While President Trump plays symphonies of chin music about his role in today’s strong economy, he has far less to do with the economic policies that affect peoples’ lives than the Federal Reserve. They set the benchmark interest rate that both regulates the economy’s heat and moves other key rates we face every day, including mortgage, car, and bank loans.
In that regard, it’s useful to know what they’re up to. But instead of analyzing their next interest rate move (the prediction market is virtually certain that they’re going to raise rates another quarter point when they meet later this month), let’s delve into a new and important paper by 5 Fed economists: Chris, Jim, Mike, Dave, Bob. Call them the Fed 5.
To understand why the Fed 5’s key result is counterintuitive, you need some background. The main job of the Fed is to balance the tension between full employment in the job market and price pressures in the product market. Historically, these two variables have been negatively correlated, meaning price pressures go up in tight labor markets. But this correlation has long been low: in recent years, low unemployment has not much led to accelerating prices.
Now, most of us tend to think low unemployment, good; high prices, bad (though that’s not quite right because you actually want some inflation in your economy). So, if inflation is low, and especially if low unemployment isn’t driving prices up too much, we’d like the Fed to ingest a time-release chill pill and keep their feet off the economic brakes. After all, if we don’t need to worry so much about low unemployment triggering ever-increasing inflation, why not let the jobless rate fall as far as it can?
Not so fast, say the Fed 5. According to their work, that path courts danger, specifically, under certain conditions — like unforeseen shocks to the system — the jobless rate could fall too far and trigger so much inflation or other distortions, like bubbles in financial markets, that the Fed would have to slam the growth brakes at great cost to economic growth.
The Fed 5 humbly submit that our knowledge of the key parameters in play in this inflation/unemployment drama are poorly understood. The level of one of the most important — u* (that’s “u-star,” the lowest unemployment rate consistent with stable prices) — has long eluded economists. We don’t enough about what’s driving inflation, why its correlation with unemployment is so low, and whether it could come back to life in a truly full-capacity economy.
That’s led many of us to conclude that our “best move is … to admit the uncertainty . . . and follow the data, particularly inflation.”
But — and here’s the paradox — the Fed 5 argue that even amid the uncertainty, we’re still better off targeting u*, even though the Fed’s guesstimate of it — 4.5 percent — seems clearly too high, as unemployment has long been below that rate and their key inflation gauge has hardly accelerated. The Fed’s own forecasts have unemployment falling a point below their u* and inflation still remaining tame!
Here’s their logic. Suppose the central bank takes my advice and targets inflation instead of unemployment. Next, suppose inflation is, as has been the case, undershooting the Fed’s 2 percent target (remember, too little inflation is also a problem). Then, because the correlation between inflation and unemployment is so low, it would take very low unemployment to juice inflation. Conversely, suppose some shock to the system … like, um, a trade war … led inflation to spike; then it would take really high unemployment to bring inflation back down. The authors’ concerns are thus that if the Fed targeted inflation, unemployment could fall so low or climb so high that it could generate “risks to financial stability and more generally to the sustainability of macroeconomic outcomes.”
To their credit, the Fed 5 are explicit that this result is baked in with an assumption that reasonable people may not accept. They use a formula (a “loss function”) that heavily penalizes policymakers for letting the jobless rate fall below u*.
To understand why this matters, consider current conditions. As noted, unemployment is well below the Fed’s u* but inflation is just now, after years of downside misses, hitting their target. Thus, an inflation-targeting rule would say to today’s Fed: “don’t worry about that low jobless rate; leave rates alone for now until inflation firms up a bit more.”
The Fed 5’s loss function shows this to be a risky strategy but there are reasons they might be wrong. First, if, as I believe to be the case, u* is lower than they assume, their analysis doles out undeserved penalties for actual unemployment being “too low.” Second, because of a) the extremely valuable benefits of super-tight labor markets to less advantaged workers, and b) the low inflation correlation, the loss function shouldn’t be symmetrical. Being below u* should be treated as better than being above it!
For the record, I think Fed Chair Powell is broadly sympathetic to my conclusion. He’s not ignoring his staff and has argued that as the economy closes in on full capacity, it makes sense for interest rates to slowly climb back to their normal perch. But his comments show an awareness of the benefits of full employment, a humility about our ignorance of key parameters and processes, and the asymmetric costs of brake slamming vs. brake tapping.
In fact, given the recent relative gains of minorities, the steady job growth, and the need for faster real wage gains among middle-wage workers, it’s more important than ever to sustain our favorable labor market conditions for as long as possible. We shouldn’t ignore the macroeconomic risks of very low unemployment. But until those risks are clearer and better understood, the bar to sacrificing even a smidgen of the benefits of full employment should be extremely high.