Double counting and the dark arts of government spending reviews
Everybody should look very closely at how many of the infrastructure projects announced are genuinely new, and how many are reannouncements or reheated projects
It is nearly 30 years since I was introduced to the dark arts of government spending reviews, of which we are about to see Rachel Reeves’s first effort. Let me take you back to July 1998, and the first of these three-year comprehensive spending reviews, under the chancellorship of Gordon Brown.
The 1998 review was important. The Labour government, under Tony Blair, was still in the process of convincing the financial markets that it could be trusted with the economy. After the 1997 election, it had pledged to stick to spending that the previous Tory chancellor, Kenneth Clarke, had described as “eyewateringly tight”.
• Spending review: follow live
It was a surprise, therefore, that the 1998 spending review, which had a title, “Modern Public Services for Britain: Investing in Reform”, highlighted what sounded like some very big figures for additional public spending, notably £19 billion more for education and £20 billion extra for health, over three years.
I was of course very, very young at the time compared with now, but I was old enough to realise that this sounded a bit rum. This, after all, was during Brown’s “iron chancellor”, prudent period. And, as conventionally measured, with the additional spending in the third year of the spending review period, these were massive increases.
In 1998-99, annual health spending by government was £37.2 billion, a fraction of this year’s UK health spending of more than £200 billion, so £20 billion would have represented an increase of more than 50 per cent in just three years, which seemed highly implausible, as indeed it was.
The £20 billion figure had been arrived at by what became known as triple counting, so the first year’s increase was counted three times, the second year’s twice, and so on. There was still quite a big increase in health spending — £8.8 billion in cash terms, or more than 20 per cent, added to the annual budget after three years — but well short of what was presented.
I would like to tell you at this point that lessons have been learnt, and that nobody would dream of doing things like this these days. Sadly, no. When Sir Keir Starmer announced recently that defence spending would rise from 2.3 to 2.5 per cent of gross domestic product (GDP) in 2027, he claimed that this would mean £13.4 billion a year of additional spending. In doing so, the prime minister appeared to have discovered some magic numbers, finding £13.4 billion of additional defence spending from £6 billion of cuts in overseas aid.
The truth was that the figure was only arrived at by assuming that defence spending would be frozen in cash terms at this year’s level until 2027, which was never going to happen. Instead, increasing defence spending by 0.2 per cent of GDP, means raising it by about £6 billion, which is the true figure, rather than Starmer’s dodgy one.
• Chancellor should focus spending on functions only government can provide
In the run-up to this year’s spending review, there is an explicit example of multiple counting, used by Reeves in a recent speech and highlighted by the Institute for Fiscal Studies director Paul Johnson in these pages on Monday.
As he put it: “You may already have seen reports of Treasury briefing about an additional £300 billion of spending over the parliament. Now, £300 billion is a very big number indeed. It is about half of total annual departmental spending. But what the Treasury does not mean is that annual spending will be 50 per cent, or £300 billion, higher by the end of the parliament than it was at the beginning.
“Rather, what they are doing is taking total spending in each year under current plans, comparing that with — essentially fictitious — spending plans set out by the last government, and cumulating the difference over a five-year period.”
Within that highly questionable £300 billion is an extra £113 billion of additional infrastructure investment, which suffers from the same problem. The big picture, as set out by the Office for Budget Responsibility (OBR) in March, is that there will be no infrastructure bonanza this week. Public sector net investment is officially projected to be slightly lower in real terms in 2028-29 than last year, 2024-25, and quite a bit lower in 2029-30, which is why the government is so keen to tap into pension fund investment.
While on infrastructure projects, this is a rich vein in spending reviews. We will hear a long list of announcements, covering the length and breadth of the country. In the days when there were more local papers, this was essential for coverage.
Everybody should look very closely at how many of these are genuinely new, and how many are reannouncements or reheated projects. Improvements to the A303 near Stonehenge, cancelled by Reeves last July, have been announced and cancelled so many times, that if you looked back far enough you might find improvement plans from the prehistoric folk who built the monument. As for the rest, from the regeneration of Tipton town centre to the electrification of the Cardiff-Swansea railway, more things are announced than ever see the light of day.
The curious thing about that 1998 spending review, which used cash figures (another dodgy device) as well as triple counting, was that the spin was not needed. The settlement, an average increase in current spending of 2.25 per cent a year adjusted for inflation, with much bigger real increases for health and education, was quite generous. Bigger increases were to follow.
This week the chancellor is working within a much tighter framework, real increases in spending of fractionally more than 1 per cent a year, much of which will be gobbled up by the National Health Service.
The necessary U-turn on winter fuel payments, which was very messy, has emboldened Labour MPs who are looking for further welfare concessions. The labour market is softening, partly in response to Reeves’s tax rises, except for public sector employment, which is still going up, including the NHS, which now employs an astonishing 2.06 million people.
She will need all the spin she can get to make a silk purse out of this sow’s ear. Keep an eye out for that spin.
David Smith is Economics Editor of The Sunday Times
david.smith@sunday-times.co.uk
Rachel Reeves thought she was being clever: punishment has been swift
The Chancellor’s tax-and-spend Budget has paved the way for an illusory boomlet to become a very real bust