Scott Sumner  

The Fed can do what King Canute could not

Saving Labor with Disposables... No Solutions, Only Trade-offs:...

Alex Tabarrok has a new podcast on monetary policy, over at Marginal Revolution University. He presents a fairly standard view of the Great Recession, emphasizing how an easy money policy during the early 2000s led to low interest rates and helped to inflate the housing bubble.

My view is different. I don't think the low rates were caused by easy money---rather it was weak investment in the US and increased saving in Asia. And I don't think low rates caused a housing bubble. But I'd like to focus on something else that caught my interest.

At the end of the video, Alex mentions NGDP targeting as a policy that some have suggested might have helped to prevent the Great Recession. He then suggests that it's not clear whether the Fed would have been able to push back against the powerful contractionary forces that were hitting the economy in late 2008. He notes that the monetary base increased rapidly in late 2008, and thus presumably even more rapid increases would have been required to stem the tide of recession. Finally, he suggests that it's not clear whether the Fed could have produced a substantially larger increase in the base (alluding to possible political barriers.)

I believe that's the wrong way to think about the problem. A truly effective NGDP targeting policy, with level targeting, would have required a much smaller increase in the base, as we saw in Australia. That's because the nominal interest rate would probably have stayed well above zero.

People often think about this issue in terms of nautical analogies. The Fed is like a ship captain fighting valiantly against the powerful winds and waves of a storm. I've used that metaphor myself. But in some ways it's more useful to think of the Fed as causing the storm. NGDPLT doesn't give you a stronger engine or rudder, it actually causes the ocean to become placid.

Screen Shot 2017-08-24 at 11.23.16 AM.png
The expectation that NGDP will return to the trend line in the future causes the current drop in velocity and NGDP to be much milder than otherwise.

This relates to the fireman/arsonist metaphors. Economists such as Paul Krugman tend to view the Fed as a fireman, which may or may not have the power to put out a fire (i.e. the unstable free market economy.) In this video, Alex suggests the Fed might have been an arsonist in the early 2000s, and a fireman who failed to contain the resulting conflagration in late 2008. I see the Fed as being broadly neutral in the early 2000s and an arsonist in 2008.

Does a fireman have enough power to prevent a conflagration? Sometimes yes, sometimes no. Does an arsonist have it in their power to prevent a fire? Yes, 100% of the time. They simply need to refrain from lighting the fire. NGDPLT is a way of preventing fires by not lighting them. Recessions are typically caused by sharp declines in expected 2-year forward NGDP. Under NGDPLT, expected 2 year forward NGDP stays on trend. Recessions would still occur on occasion, but they'd be far milder.

Somehow I wandered from nautical to firefighting metaphors, but hopefully you get the point.

PS. The video is entitled "When the Fed does too much". That's not a good title. The question is whether the Fed has the right or wrong policy. When it comes to monetary policy, it's not at all clear what "doing more" actually means. More money? More volatile money supply? More volatile interest rates? More aggregate demand? That sort of title leads to confused thinking about what monetary policy actually does.

HT: Ben Klutsey

Comments and Sharing

COMMENTS (8 to date)
Philo writes:

Another great post, Scott. Keep those metaphors coming!

DMXRoid writes:

I'm curious about this statement:

Recessions are typically caused by sharp declines in expected 2-year forward NGDP.

That's a very "animal spirits", universal explanation of recessions. Do you actually mean "caused by", or just "preceded by"? If the former, could you point me to some support for that position?

Scott Sumner writes:

Thanks Philo,

DMXRoid, In recent macro models, current aggregate demand is strongly linked to, and caused by, changes in future spending. The intuition is as follows:

1. NGDP is composed of real output and prices. The efficient markets hypothesis tells us that changes in current prices depend strongly on changes in expected future prices.

2. Real output is quite sensitive to changes in investment spending. If firms expect future real output to be lower, they will tend to cancel current investment projects.

This is what happened in late 2008, as markets recognized that future NGDP would be lower than previously expected. This led to lower current prices and lower current output

Yes, it's a sort of "animal spirits" explanation, but unlike Keynes I think these changes in animal spirits happen for a reason---unstable monetary policy.

Thaomas writes:

As a long run equiibrium I suppose that NGDP targeting could very well make for "more tranquil seas." But to get there in 2008 Would I think have implied that the Fed being willing to have a much larger balance sheet (and not IOR would have helped). And the only way I see as convincingly showing markets that the Fed was willing to have a larger balance sheet was by actually having one, ie by NOT pre-announcing a time limit or an amounts limit for QE. If there were political limits on the Fed's ability to do that, THOSE are the political forces that have to be overcome to have an NGDP target or even a PL target.

BTW, I endorse your idea that seeing the Fed fail to act vigorously to maintain NGDP growth on target or even PL gorwth on target was depressing (and still is) to "animal spirits."

marcus nunes writes:

If anything, mon pol was too tight in 2001-03, after being too loose in 98-00. Mon pol after 2003 was "correct" (not "too loose"). With Bernanke, in 2006 things started to unravel...

Jose writes:

I keep listening the LT is important as a tool to manage expectations. But if the monetary authority were credible, would it need the extra help from LT? If not a tool to manage expectations, LT seems to me to be some tentative to "rewrite" NGDP history, "keeping things even", when there is nothing to balance. A past negative shock is a past negative shock. Trying to regain some growth seems to me procyclical in a policy that is said to be counter cyclical.

Scott Sumner writes:

Thaomas, That's not an interesting question. Monetary policy will never work if it's ad hoc. You either go into a crisis with the right regime, or you fail.

If they went in to 2008 with NGDPLT, then a big balance sheet would not have been needed.

Jose, That's a debatable point. But I lean toward the view that level targeting would make policy more effective, even if credible.

Bob Murphy writes:

Scott, not trying to troll, just trying to make sure I understand your view.

You wrote in the comments above:

"Yes, it's a sort of "animal spirits" explanation, but unlike Keynes I think these changes in animal spirits happen for a reason---unstable monetary policy."

But you *define* monetary policy in terms of the market's expected path of NGDP, right?

So e.g. you would say, "The reason NGDP growth expectations collapsed in 2008 was because of tight monetary policy, not because of some ill-defined 'animal spirits' deux ex machina like Keynes posits."

Then someone says, "What exactly do you mean by tight monetary policy?"

and you answer

"NGDP growth expectations collapsed."

That's your view, right?

Comments for this entry have been closed
Return to top