The Centre for Economic Policy Research (CEPR) released a report on the 16th concluding that the sharp rise in US labor productivity since late 2022 stems not from AI-driven technological innovation but from companies intensifying existing labor and equipment utilization—a phenomenon the think tank describes as 'squeezing.' Average labor productivity growth jumped to 2.5% from early 2023 through the first quarter of this year, up from the 1.5% annual rate recorded between 2005 and 2019. CEPR attributes this acceleration primarily to total factor productivity (TFP) gains of 0.8 percentage points and capital deepening contributions of 0.3 percentage points, while labor composition remained virtually unchanged. The European think tank argues that uncertainty surrounding AI's long-term impact has led firms to avoid major hiring or capital investments, instead extracting more output from current resources to meet order increases—a pattern that risks stoking inflation rather than delivering sustainable growth.
CEPR Attributes Productivity Surge to TFP and Capital Deepening
CEPR's decomposition analysis examined labor, capital, and total factor productivity contributions to the productivity acceleration observed from 2023 through the projected 2026 period. TFP contributed a 0.8 percentage point increase compared to the pre-pandemic baseline, while capital deepening added 0.3 percentage points. Labor composition showed negligible change during this window. The report notes that labor productivity growth exhibited sharp volatility during the pandemic years before settling into the accelerated trajectory beginning in 2022, coinciding with ChatGPT's launch in late 2022.
![Labor Productivity Trends]()
Chart illustrating US labor productivity growth patterns from 2005 through projected 2026 period
Corporate Surveys Show Minimal AI Productivity Gains
CEPR's report highlights a disconnect between public expectations of generative AI's productivity benefits and actual corporate experience. Survey responses from numerous companies indicated that productivity improvements attributable to AI adoption remain minimal, with no clear trend emerging. Firms reported that meaningful technological transformation requires extended implementation periods. The think tank notes that while the assumption of AI-driven productivity gains following ChatGPT's late 2022 release appeared logical, enterprise data does not support this narrative.
CEPR Warns Labor Intensification Risks Inflation Pressure
The report distinguishes between genuine technological innovation and the current productivity pattern. CEPR states that authentic AI-driven efficiency gains could support economic growth without triggering inflation, as improved technology lowers per-unit production costs. In contrast, the observed 'squeezing' method—where companies increase output by raising labor intensity and equipment utilization rates without expanding total labor or capital stock—creates upward pressure on labor costs. This distortion, where output rises while input quantities remain static, artificially inflates TFP measurements. CEPR concludes that this mechanism poses significant inflationary risk, as intensified labor usage ultimately drives wage cost increases that firms pass through to prices.
FAQ
What did the CEPR report conclude about US productivity growth since late 2022?
The CEPR report released on the 16th concluded that the acceleration in US labor productivity from 1.5% annually (2005-2019) to 2.5% (early 2023 through Q1 of this year) resulted from companies intensifying existing labor and equipment use rather than from AI-driven technological efficiency gains. The think tank found that total factor productivity contributed 0.8 percentage points and capital deepening 0.3 percentage points to this increase, while labor composition remained unchanged.
Why do companies report minimal productivity gains from AI adoption according to CEPR?
CEPR's survey data showed that companies experience minimal productivity improvements from AI implementation, with no clear trend emerging. Firms stated that meaningful technological change requires extended periods to materialize. The report attributes companies' reluctance to invest in large-scale AI deployment to uncertainty about the technology's long-term impact, leading them instead to meet increased demand by raising the intensity of current labor and equipment usage.
How does CEPR differentiate between genuine AI innovation and current productivity patterns?
CEPR distinguishes authentic technological innovation—which improves efficiency and supports non-inflationary growth—from the observed 'squeezing' approach where firms extract more output from existing resources without expanding labor or capital. The report warns that the latter method inflates TFP measurements artificially and creates inflation pressure by driving up labor costs, whereas true AI-driven efficiency gains would reduce per-unit production costs and enable sustainable economic expansion.
