Scott Sumner  

The interest rate fallacy

PRINT
Reply to Kocherlakota... Honor Laborers...

There is a common fallacy that lower interest rates are expansionary. Where does this fallacy come from? I can think of two true facts, each of which contributes to this fallacy.

1. The true fact that if I am thinking about building a new house, a lower interest rate might make me more inclined to do so. Of course a lower rate would also make my neighbor less likely to lend me the money to build a new house, even if done with a bank as an intermediary. Never reason from a price change.

2. The true fact that central banks tend to lower their target interest rate, when they adopt an expansionary monetary policy. However in that case the lower rate itself is contractionary, whereas the thing that caused the lower rate is even more expansionary than the lower rate is contractionary.

Here's Tyler Cowen:

True, if we ratcheted everything up in nominal terms with a four percent inflation target, maybe during the next recession the Fed could, by moving to a near-zero nominal rate, make short-term real rates negative three or negative four instead of a mere negative 1.5, or however the numbers would work out exactly.

Is that going to prove so beneficial? I wonder. I get that raising rates in bad times is harmful because of its contractionary impact. I am far more skeptical that a short-term real rate of negative four is much more useful than a real rate of negative two. How many investors have said "I won't borrow at negative two but I will at negative four!" Maybe some, but I wonder. Just try to square that with a rational theory of private sector hurdle rates.

Keep in mind also that the Fisher effect does not work with any kind of one-to-one offset. So raising the inflation target by two percentage points probably would not mean that nominal interest rates go up by two percentage points. Nominal rates for instance might go up by only one percentage point. And so when the next recession would come, there would still be room for a greater cut in nominal rates, but not by nearly as much as the boost in the inflation target.


A few comments:

1. I agree with Tyler that a 4% inflation target is not a good idea. However I do think it's better than the actual monetary policy of the past decade, just not as good as alternatives.

2. Room to cut rates does not depend on the actual level of interest rates; it depends on the Wicksellian equilibrium rate. Thus in the spring of 2011 the ECB raised actual interest rates, but this reduced the eurozone Wicksellian equilibrium rate. Hence the ECB actually ended up with less room to cut rates than before.

3. Unlike Tyler, I believe a 2% rise in the inflation target would raise the Wicksellian equilibrium rate by 2%, and give central banks 2% more room to cut rates. (Of course keep in mind that what really matters is NGDP growth, not inflation, which is why Australia never hit the zero bound.)

4. I don't agree that a 2% cut would have little impact on people's inclination to invest, but much more importantly that claim is totally irrelevant even if true. Interest rate cuts do not cause the economy to expand because they boost investment. Indeed interest rate cuts are not expansionary at all. Rather the thing that causes interest rate cuts is expansionary, and it has nothing to do with investment. People need to stop thinking in terms of the Keynesian worldview.

The equation of exchange will help to clarify this issue:

M*V = P*Y

Prior to 2008, when people talked about the Fed cutting interest rates, they meant the Fed increasing the money supply, and the fall in interest rates was a side effect---dubbed the "liquidity effect". The reduction in interest rates reduces the opportunity cost of holding base money, and hence increases base money demand and reduces velocity. But the monetary base itself increases by more than V declines, and hence NGDP (PY) increases.

The increase in NGDP occurs due to the hot potato effect. The public has more base money than they want to hold, and in attempting to get rid of these excess cash balances they spend them of goods, services and assets, pushing up NGDP.

Even if the measly extra two percent makes little difference in terms of investment, as Tyler suggests might be the case, it has no bearing on this transmission mechanism. That's because the hot potato effect is very sensitive to whether you are at the zero bound or not. If you are at the zero bound, bank demand for base money soars much higher, and velocity plunges. In contrast, if you are just a measly 2% above the zero bound, then bank demand for base money plunges, to extremely low levels---basically required reserves plus a minuscule amount of excess reserves. That's the pre-2008 world of central banking.

In that "conventional" world, even a small injection of new reserves into the banking system is more than they want to hold, so banks start rearranging their affairs in such a way that the excess reserves flow out into cash in circulation. But the public also faces a hot potato effect, and in an attempt to get rid of this cash, NGDP rises. Even if that extra two percent matters little for investment, it matters a great deal for the hot potato effect.

Actually it's a bit more complicated, as I've described the long run effect of a monetary injection. In the short run, the hot potato effect is restrained by a rise in asset prices (a fall in interest rates), which temporarily holds the hot potato effect at bay. But the expectation of the future hot potato effect tends to drive up aggregate demand today, and thus it is immediately expansionary.

But not because of lower interest rates, rather despite lower interest rates.

In a world without interest rates restraining the hot potato effect, prices would rise almost immediately. Imagine a cruise ship wrecks on a tropical island, with 78 survivors and a Monopoly game that washed up on the beach. They agree to use the Monopoly money as a medium of exchange, despite the fact that a survivor named "Mike" warns them that the Monopoly money is not backed. There is no financial system, the money is used to buy and sell commodities like fish and coconuts. Three weeks later, a second Monopoly game is discovered washed up on the beach. The money supply has doubled. Without interest rates to restrain the rise in prices, NGDP doubles almost immediately.

