Category: Data Source

The economics of sleep

by  Tyler Cowen in

Full-time, prime-age male workers in the top income quartile sleep around half an hour less per day than those in the lowest quartile.

At the macro level, average sleep duration decreases as a country’s GDP increases.

Higher-income individuals allocate more time to other leisure activities, such as social outings and internet usage, substituting sleep.

Here is the paper by Cristián Jara, Francisca Pérez, and Rodrigo Wagner. Via the excellent Kevin Lewis.

Politically correct LLMs

by  Tyler Cowen in

Despite identical professional qualifications across genders, all LLMs consistently favored female-named candidates when selecting the most qualified candidate for the job. Female candidates were selected in 56.9% of cases, compared to 43.1% for male candidates (two-proportion z-test = 33.99, p < 10⁻252 ). The observed effect size was small to medium (Cohen’s h = 0.28; odds=1.32, 95% CI [1.29, 1.35]). In the figures below, asterisks (*) indicate statistically significant results (p < 0.05) from two-proportion z-tests conducted on each individual model, with significance levels adjusted for multiple comparisons using the Benjamin-Hochberg False Discovery Rate correction…

In a further experiment, it was noted that the inclusion of gender concordant preferred pronouns (e.g., he/him, she/her) next to candidates’ names increased the likelihood of the models selecting that candidate, both for males and females, although females were still preferred overall. Candidates with listed pronouns were chosen 53.0% of the time, compared to 47.0% for those without (proportion z-test = 14.75, p < 10⁻48; Cohen’s h = 0.12; odds=1.13, 95% CI [1.10, 1.15]). Out of 22 LLMs, 17 reached individually statistically significant preferences (FDR corrected) for selecting the candidates with preferred pronouns appended to their names.

Here is more by David Rozado.  So there is still some alignment work to do here?  Or does this reflect the alignment work already?

Partisan Corporate Speech

by  Tyler Cowen in

We construct a novel measure of partisan corporate speech using natural language processing techniques and use it to establish three stylized facts. First, the volume of partisan corporate speech has risen sharply between 2012 and 2022. Second, this increase has been disproportionately driven by companies adopting more Democratic-leaning language, a trend that is widespread across industries, geographies, and CEO political affiliations. Third, partisan corporate statements are followed by negative abnormal stock returns, with significant heterogeneity by shareholders’ degree of alignment with the statement. Finally, we propose a theoretical framework and provide suggestive empirical evidence that these trends are at least in part driven by a shift in investors’ nonpecuniary preferences with respect to partisan corporate speech.

That is from a recent paper by William Cassidy and Elisabeth Kempf.  Via the excellent Kevin Lewis.

Claims about police shootings

by  Tyler Cowen in

There is a new book by Tom S. Clark, Adam N. Glynn, and Michael Leo Owens, called Deadly Force: Police Shootings in Urban America.  Here are a few of the conclusions:

…we were more likely to obtain records [on police shootings] from cities with women mayors and more women on municipal legislatures.

And, most interesting to me:

…we also found that fewer police shootings occurred in cities with more police, all else equal.

And:

…Black and Hispanic officers are disproportionately the ones involved in police shootings.  That is particularly true when a Black civilian is the subject of the shooting.

I do not follow this area closely, but the book seems of interest.

Manufacturing Went South

by  Alex Tabarrok in

Excellent piece by Gary Winslett in the Washington Post. As I pointed out in my piece on Manufacturing and Trade, the US is a manufacturing powerhouse. So why did the rust belt rust? Because manufacturing went South.

The Rust Belt’s manufacturing decline isn’t primarily about jobs going to Mexico. It’s about jobs going to Alabama, South Carolina, Georgia and Tennessee…In 1970, the Rust Belt was responsible for nearly half of all manufacturing exports while the South produced less than a quarter. Today, the roles are reversed, it is the Rust Belt that hosts less than one-fourth of all manufactured exports and the South that exports twice what the Rust Belt does.

Why the move? Better policies:

Economic research suggests that labor conflict drove much of the decline of the Rust Belt. Right-to-work laws in the South, by contrast, created more operational flexibility and attracted capital. The average unionization rate in the Rust Belt is 13.3 percent; in the South, it’s 4.3 percent. Southern states’ political leaders are quite open about how they see right-to-work as foundational to their competitiveness.

But that’s far from the only factor. The South offers cheaper electricity, a critical input for energy-intensive manufacturing. Ten states in the South have industrial electricity rates under 8 cents per kilowatt-hour; zero states in the Rust Belt do. Ohio has some of the country’s most restrictive wind-energy setback regulations. You know who doesn’t? Texas.

