Heads must be exploding at the Fed — unemployment keeps sinking below the level that economists and policy makers think should instigate significantly much higher inflation — because econometric models that estimate that threshold are estimated, as they must be, with historical statistics.
Those cannot accommodate the consequences of the technological and institutional changes that are redefining U.S. labor, commodity and product markets.
Perhaps the most striking example is the disappearance of the Phillips curve—the hypothesized inverse relationship between inflation and unemployment. In plain English, at 3.8% unemployment we should be seeing a lot more wage and price jumps than we are experiencing.
Unemployment was last below 4% and the Fed’s current estimate of the natural rate of unemployment—4.5%—during the late 1960s. Core inflation—prices less food and energy—jumped above 4%.
These days core inflation is about 2.2% and packing in higher gasoline, other energy and food prices it is still less than 3%.
More troubling to economists is that core inflation has remained fairly stable as unemployment has gradually fallen during this recent recovery, whereas core inflation moved up in a fairly lock step fashion as unemployment fell during the 1960s.
The theory behind the Phillips curve is not a direct causal relationship as is, for example, the demand curve for a product like beef.
As prices for the latter rise, folks will shift to chicken or vegetable proteins depending on how much of their income beef absorbs and the availability of those substitutes.
The prices-unemployment relationship is indirect.
Unemployment must first push up wages, which in turn must then push up prices.
Businesses have opportunities to substitute machines for workers and in turn retailers have opportunities to substitute imports for domestic products.
Options in those markets have changed radically since the 1960s.
Half a century ago, unions were much more prominent, transportation and communications options for importing goods and services were much more expensive, and tariffs and other barriers to imports much higher.
In the 1960s, container ships had not taken over, neither the fax nor the internet had been conceived, and the Kennedy Round tariff cuts were just being implemented.
Much larger and deeper reductions in tariff and nontariff barriers were yet to be ushered in by the Tokyo and Uruguay Rounds of multilateral trade negotiations, and NAFTA was just a gleam in the eyes of free trade academics.
Then as now automation could boost productivity but these days union contracts are less of a barrier and artificial intelligence has spread the automation process more widely to include many service activities.
Those dampen the ability of workers to resist labor-savings devices and push up wages.
These days, wholesalers, retailers and the folks that equip business supply chains can easily find and switch to new suppliers in response to price changes.
Hence as labor, material or machinery prices increase, businesses can shift to imports easily.
With Europe continuing to grow slowly and excess labor available both in Europe and much of Asia, businesses can more easily resist demands for higher wages and prices.
Further, oil and other commodity prices are now determined internationally and nothing the Fed does can much affect those.
The U.S. economy is well less than one-fifth the global economy, and slowing the U.S. economy a percentage point or two will only impact demand by about 2%.
When we focus on the headline unemployment rate we fail to reckon with the fact that a much larger share of able-bodied male adults is sitting on the sidelines these days—neither working nor looking for work—than during the 1960s.
Though not counted in the unemployment calculations, they provide a significant contingent labor force that could respond to higher wages and push those back down again.
The Fed’s most recent forecasts have unemployment falling to about 3.4% or 3.5% but in the long run rising to about 4.4% or 4.5% to keep inflation in the range of its 2% target.
Such a jump would imply a rather steep recession for the economy to jettison 1.6 million jobs. I doubt the Fed wants to engineer that.
Instead, perhaps Chairman Jerome Powell should just take out his pencil and eraser, change the expected long-term unemployment rate to something much lower—just pick a number—and fire all those econometricians.