Edward Conard

Top Ten New York Times Bestselling Author

  • “…a must-read for serious students of economic policy.” - Glenn Hubbard, Dean, Columbia Business School, and former Chairman of the Council of Economic Advisers
  • “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
  • “…a must-read for serious students of economic policy.” - Glenn Hubbard, Dean, Columbia Business School, and former Chairman of the Council of Economic Advisers
  • “Unintended Consequences represents the most cogent and persuasive analysis of the Financial Crisis to date.” - Andrei Shleifer, 1999 John Bates Clark Medal Winner
  • “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 comprehensive explanation of the modern economy.” - Julian Robertson, Founder, Tiger Management
  • “A full-throated defense of economic dynamism.” - The Wall Street Journal
  • “…serious thinking for serious thinkers. …a thought-provoking blueprint for growing middle- and working-class incomes.” - Mitt Romney, former Governor of Massachusetts
  • “…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 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
  • “…challenges misconceptions that distort our economic debates.” - Arthur Brooks, President of the American Enterprise Institute
Upside of Inequality Oxford Unintended Consequences

Macro Roundup


Tuesday, January 13, 2026

AI Summary. The Congressional Budget Office projects the US population rises from 349m to 364m between 2026 and 2056, with annual growth slowing from 0.3% over the next decade to 0.1% thereafter before reaching zero and then shrinking. Declining fertility is projected to push deaths above births starting in 2030, increasingly offsetting net immigration.

Daniel Crown and Katherine Keisler-Starkey Congressional Budget Office
Date Posted:

CBO forecasts US population growth over the next decade will be 0.3% annually. Starting in 2030, net immigration will be the sole driver of population growth; after that date, births will be overtaken by deaths. CBO estimates net immigration of 410,000 in 2025.

Will US population growth slow to zero within three decades?

Core argument: CBO projects the U.S. population rises from 349 million in 2026 to 364 million in 2056, with growth slowing from an average 0.3% a year over the next decade to 0.1% a year thereafter.

In CBO’s projections, the population increases from 349 million people in 2026 to 364 million in 2056. Population growth slows during that period—from an average rate of 0.3% a year over the next decade to 0.1% a year thereafter. Through 2029, the population grows because more people are projected to migrate to the United States than to leave it (positive net immigration) and because the annual number of births is projected to exceed the annual number of deaths. Starting in 2030, declining fertility rates are projected to cause births to fall short of deaths. Over time, the negative net contribution of births and deaths increasingly offsets the positive contribution of net immigration, until population growth slows to zero in 2056. The population is projected to shrink thereafter.

Takeaways by Macro Roundup® AI

  1. CBO projects the U.S. population rises from 349 million in 2026 to 364 million in 2056, with growth slowing from an average 0.3% a year over the next decade to 0.1% a year thereafter.
  2. CBO projects that starting in 2030, declining fertility rates cause annual births to fall short of annual deaths, after a period through 2029 in which births exceeded deaths and net immigration was positive.
  3. CBO projects U.S. population growth slows to zero in 2056 as the negative net contribution of births and deaths increasingly offsets the positive contribution of net immigration, with the population shrinking thereafter.

Related Articles:

  • The Case for Gradual Population Decline — South Korea's fertility rate fell below 1.7 in 1985 and its share of population aged 20-40 has fallen 20% since 2000, yet it ranked first in Bloomberg's Innovation Index and twelfth in a global soft power ranking. The author argues low fertility does not inevitably cause technological and cultural stagnation.
  • Foreign-Born Number and Share of U.S. Population at All-Time Highs in January 2025 — The U.S. foreign-born population reached 53.3m, or 15.8% of the population, the highest level in American history, exceeding the shares recorded in 1890 (14.8%) and 1910 (14.7%) during the "Great Wave" of immigration. A four-year increase of 8.3m is larger than that of any four-year period in American history and exceeds the 7.9m increase over the prior 12 years.
  • America’s Immigration Mess: An Illustrated Guide — Current Population Survey estimates indicate America's foreign-born population surpassed 50m for the first time and exceeded 15% of the national population, above the prior peak of 14.8% recorded in 1890. Mass immigration during that earlier wave was associated in contemporary opinion with widening income and wealth gaps and depressed wages for less-skilled native-born men, and sweeping national restrictions followed.

AI Summary. Investment in information processing equipment, software, research and development, and data centers contributed 0.97 percentage points to real GDP growth over the first three quarters of 2025, or 39% of total growth, compared with 0.81 percentage points and 28% during the dot-com investment boom of 2000.

