Arnold Kling  

Liquidity Traps and Unicorns

PRINT
Kinsley on Budget Cutting... A Review of the Widely Unread...

The difference being that you might one day see a unicorn.

A liquidity trap requires

1. Low nominal interest rates. (The nominal interest rate is the interest rate as stated.)

2. High real interest rates, due to low or negative inflation.* (The real interest rate is the nominal interest rate minus the rate of expected inflation. If the nominal interest rate is 5 percent and the expected inflation rate is 2 percent, then the real interest rate is 3 percent.)

3. No way for the monetary authority to escape from (1) and (2). Hence the term trap.

(*In this post, Paul Krugman surprises me by stating implying (as far as I can tell) that high real interest rates are a sign that we are not in a liquidity trap. That is completely the opposite of my understanding. Anyway, in the Reagan years, it is not the high real interest rates that tell us we were not in a liquidity trap. It is the fact that nominal interest rates, although they declined, were still high also.

Paul seems to be saying that currently real interest rates are low, in which case I would say that condition (2) for a liquidity trap is not satisfied. Either the low real interest rates are going to trigger an economic expansion or the interest elasticity of spending is low. Low interest elasticity of spending--meaning that consumers are not induced by low interest rates to increase their purchases of durable goods--could certainly be a problem in today's economy. But that is not the same thing as the liquidity trap. In textbook terms, a liquidity trap is a flat LM curve. A low interest elasticity of spending is a vertical IS curve.)

It is certainly possible for (1) and (2) to be satisfied. We can have a nominal interest rate of 1 percent, an inflation rate of -3 percent, and hence a real interest rate of 4 percent, which would be pretty high, certainly higher than you would want to see in a recession. However, that does not make it a liquidity trap.

Look, I can't stop you from saying that because you have seen a beast with four legs and a horn in the middle of its forehead you have seen a unicorn. Even though it was a rhinoceros. Similarly, if you want to call any period of low economic growth, low nominal interest rates, and high real interest rates a liquidity trap, I cannot stop you. (I suppose I cannot stop you from calling a period of low real interest rates a liquidity trap, either, but, again, that seems contrary to everything that used to be taught in intermediate macro courses.)

But the textbook definition of a liquidity trap is an infinite elasticity of the demand for money. The central bank can expand the money supply to an unlimited degree, without affecting interest rates, output, or prices. (As one prominent economist once put it, "no matter how much the Fed increases the monetary base, it has no effect, because it just substitutes one zero-interest asset for another.") In a liquidity trap, the monetary authority cannot cause inflation even if it wants to. If that is true, then Ben Bernanke is a unicorn.

My claim is that if you have the power to increase the amount of money in circulation, then you have the power to cause inflation. It's as simple as that.

Let us not even bother talking about the proverbial helicopter drop, in which the central bank prints currency and distributes it among the population. That might not count as pure monetary policy--I can see making a case that it is a sort of fiscal policy as well.

No, let's restrict ourselves to the central bank exchanging money for other assets. In the textbooks, the central bank buys short-term government bonds. But it could buy long-term government bonds, it could buy corporate bonds, or it could buy foreign bonds.

If these operations are done at small scale, then I can see them having no effect. In fact, I am inclined to err on the side of believing that small-scale central bank operations have no effect.

At some point, however, these operations simply must have effects on prices and/or interest rates. Think about buying foreign bonds. If the central bank purchases enough foreign bonds, the exchange rate has to depreciate. With enough depreciation, the rate of domestic inflation has to rise.

Once again, I assert that in a fiat money economy, it is never impossible for the monetary authority to debase its currency. Undesirable, usually. But impossible, never.

If the monetary authority is always in a position to cause inflation, then there can be no liquidity trap. Once inflation gets above, say, 3 percent, then as a matter of arithmetic either:

--the nominal interest must be higher than 3 percent, in which case we remove condition (1) above; or

--the real interest rate must be negative, removing condition (2) above. For example, if the nominal interest rate stayed at 1 percent and the inflation rate reached 3 percent, then the real interest rate would be negative 2 percent.

