Scott Sumner  

Never reason from an interest rate change

PRINT
Krugman's Misleading Post on C... Arnold Kling on Well-Being Ove...

I occasionally point out that the European Central Bank raised interest rates several times in 2011, triggering a double dip recession. At other times I argue that interest rates are not a reliable indicator of the stance of monetary policy. So which is it?

Both. Ben Bernanke and I believe that the only reliable indicator of the stance of monetary policy is economic aggregates such as inflation and NGDP growth. (I favor NGDP.) But we live in a world where most people don't accept this view, and cling to faulty indicators like interest rates and money supply growth. So when Keynesians deny that tight money could have caused the double dip recession in the eurozone, because Europe was at the zero rate bound, I like to point out that monetary policy tightened in 2011 even according to the preferred Keynesian indicator.

It just so happens in that in 2011 rising interest rates reflected the "liquidity effect" of tight money, and hence were a good indicator of tight money at that moment. But our textbooks also tell us that the long run effect of tight money is actually lower nominal interest rates, due to the income effect, the Fisher effect and the price level effect. For instance, the Fed raised interest rates to relatively high levels in late 1929, and then cut rates sharply over the next 3 1/2 years. But no one would claim that money was easy during 1929-33 because of falling interest rates, after all, NGDP fell nearly in half, and monetary aggregates like M1 and M2 also plunged sharply.

Here's Simon Wren-Lewis:

The ECB raised rates from 1% to 1.5% in 2011, and compared to the US there was no Quantitative Easing. Combining the two, monetary conditions tightened considerably in the periphery as a result of the crisis. There was no comparable crisis in the US, which allowed investment to increase by 5.5%.

Explaining 2013 seems more difficult. Monetary policy had eased in the EZ (although of course not by as much as it should have), and OMT had brought the crisis to an end. In the US there was considerable fiscal tightening. So why did the US continue to grow and EZ GDP continue to fall?


Actually, explaining 2013 is not more difficult. Monetary policy remained tight, no easing occurred. The fall in interest rates represented the long run effects of the tightening of 2011, just as the fall in rates during the early 1930s reflected the long run effects of a tight money policy adopted in late 1929.

If I was a Keynesian who looked at monetary policy through the lens of interest rates I'd be very sympathetic to those who scoff at the omnipotence of the Fed. I'd see the Fed as playing at most a bit part in the grand macroeconomic drama. And that's because nominal interest rates tend to be highly procyclical. They tend to rise during booms and fall during recessions, even before the Fed existed. If your view of the world equates low rates with easy money, then of course you won't see monetary policy as playing a decisive role in long drawn out recessions. Perhaps the recession is triggered by a rise in rates, but then rates quickly fall, making monetary policy look ineffective if there is no immediate recovery.

One criticism of the "falling NGDP means tight money" view is that it seems non-refutable. If NGDP falls then market monetarists can simply blame the Fed. And if someone claims the Fed did all sorts of monetary stimulus, we can argue that they must not have done enough. So what would be a reliable indicator of the stance of monetary policy that is definitely controllable by the Fed?

Create a NGDP futures market and use NGDP futures prices as the policy indicator.

It's also worth noting that the Fed itself insists that it's never out of ammo. When NGDP falls they do not say "we've done all we could." The say "we could have done more to boost AD, but it would have been unwise."

The ECB interest rate increases of 2011, as well as the 2000 and 2006 interest rate increases in Japan, as well as the tapering currently taking place in the US, are very important because they show central bank intent. It's pretty hard to argue that a central bank is incapable of boosting AD when it is engaged in policies explicitly intended to slow the growth rate of AD. This means that many of the real world cases that are widely assumed to show central bank impotence, do nothing of the kind.

Mark Sadowski has an excellent critique of many of the specific empirical claims made in the Wren-Lewis post.

Update: Giles Wilkes has an elegantly written post on the differences between the way that Keynesians and monetarists think about aggregate nominal spending.


Comments and Sharing





COMMENTS (10 to date)

Martin Weale and Tomasz Wieladek, say that announcing makes it so;

For the UK, asset purchase announcements have an impact on interest rate futures in the UK and measures of financial market uncertainty, suggesting that ‘signalling’ is an important transmission channel. For the US, only long-term yields and the real exchange rate react to asset purchase shocks, which implies a relatively greater role for the portfolio rebalancing channel. But asset purchases in both countries affect measures of financial-market uncertainty, such as the VIX. This suggests that asset-purchase announcements may also provide economic stimulus through a reduction in uncertainty, which is a channel that has not received much attention in previous work.

