Little unites Rachel Reeves, the Centre for Policy Studies and the Tony Blair Institute. But a shared desire for more robots in Britain somehow has.
In 2019, Reeves stated: ‘The real danger for the UK economy and for future jobs growth is not that we have too many robots in the workplace but that we have too few.’ This is now conventional wisdom among Britain’s pro-growth elite.
And no wonder: Britain is woefully behind in robot density rankings. According to the International Federation of Robotics, we had just 119 robots per 10,000 manufacturing workers in 2023, compared with Germany’s 429 or South Korea’s staggering 1,012. In fact, our robot intensity trails even the global average of 162, leaving us behind most of Europe, and even Mexico and Turkey. Is it a surprise our productivity growth has been anaemic?
The most extreme example is perhaps the ‘reverse automation’ seen in Britain’s car washing industry. In 2018, a House of Commons report found that the number of automated car washes fell by 1,110 in the decade to 2016.
Traditional hand-wash services had undercut their mechanical counterparts on price, not least because of an inflow of migrant labour.
To many, this encapsulates Britain’s problem: why automate when cheap human labour abounds? One Tory frontbencher suggested to me recently that a liberal migration policy – as a deterrent to capital investment – was the main reason for our relative stagnation since 2008.
Now, it’s worth unpacking the economics here, because that story is simplistic and wrong. Yes, profitable automation clearly has driven productivity growth through history: mechanisation freed labour from agriculture, with higher productivity boosting incomes and demand for services. Robots and AI will undoubtedly repeat this process – when the tech is ready.
Yet as Goodhart’s Law states, ‘when a measure becomes a target, it ceases to become a good measure’. And a hyper-focus on this sort of crude robot statistic could lead to both muddled thinking and bad public policy.
First, there’s no reason robot intensity should be identical across nations, nor can governments reliably judge the ‘correct’ robot-to-worker ratio for their country. Germany’s manufacturing sector comprises 19% of its GDP; in South Korea, it’s even higher. By contrast, Britain’s manufacturing share languishes at 8.6%. Robot usage is naturally higher in large-scale, repetitive industries like automotive and electronics production, less common in Britain’s more specialised and services-intensive economy.
Moreover, single-mindedly pushing robot adoption by artificially inflating labour costs – either through restrictive immigration policies or raising minimum wages is a fool’s errand.
While this does create a substitution effect (encouraging firms to replace labour with machines), it also triggers a damaging scale effect. Higher wages and uneconomic capital investments raise production costs and consumer prices. As prices rise, consumer demand falls, reducing production scale. Thus, even though automation increases productivity at the firm level, industry output more broadly shrinks. This isn’t just theoretical. Ask yourself: how successful has the aggressive rise in the minimum wage we’ve seen since 2016 been in raising aggregate productivity?
The proper free market response to this issue, then, is not to engage in blind robot boosterism, pulling any levers to get more. It’s time to ask: are there broader regulatory and policy barriers that might make such automation unprofitable or even illegal, depriving firms of opportunities to invest productively in new technologies that really are ready?
There is reason to think that such barriers do play a role. After all, even within industries like car production, the UK has 808 robots per 10,000 employees in the automotive sector vs Germany’s 1,492. This at least suggests there’s something preventing us from reaching the technological frontier.
In some sectors, British regulations don’t merely discourage automation – they effectively ban it. Britain was once at the cutting edge of autonomous vehicle research. Yet, while driverless taxis already cruise streets in multiple US cities, our requirement that a human driver remain ready to seize control at a moment’s notice effectively bans autonomous ride-hailing services. Although the Government has now announced a limited robotaxi trial in London next year, it has also pushed legislation allowing wider use of self-driving cars back to late 2027.
Similarly sluggish regulatory approvals hobble drone delivery. Australia enjoys thriving commercial drone networks, while Britain’s cautious regulatory stance means our most frequent drone flights deliver contraband to prisons, not Amazon parcels to homes.
Historically, the tax system has been a problem for traditional manufacturers, though recent reforms have started to rectify this. Britain long occupied the bottom spot in the G7 for private capital investment, exacerbated by unfavourable tax treatment. The short-lived ‘super-deduction’, allowing companies to immediately expense 130% of machinery costs, briefly boosted robot installations significantly. Now, permanent full and immediate expensing for machinery dramatically lowers capital investment costs, encouraging firms to accelerate their adoption of robots and other advanced machinery.
However, we still lack full and immediate expensing for critical structures or integral infrastructure like electrical and heating systems – essential complements for many major machinery-intensive investments. Instead, these vital investments face a sluggish annual deduction rate of just 3%, spread over 33 years, making substantial structural investment unattractive and hampering long-term productivity potential. This imbalance in expensing treatment distorts investment decisions, favouring machinery-only solutions over integrated, long-term expansions that could deliver greater productivity gains.
Then there’s energy policy. Britain suffers among the highest industrial electricity costs globally: UK industrial electricity prices for medium users are more than 89% higher than the median for the EU27; and it’s worse for larger users. Such eye-watering costs dramatically raise the lifetime running costs of automated machinery, discouraging firms from investing in energy-intensive automation projects.
And, yes, our restrictive land-use planning regime creates another powerful barrier. Limited commercial land availability sharply inflates property costs, deterring businesses from making significant investments in larger, automation-capable premises. Economists at the Centre for Cities have repeatedly highlighted how Britain’s planning system penalises capital-intensive operations.
One can add economic uncertainty to this list. Post-Brexit ambiguity around trade and regulatory frameworks, combined with constant shifts in tax policy, have left businesses reluctant to commit to long-term automation investments. Robotics & Automation Magazine recently found uncertainty ranked among the highest barriers to UK firms adopting robots, second only to upfront installation costs.
In other words, Britain shouldn’t pull out all the stops to increase the number of robots in use by trying to distort the price of labour or subsidise otherwise inefficient automation or delivering robot industrial policies. What the Government should do, however, is work through the supply-side reforms that we need anyway to remove the barriers to investment, firm scale and competition that will encourage truly efficient automation to be adopted. It’s a subtle distinction. But only one path leads to a more productive economy, which is surely the end goal.
Click here to subscribe to our daily briefing – the best pieces from CapX and across the web.
CapX depends on the generosity of its readers. If you value what we do, please consider making a donation.