Once you admit that the central bank can cause inflation by expanding money growth, then you have admitted that there is a way out of the liquidity trap, in which case there is no trap.

I wonder if there are any prominent macroeconomists who share Krugman's view of liquidity traps.


Comments and Sharing


CATEGORIES: Macroeconomics



COMMENTS (27 to date)
A dude writes:

I think you are making an implicit assumption that the current scale of Fed's LARGE SCALE ASSET PURCHASE program is indeed large...

How do you know that 100s of $bn is large and not 100s of $trln?

Remember that intergallactic battle fleet in Hitchhiker's Guide to the Galaxy that was swallowed by a small dog when they came to kill all humans but miscalculated the scale?

Real rates may be negative, but expected risk-adjusted returns on any capital investments may be more negative still...

wintercow20` writes:

I have a feeling it will not be long before someone starts advocating the Fed to print money to make asset purchases itself. I mean real, physical, assets like cars and office furniture, and gosh knows what else.

Alex B writes:

Isn't the Fed trying boost inflation to no avail? To say that they could, are trying to, but are still not having that effect seems to be a problem.

MichaelM writes:
I wonder if there are any prominent macroeconomists who share Krugman's view of liquidity traps.

Of course not, because Krugman's view is informed by political necessity, rather than any real economic theory.

A dude writes:

Well the Fed is facing a real political constraint and the uncertainty of operating in uncharted waters. So the odds of them actually being too tight in the face of boomer demographics-induced debt burden are not zero.

ebh writes:

I think 2) is wrong. My understanding is that a LT requires low real rates, not high ones. And I think the story is about precautionary savings being high, not savings being high due to a high real return, i.e. savings are high in spite of low real returns. It is all about expectations; economic agents hoard cash and this hurts output. FF rate at zero means the Fed cannot entice them to spend by lowering it more. Another way of looking at it is to use the Taylor Rule. If the Taylor Rule suggests a negative target FF rate, then we might be in a LT.

intrinsic value writes:

wintercow20's, thats why the BOJ has been doing, purchasing ETF's of J-REITS. However the effect of the real asset purchase looks like to be very limited.

Philo writes:

Krugman simply failed to see that the Fed's OMOs can consist in more than simply buying T-bills; he somehow overlooked the possibility of "quantitative easing."

You write: "At some point . . . these operations [buying long-term government bonds, corporate bonds, or foreign bonds] simply must have effects on prices and/or interest rates." That point is when the public becomes convinced that the Fed intends to make such purchases--when the market expectation is that it will do so.

Did you farm the writing of this post out to Scott Sumner?

Steve Roth writes:

I do find this:

""The U.S. economy is best described as being in a bona fide liquidity trap," and given the challenges now faced by the nation, "much more policy accommodation is appropriate today," Federal Reserve Bank of Chicago President Charles Evans said."

http://online.wsj.com/article/SB10001424052748704706904575556283044368588.html

Andy Harless writes:

A liquidity trap does not require high real interest rates, nor is it likely to be associated with them, except in the advanced stages. What matters is the difference between the actual real interest rate and the natural real interest rate (in the Wicksellian sense). A liquidity trap is characterized by a positive difference between the two despite the fact that the actual nominal interest rate is near zero. In theory, this could be because the actual real interest rate is high (severe expected deflation), but in practice it is usually because the natural real interest rate is low or negative (perhaps due to a very high risk premium on real assets).

I dealt with this issue in a blog post last year. Although I didn't use the phrase "liquidity trap," it is essentially the same issue.

Desolation Jones writes:

"Once you admit that the central bank can cause inflation by expanding money growth, then you have admitted that there is a way out of the liquidity trap, in which case there is no trap. "

I know how to get out of a Chinese finger trap. It's still a trap.