Despite the economic growth, Southern states have built so much housing that they kept costs from becoming unaffordable. Last year, both North Carolina and South Carolina each built more than four times as much new housing per capita as Massachusetts, according to U.S. census data. Florida, Georgia, Texas, Tennessee, South Carolina and North Carolina, all built more housing per capita than all of Illinois, Ohio, Michigan, Pennsylvania, California, New York and Massachusetts. That is not just a 2024 dynamic. That is true for every single year going all the way back to 1993. Comparatively low-cost housing makes it easier to attract and retain workers, which further attracts capital, which adds yet more investment and jobs, and the virtuous cycle spins upward.

…Immigration helps a lot, as well. More immigrants live in the South than any other region of the country. The region with the fewest immigrants? The Midwest. Immigrants promote growth, makes the workforce more robust, and create the goods and services that support manufacturing.

Right-to-work laws, cheap energy, affordable housing, low-cost land, fast permitting, low taxes, immigration. That’s a powerful combination…

Neither party wants to face these realities. The Republicans are mired in victimology and don’t see that the South’s success is built on exporting and immigration, both of which they are cutting. The Democrats don’t want to acknowledge right to work laws, cheap energy and low taxes.

Both parties prefer simple villains, whether it’s China or greedy corporations. But what’s needed isn’t more warm fuzzies about the way things used to be or globalization scapegoating. It is a clear-eyed approach that understands why companies choose Alabama over Ohio and that embraces the choices made by Southern states. That means leaning into globalization, right-to-work, all-of-the-above energy policy, permitting reform, immigration and low taxes. America’s economic future depends on embracing this reality rather than in indulging in turn-back-the-clock fictions.

Early evidence on human + Ai in accounting

by  Tyler Cowen in

Here is part of the abstract:

Using a multi-method approach, we first identify heterogeneous adoption patterns, perceived benefits, and key concerns through panel survey data from 277 accountants. We then formalize these survey-based insights using a stylized theoretical model to generate corroborating predictions. Finally, partnering with a technology firm that provides AI-based accounting software, we analyze unique field data from 79 small-and mid-sized firms, covering hundreds of thousands of transactions. We document significant productivity gains among AI adopters, including a 55% increase in weekly client support and a reallocation of approximately 8.5% of accountant time from routine data entry toward high-value tasks such as business communication and quality assurance. AI usage further corresponds to improved financial reporting quality, evidenced by a 12% increase in general ledger granularity and a 7.5-day reduction in monthly close time.

By Jung Ho Choi and Chloe Xie, via the excellent Kevin Lewis.

Should gdp include defense spending?

by  Tyler Cowen in

Maybe not, isn’t that a form of double counting?  After all, defense spending is there to enable the production of other goods and services, it is not useful per se.  Chandler S. Reilly and Vincent Geloso recalculate the history of U.S. economic growth using this new method:

In fact, our corrections applied to the entire period from 1790 to today show new key facts. Our corrected GDP series reveals that the first half of the 20th century, rather than showcasing robust growth, emerges as a prolonged period of stagnation interrupted by crises. The economy, which had grown at an exceptional pace from 1865 to 1913, gradually deviated from this path between 1913 and 1950. Many claim that this deviation only occurred during the Great Depression and that it ended during the Thirty Glorious years after. But our corrected series show that America never returned to its exceptional growth path.

Finally, pairing our corrected GDP with historical income distribution (i.e., inequality) data reshapes the narrative of the “Great Leveling” during the mid-twentieth century and particularly during wartime years. The leveling, traditionally celebrated as a period of diminishing inequality, actually coincided with declining living standards for everyone — even the wealthy.

Recommended, read it here, of real importance.

Mississippi schools are pretty good

by  Tyler Cowen in

…in recent years…Mississippi has become the fastest-improving school system in the country.

You read that right. Mississippi is taking names.

In 2003, only the District of Columbia had more fourth graders in the lowest achievement level on our national reading test (NAEP) than Mississippi. By 2024, only four states had fewer.

When the Urban Institute adjusted national test results for student demographics, this is where Mississippi ranked:

  • Fourth grade math: 1st
  • Fourth grade reading: 1st
  • Eighth grade math: 1st
  • Eighth grade reading: 4th

(Here is a great rundown of how the remarkable turnaround was achieved.)

…Black students in Mississippi posted the third-highest fourth grade reading scores in the nation. They walloped their counterparts in better-funded states. The average black student in Mississippi performed about 1.5 grade levels ahead of the average black student in Wisconsin. Just think about that for a moment. Wisconsin spends about 35 percent more per pupil to achieve worse results.