Hannah Rubinton and Bontu Ankit Patro Federal Reserve Bank Of St. Louis
Date Posted:
Is Database:
Database

Investment spending on information processing equipment, software, data centers and R&D drove 39% of American GDP growth in the first 3Q of 2025. Those same categories were responsible for 28% of GDP growth in 2000, during the dot-com boom.

How much of real GDP growth does AI-related investment contribute?

Core argument: AI-related investment categories — information processing equipment, software, R&D and data centers — contributed 0.97 percentage points to real GDP growth over the first three quarters of 2025, above the 0.81 percentage points contributed in 2000.

The table provides the contributions of identical categories [Information Processing Equipment, Software, R&D and Data Centers] of investment to real GDP growth for the years 2000 and 2025. The tables are calculated by taking the average of data available over the first three quarters of 2025 and over the four quarters of 2000, during the height of the dot-com era’s investment boom. IPE’s contribution is similar across the two periods: 0.58 percentage points (20% of GDP growth) in 2000 versus 0.42 percentage points (17%) in the first nine months of 2025. However, the contribution of software is significantly higher in 2025 at 0.35 percentage points (14%) compared with 0.11 percentage points (3.7%) in 2000. Together, the AI categories contributed 0.97 percentage points to real GDP growth in the first three quarters of 2025 (0.90 percentage points, excluding data centers, for which data were not available in 2000), higher than the 0.81 percentage points in 2000. Through the third quarter of 2025, these categories made up 39% (36% excluding data centers) of total GDP growth versus 28% in 2000.

Takeaways by Macro Roundup® AI

  1. AI-related investment categories — information processing equipment, software, R&D and data centers — contributed 0.97 percentage points to real GDP growth over the first three quarters of 2025, above the 0.81 percentage points contributed in 2000.
  2. Software's contribution to real GDP growth reached 0.35 percentage points (14% of growth) in the first nine months of 2025 versus 0.11 percentage points (3.7%) in 2000, while information processing equipment contributed 0.42 percentage points (17%) versus 0.58 points (20%).
  3. Information processing equipment, software, R&D and data centers accounted for 39% of total GDP growth through the third quarter of 2025 — 36% excluding data centers, for which 2000 data were unavailable — compared with 28% in 2000.

Related Articles:

  • Financing The AI Boom: From Cash Flows To Debt — Spending on semiconductor plants and data centers, negligible before recent years, has averaged 0.4 percentage points of US GDP growth, and total IT-related investment has reached 5% of GDP, exceeding its dot-com-era peak. Unlike that earlier boom, which was driven by firms using IT products, the current one is driven by IT producers.
  • OECD Economic Outlook, Volume 2025 Issue 2: Resilient Growth but with Increasing Fragilities — U.S. real GDP grew at a 1.1% annualized rate over the first half of 2025, but excluding booming AI-related investment output contracted 0.1%, with non-AI investment weakening. The OECD identifies a correction in equity markets lifted by expectations of high AI returns as a key downside risk to its projections.
  • Honey, AI Capex is Eating the Economy — Capital spending on AI data centers is around 20% of railroads' peak spending as a share of GDP and is rising quickly, having already surpassed the dot-com era peak in telecom investment. Unlike long-lived rail infrastructure, data centers are short-lived assets requiring frequent hardware replacement, and the capital is being reallocated from other sectors.

AI Summary. In an experiment with over 500 professionals across management, data analysis, and consulting using 13 different large language models, each year of model progress cut task time by 8%, and a 10x increase in training compute is associated with a 6.3% time reduction. The authors project continued scaling could raise US productivity by about 20% over a decade.

Ali Merali Yale University
Date Posted:
Is Database:
Database

An experiment with over 500 professionals found that each year of model progress cut task time ~8%, with 56% of gains driven by increased compute and 44% by algorithmic progress, suggesting that US productivity could rise 20% over the next decade.

How much does AI model scaling raise professional productivity?

Core argument: In experiments with over 500 professionals across management, data analysis, and consulting using one of 13 LLMs, each year of model progress reduced task time by 8%, with 56% of gains from added compute and 44% from algorithmic progress.

