In a post this morning, I noted the Fed’s theory of the case as to why inflation isn’t accelerating in response to the tighter job market: temporary factors, including low, low oil prices and the strong dollar, are blocking the usual signal.
While it’s certainly the case that pricier oil and a weaker dollar would be inflationary, there are reasons to doubt the Fed’s explanation. First, it’s not just that inflation isn’t picking up as output gaps close and unemployment falls. Inflation didn’t fall as much as expected when such activity gaps were much wider.
Second, there’s this figure from a recent paper by Blanchard, Cerutti, and Summers (BCS), showing that a flat slope of the Phillips curve (PC) across a bunch of economies ain’t exactly a new development (the PC measures the correlation between labor market slack and inflation). By this measure, the slope of the PC has been low for a decade. BCS show that the US coefficient is now about 0.2, well below its historical levels in the 70s and 80s, and not statistically significant (from what I can tell–see Figure 9). The German PC slope is 0 and insignificant.
As you can see, these estimates allow the PC slope to change over time. Do more traditional approaches yield different results? In the same paper, BCS provide such results by country (Table 6). Many slope coefficients are insignificant (at the 0.05 level or below), including the US using the more recent sample (2007-14).
To be fair, Larry Ball, in commenting on BCS, runs particularly transparent models and finds more stable, significant PC coefficients (though they are of the same magnitude as BCS) . He argues, as he has in earlier papers, that “the biggest change in the Phillips curve is the anchoring of expectations, and that makes it easier to stabilize inflation. If a shock pushes inflation up, anchored expectations push it back down without the need for a monetary tightening and higher unemployment.”
These distinctions, however, do not give much support to the view that the flat PC is temporarily low as a function of a few unique factors. Whether it’s anchored expectations, globalization, demographics, or the other factors I cited in my earlier post, recent PC dynamics suggest that monetary policy can safely accommodate low unemployment without a lot of angst about price pressures.
Inflation hawks, pull in your talons!