Arnold Kling writes:

Andy,
In your post, you write


nearly everyone agrees that a sufficiently reckless central bank will always be able to produce a high inflation rate. (Imagine the Fed buying up the entire national debt, along with all the private sector’s offerings of commercial paper, mortgages, corporate bonds, and so on. Eventually, there will be inflation.)

Sounds like we are saying the same thing.

Jeff Halllman writes:

Just about everyone agrees that a central bank can create a hyperinflation. A liquidity trap requires extremely low inflation or deflation that you can't get out of. That's a contradiction with the first sentence.

The fallback position is that "Yes, you can hyperinflate your way out of the trap, but the cure is worse than the disease!"

But if the central bank can produce zero inflation or 1000 percent inflation, why can't it produce, say, 4 percent inflation? Those who say it can't are saying there's a discontinuity in the response of inflation to money creation. That's really an extraordinary claim, and as such it would seem to require extraordinary evidence to support it. No one's ever presented any, as far as I know.


Jeff Hallman writes:

Wow, the fingers are really fat today. I actually do know how to spell my own name, evidence above to the contrary.

JazzBumpa writes:

Point 2 is completely bogus, and your whole argument falls apart.

From your link to Krugman, it's clear that he puts NO stock whatsoever in your point 2.

A valid point 2 would be that interest rates are insensitive to money supply. (Hence point 3, BTW.)

If you take an honest look at rates - nominal or real, not much difference these days - and MZM, it's mysterious how you could conclude otherwise.

But you do seem to know quite a bit about unicorns.

Cheers!
JzB

Andy Harless writes:

Jeff Hallman:

But if the central bank can produce zero inflation or 1000 percent inflation, why can't it produce, say, 4 percent inflation? Those who say it can't are saying there's a discontinuity in the response of inflation to money creation.

Not necessarily a discontinuity, but the variance of the response may become extremely high. It seems reasonable to call such a high variance environment a trap. You have to choose between sure deflation and a risk of hyperinflation. If someone points a gun at you, they've pretty much got you trapped even if you don't know whether the gun is loaded.

Also, it seems reasonable to me to use the word "trap" even if the variance is not high, as long as there is an "excluded middle" inflation range: you're trapped in a limited set of feasible inflation rates, which may not include your preferred inflation rate. That is different from the ordinary, untrapped condition in which the central bank can choose any inflation rate it wants. If you use both monetary and fiscal policy, you can always get to your preferred inflation rate (in the standard New Keynesian paradigm, at least). So one can define a liquidity trap as an environment in which monetary policy alone cannot produce the desired inflation rate.

JR writes:
But if the central bank can produce zero inflation or 1000 percent inflation, why can't it produce, say, 4 percent inflation? Those who say it can't are saying there's a discontinuity in the response of inflation to money creation.
If your only instrumental variable is interest rates, and you're working in a small time frame, maybe it can't... like now.
Alex Arnon writes:

Krugman's liquidity trap is not Keynes's liquidity trap. It isn't reflected in the demand for money but in the central bank's reaction function. As I understand him, he sees some kind of institutional restriction on the central bank's actions. He's claimed that higher inflation expectations would get us out of the trap but seems to think that the Fed doesn't have the power (the will) to convincingly raise expectations. Given this, all its actions will be perceived as temporary, so "no matter how much the Fed increases the monetary base, it has no effect."

Looking at what's gone on over the past few years, I think its a defensible position, though that doesn't justify just accepting it and giving up on the better policy tool.

JC writes:

Krugman responds.

It's apparent you really aren't dealing with what he's talking about.

AC writes:

It appears that Krugman is changing his definition and doesn't want to stoop to actually addressing any of your points.
http://krugman.blogs.nytimes.com/2011/03/18/liquidity-traps-once-again/

GC writes:

From Krugman:

"In the model, deficit-financed government spending can, at least in principle, allow the economy to avoid unemployment and deflation while highly indebted private-sector agents repair their balance sheets.."

So I'm seeing you twisting in the wind here. Krugman himself posits a way out of the trap, which is increased deficit spending up until the volume of private sector monetary demand begins to push up interest rates, which is what we want, at which point government spending decreases in kind. You however seem to say that "things will just happen when we find the magic ponies..."

