivity, as will the detailed plan to hack back bureaucracy.