This paper derives “Scaling Laws for Economic Impacts”—empirical relationships between the training compute of Large Language Models (LLMs) and professional productivity— [from] over 500 professionals across three high-skill domains: management, data analysis, and consulting [using] one of thirteen different LLMs. Each year of model progress reduced task time by 8%, with 56% of gains driven by increased compute and 44% by algorithmic progress; a tenfold (10x) increase in model training compute is associated with a 6.3% reduction in time taken. Access to any AI model increased base Earnings Per Minute (EPM) by 81.3% (p = 0.001) [and] resulted in a 146% increase in Total Earnings Per Minute (TEPM), with AI assistance delivering total earnings gains of $1.58 per minute on non-agentic, analytical tasks [versus] $0.34 per minute for “agentic” tasks requiring multi-step interactions with external tools. While the quality of autonomous model output scales linearly with compute, [for human-assisted work] the scaling law for quality completely disappears. Multiplying the share of tasks that can be automated (19.9%), multiplied by the average productivity effect derived from our scaling projections (175.1%), multiplied by the labor share of costs (57%) yields Total GDP Gain ≈ 19.9%, [meaning] continued model scaling could boost U.S. productivity by ~20% over the next decade.

Takeaways by Macro Roundup® AI

  1. In experiments with over 500 professionals across management, data analysis, and consulting using one of 13 LLMs, each year of model progress reduced task time by 8%, with 56% of gains from added compute and 44% from algorithmic progress.
  2. Access to any AI model raised base Earnings Per Minute by 81.3% (p = 0.001) and Total Earnings Per Minute by 146%, with gains of $1.58 per minute on non-agentic analytical tasks versus $0.34 per minute on agentic tasks requiring multi-step tool use.
  3. The authors multiply the automatable task share (19.9%) by the projected average productivity effect (175.1%) and the 57% labor share of costs to project that continued model scaling could boost U.S. productivity by roughly 20% over the next decade.

Related Articles:

  • US Productivity Accelerates to Fastest Pace in Two Years — US nonfarm productivity, or employee output per hour, rose at a 4.9% annualized rate following an upwardly revised 4.1% gain the prior quarter, according to the Bureau of Labor Statistics. Unit labor costs fell 1.9% after a prior-quarter decline, the first back-to-back drop since 2019.
  • Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence — Using payroll records covering millions of US workers, the authors find employment for workers aged 22–25 in occupations most exposed to artificial intelligence, such as software developers and customer service representatives, declined 6%, while employment for older workers in those occupations rose 6–9%. Employment in less-exposed occupations grew comparably across age groups.
  • Weekend Thinking: A Cul-de-Sac With a View — Leading AI developers are shipping smaller, more frequent model updates, creating an impression of momentum through release frequency rather than the magnitude of change. The author argues the bet that scaling data, compute, and model size would produce artificial general intelligence has not paid off, pushing labs back toward harder problems like continual learning and new architectures.

AI Summary. California now draws more than 60% of its revenue from personal income taxes, compared with about 11% in the early 1950s when sales and use taxes supplied nearly 60%. The state's top 1% of taxpayers — roughly 170,000 filers — paid about 40% of personal income tax revenue, payments that depend heavily on unpredictable capital gains.

Michael Moritz Financial Times
Date Posted:
Is Database:
Database

In 1950, sales taxes accounted for ~60% of California’s revenue, and personal income taxes accounted for 11%. In recent years, the personal income tax share has risen to ~60%, with the top 1% of California taxpayers responsible for ~40% of the revenue.

How dependent is state revenue on a small group of taxpayers?

Core argument: In 2022, California's top 1% of taxpayers — about 170,000 payers — paid roughly 40% of the state's personal income tax revenues, payments that are highly contingent on unpredictable capital gains.

In 1957, Fairchild Semiconductor, the company that spawned Silicon Valley, was formed by eight men — only one of whom was born in California. At the start of that decade the state’s population was about 10.5mn and only about 11% of its revenue of $1.1bn came from personal income taxes. The sales and use tax (levied on products not services) accounted for almost 60%. Today, California’s population is about 39.5mn while its budget is about $320bn (not including the $174bn the state receives from Washington for social welfare programmes). Since 1990, the budget has grown at an average rate of 6% a year. Compared to the 1950s, the source of the state’s revenue has entirely flipped. In recent years personal income taxes have contributed more than 60% of the total. This revenue rests on a foundation of jelly. In 2022, the top 1% of California taxpayers — just 170,000 payers — coughed up about 40% of all the state’s personal income tax revenues. Worse still, these payments are highly contingent on unpredictable capital gains.

