Arnold Kling  

Kling, Krugman, and Krugman on Liquidity Traps

PRINT
Asymmetric Loss Functions... How to Help Poor People in Poo...

Kling wrote,


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.

Paul Krugman #1 wrote,

the liquidity trap is real: no matter how much the Fed increases the monetary base, it has no effect, because it just substitutes one zero-interest asset for another.

Paul Krugman #2 wrote,

The economy is in a liquidity trap when even a zero nominal interest rate isn't enough to restore full employment. That's it.

Since Krugman is Mr. Liquidity Trap, the concept is his to define at this point. If we are to go with what Krugman #2 wrote, so be it.

I just want to put it out there that the definition used by Kling and by Krugman #1 did not just come out of nowhere. I apologize if it seems as if I am repeating myself. What follows is entirely macroeconomics as I recall it from a Keynesian textbook, circa 1973. In the textbook, it would be done using the IS-LM diagram and the AS-AD diagram, but I will try to not to expose the young and innocent to those. This is supposed to be a family-friendly blog.

Here are the ways that an economy can be at zero nominal interest rates and have high unemployment. These are taught in the textbook as different cases.

1. A deep, deep recession caused by low aggregate demand. Suppose an event (say, a financial crisis) has lowered the propensity of consumers and businesses to spend. Nobody wants to borrow, and nobody wants to hire. Hence, interest rates are zero and unemployment is high. Things are so bad, in fact, that you can have output rise without causing any increase in the price level. Thus, when output increases, real money balances do not fall. As output goes up, it takes only a small increase in the money supply or a small increase in interest rates to equate supply and demand in the money market.

2. Interest insensitivity of spending. Regardless of whether interest rates are high or low, consumers and businesses do not care about interest rates when making their decisions about spending. As a result, if the economy goes into recession, lowering the interest rate to zero does little or nothing to stimulate a recovery.

3. Infinite elasticity of money demand. Regardless of how much the central bank expands the money supply, in the words of Krugman #1, "it just substitutes one zero-interest asset for another." When the central bank expands the money supply, consumers and businesses just keep doing what they were doing before, as if nothing had happened. This is what the textbook would have called the Liquidity Trap.

Also, let me add one more case, which would not have made it into the 1973 textbook but which is in the same spirit. Note that it would not apply to the United States.

4. Small open economy with a world interest rate of zero. An open economy is one that has no impediments to international capital movements and no overwhelming impediments to international trade in goods and services. By small, we mean that it is a drop in the bucket in world capital markets, so that it does not affect the world interest rate. If our small open economy is in a recession, and the world interest rate happens to be zero, then so be it. Nothing the country can do can raise its interest rate above the world rate.

The point is that once upon a time, only (3) was considered a liquidity trap. We would not have said that any time you are in a recession with a nominal interest rate of zero you are in a liquidity trap. Instead, we would have asked why the economy was in that situation. Only if the diagnosis were "infinite elasticity of money demand" would we have said "liquidity trap." (As I read it, Tyler Cowen also sees the elasticity of money demand as the defining feature of a liquidity trap.)

What difference does it make, for policy purposes? On the one hand, for fiscal policy it makes little or no difference. In all four of the cases listed above, a fiscal expansion involving deficit spending will do little or nothing to crowd out private investment. In the first case, the economy is so far from full employment that the excess capacity in the economy absorbs the deficit spending with very little increase in interest rates. In the second case, interest rates may rise, but investment will not be affected. In the third case, interest rates will not rise until the economy gets closer to full employment. In the fourth case, foreign capital will finance the deficit without raising the interest rate.

However, there is a difference with regard to monetary policy. In case (1), monetary policy can still get you to full employment. Thus, one could argue that fiscal expansion is still a last resort.

Krugman #2 would say that all four cases constitute the liquidity trap. Again, if he wants to use that terminology, that is ok. It just seems to me that he ought to be a bit less disrespectful toward those of us who, like Krugman #1, use the terminology with which we grew up.


Comments and Sharing


CATEGORIES: Macroeconomics



TRACKBACKS (1 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/4784
The author at Eli Dourado in a related article titled Forensic Semantics: The Meaning of Liquidity Trap writes:
    I promise not to do too many more posts about a) macro or b) Paul Krugman. I don’t just love macro, these are not my most popular posts, and Krugman is too shrill to read on a regular basis. Nevertheless, I think I can sort through some of the re... [Tracked on March 19, 2011 1:16 PM]
COMMENTS (6 to date)
Jason writes:

Lets start with the quantity equation MV=PY, assume sticky prices and a fixed monetary policy. Then government spending only affects output if it moves velocity. The way it does this is by increasing interest rates.

At the zero lower bound if fiscal policy literally does not move interest rates and therefore velocity then it cannot move output. Standard IS-LM in a "liquidity trap" says that the LM curve is NEAR horizontal. In this case, the government increases spending and interest rates go up very slightly, but this leads to a large decrease in money demand because money demand is so sensitive to interest rates. However, if fiscal policy literally does not increase interest rates at all, its kind of a stretch to see why it moves velocity and therefore output.

Eli writes:

I read Krugman #1 as consistent with Krugman #2, but both are wrong. Krugman #1 is saying that monetary policy is a priori ineffective at the zero bound because it substitutes one zero-interest-rate asset (cash) for another (bonds).

Krugman #1 and Krugman #2 are both wrong because it's not as if there is only one interest rate. Monetary policy can be conducted with riskier assets (or there can be a helicopter drop, or whatever). Conceptually, the only reasonable way of interpreting the concept of a "liquidity trap" is as infinitely elastic demand for money. That would be the only way that monetary policy would be ineffective.

c141nav writes:

Remarks of Krugman’s former ombudsman at the New York Times, Daniel Okrent:
"Op-Ed columnist Paul Krugman has the disturbing habit of shaping, slicing and selectively citing numbers in a fashion that pleases his acolytes but leaves him open to substantive assaults."

David C writes:

Alternatively, Krugman could be arguing that situations 1 and 2 don't ever exist or that situations 1 and 2 cause situation 3 in all cases.

Charles R. Williams writes:

The notion of a liquidity trap presumes that the government controls money directly. But if banking - understood in its broadest sense as the process of financial intermediation - has slipped out of the central bank's control, money has become endogenous, unless we define money in such a narrow fashion that it becomes meaningless or unless some financial panic damages the broader banking system.

So the impotence of monetary policy arises not from the fact that the central banks converts one zero return asset into another but from the fact that the banking system is perfectly capable of offsetting any central bank operations as the markets require. If the banks can freely monetize T-bonds, what difference does it make when the fed monetizes T-bonds? The interest rate is irrelevant.

There are two situations where the central bank can make a difference: first, they can step in in the event of some kind of financial panic to assume some otherwise private banking functions (e.g., QE1); second, the regulated banking sector is big enough that if the central bank damages it through gross mismanagement, it can damage the economy.

Doc Merlin writes:

1) If you can print infinite money without raising the price level, then you can generate infinite real wealth by printing currency. This strikes me as absurd.
2) Increases in monetary aggregates for sovereign currencies tend to hold down or lower interest rates, so if Paul Krugman is correct and liquidity traps exist, you can then make infinite wealth just by printing enough money.

This is frankly absurd.

Comments for this entry have been closed
Return to top