Every lengthy economic cycle has a point at which official statistics no longer reflect how people feel on the street, and you begin to see it in subtle ways. Photographed for a wire service at a career fair in Seattle where there were more recruiters than candidates. a LinkedIn feed that is overflowing with “open to work” banners from people who shouldn’t be looking at all. While the monthly jobs report shrugs and says, basically, nothing happened, the GDP print lands firmly in the green. The Wall Street Journal attempted to address that gap this April, and the story is more bizarre than the headline implies.
The correlation between hiring and economic growth has been so consistent for the majority of the last century that economists hardly gave it much thought. The economy grows, businesses hire, the unemployment rate declines, and wages fluctuate. It was the closest thing macroeconomics had to a physical law. Then, in 2025, the law ceased to be applicable. GDP increased. Jobs didn’t. At an economics conference in late February, Christopher Waller, a member of the Federal Reserve Board, said it aloud. The statement stuck because it was so casual. In his whole career, he had never witnessed anything like this. He was at a loss for words. A Fed governor’s admission of that nature is uncommon enough to be considered a data point in and of itself.
| Topic | GDP–Jobs Divergence in the U.S. Economy |
| Period analyzed | 2025 through early 2026 |
| Key data source | Bureau of Labor Statistics |
| Headline observation | Economy expanded while job creation flatlined |
| Reporting publication | The Wall Street Journal (Harriet Torry, April 13, 2026) |
| Federal Reserve commentary | Governor Christopher Waller, late February 2026 |
| Productivity context | U.S. labor productivity showing notable rebound |
| Likely contributing factors | AI deployment, slowing immigration, smaller labor force |
| Public sentiment | 64% of Americans expect higher unemployment within a year |
| Macro reference | Federal Reserve policy stance |
| Industries seeing biggest productivity boost | Knowledge work, finance, software, customer service |
| Industries seeing biggest job risk | Same as above |
Harriet Torry’s preferred explanation, published in the Journal, focuses on productivity. The argument goes that because workers are just producing more per hour than they used to, the economy can expand without hiring more people. A portion of the labor pool is reduced by slower immigration. The demand for it is further reduced by AI. The two trends come together in the middle to create what appears to be a robust economy that doesn’t really require more workers. It’s a neat tale. Maybe it’s the right one.
However, the longer you look at it, the more difficult it is to ignore the softer reading that sits beneath the official one. The same businesses that are reporting high profits are also quietly laying off employees. Over the past 12 months, large firms have eliminated middle management, reorganized customer service, and automated software engineering and legal review tasks that were previously completed by entire teams. Last week, Bryan Alexander made the awkward observation in a post on his AI substack that the professions with the greatest productivity gains from AI are also the ones with the highest risk of job displacement. These things typically appear as bad news in a kitchen and as good news in a quarterly report.
The consumer sentiment data has moved slightly in the wrong direction as a result. The percentage of Americans who anticipate higher unemployment over the next year has reached a level we have never seen outside of an actual recession, according to a different Journal article by Justin Lahart and Owen Tucker-Smith. The rate of unemployment is low. The S&P 500 is doing well. Spending is acceptable. Nevertheless, 64% of respondents believe that things will soon worsen. Watching the two stories together gives the impression that one country is described by the official data, while another is described by the lived experience, with the official data updating more slowly.
The optimistic framing in the WSJ might be accurate. Historically, increases in productivity have been beneficial. There was one in the late 1990s. There was one in the years following the war. Even though the transition is difficult, they usually result in increasing real wages over time. Even if individual careers in the middle of it are not, the long arc may be acceptable if AI is doing for office work what mechanization once did for farming. That’s a fair wager. It may even be the most likely result.
What happens to the social contract in the interim is the more difficult question. When an economy expands without hiring, it is also subtly shifting the benefits of growth away from workers and toward the owners of the capital that boosts productivity. Warren Buffett, who has observed several of these cycles, has been remarkably silent about this one, which is a telltale sign in and of itself.
Corporate margin calculations get better. The household budget math does not, at least not yet, at least not equally. Additionally, it seems as though the economists at the Fed and the Treasury are looking at one set of numbers while the attendees at the career fair in Seattle are looking at a completely different one as the data continue to diverge month after month. No one, not even the Journal, is able to definitively answer whether those two sets of numbers will eventually cross paths again or simply continue to walk past one another.