At least Krugman has a method to his madness. You're just twisted knickers.

Max writes:

[Comment removed for supplying false email address. Email the webmaster@econlib.org to request restoring this comment and your comment privileges. A valid email address is required to post comments on EconLog.--Econlib Ed.]

Jeff Hallman writes:

Andy Harless:

We agree that inflation is a monetary phenomena, I think. We also agree that the central bank, if it tries, can produce either very high or very low inflation. We even agree that it can produce rates in between, but you think it can only do so with with great imprecision.

I have two answers to this: one empirical and the other more academic. The empirical point is that never in history has a central bank announced and seriously pursued a sensible nominal GDP, price level, or inflation target and failed to come close to it. (Please do note the qualifiers "seriously" and "sensible". It's not serious, when, like the Bank of Japan, your actions clearly show that the announced and actual targets are very different thing. Nor is it sensible to do what many post-WWI countries did when they tried to go back on the gold standard at rates that greatly overvalued their currencies.)

My more academic answer is that it's expected, not actual, inflation that is really important to economic decision makers. Quasi-monetarists think that if the central bank clearly communicates what it's trying to do, it can move expected inflation to just about any number it wants to. This is particularly true when TIPS and other asset prices provide nearly instantaneous feedback about the effect of bank actions and announcements on expectations.

So why do you think it's so hard?

John writes:

Arnold: Krugman is right. We are in a liquidity trap. It may not be as serious as he implies, but there is no question his description is correct.

grooft writes:

Jeff Hallman writes:
"Quasi-monetarists think that if the central bank clearly communicates what it's trying to do, it can move expected inflation to just about any number it wants to. This is particularly true when TIPS and other asset prices provide nearly instantaneous feedback about the effect of bank actions and announcements on expectations."

Nobody (except perhaps academic "monetarists") believes that the Fed would or could ever "clearly communicate what it is trying to do". The financial markets don't act on statements from the Fed alone; everything is interpreted through filters. Even if the Fed was to announce a 4% inflation target there is no expectation that they would be able to fight off the opposition and maintain such a target. As Krugman noted -- you have to be credible that there is a real commitment to the inflation target in order for it to be effective.

Given the amount of 'unreality' in the discourse on whether we are in a liquidity trap, it is hard to imagine the Fed either 1) clearly articulating a policy that will get them hammered or 2) maintaining such a policy if they had articulated it.

Now if Ben Bernanke wasn't a Republican in a position the eschews politics he might note in Congressional testimony that monetary policies are not as effective as fiscal policies in getting the economy back on track and that cutting government spending is exactly the wrong policy to implement right now. This would, of course be 'entering the political arena' and on the other side of Greenspan's argument that running a surplus is dangerous and tax cuts are the answer. (Some might argue that the Fed should stay out of the political arguments on taxes and government spending. These folks usually start from the premise that Greenspan was right in his testimony in 2003.)

pgl writes:

If the IS curve shifts inward, we can be in a liquidity trap even with low real interest rates. Yet - you seem to deny this. I think Paul's frustration with your lack of understanding of a rather simple point is quite warranted.

Jonnan writes:

Your explanation of a liquidity trap doesn't even make basic intuitive sense, particularly in comparison to the Reagan era info Krugman was debunking?

At it's core, a liquidity trap requires having nowhere for the Fed to lower the interest rate to - i.e. a Taylor rule result indicating you need to have a lower interest rate than zero - how *bad* a trap this is is of course dependent on the degree of discrepancy, but your argument doesn't seem at all calibrated to debunk Krugman.

Certainly the Reagan Era, whichever interest rate you went by, had plenty of room with which to lower it.

I'm sorry, the concept is pretty simple and obvious, and even allowing for the existence of "Simple, obvious, and wrong" your counterargument just doesn't even seem related.

Jonnan

Comments for this entry have been closed
Return to top