Takeaways by Macro Roundup® AI

  1. In 2022, California's top 1% of taxpayers — about 170,000 payers — paid roughly 40% of the state's personal income tax revenues, payments that are highly contingent on unpredictable capital gains.
  2. California's revenue mix has flipped since the 1950s, when personal income taxes supplied about 11% of $1.1bn in revenue and sales and use taxes nearly 60%; in recent years personal income taxes contributed more than 60%.
  3. California's budget has grown at an average 6% a year since 1990 to about $320bn, excluding $174bn received from Washington for social welfare programmes, with population up from about 10.5mn in the early 1950s to about 39.5mn.

Related Articles:

  • Ed Conard Debates Furman On “The Expected Value of Risk Taking” — A debate on U.S. fiscal policy argues that spending increases, rather than the tax increases needed to fund them, slow growth, and that higher taxes would reduce the expected value of America's innovative risk-taking. The argument holds that cutting debt below 100% of GDP is critical.
  • California Keeps Losing Tech Jobs — California's share of US tech jobs has fallen from nearly 19% before the pandemic to nearly 16%, including a further 0.4% decline over the past year. The remaining jobs skew high-paying: California's share of information-sector compensation is slightly above its pre-pandemic peak, with nearly a third of all tech wages and benefits going to state residents.
  • Behavioral Responses to State Income Taxation of High Earners: Evidence from California — A California ballot measure raising top marginal income tax rates by up to 3 percentage points prompted an additional 0.8% of the affected tax base to leave the state, with an estimated income elasticity of 2.5–3.2 for high earners who stayed. The authors estimate these responses eroded 45.2% of windfall revenues within one year and 60.9% within two.

AI Summary. Adjusted for poverty and student demographics, Mississippi ranks first nationally in fourth-grade reading and math and at or near the top in eighth grade, according to the Urban Institute. The state changed reading instruction and adopted accountability policies that credit schools for student progress, with double credit for the bottom 25%.

Sarah Mervosh New York Times
Date Posted:
Is Database:
Database

Controlling for poverty and demographics, Mississippi has the best outcomes for fourth-grade reading and math, and is near the top in eighth-grade scores. In absolute terms, “eighth graders still rank 41st on national tests for reading, and 35th for math.”

How did one state's reading instruction and accountability reforms raise test scores?

Core argument: Adjusted for poverty and other student demographics, Mississippi ranks No. 1 nationally in fourth grade reading and math and at or near the top in eighth grade, according to the Urban Institute.

Adjusted for poverty and other student demographics, Mississippi is No. 1 for fourth grade reading and math, and at or near the top in eighth grade, according to the Urban Institute. Mississippi’s gains have been confined mostly to the early years, where the state has focused its efforts. The system revolves around standardized testing, and the gains have not always persisted. By raw scores, Mississippi eighth graders still rank 41st on national tests for reading, and 35th for math. The state pushed through a vast list of other changes from the top down, including changing the way reading is taught, in an approach known as the science of reading, but also embracing contentious school accountability policies other states have backed away from. Unlike No Child Left Behind, the Bush-era federal law that set a goal of having every child in America be proficient in reading and math, Mississippi also measures how much students progress toward proficiency. Schools get credit if students show improvement — and double credit for the improvement of students in the bottom 25%.

Takeaways by Macro Roundup® AI

  1. Adjusted for poverty and other student demographics, Mississippi ranks No. 1 nationally in fourth grade reading and math and at or near the top in eighth grade, according to the Urban Institute.
  2. Mississippi's gains are concentrated in the early grades where the state focused its efforts, and by raw scores its eighth graders still rank 41st in reading and 35th in math on national tests.
  3. Mississippi's accountability system, unlike No Child Left Behind's proficiency goal, credits schools for student progress toward proficiency and gives double credit for improvement among students in the bottom 25%.

Related Articles:

  • Many Children Left Behind: The 2024 National Assessment of Educational Progress Results Indicate a Five-Alarm Fire — The share of US fourth-grade students meeting the National Assessment of Educational Progress reading proficiency standard fell from 35% in 2019 to 33% in 2022 and 30% in 2024. More fourth-graders now score below the basic level than score proficient.
  • Twelfth-Grade Math and Reading Scores in U.S. Hit New Low — U.S. twelfth-graders' average math and reading scores on the National Assessment of Educational Progress fell to their lowest levels on record, with math the worst since that test began in 2005 and reading below any point since 1992. The share rated proficient slid 2 percentage points to 35% in reading and 22% in math.
  • 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.

AI Summary. Yields on long-term Japanese government debt have risen to their highest since the 1990s while the yen remains cheap relative to its history. One benign reading is that Japan has exited post-bubble stagnation, with modest inflation and faster wage and income growth aligning with higher interest rates.

