Well, what a coinky-dink! Just in time for the Federal Reserve’s Open Market Committee meeting tomorrow where they’ll be discussing the liftoff date for the interest rate they control, economist Josh Bivens has a smart new paper out on the importance of considering wage trends in the mix, written for our full employment project.
Here, in his words followed by a bit of annotation from moi, is his logic:
- Quantity measures of labor market slack are hard to forecast and are not useful guides to policy. In particular, the unemployment rate may understate labor market slack if labor force participation rates are cyclically depressed, as they likely are currently.
[Also, the share of involuntary part-timers remains elevated, so there’s considerable “under-employment” as well. Using Levin’s measure of slack, this and the part Josh mentioned re the labor force would add more than a point to the unemployment rate.]
- Estimates of the natural rate of unemployment (the rate below which inflation will accelerate) and the structural rate of change in labor force participation (the rate determined by demographic factors rather than economic one) have extremely large margins of error.
[Maybe 2-4 percentage points, in terms of a confidence interval around their natural-rate estimates, which is, as Josh says, large. BTW, did you see this? New report from a couple of Fed economists that backs out a “natural rate” slightly north of 4%.]
- By contrast, relatively precise wage targets can be estimated, so long as the trend of productivity growth remains consistent.
[Productivity growth and unit-labor costs loom large in the Bivens analysis, as they should. It was accelerating productivity growth that gave ol’ Al G-span the idea to let the unemployment rate fall below what they thought was the natural rate back in the 1990s: ~6%! Trend productivity growth is lower now—around 1.5%–and that’s a constraint. But there’s still considerable room for non-inflationary wage growth. The Biven’s benchmark target for non-inflationary nominal wage growth is basically the Fed’s inflation target–2%–plus the trend growth rate of productivity–1.5%. So, 3.5%.]
- Fears that exceeding wage or price inflation targets for a short period will de-anchor inflation expectations and lead to accelerating inflation are misplaced.
[Tru dat! See Larry Ball’s most recent paper from our series.]
- Wage growth that exceeds the sum of productivity growth and the Fed’s price inflation target even for years is not necessarily a problem when labor’s share of income is significantly depressed; instead, it is a necessary condition for a return to normal economic growth.
[OK, this is my fav point, one that I’d like everyone in this debate to either make a big, damn deal out of or explain to me why it’s not a big, damn deal. It’s a bit of a gnarly point but it comes down to this: thanks to job market slack and high levels of inequality, much of the income growth in this recovery has gone to profits, not wages. Another non-inflationary way in which wages can grow—even when they’re growing faster than the 3.5% benchmark Josh identifies—is through shifting some of that growth back to paychecks. In technical terms, it’s the rebalancing of factor shares, and as Josh’s Figure 7 shows, it means there’s considerably more room for wages to grow. EG, with “excess wage growth” 1.0 percentage point above the 3.5% benchmark—so, nominal wage growth of 4.5%–it would take eight years, until 2022, for the compensation share of national income to revert to its pre-recession level.]
- There is no single indicator on which the Fed should mechanically base its decisions. However, the indicator that best captures how close the economy is to labor market tightness consistent with a given inflation target is a nominal wage target that sums the trend in productivity growth and the price inflation target. By this measure, the Fed should forestall monetary tightening until substantial progress in meeting that wage target is met.
I already said “tru dat!” so how about “I feel ya, Josh!” I only hope Janet and co. are listening.