From the Desk

As Kevin Warsh takes over as Chairman of the U.S. Federal Reserve, he inherits inflation that is running hotter than he would like for a reason he can’t control. The U.S. and Iran are blockading energy shipments, driving energy and food costs higher. His models can do little to change that political stalemate.

But he has one reason to cheer: productivity is up. At the last Fed meeting, outgoing Chairman Jerome Powell effused that, “I never thought I’d see this many years of really high productivity and, by the way, expect it to continue.” His lookback period is not driven by AI. But if AI lives up to hype the gains should keep building.

Productivity seems simple to measure. Is more being done with the same head count? In the decade after the Global Financial Crisis, it became dormant. However, the last five years have brought a renaissance. Productivity has grown at the fastest rate in 20 years.

Robert Solow, the Nobel Prize-winning economist, did much of the seminal work on innovation. His original insight was that the raw inputs into growth – capital and labor – explained only 13% of the gains. The rest, or the “Solow residual” as it became known, was attributable to technical innovation. But he always felt the innovation label was too high-minded. Technical innovation was a catch-all for any improvements, such as breakthroughs in the lab, ideas from the factory floor, improved logistics or better management techniques.

But one thing that has empirically emerged is that it takes time for breakthroughs to filter into business processes. For instance, COVID forced a higher adoption of work from-anywhere technology that already existed. And it allowed others to switch professions to better match skills. Therefore, AI will likely see a staggered adoption phase. But once it becomes reliably embedded in processes, it’s likely to add another leg to productivity gains.

Sources: Bloomberg, The Financial Times, 2026.