“…a must-read for serious students of economic policy.” - Glenn Hubbard, Dean, Columbia Business School, and former Chairman of the Council of Economic Advisers
“…reminds us that inequality sends a signal of what society lacks most, in America’s case, entrepreneurship and risk taking.” - Lawrence Lindsey, CEO, The Lindsey Group, former Director of the National Economic Council
“Unintended Consequences provides a provocative interpretation of the causes of the global financial crisis and the policies needed to return to rapid growth. Whether you agree or not, this analysis is well worth reading.” - Nouriel Roubini, New York University; Chairman, Roubini Global Economics
“…serious thinking for serious thinkers. …a thought-provoking blueprint for growing middle- and working-class incomes.” - Mitt Romney, former Governor of Massachusetts
“Unintended Consequences represents the most cogent and persuasive analysis of the Financial Crisis to date.” - Andrei Shleifer, 1999 John Bates Clark Medal Winner
“…a fresh argument for the productive value of inequality.” - David Autor, Professor of Economics, Massachusetts Institute of Technology
“…challenges misconceptions that distort our economic debates.” - Arthur Brooks, President of the American Enterprise Institute
“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
“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
“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
“Unintended Consequences offers deep and well-argued analyses on almost every issue.” - The New York Times
“…a fresh argument for the productive value of inequality.” - David Autor, Professor of Economics, Massachusetts Institute of Technology
Research on education, skills, credentials, schools, and their relationship to employment, wages, productivity, and mobility. This hub includes narrower coverage of K–12, college, charter schools, and test scores.
AI Summary.An analysis of federal aid recipients' tax-reported earnings finds the highest-earning bachelor's programs are concentrated at highly selective schools and in computer science, mathematics and electrical engineering. Nursing programs at several California schools produce graduates earning around $150,000 annually, placing them in the top 100 of nearly 25,000 programs analyzed.
Arfa Momin and Ron LieberNew York Times
Date Posted:
Is Database:
Database
Is Important:
Important
A HEA Group analysis of the earnings of 2018 and 2019 college graduates four years after graduation finds earnings are widely dispersed across schools in higher-earning fields such as computer science and economics and have a narrow dispersion in lower-earning fields such as English and drama.
Which college majors and schools produce the highest-earning graduates?
Core argument: HEA Group’s analysis of nearly 25,000 bachelor’s programs found computer science, mathematics and electrical engineering majors top the earnings list, with graduates of highly rejective schools including the Ivies, M.I.T., Stanford and U.C.L.A. dominating.
HEA Group tracked students who received federal financial aid and graduated during the 2017-18 and 2018-19 periods. The top of the earnings list is dominated by graduates of highly rejective schools — the Ivies and others, like M.I.T., Stanford and U.C.L.A. The majors that top the list are computer science, mathematics and electrical engineering. The income data come from whatever people reported on their federal tax returns. That means certain stock awards and other potential deferred compensation aren’t part of the totals. Nursing programs at a handful of California schools, including California State University, East Bay; San Francisco State University; and Samuel Merritt University, produce graduates earning around $150,000 annually, putting them in the top 100 out of the nearly 25,000 bachelor’s programs the HEA Group analyzed nationwide. They stick out amid quantitative and science majors from institutions like the California Institute of Technology and Cornell.
Takeaways by Macro Roundup® AI
HEA Group’s analysis of nearly 25,000 bachelor’s programs found computer science, mathematics and electrical engineering majors top the earnings list, with graduates of highly rejective schools including the Ivies, M.I.T., Stanford and U.C.L.A. dominating.
Nursing graduates of California State University, East Bay; San Francisco State University; and Samuel Merritt University earn around $150,000 annually, placing those programs in the top 100 amid quantitative and science majors from institutions like the California Institute of Technology and Cornell.
HEA Group tracked federal financial aid recipients who graduated during the 2017-18 and 2018-19 periods, using income reported on federal tax returns, which excludes certain stock awards and other potential deferred compensation.
Related Articles:
Do Adults Have the Skills They Need to Thrive in a Changing World?— An international assessment of adult skills describes numeracy proficiency levels ranging from counting objects and adding small whole numbers at the lowest level, to extracting information from tables and graphs and performing basic arithmetic with money and percentages at Level 1, up to analysing complex mathematical information and critically evaluating whether data support a claim at Level 5.
Have Humans Passed Peak Brain Power?— The share of Americans reporting they read a book in the past year has fallen below half, alongside declining performance in numeracy and problem-solving in most countries. The share of adults unable to use mathematical reasoning to evaluate the validity of statements averages 25% in high-income countries and 35% in the US.
