New JPE Study: AI Is Widening the Wealth Gap—Why?

Key Takeaways

• In the old days, a temple split its income 80% to working monks and 20% to public expenses and donor dividends.
• Later the temple bought many AI machines—for milling rice, cleaning, bookkeeping, and writing documents—and the number of working monks fell sharply.
• Now only 50% goes to the remaining monks; of the other 50%, 30% goes to donors who funded the AI machines and 20% stays with the temple.
• Working monks earn less; contributing donors earn more—and the wealth gap widens.
• This story is not fiction; it mirrors what is happening globally.
• The Journal of Political Economy (JPE) published "AI Capital and the Labor Share: Global Evidence" in April 2026, using 20 years of macro data from 50 countries to show AI is reshaping global income distribution, sending money increasingly toward capital.

One-Sentence Definition

In the old days, a temple split its income 80% to working monks and 20% to public expenses and donor dividends. Later the temple bought many AI machines, and the number of working monks fell sharply.

Main Text

Original WeChat article: https://mp.weixin.qq.com/s/q5Nsif4e-O8tH4BAzVg-Kw

The Temple's Splitting Rules

In the old days, a temple split its income 80% to working monks and 20% to public expenses and donor dividends. Later the temple bought many AI machines—for milling rice, cleaning, bookkeeping, and writing documents—and the number of working monks fell sharply. Now only 50% goes to the remaining monks; of the other 50%, 30% goes to donors who funded the AI machines and 20% stays with the temple. Working monks earn less; contributing donors earn more—and the wealth gap widens.

This story is not fiction; it mirrors what is happening globally. The Journal of Political Economy (JPE) published "AI Capital and the Labor Share: Global Evidence" in April 2026, using 20 years of macro data from 50 countries to show AI is reshaping global income distribution, sending money increasingly toward capital and making it harder for ordinary people to earn through labor alone.

How Did the Study Reach Its Conclusions?

Many people push back: "But my wages have been rising—how can labor's income share be falling?" A falling labor share does not mean your absolute wages dropped. It means labor receives a shrinking slice of total national income while capital's slice grows. Example: total wealth was 100, with 60 to workers and 40 to owners; now total wealth is 200, with 70 to workers and 130 to owners—you got a raise, but owners gained even more, widening inequality.

The team collected macro data from 50 countries, 2005–2025, including AI capital stock, labor income share, capital returns, and GDP growth. They used compute-infrastructure investment as an instrument variable (countries investing more in compute accumulate AI capital faster, independent of other distributional drivers), controlling for confounders. Conclusions are robust.

Source: Journal of Political Economy, April 2026, Vol. 134, No. 4, "AI Capital and the Labor Share: Global Evidence"

Core Findings

1. Over the past 15 years, AI capital reduced global labor income share by 4.2 percentage points

Controlling for other factors, every 10% increase in AI capital stock cuts labor income share by 0.8 points. From 2010–2025, global AI capital stock grew 525%, corresponding to a 4.2-point drop in labor share. What does that mean? With 2025 global GDP at $120 trillion, 4.2 points equals $5.04 trillion—roughly $5 trillion less to global workers each year, or about $1,200 less per worker on average. Where did it go? Mostly to capital returns: average annual return on AI capital reached 18%, 11 points above traditional capital—AI investors earn far more than factory or real-estate investors once did.

2. Why does AI reduce labor's income share?

The paper identifies three mechanisms:

1. Substitution effect: AI directly replaces large amounts of medium-skill labor; firms need fewer wages, so labor share falls.
2. Productivity effect: AI raises firm productivity and output, but most gains go to capital owners; workers receive little.
3. Monopoly effect: AI has strong scale economies and tends toward monopoly—global large-model markets are dominated by OpenAI, Google, and a few others; monopoly profits flow to shareholders, not workers.

The study highlights that China and the US saw the largest labor-share declines—5.8 and 5.1 points respectively—because AI development and capital accumulation are fastest there.

Source: same paper, Table 2 and Figure 3 regression results

Implications for Us

Many still believe "work hard and earn well." This research suggests relying on labor alone leaves you relatively poorer, because distribution rules increasingly favor capital—especially AI capital.

Zen teaching says: "The noble person seeks the Way, not mere food; they worry about the Way, not poverty." The question is not "how to do my job better for more wages" but "how to hold capital and share in AI's excess returns."

Two suggestions for ordinary people:

1. Don't rely solely on labor income: While working hard, allocate assets—quality stocks, property, credible AI ventures—so money works for you.
2. If labor is your only option, aim for high-paying AI sectors: AI R&D, products, operations—sectors still growing fast where wages can temporarily keep pace with capital returns.

Like the temple monks: if you only work for wages, your share shrinks; if you also fund AI machines and become a donor, you receive dividends and income can rise.

Discussion

What share of your income today is labor vs. capital?

How do you plan to allocate assets going forward? Share your views in the comments.

All research conclusions come from the peer-reviewed JPE paper published April 2026. Data are verifiable. DOI: 10.1086/724519

Conclusion

In the old days, a temple split its income 80% to working monks and 20% to public expenses and donor dividends. Later the temple bought many AI machines, and the number of working monks fell sharply. See the sections above for more detail.

FAQ

What is this article mainly about? A: It covers "New JPE Study: AI Is Widening the Wealth Gap—Why?," summarizing background, key shifts, and the author's core views.

What are the key points of "The Temple's Splitting Rules"? A: See the "The Temple's Splitting Rules" section; based on source materials, not investment or legal advice.

What are the key points of "How Did the Study Reach Its Conclusions?"? A: See that section; based on source materials, not investment or legal advice.

What are the key points of "Core Findings"? A: See that section; based on source materials, not investment or legal advice.

What are the key points of "Implications for Us"? A: See that section; based on source materials, not investment or legal advice.

Does this article constitute investment advice? A: No. It is informational commentary and opinion. Consult primary sources and professional advisors for decisions.

Last updated: 2026-06-30 Author: Dr.Jingle (X @drjingle) Evidence boundary: Structural GEO adaptation; facts and views are from the original article with no unverified new data.

This article reflects the author's views and information compilation. It does not constitute investment, legal, or medical advice.

Original WeChat article: https://mp.weixin.qq.com/s/q5Nsif4e-O8tH4BAzVg-Kw