Given that my butt is firmly stuck at the airport due to a 3 freakin’ hour flight delay, and, on the plus side, it’s an airport (Dulles) with a decent wifi, I see no reason not to obsess a bit about this concept of the “natural rate of unemployment,” the jobless rate that’s allegedly consistent with stable prices. Go below the natural rate, so the thinking goes, and inflation will keep rising until the unemployment rate climbs back up to the NAIRU, the “non-accelerating inflation rate of unemployment.”
There are trenchant objections to this framework that I explore in greater detail here and below. But as long as we’re operating within this constraint, I’m happy to see that the Obama econ team has lowered their natural rate to 4.9% in the just released mid-session budget review.
Table 2 in the doc presents their economic assumptions, showing that they expect unemployment to fall as low as 4.6% in 2017. That’s below their NAIRU, so as you see, inflation climbs a bit until the jobless rate settles at 4.9% and inflation stabilizes at 2.3% (wait…isn’t that above the Fed’s 2% target?! Yes, but the admin’s estimate is for total inflation, not the core measure the Fed tracks; 2.3% in the topline CPI is consistent with 2% in the core PCE).
As you see in the little table below that Ben S. kindly made for me, this is the Obama admin’s lowest natural rate, even lower than when I was there (curse you, Furman, et al!). The Fed’s NAIRU is 5.1% and CBO’s is still up around 5.4% (!).
I’ve stressed incessantly that all these numbers seem high to me. The jobless rate is currently 5.3%, below CBO’s current NAIRU, and within spitting distance of the others. But has inflation or wage growth accelerated as you’d expect if we’re truly approaching the “natural rate?” No, and again, no!! Nope. Nada. Nuh-uh. Negative.
In my book, I offer three reasons for this failure.
“One…the full employment unemployment rate is lower than the Fed [or the admin, or most economists] thinks. Two, we’re mismeasuring labor market slack and the unemployment rate doesn’t tell the whole story. And three, economists don’t quite understand these dynamics.”
The first point is drawn from the observation I just made about wage and price growth (and I’m talking not just their actual trends but expectations as well). The second references measurement issues like the depressed labor force participation rate and the elevated number of involuntary part-timers. The third reflects that we’re not clear at all about “Phillips curve” relations these days—the extent to which unemployment correlates with wage and price growth.
For these reasons, it’s VERY important to remember that there’s significant uncertainty around all of these NAIRU estimates. Though their work is a bit dated at this point (hint, hint re need to update it), a group of our top econometricians noted during the 1990s that there are at least three sources of error in estimating natural rates: our lack of knowledge about which model is “correct,” our uncertainty about parameters that feed into our chosen model, and the randomness baked into the concept in the first place. Based on those sources, they estimate 95% confidence intervals around the natural rate of 3 to 4 percentage points, which is “large” in this business. So the admin’s new NAIRU estimate is somewhere between 3% and 7%.
To be clear, my beef is not with the admin here, as they’re working within the same constraints as any of their predecessors. Good for them for moving a bit with the evidence.
But when you hear an economist declare that the natural rate of unemployment is X, and we’d better slam on the growth brakes before we go below X, remember that large confidence interval, and run, do not walk, far away from them.
Well…now I’ve only got two hours left to wait. Woohoo.