Standardized Test Scores and Academic Performance at a Public University System— A study of a large public urban university system finds that high school grade point average is a far stronger predictor of student academic success than SAT scores, contrasting with analyses of highly selective colleges that found the SAT superior. Under-represented minority students were less likely to graduate within six years at every grade point average level.
AI Summary.China has extended travel restrictions on top artificial intelligence and chip talent to their spouses and children, who must now obtain government approval before going abroad even for short trips. The requirement broadens existing curbs on the movement of influential technology entrepreneurs, researchers and executives.
Bloomberg StaffBloomberg
Date Posted:
Beijing has expanded overseas travel restrictions for top AI professionals to include their families. While not an outright ban, foreign travel will now require state approval.
Is China extending travel approval requirements to tech talent's families?
Core argument: China now requires spouses and children of certain AI and chip executives whose work is deemed paramount to state security to obtain Beijing’s approval before traveling abroad, even for short trips, according to people familiar with the matter.
Government agencies have recently begun to share the wider restrictions with affected people, which include prominent startup founders as well as the heads of strategically important companies in AI-related fields. Direct relatives such as spouses and children of certain AI and chip executives, whose work is considered paramount to state security, are now required to obtain approval from Beijing before going abroad even for short trips. That marks a broadening of existing curbs on the freedom of movement of China’s most influential tech talent. While not an outright ban on travel, the new requirement is likely to exert a further chilling effect on a tech sector already facing unprecedented curbs. China has restricted foreign travel for top AI professionals in private firms such as Alibaba and DeepSeek since earlier this year.
Takeaways by Macro Roundup® AI
China now requires spouses and children of certain AI and chip executives whose work is deemed paramount to state security to obtain Beijing’s approval before traveling abroad, even for short trips, according to people familiar with the matter.
The restrictions extend to prominent startup founders and heads of strategically important companies in AI-related fields, broadening existing curbs on the movement of China’s entrepreneurs, researchers and executives.
China has restricted foreign travel for top AI professionals at private firms such as Alibaba and DeepSeek since earlier this year, and the new family requirement is likely to exert a further chilling effect on the tech sector.
Related Articles:
Who’s Ahead in the Global AI Talent Race?— An analysis of authors at a leading AI research conference finds China is the largest originating source of elite AI talent, with its share of researchers holding Chinese undergraduate degrees rising 11 percentage points to 57%, while the US share fell to 13%. The US remains a net importer of AI researchers, gaining 2,145 while China lost 1,729.
China Tightens Control Of Overseas Travel In Sweeping New Law— China has imposed new legal controls on overseas travel that formalise restrictions on mid-to-senior civil servants, Communist Party cadres, state-owned enterprise employees, and private citizens working in sensitive areas, as part of efforts to protect state secrets, advanced technology, and highly skilled workers. Violators face steep fines and exit bans ranging from three months to indefinite.
The Future, Made In China— China's economy runs on parallel tracks: a property crash, a booming high-tech manufacturing sector, and university graduates trained for traditional white-collar jobs entering a labor market geared toward advanced manufacturing. Researchers at Stanford and Harvard found a swing from optimism to pessimism, with about 28% of citizens expecting living standards to improve.
AI Summary.An analysis of authors at a leading AI research conference finds China is the largest originating source of elite AI talent, with its share of researchers holding Chinese undergraduate degrees rising 11 percentage points to 57%, while the US share fell to 13%. The US remains a net importer of AI researchers, gaining 2,145 while China lost 1,729.
Damien Ma and Binyi YangCarnegie Endowment for International Peace
Date Posted:
Is Database:
Database
Surveying education/employment histories of authors at the Neural Information Processing Systems conference, Ma and Yang infer China educated 57% of elite AI talent as of 2025, versus 13% for the US. Yet the US gained 2,145 researchers y/y while China lost 1,729.
Does China produce the most elite AI talent while America attracts it?
Core argument: Damien Ma and Binyi Yang find China’s share of elite AI talent, defined by undergraduate degrees earned in China, rose 11 percentage points to 57% in 2025 versus 2022, while the U.S. share fell to 13%.
[Using a methodology based on paper authors from the Neural Information Processing Systems conference (NeurIPS) Damien Ma and Binyi Yang find] China has established a commanding lead as the largest originating source of elite AI talent globally. Compared to 2022, the share of Chinese-origin AI talent, defined as those who received undergraduate degrees in China, increased 11 percentage points to 57% in 2025, while the U.S. share fell to 13%. The United States remains a magnet for global AI talent, particularly Chinese talent. Despite U.S.-China tensions, among AI researchers working in the United States, the 2025 share of those holding undergraduate degrees from China rose by 4 percentage points. When it comes to brain gain vs. brain drain, the United States saw a net gain of +2,145 researchers in 2025 while China registered a net loss of -1,729 researchers.
