Scott Sumner  

Murder on the Orient Express

Did Republicans Sabotage Obama... Ideological Turing Test: Ca...

Here's Bloomberg:

President Donald Trump hailed his appointment of Jerome Powell to be the next Federal Reserve chair, citing Powell's "considerable talent and experience." Given the realistic alternatives, Powell was one of Trump's better appointments.

If the economy goes south in the next three years, however, Trump's tune will change. He'd surely lash out at Powell and the Fed with vitriol.

Let's put aside the question of whether it makes sense for a person to blame one of their own appointees for a policy failure. I'm interested in another question; would an economic downturn be the Fed's fault? And if the answer is "maybe", then how would we determine fault?

1. If the downturn in real GDP were not accompanied by a sharp decline in NGDP growth, then it would not be the Fed's fault.

2. If NGDP also fell sharply, then it would be the Fed's fault.

Surely it can't be that simple? Don't we need to consider why there was a decline in NGDP? Actually no. If NGDP growth falls sharply, it's always the Fed's fault.

Here's why people get confused on this issue. Fed policy errors don't happen in a vacuum, they occur when the Fed takes it's eye off the ball, say due to an oil shock, or a financial crisis. These distractions may cause the Fed to lose focus on maintaining adequate NGDP growth (as both shocks did in 2008), triggering a recession. But even in that case the recession would probably be caused by the fall in NGDP growth, not the oil shock or the financial crisis. Real shocks by themselves can cause a mild recession in a large diversified economy like the US, but it happens very rarely.

Consider this analogy. Someone has given you slow acting poison, which may or may not be a fatal dose. Then someone comes up and shoots you in the heart. That gunshot wound is the cause of death, regardless of the poison in your system. In this analogy, the poison is like a real shock and the gunshot wound is like a monetary shock. With sticky wages, a sharp fall in NGDP growth means higher unemployment.

Some people have a very odd belief that a tight money policy instituted during a period when the economy is healthy can cause a recession, but a tight money policy instituted during a period of financial turmoil cannot cause a recession. In the latter case, they see the financial turmoil as the "real cause" of the recession. That's like claiming that a gunshot to the heart will kill a healthy person, but will not kill someone with poison in their system.

A sharp fall in NGDP growth is always contractionary, regardless of what else is going on in the economy. If NGDP falls sharply then the Fed has fired a gun into the body of the economy. That body might or might not also have poison in its system, but if NGDP growth plunges then the Fed fired a gun into the body of the economy. It's that simple.

Screen Shot 2017-11-20 at 11.28.09 AM.png
PS. Stephen Williamson had a good post on the choice of Powell for Fed chair. My first choice was Goushi Kataoka; I'll explain why in a future post.

Comments and Sharing

COMMENTS (9 to date)
bill writes:

I think that under their current target, the Fed would disagree with #2. For instance, if RGDP fell 1% and inflation was 2.5%, then we'd have NGDP growth of 1.5% and a recession. I would agree that NGDP growth fell sharply (from current 4% to 1.5%) and that the Fed is at fault. But I think they would say, "Our target is 2% inflation. We even provided a counter-cyclical inflation boost to 2.5% which helped mitigate the recession. We have done our job." They are fighting against NGDP targeting because they really don't want to have to target even a 4% inflation rate if RGDP growth fell to 0%.

bill writes:

PS - Obviously 2008/2009 was a total debacle with both falling RGDP and falling prices. Yet I still don't think the Fed has admitted much, if any, fault then either.

Ben writes:

The problem I see with this is you seem to assume that the Fed possesses perfect information about NGDP as it's happening.

All economic data is delayed. In 2008, for example, we didn't know NGDP was falling until at least a few months afterwards (I can't remember precisely as I was in college). Not to mention the huge corrections made to GDP figures months after they're initially published.

Their reaction to a shock can be improved sure, but it's not realistic to suggest that a central bank can completely negate all negative shocks as they are not some sort of omniscient entity.

bill writes:

Mor timely data would certainly help. But looking at the PCE and CPI numbers available to them at the time, the Fed really messed up in 2008. The later revisions changed their mistakes from bad to truly awful.

Scott Sumner writes:

Bill, Yes, the Fed would disagree---but they'd be wrong.

Ben, The problem with your argument is that you are assuming that shocks are exogenous. But the decline in NGDP was caused by bad monetary policy. Under a NGDPLT regime there would not have been the sharp fall in NGDP during 2008 (which I agree began before the Fed was aware of it, due to lags.) NGDPLT helps prevent the shocks from occurring. It stabilizes expectations. A near term fall in NGDP is caused by expectations of 2 year forward NGDP falling sharply.

Also, they should not be reacting to past data, but rather to forecasts of future NGDP, which were falling sharply during 2008.

bill writes:

Back in 2008/2009, when Bernanke was calling for help from the fiscal side, do you think that was because he agreed with you and me (that the falling NGDP was a Fed mistake) and he realized that he couldn't get the votes to rectify it? Or was that an additional, uncorrelated mistake?

Bob Murphy writes:

Scott, you wrote: "Real shocks by themselves can cause a mild recession in a large diversified economy like the US, but it happens very rarely."

If you have the time--either here in the comments are ideally in a future post--I would love to see you elaborate on that. I mean, there is a whole school of thought, including Nobel laureates, that thinks businesses cycles *can* be explained by real factors. Are you just flatly saying they are obviously wrong, or are you saying that even RBC people ultimately rely on sticky wages / inadequate demand but think a "real shock" is the thing that gets the ball rolling?

Scott Sumner writes:

Bill, I do think he felt it was at least partly a Fed mistake, and that the Fed was not ready to do the sort of really dramatic policy regime change that would be needed to fully address the problem.

Scott Sumner writes:

Bob, I'm saying they are flat out wrong. And there are far more Nobel Laureates that agree with me on this issue.

Comments for this entry have been closed
Return to top