Top Ten New York Times Bestselling Author
Efraim Benmelech, Janice Eberly, Joshua Krieger and Dimitris Papanikolaou, "Private and Social Returns to R&D: Drug Development and Demographics," National Bureau Of Economic Research, January 2021, https://www.nber.org/papers/w28355
AI Summary. S. 575 between 2022 and 2026, the first sustained decline in relative demand for college-educated labor in four decades. AI exposure in white-collar occupations accounts for roughly 28% of that drop, as wage growth slowed disproportionately in high-AI-exposure jobs where college graduates are concentrated.
José Azar, Mireia Gine and Javier Sanz-Espín Social Science Research NetworkCore argument: The U.S. college wage premium fell from 0.626 to 0.575 between 2022 and 2026—the first sustained decline in relative demand for college labor after four decades of uninterrupted expansion.
After expanding for four decades, the U.S. college wage premium [dropped] sharply from 0.626 in 2022 to 0.575 in 2026. Current Population Survey data through 2026 implies an unprecedented drop in relative demand for college labor—the first sustained negative relative demand growth. Post-2022 wage growth slowed disproportionately in high-exposure occupations, which employ a disproportionate share of college graduates. By 2026, going from zero occupational AI exposure to full exposure had a negative effect on wages of−0.086. Combined with the college–non-college exposure gap, this mechanism accounts for roughly 28% of the total drop in the college wage premium from 2022 to 2026. While non-causal, these patterns indicate that task displacement in AI-exposed white-collar occupations plays a quantitatively meaningful role in the recent compression of the aggregate skill premium.AI Summary. Net investment has fallen from ~40% of gross investment in the 1970s to ~25% today, meaning three-quarters of gross investment merely replaces depreciating assets. The shift toward faster-depreciating information technology assets requires larger gross investment increases to achieve any given gain in productive capital per worker.
Timothy Taylor Conversable EconomistCore argument: Net investment has fallen from ~40% of gross investment in the 1970s to ~25% today, meaning three-quarters of gross investment now merely replaces depreciating capital rather than expanding the productive stock.
The figure divides net investment by gross investment. Back in the 1970s, net investment was often around 40% of gross investment, but the share has been slumping over time. For the last decade or so, net investment has been about 25% of the gross–that is, about three-quarters of gross investment is just making up for depreciation of the pre-existing capital stock. The likely reason for the growing gap between gross and net investment is that modern investment is more likely to be related to information technology [which] depreciates more rapidly and thus needs to be replaced and updated more often. If we want the average US worker to be using a greater amount of capital on the job–which was one of the key drivers of rising labor productivity in the past–it now takes a bigger rise in gross investment to lead to a given rise in net investment.AI Summary. 615). Claims of re-acceleration result from cherry-picking frontier observations, selecting a breakpoint, ignoring variance collapse, and fitting separate lines on either side.
Paul Kedrosky Applied ComplexityCore argument: Epoch’s Capabilities Index shows no statistically significant AI performance acceleration when controlling for developer and model family, with a breakpoint test returning p = 0.615 and a slope-change confidence interval of -8.4 to +23.4 pts per year.
Using all Epoch’s Capabilities Index observations, and controlling for developer and model family, there is no statistically significant breakpoint. A piecewise model—which splits the series into intervals and applies a sub-function to each segment—does not improve on a purely linear trend: p = 0.615, The estimated change in slope has a confidence interval of -8.4 to +23.4 points per year. In short, the maths shows there is no model acceleration, contrary to claims, and as expected. The result comes from selecting frontier observations only, choosing a breakpoint, ignoring variance collapse, and fitting separate lines on either side.AI Summary. U.S. productivity growth has accelerated to ~2.2% annually since mid-2022, above the 2010s baseline, though pandemic-era labor market and business formation dynamics likely contributed alongside AI. Historical general-purpose technology booms sustained labor productivity growth above 2.5% for a decade or more, making the current acceleration substantial but not unprecedented.
Chad Syverson Economic Innovation GroupCore argument: U.S. labor productivity has grown at roughly 2.2% annually since mid-2022, a pace exceeding the 2010s trend and, if sustained, implying GDP per capita roughly 7% higher within a decade than the prior trajectory.
