A new analysis finds only a 21 percent chance that the U.S. economy has entered a sustained high-productivity boom when measured by total factor productivity, even as artificial intelligence investment surges and worker output accelerates. The report, "Have We Entered an Era of High Productivity Growth?" by San Francisco Federal Reserve economists Hamza Abdelrahman and Andrew Foerster, suggests AI may still be in the "better tools" phase rather than the economy-transforming stage that optimists predict. The analysis warns that while labor productivity has picked up solidly in recent years, a more fundamental measure of innovation remains far more modest.

The report reveals a striking divergence between two key productivity measures since 2022. Labor productivity—worker output per hour—has accelerated notably, while total factor productivity gains have been much more modest over the same period. When the economists ran a statistical model based on labor productivity alone, it put the odds of a sustained high-productivity boom at about 57 percent. But when they used total factor productivity as the measure, those odds dropped sharply to just 21 percent. This gap matters because labor productivity simply measures how efficiently workers use available capital like equipment and software, while total factor productivity captures how efficiently the economy uses all inputs together—both labor and capital—making it the standard proxy for genuine innovation.

The report finds reason for "cautious optimism that the U.S. economy could be starting to experience a period of sustained high productivity growth." When Abdelrahman and Foerster ran their model using only data available through 1997, the signals looked remarkably similar to today, including the same divergence between the two productivity measures. According to the analysis, "The dynamic then was similar to what we have seen in recent years. If today mirrors what we experienced in the mid-1990s, we may be in the early stages of a productivity boom driven by AI that will only become clear in retrospect." That earlier period, which showed no clear confirmation at the time, turned out to be the start of a massive productivity surge that defined the "New Economy" internet era.

The divergence between the two measures hints at what stage of the AI transition we're actually in. If generative AI were already spreading broadly across the economy—not just concentrated in tech or information-heavy sectors—both productivity measures would likely be keeping pace with each other. Instead, only labor productivity is moving strongly, which suggests businesses are investing enormous amounts in AI infrastructure and technology, creating better tools for workers through what economists call capital deepening. But total factor productivity reflects genuine improvements in processes and technology across all sectors—the kind of fundamental innovation that happens when a general-purpose technology like electrification, the internal combustion engine, or the computer truly reshapes an entire economy. Right now, that deeper transformation hasn't shown up clearly in the data.

The San Francisco Fed economists will be watching for more evidence as new data become available. The comparison to the mid-1990s offers hope that today's uncertainty could be masking the early stages of a genuine boom that will only become obvious years from now. But for now, the expected AI productivity revolution is still waiting for its proof.