John Cochrane on QEII

|Peter Boettke|

John Cochrane has a piece up entitled Sense and Nonsense in the Quantitative Easing Debate, and he offers some straightforward arguments most of which I buy completely (e.g., non-effectiveness of the policy in stimulating the economy) and others that I am not completely persuaded of (the non-inflationary danger of the policy for the economy).  But I believe the most important point he raises is:

"Unemployment is not high because the maturity structure of government debt is too long, thank you, nor from any lack of “liquidity” in a banking system that is sitting on a trillion dollars of cash. It’s time to focus on the real, microeconomic, tax, and regulatory barriers to growth, not a policy that creates a lot of noise but no real effect."

What do you think of his analysis and his argument?

 

Comments

[Andy]

I think much of it makes sense, but is it so obvious that "money and short-term bonds are the same thing" even if rates are near zero?

"If everyone knows inflation is coming, they raise prices and wages immediately and are not fooled into a little boost of output." You don't have to be a Keynesian to think there is some degree of wage stickiness in the short-run, do you?

[Mark Thomson]

The thing that puzzled me was his suggestion that QE2 increases the risk of a debt crisis by increasing the need for the Government to continually roll over short term debt.

Maybe I've misinterpreted him but I read this as a reference to the shortening of the maturity structure that he discussed earlier. But as I understand it, QE2 does nothing to the _amount_ of short term debt outstanding. It does increase bank reserves, but these are held by the Fed not the Treasury, and don't need to be 'rolled over'. Here the equivalence of money and short term bonds seems to break down.

Is he perhaps referring to the possibility that banks might withdraw reserves in a hurry? Or is he concerned not about QE2 per se, but rather just that the government is not issuing more long term debt? Econbrowser had an interesting post yesterday showing that average maturity is actually increasing and that QE2 is not changing that - http://www.econbrowser.com/archives/2010/12/evaluating_qe2.html, though he (correctly I think) doesn't consider Fed reserves to be part of short term debt.

Plainly I'm not an economist, but I am interested in understanding this stuff. I would be interested in any light you can shed.

[Steve Horwitz]

What's nice about this piece, and esp. the bit you quoted, is that we can agree that those are the real problems even if we don't agree about the inflationary threat. Even if Cochrane is right that the attempted cure won't make the patient sicker (and I don't think he is), he's quite correct that it won't be the cure and that it misses what the underlying disease is.

[Jerry O'Driscoll]

The critical thing about the Cochrane critique is that he focuses on the microeconomics. I don't agree with his version of RBC, but I do agree with the micro approach.

The shocking thing about Bernanke's justification for QE2 is that he is now falling back on the crudest version of the Phillips curve. Unemployment is high so we need a "little" inflation: anticipated inflation, not unaticipated. How does that work?

Bernanke has thrown out 50 years of monetary theory. Including the good stuff on expectations and inflation dynamics.

[Pietro M.]

O'Driscoll:

I'm not sure Bernanke is out-of-date with his economics, because the basic New Keynesian macromodel, Woodford-style, can be used to defend Bernanke's choices, especially if one considers the zero-interest-rate-policy variety of these models.

In these models there is neither finance nor banking, there is an exogenously set interest rate which determines the behavior of an economy comprising a continuum of monopolistically competitive firms subject to nominal rigidity and affected to random macroeconomic shocks.

The natural rate of interest is the rate which is to be set by the Fed in order for the nominally rigid economy to behave as efficiently as in a real-business cycle model. The money rule is a feedback rule achieving this result.

There is only one source of instability and inefficiency: nominal rigidities. There is thus only one diagnosis for a recession: interest rates are too high and shall be cut down to the new lower natural rate because of a shock. There is no structure which can become unbalanced. Thus there is no cost in that monetary policy because there cannot be any structural imbalances fostered by "stabilization".

When the economy touches the zero interest rate limit, special policies are required (and Bernanke has published on this subject, both theoretically and empirically).

These policies are monetization of term assets, monetization of private assets, commitment to long-term low-interest rates, expansion of the central bank balance sheet, devaluation of the currency.

Bernanke did what he, together with many other new-keynesian economists such as Krugman, Eggertson, Woodford, Clarida, Gali, Svensson, said the Fed should have done in that situation.

It's not a failure of some 100 years old vintage amateur economics, but of very recent fields of professional research in new-keynesian theories.

These things got published on the most important journals. And I can't believe it.

[Pietro M.]

To more closely comment on Cochrane, Bernanke is likely to think in these terms:

"Unemployment is high because the natural interest rate is negative. It’s time to focus on the nominal barriers to growth, by further reducing the cost of funding despite the zero interest rate barrier."

