Scott Sumner  

Low interest rates do not cause low inflation, or high inflation

Third Parties in Congress: Pro... Richard McKenzie on Gordon Tul...

I often argue that people should never reason from a price change. Interest rates are also a price. Here Noah Smith gets into trouble by considering the impact of a change in interest rates:

But what if QE had the opposite of the intended effect? That is the claim of a small but well-credentialed group of macroeconomists that I once labeled the "Neo-Fisherites," after the famous monetary economist Irving Fisher. These economists wonder if quantitative easing reduced inflation, instead of increasing it as many feared it would. The Neo-Fisherites go even further than that -- they wonder if low interest rates, which we usually think of as being inflationary, are actually deflationary!
There are two big problems with this quotation. First, the idea that QE might have reduced inflation is not an interesting hypothesis. We know that QE slightly increased inflation. What economists need to do is figure out why QE slightly increased inflation, and by how much. Raising it as a question gives uninformed readers the false idea that there is serious debate about whether QE raised or lowered the price level.

More importantly, this quotation gives readers the impression that standard macro theory tells us that low interest rates are inflationary. This is not true. Low rates caused by expansionary monetary policies are inflationary, and low rates caused by contractionary monetary policies are deflationary. But in that case it's a waste of time to even talk about the impact of interest rates on inflation. Always focus on the effect of the thing that caused interest rates to change.

Comments and Sharing

CATEGORIES: Monetary Policy

COMMENTS (10 to date)
Andrew_FL writes:

Low is a relative term (putting aside, in any case, the common confusion of real and nominal rates) so rates are low relative to what?

One thing I disagree with other self styled Austrians on, is their apparent implicit belief that they know what the Wicksellian natural rate should be at any given time, and therefore know when it is too low or too high. But that can't be true, or else low rates wouldn't fool anyone into making malinvestments.

Of course, this is why it makes much more sense to focus on MV (NGDP). If MV is constant, investment should equate to voluntary savings by definition. This was the source of Hayek's preferred policy norm, at least until practical considerations made him retreat to stable P (in The Denationalization of Money). But these are already familiar facts to fans of both Hayek and the conceptual framework of NGDP targeting.

Of course, there are also people like me who believe a laissez faire system could achieve this result "automatically." You could call that targeting or monetary policy but I'd find it an odd use of either word.

Pioterk writes:

How can you prove that QE slightly increased inflation?
Is it inflation as measured by CPI, SP500, Case-Shiller, wages?
Do we have a control subset of reality that did not have QE and shows lower inflation?

Scott Sumner writes:

Andrew, Good points.

Pioterk, There are all sorts of market indicators that respond to QE announcements as if the policy is inflationary. Not just in the US, but other countries as well. I don't think there is any controversy on that point. To take just one example, currencies fall sharply in forex markets on unexpected news of QE policies.

Yancey Ward writes:

"Never reason from a price change unless it is a price change you want to reason from", is what this sounds like to me.

Noah Smith 'isn't even wrong'. But he's in a long standing tradition of confusing himself by talking about monetary policy in terms of interest rates.

It wouldn't be worth repeating this, yet again, to Scott, if we hadn't just been discussing this very problem with David Henderson, beginning with the post on Jeffrey Rogers Hummel's paper. As I pointed out with the example of Walter Heller in 1968; interest rates are not the price of money.

Here's Noah Smith's almost pure example of that fallacy (that never dies);

Most monetary economists -- and most people who watch the news -- think that if the Federal Reserve pushes r^nominal down, by printing money and buying bonds, that inflation goes up and r^real goes down by even more.

Hunter writes:

Wrote this on Arnold Kling's page. It works here too.

I prefer tautology. Inflation is when money is a liability. Deflation is when money is an asset. Zero is when money is a store of value. The circumstances under which any of these states occur will vary.

RPLong writes:

I agree with this blog post, except the last sentence, and Yancey Ward neatly articulates my objection to that last sentence.

Scott Sumner writes:

RPLong, I don't see any problem with the last sentence, and I have no idea what Yancey is talking about so I can't comment on that.

RPLong writes:
Always focus on (1) the effect of (2) the thing that caused (3) interest rates to change.

Focusing on (1) the effect of (2) the cause of (3) an effect sounds a lot like reasoning from a price change to me, especially when the second sentence in your blog post defines (3) to be a price change.

I can understand if you didn't mean to say this, but it kind of gives that impression. Perhaps I am nitpicking in this case, because I agree with the rest of the post.

Michael Byrned writes:

A price change can happen for any number of reasons. Thus, the price change itself is not evidence of what caused it.

Comments for this entry have been closed
Return to top