The UK has a long-running productivity problem. Since the financial crisis, growth in output rates has been weak to flat and generally lower than for similar economies.
Output per hour was growing at an annual rate of around 2 per cent before 2008, but has fallen to 0.5 per cent since then, according to EY. But having been the laggard at the back of the pack for so long, the consensus view is that a rush of AI advances will help the UK economy pick up some serious pace.
One reason for this is the prevalence of knowledge-intensive industries of the type expected to experience strong productivity gains from the new technology.
According to the International Monetary Fund, obvious AI beneficiaries include business services, finance, IT and pharmaceuticals. The lender believes that if Britain’s regulatory constraints can be eased and AI skills developed, output gains could rise by two-thirds.
If that happens, AI will achieve a feat that has been beyond the ability of a succession of governments and assist the current one to create the exact outcome it wants from its drive for growth – a real improvement in standards of living.
There is evidence that productivity is rising. According to Berenberg, the past 18 months have been one of the best periods for productivity growth since 2008, with output per job growing by 1.4 per cent year on year, based on Office for National Statistics (ONS) figures.
With jobless numbers rising (ONS estimates for July indicate that the number of payrolled employees has fallen by 94,000 over the past year), growth is not coming from businesses expanding their workforces; it’s higher output from fewer employees. Berenberg links one-third of the increase in output last year to computer programmers’ adoption of AI, noting that an 8 per cent surge in information and communication sector output coincided with an 8 per cent drop in employment in the industry since 2023.
It puts another third of the growth down to low-pay sectors (such as hospitality) adapting to higher costs imposed by the government, with the final third due to productivity gains elsewhere in the economy.
Not everyone agrees that AI is behind the productivity boost. Other economists put more emphasis on workforce reductions, and cost-based cuts at that.
Although AI is expected to destroy jobs, with automation and process efficiencies one factor behind expected productivity gains, nearly 90 per cent of the Bank of England’s decision maker panel (DMP) report no material impact from AI on employment over the past three years in the July survey.
Pantheon Macroeconomics, which sees few indications of an AI-driven boom, says it has found limited evidence that AI has been displacing workers. Instead, it says job cuts resulting from the government’s national insurance increase and minimum wage hikes lie behind the bulk of the productivity surge. It would be deeply ironic if Labour government policies driving businesses to cut their workforces turn out to be the solution to this particular crisis.
Pantheon agrees that AI adoption is growing and will eventually boost productivity, but argues that it will take time for the new technology to show up in “widespread efficiency gains”.
According to the ONS, self-reported use of AI in UK businesses with 10 or more employees has increased from around 12 to 35 per cent since late 2023, while the DMP survey shows increasing intensity of AI use and an expectation that AI will raise productivity by around 0.9 per cent a year in the next three years.
One drag on UK output, however, is the much weaker productivity gains in the public sector. EY estimates that the public sector’s poor performance is costing the economy £80bn a year, and that if its productivity growth had kept pace with the private sector between 2019 and 2024, UK GDP would have been 3 per cent larger. It warns that without improvement, the economic shortfall could hit £170bn a year by 2030.
Public sector output has improved recently, but the input rate is considerably higher (for example, in healthcare), which suggests continually rising levels of spending may be required to drive improvements. EY suggests part of the solution must involve more tech-enabled and AI tools, if only for the most straightforward queries.
As AI adoption continues to rise, AI-driven productivity gains may be the UK’s best hope in its current economic bind. An ongoing boost will go a long way to easing the chancellor’s fiscal problems. Berenberg says Prime Minister Andy Burnham may get lucky: it estimates that each 0.1 percentage point upgrade in the annual trend GDP growth rate means £10bn more tax revenue in five years.