Takeaways by Macro Roundup® AI
Damien Ma and Binyi Yang find China’s share of elite AI talent, defined by undergraduate degrees earned in China, rose 11 percentage points to 57% in 2025 versus 2022, while the U.S. share fell to 13%.
Among AI researchers working in the United States, the share holding Chinese undergraduate degrees rose 4 percentage points in 2025, despite U.S.-China tensions.
The United States recorded a net gain of 2,145 AI researchers in 2025, while China registered a net loss of 1,729, under Ma and Yang’s methodology based on NeurIPS paper authors.
Related Articles:
China Lures Home Its Top AI Talent From Silicon Valley— Elite artificial intelligence researchers have left US firms including Google DeepMind and OpenAI for Chinese companies such as ByteDance, Tencent, and Alibaba. Three China- and San Francisco-based headhunters say they relocated more than 30 US-based researchers to China over 12 months, up from low single digits the prior year.
China and the Future Of Science— China graduates about 5x as many medical and biomedical students as the US each year, 7x as many engineers, and nearly double the number of science and engineering PhDs, while one-sixth to one-fifth of US science and engineering graduates are themselves Chinese. One study found returning Chinese scientists lead-author 2.5x more papers than peers who stay in the US.
Why Europe Produces Nobel Prize Winners but Not Elon Musks— Europe trains more engineering graduates per capita than China and more STEM PhDs than any region outside the United States, yet the only EU-born companies to reach $100bn market value in the past fifty years, Spotify and BioNTech, listed on US exchanges. Europe captures 6% of global AI startup funding versus 61% for the United States.
AI Summary.College graduates in the majors most exposed to artificial intelligence show statistically significant employment declines following the arrival of AI chatbots, with the most-exposed 10% of majors down about 5 percentage points and the seventh and eighth deciles down about 2 points. The authors note flat pre-2020 trends and precise timing argue against alternative explanations such as uneven post-pandemic recovery.
Cody Orr, Lee Tucker and Lawrence WarrenU.S. Census Bureau
Date Posted:
A Census Bureau analysis of 29% of recent US college grads ranked majors by AI exposure using pre-ChatGPT occupational patterns. After ChatGPT’s November 2022 release, employment fell across the top 4 exposure deciles, with the top decile down 5pp compared to the bottom 6 deciles.
Are graduates in AI-exposed majors seeing employment declines?
Core argument: Employment for graduates in the most AI-exposed decile of college majors declined by an average of five percentage points relative to 2022 after ChatGPT’s release, with statistically significant declines of about two percentage points for deciles seven and eight.
In the quarters prior to 2020, the event study coefficients are usually statistically indistinguishable from zero across all deciles. Post-ChatGPT, there is a small to moderate decline in employment relative to the control group across the top four deciles with a particularly large effect for the tenth decile. For deciles seven, eight, and especially ten, these are statistically significant declines: an average of two percentage points for deciles seven and eight and five percentage points for decile ten relative to 2022. Decile nine also shows evidence of statistically significant declines, but this might reflect reversion to long-run trends rather than a negative effect of AI. Employment effects are clearly most acute among the most-exposed decile of majors. The relatively fat pre-trends and precise timing help to rule out some alternative explanations, such as heterogeneity in the post-pandemic recovery or the effects of a tight labor market in the reference period.
Takeaways by Macro Roundup® AI
Employment for graduates in the most AI-exposed decile of college majors declined by an average of five percentage points relative to 2022 after ChatGPT’s release, with statistically significant declines of about two percentage points for deciles seven and eight.
Event study coefficients were usually statistically indistinguishable from zero across all exposure deciles in the quarters before 2020, with small to moderate employment declines concentrated in the top four deciles afterward.
Flat pre-trends and the precise timing of the employment declines help rule out some alternative explanations, such as heterogeneity in the post-pandemic recovery or a tight labor market in the reference period, while decile nine’s decline may reflect reversion to long-run trends.
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.
Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence— Employment for workers aged 22–25 in the occupations most exposed to artificial intelligence has declined roughly 18pp relative to less-exposed peers, with no comparable decline in other age groups. The authors find this operates mainly through reduced hiring, concentrated where AI substitutes for human tasks rather than complements them.