Matt Klein The Overshoot
Date Posted:

Klein notes the “relentless” rise in Japan’s long-term yields alongside continued yen weakness. Pessimists cite unsustainable debt requiring default or debasement, but a “more benign explanation” is that Japan has finally exited its deflationary slump.

Do rising long-term yields signal Japan's exit from post-bubble stagnation?

Core argument: Yields on long-term Japanese government debt have risen over the past few years, especially at longer maturities, and now stand higher than at any point since the 1990s.

Japanese bond yields have been rising relentlessly over the past few years, especially at the longer maturities. Yields on long-term Japanese government debts are now higher than at any point since the 1990s. Yet the yen has not only failed to appreciate against the currencies of its trading partners - it still looks cheap relative to its history. Those who believe that the debt is unsustainable believe that the government will eventually be forced to choose between outright default and “debasement” that would cheapen nominal yen claims relative to assets in other currencies. As of now, however, there is a more benign explanation: Japan may have finally exited its post-bubble stagnation. The return of modest inflation alongside persistently faster wage and income growth should align with higher interest rates than those that prevailed when the economy experienced essentially no growth in yen terms between 1997 and 2019. Moreover, if the recent growth in nominal incomes is sustained, Japanese government borrowing costs would still be lower, relative to expected revenue increases, than in much of the past few decades. (That helps explain why the yen has not appreciated recently despite the apparent convergence of nominal yields with the U.S. and other G10 economies.) Far from indicating trouble, Japanese bond prices are implying that Japan has converged, in at least one important way, with the rest of the rich world.

Takeaways by Macro Roundup® AI

  1. Yields on long-term Japanese government debt have risen over the past few years, especially at longer maturities, and now stand higher than at any point since the 1990s.
  2. The yen has failed to appreciate against trading partners' currencies and still looks cheap relative to its history, despite the apparent convergence of Japanese nominal yields with those of the U.S. and other G10 economies.
  3. Japan's return of modest inflation alongside persistently faster wage and income growth should align with higher interest rates than those prevailing when the economy saw essentially no growth in yen terms between 1997 and 2019.

Related Articles:

Wednesday, January 14, 2026

AI Summary. 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.

Zanna Iscenko and Fabien Curto Millet Agglomerations
Date Posted:
Is Database:
Database
Is Important:
Important

The downtick in hiring in AI-exposed occupations started 6 months prior to the release of ChatGPT, and is “perfectly” aligned with the start of Fed rate hikes in Spring 2022. Sectors in the most AI-exposed quintile tend to be interest-sensitive.

Is the hiring slowdown in AI-exposed jobs driven by interest rates?

Core argument: Roughly 38% of workers in the most AI-exposed quintile of occupations work in Information, Finance and Insurance, or Professional and Technical Services, versus under 2% in the least-exposed quintile, according to an analysis of 2023 Census data.

The sharp downturn in hiring for AI-exposed occupations aligns perfectly with the Federal Reserve’s monetary policy timeline. The key to understanding the misattribution of this downturn to AI is that “AI exposure” and “interest rate sensitivity” are deeply correlated variables. The occupations rated as most exposed to generative AI are not randomly distributed across the economy. An analysis based on 2023 Census data reveals that occupations in the top quintile of AI exposure are overwhelmingly concentrated in sectors like “Information”, “Finance and Insurance”, and “Professional and Technical Services”. Approximately 38% of workers in the most AI-exposed quintile are employed in these sectors, compared to less than 2% in the least-exposed quintile. These are precisely the sectors most sensitive to capital costs and broad economic uncertainty.

Takeaways by Macro Roundup® AI

  1. Roughly 38% of workers in the most AI-exposed quintile of occupations work in Information, Finance and Insurance, or Professional and Technical Services, versus under 2% in the least-exposed quintile, according to an analysis of 2023 Census data.
  2. The author argues that the hiring downturn in AI-exposed occupations has been misattributed to AI because "AI exposure" and interest rate sensitivity are deeply correlated, with the most exposed sectors also the most sensitive to capital costs and economic uncertainty.
  3. The author contends that the sharp downturn in hiring for AI-exposed occupations aligns with the Federal Reserve's monetary policy timeline.

