The British bank’s new note estimates humanoid robots can offset 60% of the projected labour-force decline by 2035, with China’s workforce shrinking by 37 million people over the next ten years. Up to 24 million humanoids would be needed.
Barclays published a research note estimating that China’s deployment of humanoid robots could offset as much as 60% of the country’s projected labour-force decline by 2035.
The arithmetic underneath the note is the part that gives the number its sharpness. China’s working-age population is, on Barclays’ demographic forecasts and a 65% participation-rate assumption, expected to contract by 37 million people over the next ten years.
Manufacturing is roughly a quarter of China’s economy, so a 37-million-worker decline is not, on the bank’s framing, a slow-acting demographic question. It is an industrial-based risk that arrives within a single decade.
The 💜 of EU tech
The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!
Barclays’ broader ‘Robots Roll Out, Economies Rewire’ research quantifies the trade as needing up to 24 million humanoid robots deployed across the Chinese workforce by the mid-2030s, equivalent to roughly 4% of the country’s current labour force, to fill the projected gap at the level of effective output rather than at the level of nominal employment.
What makes the Barclays note operationally interesting is the underlying production reality that the analysts are leaning on. China is now the centre of the global humanoid-robot supply chain, in every available unit-shipment measure.
Unitree shipped roughly 5,500 humanoid units in 2025 (more than any other producer worldwide); Shanghai-based Agibot followed at 5,168 units; UBTech produced around 1,000. Beijing has set a 2027 MIIT target for integrating humanoids into manufacturing supply chains, and a 2035 target for a 300bn-yuan ($41bn) domestic humanoid-robot market.
The ‘up to 24 million robots’ number Barclays uses is a maximum-substitution figure rather than a base-case forecast. MERICS’s research on China’s embodied-AI strategy has the wider context.
The bank’s estimate assumes continued rapid cost-curve decline on humanoid units (Unitree’s industrial-grade products already retail below $20,000 in some configurations), continued government subsidisation of pilot programmes between humanoid-robot firms and state-owned factories, and the maintenance of the current 6-8% annual manufacturing wage growth in eastern China that has been driving the underlying automation case.
Where the note’s framing becomes more cautious is on the customer side of the supply chain. The Chinese humanoid market is, on current independent surveys, supplied by more than 150 active manufacturers chasing a buyer base where only 23% of customers report satisfaction with the robots they have purchased.
The reality-check the operational data forces is that the 24-million-unit deployment Barclays needs to land its 60%-offset estimate runs into commercial-readiness questions that the current supply curve does not, on the available evidence, cleanly resolve.
The geopolitical and corporate context is the third leg. Tesla’s Optimus programme is, on its own reporting, scaling production at the company’s Shanghai Gigafactory, with the company’s third-generation humanoid debuting at AWE 2026 in Shanghai earlier this year.
The competitive frame Barclays sketches in the note is one in which Chinese humanoids displace not just human Chinese workers but also the Western humanoid offerings (Tesla Optimus, Figure 02, 1X Neo) that would otherwise compete for the same factory-floor deployment slots.
The cost differential at industrial scale, on the unit-economics figures circulating across the past two quarters, is the primary lever.
On the macroeconomic read, Barclays is making a specifically Chinese argument. The labour-force-decline-versus-automation trade is being run in parallel in Japan, South Korea, Germany, and most of Eastern Europe, but China is the only major economy combining (a) the demographic decline at this absolute scale, (b) the domestic robot-manufacturing capacity, and (c) the policy framework actively coordinating between the two. The Barclays note is the formal sell-side articulation of why the combination matters at the index-level economic forecast scale.
What the note does not resolve is the absorbed-versus-displaced question. A humanoid robot offsetting a worker on a production line is, in formal terms, a productivity gain. The 60%-offset figure assumes the rest of the Chinese economy can absorb the 40% of labour-force decline that the robots cannot.
The two-decade demographic frame Barclays uses is wide enough to make the assumption defensible at the macroeconomic level. At the social-policy level, the same assumption is the question Beijing’s planning institutions will be answering for the rest of the decade.
