Right, so former Minneapolis Fed chief Narayana Kocherlakota is fun.
This is a guy who, if you ask him, will tell you the story about his dramatic transformation from hawk to dove.
According to Kocherlakota, that shift was inspired by a realization that the real world is different from academia. See, it turns out that this thing called “unemployment” that Kocherlakota used to read about in his economics books is real. That is, there are actually people who don’t have jobs. Joblessness isn’t a hypothetical.
So when Kocherlakota became a policymaker, he had to come to terms with the fact that if he’s a hawk and he’s wrong, it’s possible he could literally be keeping people from having jobs.
Kocherlakota didn’t have the stomach for that.
“The incentives in academia are to create a set of ideas and stick to them. You aren’t responsible for making decisions that affect everyday lives,” he told WSJ back in 2015. “I was just wrong,” he added.
Well, if you agree that Kocherlakota was “just wrong” to be hawkish, don’t say he didn’t try to remedy the situation because goddammit he floated some pretty silly ideas of the dovish kind before putting himself out to pasture and abdicating to “crazy eyez” Neel Kashkari (Curb Your Enthusiasm fans will appreciate the reference).
Anyway, Goldman is out with a new interview featuring Jan Hatzius and Kocherlakota (moderated by Allison Nathan).
Unsurprisingly, Kocherlakota thinks the Fed is on its way to making a policy mistake by hiking into lackluster inflation prints. This is, in some ways, an extension of what we wrote on Saturday in “A Contrarian View On The Next Recession Consistent With Historical Precedent.”
Read the full interview below…
Allison Nathan: Despite impressive improvement in the US labor market, inflation has disappointed for the last three months. How concerned are you about this weakness in inflation?
Narayana Kocherlakota: I see low inflation as less of a disappointment and more of an opportunity. We have seen improvement in the labor market, but I think there’s room for even more. The number of people in the prime of their working life—age 25 to 54—who have jobs has steadily risen in the last three or four years but is still lower than at the beginning of the Great Recession. In general, the low inflation numbers tell us that there is productive capacity left to be exploited in the US economy even though the unemployment rate is already so low. The Fed should be taking advantage of the fact that we have a surprising amount of slack left in the labor market by keeping policy accommodative.
Jan Hatzius: I agree that the key issue is whether there is still productive capacity left to be utilized. But I am skeptical that we can answer that question from the behavior of core inflation alone, because it is too loosely related to economic slack. Instead, I would consider a broader range of indicators. Some of them—for example, core price inflation or the hiring rate— suggest we are still somewhat below full employment. Others—for example, the headline unemployment rate or the proportion of people who say that jobs are plentiful rather than hard to get—suggest that we’re actually beyond full employment. If I look at all of them together, I think we are roughly at full employment. So how concerned am I about the low inflation readings? I think they are a reason to keep the normalization of monetary policy gradual, but for me they are only one piece of the puzzle. The other is the faster-than-expected improvement in the labor market this year, which says that the Fed should continue to normalize policy.
Allison Nathan: What do you make of the large impact that methodological changes, such as a recent adjustment to the cellphone category to better account for unlimited data plans, have had on inflation data? How should the Fed deal with these types of changes and, more broadly, generally flawed inflation measures?
Narayana Kocherlakota: As long as there is some independent measure of prices, the Fed is in good shape. Over the short run, there are obviously deviations in various inflation measures in addition to occasional one-off changes like the cellphone issue in March. But we would expect different inflation measures to converge over time, and any one-off changes shouldn’t materially affect one’s inflation forecast. A problem only arises when methodological changes affect the rate of inflation over very long periods, which the Fed must ultimately be held accountable for and address. I do think having one target inflation rate—in the Fed’s case, PCE—helps provide this accountability, even if no single measure of inflation is perfect. However, I personally would prefer for the Fed to use CPI as opposed to PCE because I think most Americans are more familiar with the CPI inflation rate.
Jan Hatzius: The Bureau of Labor Statistics and other agencies should try to quality-adjust price indices as best they can, and I see more room to do this in a variety of areas such as healthcare and technology. That said, quality adjustment does introduce inconsistencies over time. Even if these are one-off changes, they matter; the cellphone adjustment, for example, will hold down inflation by nearly 0.2pp until March 2018. These issues just reinforce the idea that it’s best to avoid putting too much weight on core inflation alone because it is only one part of the picture. No single indicator is perfect, so you tend to be better off taking a broader perspective in your attempt to assess where the business cycle stands.
Allison Nathan: What will generate sustainable inflation pressure that will allow the Fed to reach and maintain its 2% inflation target?