Related Articles:

  • Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence — Using payroll records covering millions of US workers, the authors find employment for workers aged 22–25 in occupations most exposed to artificial intelligence, such as software developers and customer service representatives, declined 6%, while employment for older workers in those occupations rose 6–9%. Employment in less-exposed occupations grew comparably across age groups.
  • AI and Jobs: The Final Word (Until the Next One) — Unemployment has risen by 0.30 percentage points for the fifth of workers most exposed to artificial intelligence, compared with 0.94 percentage points for the least exposed fifth, leaving no visible AI effect on jobs. The share of workers leaving the labor force within a year is lowest among the most AI-exposed workers and roughly flat.
  • Our Favorite Charts of 2025 — Wage growth is slowing more for lower-paid workers than for higher earners, limiting those households' ability to keep spending, while hiring is concentrated in select industries such as healthcare and among larger firms as small private firms shed workers. Layoffs remain low, but the number of workers unemployed for 27 weeks or more has only rarely been higher.

AI Summary. The authors estimate that job growth consistent with full employment was 20,000–50,000 a month in the second half of 2025, while reported nonfarm payroll growth averaged about 14,500 and unemployment rose 0.3 percentage points. They project immigration policy changes will keep potential employment growth low, reducing GDP growth by roughly 0.2 percentage points in 2025 and 0.1 in 2026.

Wendy Edelberg, Stan Veuger, and Tara Watson Brookings Institution
Date Posted:
Is Database:
Database

Brookings estimates that in 2025, net immigration to the US was btw -10,000 and -295,000, and will be btw 185,000 and -925,000 in 2026. They estimate that sustainable monthly job growth was btw 20k and 50k last year and will be btw -20k and 50k in 2026.

How do lower immigration flows affect job growth and GDP?

Core argument: The authors estimate job growth consistent with full employment at 20,000–50,000 per month in the second half of 2025, modestly above the roughly 14,500 average monthly non-farm payroll gain reported by the Bureau of Labor Statistics, with the October–November shutdown affecting data collection.

If we look solely at the second half of 2025, we estimate the rate of job growth consistent with full employment to have been 20,000 to 50,000 in both of our scenarios. Monthly non-farm payrolls as reported by the Bureau of Labor Statistics were just modestly below these estimates. With the caveat that the October–November government shutdown affected data collection, non-farm payroll growth averaged around 14,500 in the second half of the year. Over that same period when employment growth was modestly below our estimate of potential employment growth, the unemployment rate has risen by 0.3 percentage points. Looking toward the current calendar year, 2026, we expect the decline in potential employment growth caused by changes to immigration policy to persist. In our low-immigration scenario, monthly growth for the year will average between –20,000 and +20,000 jobs. In our high-immigration scenario, we expect payrolls consistent with full employment to average between 10,000 and 50,000 jobs. We can translate our estimates of labor force growth into effects on consumer spending and GDP growth. The direct effect on GDP growth owing to less output created by immigrants is roughly 0.2 percentage points in 2025 and 0.1 percentage point in 2026 (top row of Table 2).

Takeaways by Macro Roundup® AI

  1. The authors estimate job growth consistent with full employment at 20,000–50,000 per month in the second half of 2025, modestly above the roughly 14,500 average monthly non-farm payroll gain reported by the Bureau of Labor Statistics, with the October–November shutdown affecting data collection.
  2. The unemployment rate rose 0.3 percentage points in the second half of 2025, a period when employment growth ran modestly below the authors' estimate of potential employment growth.
  3. The authors project payroll growth consistent with full employment in 2026 of –20,000 to +20,000 monthly under low immigration and 10,000–50,000 under high immigration, with direct effects on GDP growth from less output created by immigrants of roughly 0.2 percentage points in 2025 and 0.1 in 2026.

Related Articles:

  • The Demographic Outlook: 2026 to 2056 — The Congressional Budget Office projects the US population rises from 349m to 364m between 2026 and 2056, with annual growth slowing from 0.3% over the next decade to 0.1% thereafter before reaching zero and then shrinking. Declining fertility is projected to push deaths above births starting in 2030, increasingly offsetting net immigration.
  • America’s Immigration Mess: An Illustrated Guide — Current Population Survey estimates indicate America's foreign-born population surpassed 50m for the first time and exceeded 15% of the national population, above the prior peak of 14.8% recorded in 1890. Mass immigration during that earlier wave was associated in contemporary opinion with widening income and wealth gaps and depressed wages for less-skilled native-born men, and sweeping national restrictions followed.
  • The Case for Gradual Population Decline — South Korea's fertility rate fell below 1.7 in 1985 and its share of population aged 20-40 has fallen 20% since 2000, yet it ranked first in Bloomberg's Innovation Index and twelfth in a global soft power ranking. The author argues low fertility does not inevitably cause technological and cultural stagnation.