The College Wage Premium in the Generative AI Era— 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.
AI Summary.Varsity sports participation varies with academic achievement: after adjusting for school size, 35% of students in the top quintile of eighth-grade math scores played varsity sports versus 12% in the bottom quintile. Students at the smallest schools were 2.8x as likely to participate as those at the largest.
Nat Malkus and Sam HollonAmerican Enterprise Institute
Date Posted:
A large sample of Indiana high school students shows a near-linear relationship between 8th grade math test scores and participation in varsity athletics. Academic success and participation in extracurriculars appear to be complementary.
How do academic achievement and school size shape varsity sports participation?
Core argument: After adjusting for school size, 35% of students in the top eighth-grade math achievement quintile played varsity sports in 2025, compared with 12% in the bottom quintile, with a largely linear relationship across quintiles.
We find large gaps in varsity sports participation by academic achievement, which we measure using students’ eighth-grade math test scores. Students who attended the smallest schools (first quintile) were 2.8 times as likely to participate in varsity sports as those who attended the largest schools (fifth quintile). We find that after adjusting for school size 35% of students in the top achievement quintile played varsity sports in 2025, compared with just 12% of students in the bottom quintile. In general, the relationship between achievement quintiles and varsity sports participation was remarkably linear. [The exception was football], the largest sport by participation [17% of all varsity athletes and overwhelmingly male]. Here the top quintile [in 8th grade math] had 20% lower odds of playing than students in the middle achievement quintile.
Takeaways by Macro Roundup® AI
After adjusting for school size, 35% of students in the top eighth-grade math achievement quintile played varsity sports in 2025, compared with 12% in the bottom quintile, with a largely linear relationship across quintiles.
Students at the smallest schools (first quintile by size) were 2.8 times as likely to participate in varsity sports as students at the largest schools (fifth quintile).
Football, the largest sport at 17% of all varsity athletes and overwhelmingly male, is the exception: students in the top eighth-grade math quintile had 20% lower odds of playing than those in the middle quintile.
Related Articles:
An Extra Point for Attendance: The Impact of High School Varsity Athletics on Absenteeism— High school varsity athletes are absent about 1.37 percentage points less often than non-athletes over the school year, nearly 20% lower, after controlling for factors linked to both sports participation and attendance. Athletes' absence rates fall further while their sport is in season, which the authors argue suggests participation itself improves attendance.
The Benefits of Scholastic Athletics— Using longitudinal data with extensive controls for family background, ability, and personality, the authors estimate that college varsity athletes from disadvantaged backgrounds—including women and men from low-income families and Black men—earn more after graduation than comparable non-athletes. The authors conclude that participation in college sports promotes social mobility, with no evidence of adverse outcomes.
No Revenge for Nerds? Evaluating the Careers of Ivy League Athletes— Ivy League athletes are far more likely than non-athlete classmates to enter business and finance jobs, and the authors estimate they attain higher terminal wages, greater cumulative career earnings, and more senior positions, controlling for school, graduation year, major, and first job. The authors argue this suggests non-academic human capital developed through athletics is valued in the labor market.
AI Summary.The unemployment rate for workers ages 22 to 34 without a college degree has rarely been lower in the past two decades, and relative to their own historical ranges the job market looks much better for blue-collar and manual-service workers. Degree-holders still face lower unemployment rates overall.
Theo Francis and Ray SmithWall Street Journal
Date Posted:
Is Database:
Database
In 2026, the 12-month moving-average unemployment for college-educated 22–34-year-olds is above its post-2003 mean, while the rate for non-college peers is historically low. Prime-age college grads still have lower unemployment than those with no degree.
Is the job market unusually strong for workers without college degrees?
Core argument: The unemployment rate for workers ages 22 to 34 who never graduated from college has rarely been lower in the past two decades, based on data through July.
The unemployment rate for workers ages 22 to 34 who never graduated from college has rarely been lower in the past two decades. To gauge how the job market has shifted for each cohort, [Gad Levanon, Burning Glass’s chief economist] compared current unemployment rates for the different groups with their own range of unemployment rates since 2003. The analysis included data through July. By that measure, the job market looks much better for blue-collar workers, including those in construction and on manufacturing lines, and manual-service workers. It is [however] still easier to find a job with a college degree. The unemployment rate for degree-holders in their prime working years—ages 25 to 54—averaged 2.7% for the 12 months ending in July - well below the 3.6% rate for workers with just some college education, and 4.7% for people with a high-school diploma only.