Something Scott has been claiming for years.

Scott Sumner writes:

Thanks Patrick, that's very interesting.

Tom writes:

Great post,I'm a former student of yours and a frequent reader. I work in financial advising now and one of the issues that I am constantly having to defend is the roll of interest rates on the broader economy. Its an up hill battle trying to get them to understand the variability with interest rates and why they are such poor indicators of economic conditions. Please excuse the poor writing, I'm sending this from my phone.

P.s. I would love to see PK defend his liquidity trap argument in a debate after we see how Europe does this year... I'm sure he'll think of some excuse as to why there will be a stronger recover outside of monetary policy.

Scott Sumner writes:

Tom, Thanks for the comment. Yes, it's a constant battle to deal with the perception that interest rates=monetary policy.

Justin Rietz writes:

New Keynesians, I believe, wouldn't view a low interest rate as necessarily indicative of loose monetary policy. Instead, it's the commitment to keep the future rate low, i.e. below what the policy rule would recommend at that future date. Under the NK model, this generates expectations of higher future inflation, which then impacts the current output gap in a positive manner.

The problem, as you have alluded to, is that without a credible commitment (such as a NGDP level target), this won't work. As a professor described it to me, you need to be able to tie Odysseus to the mast.

dannyb2b writes:

Is fed credibility undermined by the fact that after reaching the zlb the fed needs to create ever more MB to get an effect on ngdp? Could this be why the fed is reluctant to do higher levels of QE? Could the loss of credibility as the fed expands MB with little result on ngdp be what limits the effects of increases in MB on expectations? Lower fed credibility means lower change in expectations from increases in MB?

Arent all OMO's and QE operations non permanent increases in mb? If the fed buys assets like treasuries which draw money back out of the system from coupon and principle payments then doesn't that mean an OMO is not permanent in the long run?

Scott Sumner writes:

Justin, I'm told that new Keynesians consider easy money to be an interest rate that is below the Wicksellian equilibrium rate. In 2008-09 the interest rate was clearly above the Wicksellian equilibrium rate. So why did new Keynesian economists think money was easy?

danny, You asked:

"Is fed credibility undermined by the fact that after reaching the zlb the fed needs to create ever more MB to get an effect on ngdp?"

It could be, if that were true. Fortunately it does not need to create more base money, merely to do forward guidance.

dannyb2b writes:

"Fortunately it does not need to create more base money, merely to do forward guidance."

But is it the case that expectations are formed based on the central bank's ability to actually affect NGDP? Even if forward guidance clarifies the future path of policy it may not budge expectations because people think the fed is incapable of reaching said targets. The fed may be unable to reach its targets if people are deleveraging at the ZLB. IOR and negative rates may not make more private lending happen if debt to gdp is at excessive levels.

Mike Rulle writes:

I have never seen you admit the potential "non-refutability" of NGDP before----or at least as a valid critique. It was one of the very first things I wrote on the Money Illusion site, which you rebuffed (I think).

This does not make it wrong per se----but it gets pretty close!

I would like to see a futures market---but we do not have one---nor do we know if it would trade.

I like to believe you are right---you are passionately persuasive.

Justin Rietz writes:

First, I should say this may just be a case of differences in semantics, or my lack of understanding the New Keynesian position.

I don't know if all NKers were saying that monetary policy was easy in 2008-2009, but it could be. I imagine it would depend on what were their inflation/deflation and output gap expectations at the time. And at the zero lower bound, the nominal rate can be zero and the real/Wicksellian rate can still be positive, with a magnitude dependent on the level of deflation. Of course, under an NK model, that's when the central bank has to make long-run future commitments to keep rates relatively low or some other unconventional policy.

Also, to be fair, I know respected NKers who see NK models as one tool in their tool box, and a tool that may not work well as we approach the ZLB. There are definitely disagreements about the different NK models that have sprung up from the textbook model. Many understand that QE activity to boost the money supply and therefore inflation expectations can be effective, but they prefer to use that at the ZLB. I admit, it's a bit of a dichotic stance, and it's why I'm not thrilled with NK models in general.

Comments for this entry have been closed
Return to top