AI Summary. The share of foreign-born workers in US construction is now nearly twice that of other industries, after being equal in the 1970s and 1980s, and 9 of 10 construction workers are nonunion. Unionized construction workers earn $1.15 for every $1 the average American worker earns hourly, down from $1.71, while nonunion construction workers earn 86 cents.

Ronda Kaysen and Robert Gebeloff New York Times
Date Posted:
Is Database:
Database

Relative to the mean US worker, unionized construction workers earned a 71% premium in 1973, versus 15% in 2024. Nonunion construction workers earned a 7% premium in 1973, but only 86 cents on the dollar in 2024. 90% of construction workers are nonunion.

Have construction wages fallen as foreign-born and nonunion shares rose?

Core argument: Unionized construction workers earned $1.71 for every $1 the average American worker made hourly in 1973, a ratio that has fallen to $1.15.

In the 1970s and 1980s, the share of foreign-born workers in construction was the same as in other industries. But that quickly changed. Today that share is nearly twice as high as in other jobs. As of 2024, nine of 10 workers in the construction industry are nonunion. And with that decline in union strength came a decline in the economic standing for both union and nonunion workers. In 1973, the average union construction worker made $1.71 for every $1 the average American worker earned hourly. Today, unionized construction workers earn $1.15 for every $1 an American worker earns. And nonunion construction workers make just 86 cents for every dollar earned by the average American worker, often without health or retirement benefits.

Takeaways by Macro Roundup® AI

  1. Unionized construction workers earned $1.71 for every $1 the average American worker made hourly in 1973, a ratio that has fallen to $1.15.
  2. Nonunion construction workers earn 86 cents for every dollar earned by the average American worker, often without health or retirement benefits, and nine of 10 construction workers were nonunion as of 2024.
  3. The foreign-born share of construction workers matched that of other industries in the 1970s and 1980s but is now nearly twice as high as in other jobs.

Related Articles:

  • The Strange and Awful Path of Productivity in the U.S. Construction Sector — U.S. construction sector labor productivity and total factor productivity grew faster than the whole economy in the 1950s and 1960s, then fell, ending 2020 below their 1950 levels while economywide labor productivity and total factor productivity were 290% and 230% higher. Construction labor productivity fell about 1% a year from 1970 to 2020.
  • Stagnant Construction Productivity Is a Worldwide Problem — Construction labor productivity has been roughly flat or declining since the 1990s across most of Western Europe, Japan, and Korea, a pattern consistent with the flat US trend. Countries often held up as models — Sweden and Japan for prefabrication, China for infrastructure — show little or unremarkable construction productivity improvement.
  • Five Decades of Decline: U.S. Construction Sector Productivity — U.S. construction labor productivity in 2023 was essentially unchanged from 1948, with real output per worker in 2020 more than 30% below its 1970 level, even as aggregate U.S. labor productivity tripled between 1950 and 2020. Construction accounted for about 4.5% of U.S. GDP in 2024.

AI Summary. China reported a record annual goods trade surplus of $1.2tn, with monthly exports growing 6.6% year over year in dollar terms, more than double analysts' average forecast of 3.1%. Chinese customs officials argue that other countries' restrictions on high-tech exports to China are holding down China's imports.

Joe Leahy and Haohsiang Ko Financial Times
Date Posted:

China ran a ~$1.2T surplus in 2025 according to customs data, despite exports to the United States falling 20%.

Are high-tech export restrictions to blame for China's record trade surplus?

Core argument: China reported a record full-year trade surplus of $1.2tn for 2025, despite President Donald Trump's trade war, according to official customs data.

China blamed the US for growing global trade imbalances as the world’s second-biggest economy reported a record full-year trade surplus of $1.2tn for 2025 despite President Donald Trump’s trade war. China’s exports in goods grew 6.6% in dollar terms in December from a year earlier, official data released by the country’s customs administration on Wednesday showed. That was more than double the average forecast from a Bloomberg poll of analysts of 3.1% and greater than November’s growth rate of 5.9%. Wang Jun, vice-minister of the General Administration of Customs of China said “It should be pointed out that some countries politicise economic and trade issues, using various pretexts to restrict exports of high-tech products to China; otherwise, we would import more. There is vast room for import growth.”

