Alphachatterbox: Gavyn Davies and Tyler Cowen on the productivity puzzle

Episode Summary

Executive Summary: The episode examines the productivity puzzle: why advanced economies, despite rapid innovation, have seen weak productivity growth since the early 2000s or earlier. Gavin Davies and Tyler Cowen debate measurement error, welfare gains from the internet, capital deepening, demand weakness, trade/globalization, and cultural/psychological stagnation. They agree innovation continues but differs in impact and timing, while disagreeing on whether mismeasurement or real structural slowdown is the main cause.

Main Topics: What the productivity puzzle is (Priority: 5/5): The discussion defines productivity and total factor productivity as the output gained from combining labor and capital more efficiently, beyond simply adding more workers or machines. Mismeasurement vs real slowdown (Priority: 5/5): They debate whether GDP and price statistics undercount internet-era welfare and free digital services, or whether productivity really has slowed materially in the market economy. Innovation, welfare, and the internet (Priority: 4/5): Tyler argues much of the internet’s value is already reflected in spending and that free services are overhyped in productivity terms; Gavin is more open to unpriced welfare gains like Wikipedia. Investment, capital stock, and demand (Priority: 5/5): The speakers connect weak productivity to lower business investment and smaller capital stock, while rejecting a simple demand-deficiency explanation as the main driver. Long-run structural causes and social change (Priority: 4/5): Cowen emphasizes psychological, sociological, and institutional shifts toward caution, NIMBYism, and lower ambition as deeper reasons for stagnation. Globalization, trade, and labor markets (Priority: 4/5): They discuss whether globalization suppressed wages, encouraged labor substitution for capital, and contributed to the West’s productivity slowdown, with special concern about the UK and US. Sectoral differences and the future of innovation (Priority: 4/5): Healthcare, education, biotech, robotics, driverless cars, AI, and biosciences are used to illustrate both current weakness and possible future productivity surges.

Key Arguments: TFP measures the extra output from smarter use of labor and capital, not just more inputs. A significant part of the slowdown is real, not just statistical mismeasurement, especially in the US consensus view. Digital services provide welfare gains, but many are already indirectly captured through spending on devices, broadband, and related goods. Free goods like Wikipedia are valuable, but their unpriced nature does not automatically mean large hidden GDP growth. Business investment has weakened, reducing the capital stock available to workers and lowering productivity growth. The post-2000 productivity slowdown is too long-running and too broad to be explained mainly by the 2008 crisis or demand weakness. Social attitudes have shifted toward caution, blocking coalitions, and lower appetite for ambitious projects, which may suppress innovation and investment. Globalization and the arrival of hundreds of millions of new workers likely depressed Western wages and may have contributed to labor-capital substitution. Innovation has not ended, but some major breakthroughs take decades to show up in productivity statistics. Future gains may come from biosciences, AI, IoT, robotics, and driverless vehicles, but not quickly enough to solve near-term fiscal problems.

Data Points: U.S. productivity growth expectation in the 1990s: 3% annually - Tyler Cowen said he once expected U.S. productivity growth to average 3% for decades. Real wage expectation in the 1990s: At least 2% annually - Cowen said he expected real wages for most income classes to rise by at least 2% a year on average. Post-2000 growth regime: About 1% - Cowen argued the economy entered a lower-growth regime around 2000. CPI mismeasurement claim in the UK: Up to 1% per annum - Gavin Davies cited Charlie Bean’s work suggesting inflation may have been overstated by as much as one percentage point yearly. UK output per person vs France and Germany: About 30% lower - A question from the audience and Gavin’s response highlighted the UK’s significant level gap. Life expectancy trend in the U.S.: Up to last year, average life expectancy rose at a roughly constant rate for decades - Cowen used this to argue health gains were no longer accelerating. Large labor inflow from globalization: Hundreds of millions - Cowen said the arrival of hundreds of millions of new workers into the labor force affected Western wages. Time since the productivity crisis began: Starts in 2000; longer-term measures start in 1973 - Cowen said the slowdown predates the Great Recession and stretches back decades. Potential monetary obligation issue: Public and private debt - Davies argued welfare gains don’t pay down debt obligations measured in money.

Pivotal Quotes: "you can't pay your mortgage in Facebook likes" — Gavin Davies: Used to explain why nonmarket welfare gains do not solve monetary debt burdens. "Techies are wrong. The techies are wrong." — Tyler Cowen: Cowen argued that internet value is mostly already captured in GDP and productivity statistics. "I think it's a productivity event in my mind, which is really serious and dangerous for the world economy more than it is a puzzle" — Tyler Cowen: Cowen framed the slowdown as a major structural event rather than a measurement mystery.

Implications: Listeners should expect weak productivity to remain a fiscal and political problem even amid visible tech progress. The debate suggests future growth may come in bursts, but governments, firms, and households must plan for slower income growth, tighter budgets, and possible reallocation shocks.