“…a must-read for serious students of economic policy.” - Glenn Hubbard, Dean, Columbia Business School, and former Chairman of the Council of Economic Advisers
“There are an amazing number of good ideas and interesting points made in Unintended Consequences. The thinking underlying it, and the obvious depth of understanding of the author, are very impressive.” - Steven Levitt, coauthor of Freakonomics; 2004 John Bates Clark Medal
“A full-throated defense of economic dynamism.” - The Wall Street Journal
“…a fresh argument for the productive value of inequality.” - David Autor, Professor of Economics, Massachusetts Institute of Technology
“…a comprehensive explanation of the modern economy.” - Julian Robertson, Founder, Tiger Management
“…serious thinking for serious thinkers. …a thought-provoking blueprint for growing middle- and working-class incomes.” - Mitt Romney, former Governor of Massachusetts
“Unintended Consequences is full of substance, it is one of the must-read books of the year, and once I finish it I will be giving it a second read through right away.” - Tyler Cowen, Professor, George Mason University
“Unintended Consequences should be read by anyone who takes for granted the superiority of progressive taxation and has not thought carefully about the trade-offs involved.” - The New Republic
“…serious thinking for serious thinkers. …a thought-provoking blueprint for growing middle- and working-class incomes.” - Mitt Romney, former Governor of Massachusetts
“…a comprehensive explanation of the modern economy.” - Julian Robertson, Founder, Tiger Management
“Unintended Consequences is far smarter and more thought-provoking than most economics written for the general public” - Greg Mankiw, Harvard University, Former Chairman of the Council of Economic Advisors
“…challenges misconceptions that distort our economic debates.” - Arthur Brooks, President of the American Enterprise Institute
AI Summary.The U.S. college wage premium fell from 0.626 to 0.575 after four decades of expansion, implying the first sustained negative growth in relative demand for college-educated labor. The authors estimate that weaker wage growth in occupations highly exposed to artificial intelligence accounts for roughly 28% of that decline, though the patterns are not causal.
José Azar, Mireia Gine and Javier Sanz-EspínSocial Science Research Network
Date Posted:
Is Database:
Database
The college wage premium flattened in the mid-2010s and has fallen ~8% since 2022. The authors argue that this compression reflects a broad decline in the returns to formal schooling, rather than a decline in the upper tail.
Is AI exposure driving the decline in the college wage premium?
Core argument: The U.S. college wage premium fell from 0.626 in 2022 to 0.575 in 2026 after expanding for four decades, with Current Population Survey data implying the first sustained negative relative demand growth for college labor.
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.
Takeaways by Macro Roundup® AI
The U.S. college wage premium fell from 0.626 in 2022 to 0.575 in 2026 after expanding for four decades, with Current Population Survey data implying the first sustained negative relative demand growth for college labor.
By 2026, moving from zero occupational AI exposure to full exposure had a -0.086 effect on wages, and combined with the college–non-college exposure gap this mechanism accounts for roughly 28% of the 2022–2026 drop in the college wage premium.
Post-2022 wage growth slowed disproportionately in high-exposure occupations that employ a disproportionate share of college graduates, patterns that, while non-causal, indicate task displacement in AI-exposed white-collar occupations plays a quantitatively meaningful role in compressing the skill premium.
Related Articles:
Looking for the Ladder— The article argues that the hiring downturn in occupations most exposed to generative AI reflects interest-rate sensitivity rather than AI itself, since roughly 38% of workers in the most AI-exposed quintile work in information, finance and insurance, and professional and technical services, versus under 2% in the least-exposed quintile.
How Students and Recent Grads are Responding to the Rise of AI— Undergraduate enrollment in degrees tied to the most AI-exposed jobs has risen 8%, indicating students are moving toward rather than away from fields exposed to artificial intelligence. A decline in computer science degrees is more than offset by growth in other exposed fields such as engineering.
AI and Young-adult Jobs: The Real Mystery— Employment rates have declined for young adults both with and without college degrees, with non-degree holders faring marginally worse, and employment growth for young workers of all education levels has lagged that of other workers. Artificial intelligence can replace some lower-skilled jobs, but those jobs are too few to drive youth labor market weakness.
AI Summary.Artificial intelligence models' capability gains are shrinking in percentage terms, falling from about 15% relative to the baseline early on to roughly 10% for the same absolute increment, narrowing the gap between frontier and good-enough models. The author argues this points to a commoditizing market where price, not minor model differences, becomes the main differentiator, pressuring frontier firms' margins.
Paul KedroskyApplied Complexity
Date Posted:
Citing the Epoch’s Capabilities Index, Kedrosky notes “the gap from the model frontier to the 90th percentile is under 6%. The gap from the #1 to the #2 model is roughly zero. The premium that the best models command [are] collapsing with declining variance.”
Do shrinking relative AI capability gains turn models into a commodity?
Core argument: AI model gains that once delivered about 15% improvement relative to the baseline now buy roughly 10%, so progress continues while its rate declines.