Productivity from mid-2022 on has maintained a faster-than-2010s trajectory involving annual growth of about 2.2%. Could this acceleration be due to AI? Perhaps. The timing leans against AI being the sole initial cause. Additionally, there were well-documented increases in economic dynamism (labor market churn and business formation) during the pandemic emergence whose timing matches the acceleration’s start. Regardless of AI’s current effect, the longer the aggregate productivity acceleration continues, the more plausible it is that AI is an important driver. As for the magnitude, a sustained increase from 1.5 to 2.2% annual productivity growth would be substantial (after a decade, GDP per capita would be 7% higher than otherwise), but hardly unprecedented. The 1995–2004 productivity boom saw annual productivity growth of nearly 3% per year, and other past general-purpose-technology-related productivity boosts saw labor productivity growth in excess of 2.5% for a decade or longer.AI Summary. US corporate investment in equipment and facilities is projected to grow 40% in real terms by the end of next year, versus 12% in the euro area, widening a productivity gap where output per hour worked rose $14 in the US compared with $2 in Europe since 2018.
Sam Fleming, Amy Borrett and Olaf Storbeck Financial TimesCore argument: U.S. corporate investment in equipment and facilities is projected to rise 40% in real terms between 2021 and end-2026, versus 12% in the euro area and near-zero growth in Germany, sharply widening the transatlantic capital-spending gap.
Corporate spending on new equipment and facilities in the US is projected to increase 40% in real terms between 2021 and the end of next year, according to forecasts from Oxford Economics. The US surge compared with a real-terms increase of just 12% in the euro area, while German business investment is expected to have all but stagnated over the same period. Europe also faces a large and growing productivity gap with the US. “The United States has recently pulled further ahead of Europe,” Bart van Ark, a professor at the University of Manchester, told policymakers at the ECB Forum in Sintra. GDP per hour worked increased $14 in the US between 2018 and 2025, compared with just $2 in Europe. “The gap is not only a digital sector story,” added van Ark, stressing that the US outperformance extended to other sectors, including wholesale and retail as well as professional services.AI Summary. Nine major technology companies carry ~$3tn in off-balance-sheet AI commitments — 5x their ~$600bn in reported capital spending — obligations that are growing faster than traditional investment and triple their combined lease and debt liabilities.
Peter Rudegeair and Peter Santilli Wall Street Journal
Ed Comment:“Nothing more to add!”
Great new NBER using evidence from the pharmaceutical sector (which parallels your story about your friend's biotech firm you invested in that was trying to extend the life of patients receiving advanced chemotherapy) productivity spillovers from pharmaceutical research are limited as the increased life expectancy/quality of life among the elderly increases social welfare but doesn't show up in TFP.
Obviously that relationship likely wouldn’t hold in terms of Covid related investment but they argue this trend is likely true across investment in intangible capital and is contributing to slower productivity growth.
Core argument, “…a significant share of this increased intangible investment is geared towards medical R&D targeting older patients. To the extent that these patients are no longer in the labor force, their improved health and well-being would be welfare-enhancing but not directly productivity-improving…”
They walk you though it here, “….First,we show that pharmaceutical firms account for an increasing share of the total R&D spending in the economy. In the 1970s, U.S. pharmaceutical firms accounted for less than 3% of the overall R&D spending in the economy. Today, that share has risen to approximately 10%, and their share among manufacturing firms has risen from 8% to 35%.Second, we show that much of this increased spending is geared towards developing drug candidates targeting ailments typically afflicting older patients. Using detailed data on firms’ drug development pipelines,we show that a significant share of drug candidates under development are treating diseases that disproportionately affect patients that have exited the labor force (over 65 years old)....We find that the share of expenditures allocated to treating diseases common in the over-65 year group has increased by more than 50% since 2000.Though much of our analysis is focusing on the post-2000 period, many of the trends we document are also present in the pre-2000 sample…Our estimates suggest that about a third of total R&D spending by pharmaceutical firms is geared towards those aged 65 and over. While prolonging life expectancy and improving quality of life, these investments in R&D have little effect on measured productivity and output growth.Male life expectancy at birth in the U.S. increased from 70 years in 1980 to 76.3 years in 2018. However, the effective retirement age for men has been hovering around 66 since 1980 and has risen only slightly to 67.9 by 2018, while the statutory retirement age increased from 65 to 66. Absent a significant change in retirement patterns, it is unlikely that R&D spending targeting seniors will directly enhance the labor force and hence output growth….”