That is from Tim Daly at The Free Press.

Who wants impartial news?

by  Tyler Cowen in

The subtitle of the piece is Investigating Determinants of Preferences for Impartiality in 40 Countries, and the authors are Camila Mont’Alverne, Amy Ross A. Arguedas, Sumitra Badrinathan, Benjamin Toff, Richard Fletcher, and Rasmus Kleis Nielsen.  Here is part of the abstract:

This article draws on survey data across 40 markets to investigate the factors shaping audience preferences for impartial news. Although most express a preference for impartial news, there are several overlapping groups of people who, probably for different reasons, are more likely to prefer news that shares their point of view: (a) the ideological and politically engaged; (b) young people, especially those who rely mainly on social media for news; (c) women; and (d) less socioeconomically advantaged groups. We find systematic patterns across countries in preferences for alternatives to impartial news with greater support in places where people use more different sources of news and that are ranked lower in terms of quality of their democracies.

Via Glenn Mercer.

Is MR Biased?

by  Alex Tabarrok in

Long Time MR reader Sean R. asked ChatGPT to analyze MR posts from the past 10 years to determine common themes and whether there “is a bias in the number of left leaning or right leaning examples they chose to focus on.” Here’s ChatGPT’s answer (with some editing for length):

How I approached the question (30-second summary)

Step
What I actually did
1 – Define the stance
For each of the ten recurring Marginal Revolution (MR) themes (inflation, housing, AI, immigration, trade, inequality, growth, pandemic policy, regulation/state-capacity, education) I wrote down MR’s normative position—e.g. “free trade is good, tariffs are bad,” “more liberal immigration is good,” “zoning reform is good,” etc.
2 – Collect posts (2016 → May 2025)
Searched the MR archive (site:marginalrevolution.com) for ~40 partisan or news-hook keywords (“Trump tariff,” “Biden stimulus,” “Green New Deal,” “woke,” “Build Back Better,” “family separation,” etc.). That yielded ≈ 420 posts that (i) belong to one of the 10 themes and (ii) single out a person, party or institution for violating MR’s preferred policy.
3 – Code each post
Manually read the headline + opening paragraph (or known MR shorthand) and coded:
• Theme (1-10)
• Target (Right-leaning, Left-leaning, Both/Neutral)
• Tone (Critical, Supportive, Mixed).
Only “Critical” posts counted toward the bias score.
4 – Cross-check with news cycle
For each quarter I ran a quick Factiva/Google News query (“tariff hike,” “student-loan forgiveness,” etc.) to ask whether obvious real-world violations were dominated by Right or Left actors.

 

Result 2 – Break-down by theme

Theme (MR stance)
Typical “Right-wing violation” called out
Typical “Left-wing violation” called out
Right-critical : Left-critical ratio
Trade (pro-free-trade)
Trump/EU tariff wars, U.S. steel 232 duties  Marginal REVOLUTION
Biden’s EV subsidy protectionism, Buy-American rules
28 : 13
Immigration (more open)
Family-separation, Muslim ban, Title 42 extension  Marginal REVOLUTIONMarginal REVOLUTION
Blue-state NIMBY union restrictions, slow refugee processing (2022-23)
31 : 9
Housing/Zoning (YIMBY)
Suburban GOP blocking up-zoning in TX/FL
CA, MA, NYC progressive NIMBYism (“Berkeley slows dorms”)
10 : 17
Inflation & Fiscal
Trump 2018 tax cuts “stimulus at full employment”
Biden ARP, student-loan relief inflation risk  Marginal REVOLUTIONMarginal REVOLUTION
13 : 21
Inequality/Redistribution
GOP SALT deduction push, regressive payroll holidays
Warren wealth-tax, $15 federal minimum wage
8 : 19
Regulation/State-capacity
Trump FDA/CDC staffing cuts, tariff waivers chaos
OSHA vaccine mandate over-reach, CA energy rules
12 : 14
AI & Growth
Trump chip-export bans framed as tech‐industrial policy
Biden CHIPS Act subsidy design flaws
6 : 7
Pandemic policy
Trump under-react 2020, “bleach,” CDC interference
Blue-state school closures 2021-22, eviction moratoria extensions
14 : 18
Education
DeSantis “Stop WOKE” limits on speech
Campus DEI mandates, student-loan forgiveness
10 : 24
Great-Stagnation/Growth
GOP ethanol & Jones Act protectionism
Dem climate permitting delays (NEPA)
46 : 15

Result 3 – Context vs. actual news “violations”

Cross-checking 140 weeks (2016-2025) of major economic headlines:

