An interview I did last fall with Jan Hatzius:

Brad DeLong is a professor of economics at UC Berkeley, where his research focuses on financial crises and 20th century macroeconomics, as well as the political economy of monetary and fiscal policy. He has taught at Harvard University and served as Deputy Assistant Secretary of the Treasury for Economic Policy under the Clinton Administration. Below, he and Goldman Sachs Chief Economist Jan Hatzius discuss risks around liftoff and the structural downshift in rates.

The views stated by Brad DeLong herein are those of the interviewee, and do not necessarily reflect those of Goldman Sachs:

Allison Nathan: Has the US economy recovered from the Great Recession?

Brad DeLong: Yes and no.

It has not recovered from the large loss in productivity and potential output. The Great Recession knocked down our level of output by 8% compared to where we thought it would be now in 2007. We will probably never recover that loss. That is very unusual for the United States, which had a substantial recovery in lost output even after the Great Depression.

But as far as the labor market, we have mostly recovered–but not fully recovered.

Jan Hatzius: I believe that the labor market is mostly recovered and that the overall economy has come a long way toward recovery. To Brad’s point about disappointing output, the question is how much of that is exogenous weakness versus some form of hysteresis, with the effects of the Great Recession weighing progressively on economic activity. It’s difficult to know the answer, but I’ve become more sympathetic to the idea that we underestimated the extent of the exogenous slowdown. We’ve substantially revised our views on potential growth, and I don’t think it is all or even mostly due to the aftereffects of the Great Recession.

Allison Nathan: Has the time come to raise rates?

Jan Hatzius: Given how far the funds rate is below normal, how close the economy is to full employment, and my expectation of gradual increases in wage and price inflation, now seems like an appropriate time to move. But I would say there is still a good case for waiting on risk-management grounds because the future of the economy is uncertain. It seems more dangerous to lift off too early and find that the economy can’t tolerate tighter policy than to end up a bit high on inflation for a while. That said, I feel less strongly about the risk-management case than I did three or six months ago.

Brad DeLong: There are four reasons why it is not yet time to begin normalizing monetary policy.

First, we are only 300bp away from the equilibrium funds rate. Given how close we are, why not just wait until we get to full employment?

Second, there is an unknown amount of slack left, and it might be substantial.

Third, to Jan’s point on risk management, I think there’s a 50% chance that the Federal Reserve will be really sorry that it raised rates when it did. And there is no upside for proceeding with rate hikes now. The Fed could delay hikes another six months and then just raise them faster and still get to the same place. But if it raises rates now, it will have no way to catch up because rates would presumably still be very low.

Finally, we are likely to find ourselves at the zero lower bound sometime in the future again—and when we do, we want market participants to feel very certain that there will be overshooting coming out of it—more inflation and lower real interest rates over the medium term that will boost growth. We need a reputation of coming off of the zero lower bound with a roaring economy, and I believe we need to stay at zero longer for that to happen.

Allison Nathan: Why isn’t the Fed more concerned about the lack of inflation, and should it be?

Brad DeLong: Janet Yellen and Stanley Fischer really believe in their models, which predict that inflation will rise to and above 2% a year in 2017 and 2018. They see this inflation path as a tangible reality, but it is in fact only a shadow cast by their assumptions. The Fed should be much more concerned about model uncertainty and, in turn, the lack of inflation right now.

Jan Hatzius: I have a slightly more positive view on inflation. My approach is “trust but verify,” which Ronald Reagan used to say on arms control negotiations. I trust the inflation models more than Brad does, but I’d like to see more verification in terms of the numbers picking up. A recent pick-up in wage growth is somewhat encouraging; our broad measurement of wage growth has risen to 2.6% yoy—still low but the highest rate we’ve seen in the recovery. We’re only at 1.3% for core PCE, but that is roughly what we expected at the start of the year, not a downside surprise, which would be more worrying.

Allison Nathan: Do you worry about financial imbalances?

Jan Hatzius: Not really. Asset markets don’t look particularly frothy to me today, debt growth is reasonably muted, and the private sector is still running a decent financial surplus of a little over 2% of GDP. So I just don’t see sources of worries over financial imbalances in the United States. But there are certainly other places around the world that show greater cause for concern, China being among them.

Brad DeLong: I generally agree. The aggregate numbers don’t seem to suggest anything like the troubling financial imbalances we have seen in the past. People worried about imbalances today tend to say that their concern centers on who is bearing risk in the economy or the markets, especially given that risk-bearing capacity on Wall Street has declined and that low interest rates have continued to generate a “search for yield.” But we don’t have the data to know how many people are unprepared to bear the risks associated with their positions.

Allison Nathan: Given current excess liquidity, will the Fed be successful in actually lifting the fed funds rate?

Brad DeLong: Yes. It will be very interesting to see what it has to do to be successful. But if the Fed wants the rate to get somewhere, it will get it there.

Jan Hatzius: Agreed.