Jan Hatzius: First, the recent one-off effects, especially in cell phone services, will drop out of the numbers next year, which should produce a mechanical lift. Second, we expect a reduced impact from past dollar appreciation, which has been holding down core goods prices. And third, stronger wage growth should lead to a gradual increase in inflation in the services sector, where wages flow through more directly to prices than in other parts of the economy. While the latest wage numbers have been a little disappointing, wage growth has trended higher in recent years, consistent with the idea that slack has declined substantially.
Narayana Kocherlakota: Over the long haul, we’re going to have to see sufficient wage pressure. I feel very confident that if unemployment stays low and continues to fall, we’re going to see more wage inflation. I feel less confident about how companies will react. Will they take higher wages out of profits or will they pass them onto consumers? Given that profits are still high by historical standards, there is good reason to expect that companies will be willing to absorb higher wages in the form of lower margins. As long as that’s true, I think it’s going to be hard to get back to 2% inflation without much higher wage growth than we’ve been seeing.
Allison Nathan: If the labor market begins overheating but inflation remains below target, what should the Fed do?
Narayana Kocherlakota: As long as inflation remains under control, I find it hard to think about the labor market overheating, per se. Margins might fall as wage growth rises, but I don’t think the Fed should be raising rates because of that. Of course, if unemployment falls sufficiently far below where we think it’s going to be in the long run, it’s going to have to come back up at some point, and you might worry about the Fed’s ability to manage that process. But it’s difficult for me to imagine the unemployment rate falling much from here—and certainly not much below 4%—especially given the potential for further increases in labor force participation. So as long as inflation stays low and we expect it to stay low—the Fed staff forecasts inflation staying below 2% even into 2019— I don’t see any reason to be worried, regardless of what’s happening on the labor-market front.
Jan Hatzius: Labor-market overheating needs to be visible in something other than just a declining unemployment rate, or else it isn’t really overheating. But, again, I don’t think we can just rely on current price inflation to make that assessment. And based on what a number of other indicators are signaling about the labor market right now, I worry that further gains, relative to potential, will push us beyond the sustainable level of employment. And once we are beyond that level, a return to it will probably require an increase in the unemployment rate.
The problem is that such an increase is hard to pull off without a recession. We’ve found that in the post-war period, an increase in the unemployment rate by more than 0.35pp has always been associated with a recession. I would prefer to see gradual, steady employment gains that have a better chance of being sustained than employment gains that average 200,000 for a couple more years but are then followed by recession. So I think the Fed needs to stay the course of gradual tightening even if inflation remains muted, which was the approach reflected in last week’s FOMC statement.
Narayana Kocherlakota: Recessions often get lumped together, but some are quite mild. If we’re talking about a recession like 2000-2001, I’d pay that price anytime. But if it’s a severe recession like 2008-2009, then I would agree with Jan’s point.
Jan Hatzius: Those two recessions span the range that we’ve seen in the post-war period. But even if you are at the less severe end of that range—a 2pp increase in the unemployment rate—we’re still talking about millions of people losing their jobs. The typical recession actually leads to a roughly 3-3.5pp increase in the unemployment rate, which is not as severe as 2008-2009 territory, but is still a risk I wouldn’t want to take.
Allison Nathan: Jan, recession risk has historically often been associated with the Fed needing to tighten policy more aggressively once inflation picks up. How relevant is that precedent today given that inflation has been below target for so long?
Jan Hatzius: Even in the more recent 20-30-year period in which inflation has been quite low and stable, increases in the unemployment rate have always coincided with recession. Of course, the sample size is small just because there aren’t that many business cycles within that period. But this tendency also basically holds in other G10 economies, where we have found that an increase in the unemployment rate of at least 0.35pp was associated with recession 80% of the time. And if the unemployment rate increased from a level at or below estimates of the structural unemployment rate, the odds of recession increase to about 90%. While there are potential objections to this sort of analysis, such as the uncertainty of structural unemployment estimates, evidence in the US and internationally reinforces the view that soft landings from below are difficult to achieve.
Allison Nathan: Jan, what would have to happen on the inflation side to convince you to change your Fed call?
Jan Hatzius: Our forecast for core PCE inflation by the end of the year is 1.7%; if that disappointed by a couple of tenths of a percentage point, i.e., to 1.5%, then I’d probably want to take out 1-2 hikes through the end of 2018. That said, it does depend on other factors. In particular, greater-than-expected labor-market improvement could offset low inflation.
Allison Nathan: Narayana, if inflation fails to accelerate from here, is there anything that would convince you that the Fed should get more hawkish?
Narayana Kocherlakota: No.
The Fed has two mandates: price stability at 2% and maximum employment. If inflation is running below 2%, that calls for more accommodation in order to bring inflation back up to 2%. This could also lead to more job creation, which, again, I would view as healthy. And remaining accommodative is also important for inflation expectations, which have stabilized somewhat but remain meaningfully lower than they were three years ago. The Fed has to seriously consider the implications of persistently low inflation for the credibility of its 2% target.