The key thing to keep in mind is that lower nominal interest rates reduce the opportunity cost of holding base money, and thus are contractionary for any given money stock. Under the gold standard, the quantity of gold was relatively stable in the short run. Barsky and Summers (1988) showed that lower interest rates were deflationary during this period, just as the basic monetarist model predicts. And they aren't even monetarists. This may be hard for you to believe, but this post is just mainstream macroeconomic principles. (A paper by Lee and Petruzzi (1986) independently reached similar conclusions to Barsky and Summers.)

PS. With interest on reserves, the preceding story becomes more complicated, but nothing fundamental changes. Now the op. cost of holding base money is market interest rates minus IOR (for banks) and it is still the nominal rate for cash held by the public.

PPS. Any comment containing the term 'endogenous' will be ignored. Life's too short.

PPPS. On Wednesday September 7th, I will be presenting a paper at a joint Mercatus-Cato conference on monetary policy rules. The conference will be livestreamed at mercatus.org/live.


Comments and Sharing






COMMENTS (10 to date)
Mark Barbieri writes:

What are you suggesting that the Fed do? Are you suggesting that the fed increase the money supply by increasing open market purchases of securities?

bill writes:

LOL on endogenous.
Last week a commenter at another blog disparaged me with a comment that included the word exogenous.
:- )

Nick Rowe writes:

Yep. Here's the way I think about it, which (I think) is very similar to your way.

Consolidate the Fed and the commercial banks into One Big Bank, that lends to the public at an interest rate it sets. Just to keep it simple.

When the Bank wants to increase the money supply, it sets a lower interest rate. People borrow more money, not (necessarily) because they plan to hold more money, but because they plan to spend the money they borrow. In aggregate, they are surprised to find that the money they spend with one hand comes straight back to their other hand. So the hot potato starts rolling (sorry, mixed metaphor).

Yes, the lower interest rate increases the stock of money they want to hold, but the actual stock of money always increases more than the increase in the stock they want to hold, because each person plans to get rid of part of what they borrow (and they fail in aggregate to get rid of it).

Michael Byrnes writes:
PPPS. On Wednesday September 7th, I will be presenting a paper at a joint Mercatus-Cato conference on monetary policy rules. The conference will be livestreamed at mercatus.org/live.

Any chance that the audio for this conference will be put on a podcast feed for ease of after-the-fact listening?

Scott Sumner writes:

Mark, I want the Fed to set a NGDP target, level targeting. Then I want them to allow the market to set the money supply and interest rates.

Nick, The hot potato effect is interesting, because it can be approached in so many different ways. I like to think of it as a long run effect, which has a big short run impact because of expectations of that long run effect.

So if people expect money to be neutral in the long run, then the interest rate effect will go away once prices adjust, and with no long run change in interest rates, the HPE takes over.

Expectations of that future HPE cause NGDP to rise right now.

Michael, I'll try to find out, and get back to you.

Mark Barbieri writes:

Scott, I still don't understand the mechanics of what you are suggesting. The bank sets an NGDP target. Presumably they rely on a robust NGDP futures market to target progress. Assuming that, after a few months with the target in place and the futures market telling us that NGDP is on track to be below the target, what happens? How do we get back to the target? How do we grow the money supply or adjust interest rates to make it happen?

Market Fiscalist writes:

'Three weeks later, a second Monopoly game is discovered washed up on the beach. The money supply has doubled. Without interest rates to restrain the rise in prices, NGDP doubles almost immediately.'

I'm not quite following this. If everything adjusts instantaneously (prices are not sticky) then both assets prices and other prices (and NGDP) double, and interest rates stay the same. If non-asset prices are stickier than asset prices then probably interest rates will fall and the doubling of NGDP will be help up - but to my mind its something other than interest rates that makes prices sticky. Its the fact that prices are sticky that means interest rates have to fall, so that people want to hold the higher stock of money.

James writes:

Scott,

A lot of different macro theories (old Keynesian, New Keynesian, Ratex, etc.) have the idea that entrepreneurs will borrow money and spend it on additional capital stock if the loan is cheaper than their expected ROI. Lowering the interest rate could make a previously unattractive business idea attractive and the expenditure on the now profitable capital asset would be expansionary.

What exactly is wrong with this reasoning?

DZ writes:

In order for velocity to fall when interest rates drop wouldn't bank lending have to fall? Ignore the effect of increased M. In real terms, you're suggesting real lending supply falls because of credit supply restriction?

Benjamin Cole writes:

I think I agree with this post.

But there are huge structural impediments embedded in the US financial system. About 80% of commercial bank lending is against property. Commercial banks lend more and create more money when they lend on property, and more people are willing to borrow, when real interest rates are low. Remember there is a commercial property market also.

In sum, I am not sure what I am driving at, but I think institutional realities place trip-wires in front of theories.

[extraneous "w" at end of comment deleted--Econlib Ed.]

Comments for this entry have been closed
Return to top