Even if the absolute capability gain per unit of time is constant, the percentage gain is not. Early in the data, models improved by about 15% relative to the baseline. At the end of the data, the same increment buys a 10% relative gain. Progress is real, but its rate is declining. The economic implications are clear. When the variance is wide, being the frontier model meant something. You could, after all, be almost 50% better than the average model. And you could, in theory, price like it. The distance between frontier and good-enough was wide enough that it looked like an early moat. This is what maturing commodity markets look like. Price will become the main differentiator, not minor model differences. This will have huge implications, from the continuing and inexorable rise of China in AI, as we are already seeing with DeepSeek's massive pricing advantage, to the resulting margin pressure on frontier companies, and the implications for said companies' post-IPO performance.
Takeaways by Macro Roundup® AI
AI model gains that once delivered about 15% improvement relative to the baseline now buy roughly 10%, so progress continues while its rate declines.
Wide variance across AI models once let a frontier model be nearly 50% better than the average and, in theory, price accordingly, a gap that looked like an early moat.
The author argues AI is maturing into a commodity market where price, not minor model differences, becomes the main differentiator, citing DeepSeek's pricing advantage, margin pressure on frontier companies, and implications for their post-IPO performance.
Related Articles:
Chart of the Day: Small Models are Closing the Gap to Frontier AI— Achieving 60% on MMLU, a test of academic reasoning, once required a 540bn-parameter model but was reached by a 3.8bn-parameter model two years later, a 142x reduction in parameters. Models small enough to run on phones or sensors shift inference from data centers to devices, weakening the case for paying cloud providers per token.
Four Takeaways from Microsoft’s Satya Nadella at WEF Today— Microsoft's chief executive says the price of AI model usage ("tokens") falls by roughly half every three months. Commentary argues that such rapid deflation pulls demand forward, encourages overbuilding, and commoditizes models, shifting competitive advantage to orchestration and context rather than the model layer.
Commoditization, Orchestration, and the New AI Stack— The author argues that as incremental gains from larger AI training runs shrink, the return on additional computing power falls, weakening the case for the training spending that dominates data center demand. Survivors of model commoditization will be those owning the orchestration layer or the applications where outcomes are priced.
AI Summary.Firms that begin supplying a foreign multinational for the first time show productivity gains of about 8% after three years, along with higher sales to other customers. The authors find similar performance gains when firms start supplying other "superstar firms," defined as the largest or most export-intensive firms.
Mary Amiti, Cédric Duprez, Jozef Konings and John Van ReenenNational Bureau of Economic Research
Date Posted:
Is Database:
Database
Evidence from Belgian btw 2002-14 suggests that large firms generate significant TFP spillovers to domestic suppliers. Firms that start a serious relationship with a superstar firm increase their TFP by 8% three or more years after the relationship is formed.
Does supplying a large firm raise a domestic supplier's productivity?
Core argument: Firms that begin supplying a foreign multinational for the first time see productivity rise about 8% after three years, alongside gains in other performance measures such as sales to firms other than the new multinational partner.
We use an event study approach, examining what happens when a firm begins supplying a foreign multinational for the first time. We uncover a sharp increase in productivity (which rises by about 8% after three years) and other performance measures (e.g. sales to firms other than the new multinational partner). The fraction of aggregate value accounted for by multinationals declined by about ten percentage points in Belgium in our sample period (2002 and 2014), which would suggest a strong headwind against productivity growth. However, in a novel result we are also able to document that when we look at similar events of starting a serious relationship with other “superstar firms” - defined as those who are in the top thousandth of the size distribution and/or export intensively - we find very similar performance impacts.
Takeaways by Macro Roundup® AI
Firms that begin supplying a foreign multinational for the first time see productivity rise about 8% after three years, alongside gains in other performance measures such as sales to firms other than the new multinational partner.
The share of aggregate value accounted for by multinationals in Belgium fell about ten percentage points between 2002 and 2014, which would suggest a strong headwind against productivity growth.
Beginning a serious relationship with other “superstar firms” — those in the top thousandth of the size distribution and/or intensive exporters — yields very similar performance impacts to supplying foreign multinationals.
The National Economic Council Gets It Wrong on the Roles of Big and Small Firms in U.S. Innovation— U.S. small firms account for 49% of employment but 16% of business R&D spending and 18.8% of patents issued, while firms with more than 25,000 workers account for 36% of R&D and the largest firms 37.4% of patents. A survey of nearly 1,000 scientists and engineers found most internationally filed patents came from firms with more than 500 employees.
Mega Firms and Recent Trends in the U.S. Innovation: Empirical Evidence from the U.S. Patent Data— Analysis of US patent data finds that the largest firms are more likely to file novel patents even after controlling for size, industry, and patent counts, and that firms produce more novel patents as they become mega firms. Venture-backed startups also account for a disproportionate share of novel patents.