Takeaways by Macro Roundup® AI
The unemployment rate for workers ages 22 to 34 who never graduated from college has rarely been lower in the past two decades, based on data through July.
Burning Glass chief economist Gad Levanon compared each cohort's current unemployment rate against its own range since 2003, and by that measure the job market looks much better for blue-collar workers in construction and manufacturing and for manual-service workers.
College graduates ages 25 to 54 averaged a 2.7% unemployment rate for the 12 months ending in July, below the 3.6% rate for workers with some college education and 4.7% for those with only a high-school diploma.
Related Articles:
College Grads Struggle to Find Jobs. Non-Grads Are Giving Up— The narrowing gap in unemployment rates between young college graduates and non-graduates appears driven mainly by non-graduates dropping out of the labor force rather than by diplomas losing value, since people not seeking work aren't counted as unemployed. The share of young non-graduates employed is 1.7 percentage points below pre-pandemic levels.
To Fix Education, Fix The Economy First— 30% of US adults have literacy skills typical of a 10-year-old, yet US workers scoring at that low level earn an average of nearly $30 an hour and two-thirds are employed, compared with about $20 and fewer than half employed in Britain. The author argues this reflects stronger economic growth rather than vocational education policy.
The College Wage Premium in the Generative AI Era— 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.
AI Summary.International student assessment results show mathematics performance held near earlier levels for about 15 years before falling sharply, with mean scores down 9 points in math and about 14 in reading in the latest cycle. Reported student motivation and school engagement declined, and hasty incorrect answering rose to 11%.
OECD StaffOrganisation for Economic Co-operation and Development
Date Posted:
Is Database:
Database
Mean OECD math and reading scores fell roughly 25 PISA points—about a year of schooling—between 2018 and 2025, alongside reduced motivation and engagement and increased carelessness, exemplified by a near-doubling of “hasty readers.”
Are international student test scores falling alongside declining motivation and engagement?
Core argument: PISA 2025 mean performance fell by 9 score points in mathematics and about 14 points in reading between 2022 and 2025, after mathematics scores had remained close to their 2003 level through 2018.
In math, performance remained close to the 2003 level up to 2018, then dropped sharply btw 2018 and 2025. Mean performance dropped by 9 score points in mathematics and about 14 score points in reading btw 2022 and 2025. Over the 2018-2025 period, the largest drops [in reading] were among the most advantaged 25% of socio-economic status, and the socio-economic gap with the most disadvantaged students reduced somewhat, by 11 points, on average across 35 OECD countries. In mathematics, the share of disadvantaged students scoring below proficiency Level 2 increased in previous cycles and remained high in 2025; among advantaged students, the share scoring below proficiency Level 2 increased, reaching 18% in 2025, about 4pp. Students’ reports of their motivation for learning and engagement with school also declined between 2022 and 2025. The proportion of “hasty readers” – those who gave fast and incorrect responses – increased by about five percentage points, on average, from almost 7% in 2018 to 11% in 2025. Analyses for math, also show an increase in the proportion of hasty responses between 2022 and 2025.
Takeaways by Macro Roundup® AI
PISA 2025 mean performance fell by 9 score points in mathematics and about 14 points in reading between 2022 and 2025, after mathematics scores had remained close to their 2003 level through 2018.
Over 2018–2025, the largest reading declines came among the most advantaged 25% by socio-economic status, and the gap with the most disadvantaged students narrowed by 11 points on average across 35 OECD countries.
The share of "hasty readers," who gave fast and incorrect responses, rose about five percentage points, from almost 7% in 2018 to 11% in 2025, with hasty responses in mathematics also increasing between 2022 and 2025.
Related Articles:
Do Adults Have the Skills They Need to Thrive in a Changing World?— An international assessment of adult skills describes numeracy proficiency levels ranging from counting objects and adding small whole numbers at the lowest level, to extracting information from tables and graphs and performing basic arithmetic with money and percentages at Level 1, up to analysing complex mathematical information and critically evaluating whether data support a claim at Level 5.
Are China’s Students Really Number One? A Statistical Riddle— China's reported 65-point jump in mean international student test scores between testing rounds coincided with a change in which provinces represented the country, swapping the most populous province for a richer, much smaller one. The author argues such swings, including a fall in low performers from 43% to 19% in reading, are unmatched elsewhere.
Have Humans Passed Peak Brain Power?— The share of Americans reporting they read a book in the past year has fallen below half, alongside declining performance in numeracy and problem-solving in most countries. The share of adults unable to use mathematical reasoning to evaluate the validity of statements averages 25% in high-income countries and 35% in the US.
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.
