Sunday, August 02, 2026

Return to NAIRU

As I sometimes do, this week I looked at the tables of economic data in the back of the Economist. One of the most important relationships in economics for the last few decades has been that between inflation and unemployment. Ideally, both of them should be stable. The Fed has long had a stated goal of 2% inflation, and the idea has been to set interest rates to manage business cycles to keep inflation around there. When the economy is growing fast, tight labor markets drive up the cost of labor -- a big component in the cost of everything -- so banks raise rates to slow growth in order to drive unemployment up to the equilibrium rate of unemployment, the NAIRU for short ("non-accelerating inflation rate of unemployment").

Recently we got the news that inflation had run above the target rate of 2% for 63 consecutive months. Leaving aside discussions of causes, which I'd say are manifold and contain amongst them a lot of COVID-driven accident compounded by policy mistakes on both sides -- amongst the biggest being ill-targeted stimulus from Biden and too-tight borders and excessive Fed jawboning from Trump (early 2018 rate cutting cycle was a mistake -- markets should have been left to correct and settle) -- that's a long time.

But the biggest mystery to me is why public discussion of what the NAIRU should be has vanished. It sounds like a really wonky discussion but it's not. When the debate first came on my radar, back in the mid-late 90s, consensus on what the NAIRU should be hovered around the 5.5-6% range. Unemployment went down to 4% in the height of the dot-com boom then topped out at 6.1% in August 2003, by which time the Fed was lowering rates. Admittedly, we now see that may have been a mistake because it created the credit conditions which created the financial crisis. Under Obama we climbed back from the brink, and by the time Trump was elected unemployment was well below 5%, where it has stayed with the exception of the short-lived COVID megaspike. 

I'm going deeper into the weeds than I intended to but my main point is that since late in Obama's days, we have tolerated unemployment much lower than we would have historically. It sounds cruel to want higher unemployment when everybody wants to work (in theory, except when we are complaining about our jobs, of course). Maybe we actually need to get back to managing to somewhat higher unemployment.

One more thing: it's important for there to be unemployment up and down the labor value chain. Highly compensated people need to be getting let go for a bunch of reasons, both economic and social (for the growth of empathy). We need businesses to fail. Career and business failures create the seeds of creativity and growth. We may see a lot of this as AI gets integrated into the economy. We may need empathy from and for many.