Allison Nathan: Where will the Fed’s communication from the December meeting leave market expectations for future rate hikes?

Jan Hatzius: Markets will think that January is firmly off the table, and that March will be on the table if the data cooperates. The probability for March now priced into the market is roughly 50%, and my guess is that this will rise, but probably not above 60-70%. How exactly the Fed gets the market there is a bit of a dance. They will emphasize data-dependence, but the idea of March being a real possibility could cause a tightening of financial conditions and set expectations for hiking at every meeting, which the Fed wants to avoid. But I think they’ll find a way to manage this.

Brad DeLong: I agree they will emphasize data dependence and the need to assess the effects of the first hike, which is likely to push expectations to March at the earliest.

Allison Nathan: In the last few rate-hike cycles, the fed funds rate rose faster and ended up higher than the Fed initially projected. Will this time be different?

Brad DeLong: This time will be different, because in all recent hiking cycles, the Fed started out well behind the curve.

In the mid-2000s, the Fed was unhappy that its short-term rate increases were having so little traction on the long end of the bond curve, which led them to hike more rapidly than they initially intended.

In 1994, the tightening cycle began after Alan Greenspan had cut Bill Clinton slack for an entire year to get deficit reduction accomplished, so Greenspan was very eager to start raising interest rates once the Fed actually began to hike.

In 1989, Greenspan had delayed hikes out of fear that stock market crash in October 1987 would cause a recession. It didn’t, so the Fed ended up needing to catch up to where it thought it should be.

And before then, Paul Volcker definitely believed that the Fed was far behind the curve at the end of the 1970s when he became Fed Chair.

Given this pattern of the past four major tightening cycles, this time really is different.

Jan Hatzius: This time should be different, but the market is priced for something too different. If the recent trends continue, we will be at full employment by the end next year and inflation and wage growth will also likely be higher. In that environment, I don’t see the Fed taking a six-month break. I see them hiking a quarter-point every quarter, which is twice as much as what the market currently expects.

Allison Nathan: What’s the risk that the Fed will need to return to more accommodative policy?

Jan Hatzius: I’d give it a roughly 15% probability, which is not insignificant and still a good reason to delay, or at least to go more slowly and be very responsive to new information about the economy, especially in the early days of the normalization process. You could call that a second-best approach to delaying liftoff, but I do think that is where they are now. I see the odds of a shallower path than projected by the dots, but not an outright reversal, as higher, roughly in the 30% range.

Brad DeLong: I think there’s a greater than 50% chance that the rate path will be shallower than the current dots. I see a roughly 50% chance the Fed will end up wishing that they had stuck at zero or at least hiked even more gradually than what the market is currently expecting, but I am not sure they would actually dare reverse course. I could see them staying at 1% for a while, wishing they hadn’t hiked but not daring to go back.

Allison Nathan: Will the ECB and BOJ be less likely to ease further once the Fed lifts off?

Jan Hatzius: The extent to which the Fed, ECB, and BOJ are driven by each other gets very overplayed in the market. These are all economies with flexible exchange rates, and foreign monetary policy decisions almost always have offsetting effects on their own desires to move in one direction or another. For example, if the Fed tightens and that leads to a stronger dollar and a weaker US economy, the implications for the ECB are pretty ambiguous. So I don’t see a big spillover, and it is very hard to prove any such spillover empirically.

Brad DeLong: I agree that internal monetary politics are overwhelmingly primary in both the Euro area and Japan. What the Fed does is only very small noise compared to their focus on their own political-economic configuration. The only places in the developed world where monetary policy is tightly linked to the Fed are Britain and Canada.

Allison Nathan: Where will rates end up in this cycle?

Brad DeLong: There has been a structural downshift in rates. A decade ago, we wondered if rates would end up at 6%; a decade before that, it was 8%. Now, virtually everyone would be surprised if they ended above 4%. I’m more pessimistic than most in that I see them most likely to end at 3% or below, with a fairly small chance they’ll end up higher and a one-in-three chance that in retrospect we won’t really see this as a tightening cycle at all because rates will be in the 0-1% range three or four years down the road.

The primary factor behind my relatively pessimistic view is people’s loss of confidence in the capacity to bear and understand risk, and in the idea that large downward movements in asset prices are once-in-a-generation episodes like the Great Depression. In the 2000s, we suddenly had three episodes: the bursting of the dotcom bubble, the US real estate collapse, and the 2008-2009 equity and risky debt market crash, with the most sophisticated players blindsided by one or two if not all three of those events. This eroded confidence, widening the risk spreads between safe and risky assets.

Jan Hatzius: I forecast the terminal rate at 1-2% in real terms and 3-4% in nominal terms, which puts me on the slightly more optimistic side of this debate. In my view, the labor market is a much better indicator of cyclical progress than real GDP, especially in an environment where potential growth has slowed. And I am struck by the amount of labor market improvement we’ve seen. So I do think rates will end substantially higher from here.