Why has productivity growth slowed since 2005?

Question

AcadiFi · Accepted Answer

US labor productivity growth indeed slowed from 2.7% (1995-2005 IT boom) to 1.3% (2006-2020). Major explanations:

(1) Measurement error (Feldstein, Groshen): digital services (Google, Facebook, free apps) provide value not captured in GDP. Estimates suggest mismeasurement adds 0.3-0.5pp to true productivity. Plausible but can't fully explain a 1.4pp gap.

(2) Ideas getting harder to find (Bloom, Jones): doubling R&D effort no longer doubles output; fundamental innovations are exhausted or require exponentially more effort. Semiconductor density improvements required 18x more researchers in 2010s vs 1970s for similar progress.

(3) Business dynamism decline (Decker, Haltiwanger): startup rates fell from 13% of firms in 1980s to 8% in 2010s; job reallocation slowed. Less creative destruction limits productivity gains from new firms replacing stagnant incumbents.

(4) Investment weakness: intangible capital (software, brand, R&D) grew but tangible investment weakened post-2008. Corporate buybacks preferred over capex. Lower capital deepening directly reduces labor productivity.

(5) Skill-technology mismatch: educational attainment gains slowed; workers struggle to adopt new technology, creating transition frictions.

(6) Slower IT diffusion: 1995-2005 reflected enterprise IT adoption; post-2005 gains concentrated in a few superstar firms (Amazon, Google) without economy-wide spillover.

(7) Robert Gordon thesis: the major productivity-enhancing innovations (electricity, auto, indoor plumbing) are behind us; digital gains are narrower.

AI/generative AI is the 2025+ wild card—potentially reversing the slowdown. Goldman estimates AI could add 1.5pp to productivity by 2035 if widely adopted. Beacon Hill should monitor business investment in AI tools and productivity statistics for early signs.

Why has productivity growth slowed since 2005?

Master Level II with our CFA Course

Related Questions

Practice Questions