Productivity is in the news because an impending downgrade by the Office for Budget Responsibility will help determine the extent of tax rises in the November 26 budget. Credible reports suggest the downgrade will take 0.3 percentage points off expected annual productivity growth, which may not sound much but is more than the Treasury was expecting.
Productivity is important for the public finances — each 0.1-point downgrade adds £7 billion to £9 billion to expected future annual government borrowing — and also important for prosperity, for competitiveness and, to paraphrase a famous quote from Paul Krugman, the Nobel prizewinning economist, for just about everything.
Whenever I write about it, though, I am guaranteed emails questioning how productivity is measured and, in some cases, whether it is possible to measure it, and achieve productivity growth, in services. This sector, of course, dominates the UK economy. So let me try to clarify some misconceptions about productivity.
A good place to start is a typical clothes shop, part of our large services sector. Some assistants appear to spend the whole day standing and keeping an eye on the customers, occasionally offering help and advice, straightening out piles of jumpers, replenishing stock and sometimes helping on the till.
Where is the productivity there? The answer is that the shop, if successful, achieves sales and profits, value-added. Productivity is measured by dividing that value-added by the number of staff hours needed to achieve it, and retailers are good at knowing how many staff they need, and for how long, on a given day.
Or to use an example often quoted to me by readers: when they are waiting to be seen, say at a hotel reception, they are frustrated to see staff in the back office apparently glued to their computer screens rather than rushing to the desk and attending to customers. But those staff may be dealing with customer inquiries, and selling future stays, which is time better spent. Productivity is not always visible.
There is a particular issue with public sector productivity, which has been weak in recent years. But the same principle applies. Productivity improvement is achieved by ensuring that “outputs”, for example GP appointments, rise faster than inputs: the number of GPs and the hours that they work.
A problem emerged some years ago when it was realised that reducing class sizes would, other things being equal, reduce teacher productivity. A teacher who previously taught of class of 30 would suffer a drop in productivity if the class size was reduced to 25, so some way had to be found to quality-adjust the teacher’s output.
• Rachel Reeves: Spending cuts and tax rises both options in budget
None of this is straightforward. Service sector productivity is important, but statisticians would concede that it is harder to measure than productivity in manufacturing.
The other issue is how to add all this up into a macroeconomic whole. That there are different measures is not in doubt. Two similar firms will often have very different ways of measuring productivity. How does it all add up?
These complications are dealt with by simply adopting macro measures of labour productivity. They are gross domestic product (GDP) per hour worked, and GDP per person employed. GDP, sometimes called national output, measures output, while hours worked or numbers employed measure labour inputs.
If we have good and reliable measures, measuring productivity is relatively simple. It is why we know that productivity growth prior to the 2008-9 financial crisis averaged 2 per cent a year through thick and thin but has struggled to achieve a quarter of that since.
If you want to get thoroughly depressed, look at recent productivity on these measures. Output per hour worked fell by 0.8 per cent in the 12 months to the second quarter of this year, while output per worker was down by 1 per cent. Since 2019 the per hour measure has risen by a cumulative 1.5 per cent, the other measure by 1.1 per cent, even weaker than the average performance since 2008-9. No wonder the OBR is sharpening its pencil for a downgrade.
But hang on a second. I said measuring national productivity requires reliable figures. Are the statistics reliable? GDP figures are often revised and a new “blue book” of national accounts, out on Friday, may include revisions. But this has always been the case and is not where the problem lies.
• Mansion tax idea is incoherent, ex-Bank governor tells Reeves
It lies with the labour inputs, numbers employed and hours worked. Both rely on the official Labour Force Survey (LFS), which has had serious problems, mostly because of low response rates. To take an example, the LFS suggests employment rose by 725,000 between the second quarters of 2024 and 2025, while other measures, such as HMRC payroll data, say it fell.
Hours worked, on LFS figures, apparently rose by 2 per cent over the period, more than the increase in GDP. The trouble is, nobody believes the employment and hours worked figures, so recent productivity weakness is exaggerated,
The Resolution Foundation think tank has attempted to plug in some more realistic figures for employment and hours worked, based on other data. It thinks employment was falling between the second quarters of 2024 and 2025, not rising, with hours worked also down.
Instead of significant falls in productivity over that period, the Resolution Foundation thinks it rose, by about 1.6 per cent, a very different picture.
This does not, in its view, negate the need for a downgrade by the OBR, following years of productivity disappointment. It and others may, though, call into question the extent of that downgrade and it will be for the fiscal watchdog to argue its case, amid the wailing and gnashing of teeth from the chancellor and her advisers.
How do we raise productivity? The current fashion is for greater innovation, following this year’s Nobel prize for economics award to Joel Mokyr, who majors on this topic. I would add innovation, as always, to a list which includes more investment, better infrastructure and improved skills. But there is another column in all that.
David Smith is economics editor of The Sunday Times