Not a surprise, but Janet Yellen appears to have solid intuition when it comes to the labor market slack vs. tautness. Let me explain.
There’s this debate going on that says “given all the slack in the job market, we should have seen inflation (price growth) fall a lot faster than it has. Thus, maybe there’s not so much slack in the labor market.”
This position got a bit of a boost from a recent paper by Krueger at al, which argues that if you just look at the total unemployment rate, you will mistakenly conclude that there’s more slack than there really is. The short-term rate (there’s no set definition, but people generally use unemployed for less than 27 weeks), which has fallen back to its historical average (a bit above 4%), fits the inflation data better than the overall rate and suggests the job market is actually tightening up, meaning the Fed may need to act sooner than later.
When asked about this the other day, Chairwoman Yellen called this line of argument “tremendously premature.” That’s actually a strong response—one could imagine the Fed chair doing the usual “well, certainly something worth looking at but let’s be cautious…yada, yada.” Her strong response reflects her intuition that the job market is still quite slack and the extent of long-term unemployment remains an important input into their policy calibrations.
Well, over the weekend I read an interesting piece by Goldman Sachs economists Hatzius and Stehn (H&S, no link) that tackled this question by digging a bit deeper in the current dynamics between unemployment and wages/prices. Their key findings are:
–Starting around the 1990s, inflation fell and has stayed pretty low and well-anchored (meaning less responsive to temporary shocks);
–Thus, the Phillips curve—the reaction function between inflation and labor market slack flattened considerably, as I show here.
–If you fail to account for this regime shift in the rate and responsiveness of inflation, you’ll overestimate its expected reaction to any given amount of slack.
Why the shift? H&S say that “a likely explanation is that downward nominal rigidities are much more important at lower levels of inflation than at higher levels, and a given amount of slack is likely to have a smaller impact on inflation when the starting point for inflation is low than when it is high.” That’s a bit gnarly but what they’re saying is that when there’s a lot of inflation—when it’s 4 or 5% instead of 1 or 2%, like it’s been lately—and slack in the job market, employers can more easily cut real wages by just holding nominal wages pretty constant and letting inflation erode them. But when inflation is very low, your nominal wage isn’t that different from your real wage, and so it’s tougher to implement real cuts.
To test the proposition that broader measures of slack, including the total rate and the so-called U6 rate—BLS’s most inclusive measure of labor market under-utilization—are better at predicting wage and price movements than the short-term unemployment rate (once they account for the critical regime shift) they produce the following figure.
It shows actual nominal wage growth (the black line) along with various predictions based on different measures of slack. Actually, all the slack measures over-predict recent wage growth, which has been awfully stagnant (while profits have soared), but the short-term rate does notably worse (it over-predicts the most). They get a similar result for inflation.
This may seem arcane but I assure you, it’s not. Were the Fed to buy the short-term-rate-is-the-one-to-watch line, they’d conclude that instead of being biased down (as I and others have been stressing for months) the unemployment rate is biased up. Thus, they’d be more likely to tighten prematurely.
Second, and Yellen was great on this point in a talk she gave today, to buy the short-term story implies some degree of writing off the long-termers, basically assuming that they’re permanently out of the picture/labor market. As Ylan Mui reported, Yellen rejected that notion:
Yellen said there have not been “clear indications” that the short-term unemployed are finding jobs more readily than the long-term unemployed. This debate has gained traction among economists lately. The short-term unemployment rate is near its pre-recession level, raising fears that the Fed may be pushing the economy too far. Yellen clearly comes down on the other side of the argument: “This fact gives me hope that a significant share of the long-term unemployed will ultimately benefit from a stronger labor market.”
That view is one I tried to amplify here, and it’s a view that underscores both the long-term costs of such persistent slack in the job market, as well as the potential benefits of pursuing a path back to full employment.