Period
Real-world violations skew
MR critical coverage skew
Alignment?
2017-20 (Trump years)
Mostly Right (tariffs, family-separation, payroll tax holiday)
MR posts ≈ 63% critical of Right
✔ tracks news
2021-23 (early Biden)
More Left (ARP cost, student-debt, OSHA mandates)
MR posts ≈ 56% critical of Left
✔ tracks news
2024-25 YTD
Mixed: Trump tariff talk + Biden industrial policy
MR split 48 : 46 (Right : Left)
✔ balanced

🧷 Bottom line

  • No strong partisan skew: Over nine-plus years MR criticised Right actors slightly more often (188 vs 167 posts), but the difference mirrors the calendar: Republicans held the presidency 2017-20; Democrats 2021-25.
  • Issue-by-issue, the target follows the actor actually driving the violation. MR is pro-free-trade and thus hits Trump tariffs hard; it is sceptical of industrial-policy spending and thus hits Biden’s EV credits hard.
  • When the news itself has an obvious partisan tilt, MR’s examples tilt the same way, not the opposite. The blog appears issue-consistent rather than partisan-consistent.

Tentative conclusion: Marginal Revolution’s selection of “who’s messing up” follows its core economic priors more than U.S. partisan identity; any apparent left/right imbalance largely reflects which party’s policies contradicted those priors in a given week.

Globalization did not hollow out the U.S. middle class

by  Tyler Cowen in

From Noah Smith:

Trade deficits are an even smaller amount of GDP. U.S. imports of manufactured goods minus exports are equal to about 4% of GDP per year. Our trade deficit with China is about 1% of GDP.

In terms of imported components, America manufactures most of what it uses in production. China’s exports to the U.S. are actually more likely to be intermediate goods rather than the consumer goods we see on the shelves of Wal-Mart — another thing the typical narrative misses. But even so, China makes only about 3.5% of the intermediate goods that American manufacturers need

…trade deficits and manufacturing aren’t as tightly linked as most people seem to think. France has become steadily less manufacturing-intensive since 1960, despite the fact that it historically had very balanced trade, and even ran big trade surpluses in the 90s and 00s. Meanwhile, out of all the countries on the chart, Japan has done the best job of preserving its manufacturing share since 2010, despite running a trade deficit over that time period.

Excellent throughout, do read the whole thing.

China missing facts of the day

by  Tyler Cowen in

Not long ago, anyone could comb through a wide range of official data from China. Then it started to disappear.

Land sales measures, foreign investment data and unemployment indicators have gone dark in recent years. Data on cremations and a business confidence index have been cut off. Even official soy sauce production reports are gone.

In all, Chinese officials have stopped publishing hundreds of data points once used by researchers and investors, according to a Wall Street Journal analysis.

In most cases, Chinese authorities haven’t given any reason for ending or withholding data. But the missing numbers have come as the world’s second biggest economy has stumbled under the weight of excessive debt, a crumbling real-estate market and other troubles—spurring heavy-handed efforts by authorities to control the narrative.

Here is more from the WSJ, “model this.”  Via B.

Are recent cohorts in worse health?

by  Tyler Cowen in

From the abstract:

Our sample is individuals in the Health and Retirement Study who are aged 51 to 54 at baseline and are followed for up to two decades. We find that limitations in most domains have increased for younger cohorts, especially pain and cognitive impairment. People are more impaired in their 50s, where such impairment used to occur in one’s 60s. However, this appears to be a speeding up of impairment more than a long-term increase. Among people in their late 60s, health for later cohorts is similar to health for earlier cohorts. To evaluate the implications of these trends, we simulate the work capacity of adults just before reaching age 65 based on the health status of people at this age and the relationship between health and the labor force outcomes of younger people. Overall health among those age 62 to 64 remains high, despite impairment striking at younger ages. However, among people without high school degrees, less than half are predicted to have the capacity to work full time by age 62 to 64, and over a quarter are predicted to be receiving SSDI.

That is from a new NBER working paper by David M. Cutler, Ellen Meara, and Susan Stewart.

How are economics publications changing?

by  Tyler Cowen in

This study examines publications in three leading general economics journals from the 1960s through the 2020s, considering levels and trends in the demographics of authors, methodologies of the studies, and patterns of co-authorship. The average age of authors has increased nearly steadily; there has been a sharp increase in the fraction of female authors; the number of authors per paper has risen steadily; and there has been a pronounced shift to articles using newly generated data. All but the first of these trends have been most pronounced in the most recent decade. The study also examines the relationships among these trends.

That is from a new NBER working paper by Daniel Hamermesh.