Takeaways by Macro Roundup® AI

  1. China reported a record full-year trade surplus of $1.2tn for 2025, despite President Donald Trump's trade war, according to official customs data.
  2. China's goods exports grew 6.6% year on year in dollar terms in December, more than double the 3.1% average forecast in a Bloomberg poll of analysts and above November's 5.9%.
  3. Wang Jun, vice-minister of China's General Administration of Customs, said some countries politicise economic and trade issues and use various pretexts to restrict exports of high-tech products to China, and that otherwise China would import more.

Related Articles:

  • China’s Economy: Rightsizing 2025, Looking Ahead to 2026 — The authors argue China's official growth figures are inconsistent with ten consecutive quarters of deflation, estimating that real GDP is around 11% below official levels, with most of the adjustment in investment. That would imply a drop of roughly 9 percentage points in China's share of global investment, helping explain its export surge.
  • Markets Outlook 2026: Some Like It Hot — China's annual goods trade surplus surpassed $1tn for the first time, and analysts project its current account surplus as a share of global GDP will exceed levels last seen during the mid-2000s "China shock." The Chinese currency is estimated to be undervalued to a comparable degree, and the analysts expect gradual appreciation to continue.
  • Sticking with What Works: Raising GDP Forecast on China’s Manufacturing Push — Stronger Chinese growth driven by export-oriented manufacturing may not produce net positive spillovers: for every 1pp export-driven boost to China's GDP, other economies may see a 0.1–0.3pp drag, with high-tech producers such as Europe and Japan facing greater pressure. Cheaper Chinese goods continue to exert a disinflationary impulse.

AI Summary. 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.

Nicholas Eberstadt American Enterprise Institute
Date Posted:

The sharp rise in China’s PISA scores btw 2015 and 2018 coincided with “peculiar changes to the roster of provinces representing China,” further underrepresenting rural areas, while “China’s share of low-performing students plummeted” across fields.

Do China's international test score gains reflect changes in which provinces participate?

Core argument: China's PISA mean scores jumped 65 points between the 2015 and 2018 rounds, a gap roughly equal to that separating the Netherlands from Mexico in PISA 2018.

Between PISA’s 2015 and 2018 rounds of testing, “China” students reported an overall leap in mean scores of 65 points – approximately the same gap in scores as that separating the Netherlands from Mexico in PISA 2018. Note the peculiar changes to the roster of provinces representing China between 2015 and 2018. By swapping out Guangdong (the country’s most populous province) for Zhejiang (a richer and much smaller province), PISA and the Ministry of Education opted for a smaller, less representative set of Chinese provinces for 2018 than the already unrepresentative one it had been working with in 2015. A number of anomalies and curiosities from the reported results practically scream for closer attention. Consider a comparison between PISA’s sample of China and that of Massachusetts, America’s top-scoring state. In 2015, Massachusetts’ mean scores exceeded China’s in reading and science. By 2018, however, a redefined China sample vaulted ahead of Massachusetts. Where Massachusetts’s scores did not change dramatically over the short time span under consideration, China’s surged. And between 2015 and 2018 China’s share of low-performing students plummeted from 43% to 19% in reading, from 32% to 9% in math and from 38% to 10% in science. Nothing like these gyrations have been recorded elsewhere in PISA – ever.

Takeaways by Macro Roundup® AI

  1. China's PISA mean scores jumped 65 points between the 2015 and 2018 rounds, a gap roughly equal to that separating the Netherlands from Mexico in PISA 2018.
  2. Between the 2015 and 2018 PISA rounds, PISA and China's Ministry of Education swapped out Guangdong, the country's most populous province, for Zhejiang, a richer and much smaller province, yielding a smaller, less representative sample.
  3. China's share of low-performing students fell from 43% to 19% in reading, 32% to 9% in math and 38% to 10% in science between 2015 and 2018, while Massachusetts, previously ahead in reading and science, changed little.

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.
  • Exploration of “Knowledge Capital” in the PRC: “Mapping” China’s Knowledge and Skills Terrain — China's reportedly high average IQ is best treated as a stylized fact: few serious studies have measured it across the population, measured IQ levels shift over time, and testing has been concentrated in urban rather than rural areas. Research from Stanford's REAP group reports rural Chinese children average at least one standard deviation below the international norm.
  • The Power of Demography and the Fate of East Asia — East Asia's total population has peaked and is projected to decline for at least a generation, with the pace of decline likely accelerating each decade, while US population is projected to keep growing for decades. The author argues such sustained divergence can shift the global balance of power.