Cochrane doesn't tackle this issue and reasons in an "assume Bernanke is Prescott" fashion. I agree with his assessment, but he doesn't analyze and criticize Bernanke's theoretical arguments.

[Jerry O'Driscoll]

I know some in the Fed, perhaps Bernanke, are influenced by the Woodford model. But what Bernnake has been saying harkens back to the Phillips curve, which still guides much of Board staff.

To return to earlier discussions (on Alchian, etc.), nominal rigidites must be motivated not postulated. Nominal rigidities in the face of expected inflation make no sense.

[Bill Woolsey]

I thought Cochrane's argument assumed market clearing.

The current price level (and wages) is at a level so real expenditure is equal to the productive capacity of the economy. He doesn't explain how he knows that it true, put I presume he believes this because if real expenditure was less than productive capacity, then the price level (and wages) would be lower.

If inflation rises, then, perhaps, there will be some short run confusion about real wages, and people can be tricked into working more. But if it is expected, then they won't be tricked. Prices will just rise to keep the real quantity of money unchanged, real expenditures grow (or shrink) however much needed to keep them equal to the productive capacity of the economy.

If all of this is true, then increasing nominal expenditures won't increase real expendenitures and so quantitative easing is pointless.

Quantitative easing only makes since if the price level (including wages) has not fallen enough to make real expenditures equal to the productive capacity. It only makes sense if the current level of prices (and wages) are above market clearing levels.

The notion that advocates of quantitative easing imagine that there is a long run phillips curve to exploit is--well, incredible.

Suppose it is 2006 and Bernanke decided that 4.5% unemployment was too high and that it was worth while shifting from a 2% trend rate of inflation to a 3% in order to increase output to a higher growth path and the unemployment to say, 3.5 percent. That would have been the old phillips curve. More rapid growth in money expenditures, say from 5% to 6% will result in a higher growth path of output and lower unemployment permanently.

But the status quo is that money expenditures fell and are currently 13% below their growth path of the great moderation. The price level is about 2% lower and wages--well, I haven't checked lately.

If you believe in market clearing, then it _must_ be that the productive capacity of the economy fell about 13% below its trend of the great moderation. The current price level is at equilibrium. That this reduction in productive capacity happened to be associated with a massive drop in money expenditures from trend is just chance, really, and irrelevant.

Well, maybe it would adversely effect the economy for a month or so, but nothing more than that.

Anyway, you all should know all of this.

All I can say is--where are all the bottlenecks and shortages? Where is the consumer goods boom that cannot be met becaues all the capital goods are set up to produce houses?

The answer is-- no where.

[Bill Stepp]

"If everyone knows inflation is coming, they raise prices and wages immediately and are not fooled into a little boost of output." You don't have to be a Keynesian to think there is some degree of wage stickiness in the short-run, do you?"

This is the efficient markets theory run rampant.
Markets as big as the ones under consideration (e.g. the labor market) in "macroeconomics" do not adjust immediately to changes in data.
If the XYZ company gets a buy out offer tomorrow at $x/share, its stock price might well reflect that immediately (or perhaps not if another suitor is lurking, for example), but labor markets do not reflect inflation expectations immediately, whatever that might actually mean.

[Bill Stepp]

The bond market showed First Soviet Comrade Bernanke's planning hubris for what it is.

[Current]

> but is it so obvious that "money and short-term bonds
> are the same thing" even if rates are near zero?

Exactly.

Next time I read an American economist saying this I'm going to fly to America and find that economist. Then I'm going to take him to a strip club, when I'm there I'm going to stuff T-bills down the thongs of the strippers. I'll leave him to be responsible for what happens next.

[Jerry O'Driscoll]

There are no relative prices in Bill Woolsey's story. There is no meeting of the minds until he addresses that issue. His comments at the end address nobody's position.

Current monetary and fiscal policy aims at propping up unstainable nominal magitudes. Why it would be surprising then that real variables are out of equilibrium? Woolsey's "solution" is the source of the problem. His policy causes the problem he ostensibly wants to cure.

Nominal and real factors interact. Nominal stimulus affects real factors. (Money is non-neutral.) The collapse of the bubble causes both real and nominal values to fall. What is confusing about that?

That observation is not an argument for monetary stimulus, but part of the case for a rule-based monetary regime. You can't "fix" a problem by recreating it.

I'm neutral on keeping MV constant as a second-best policy in a fiat money regime. But it is decidedly a second-best policy.

Prices and wages fall all the time. Even in a hyperfinflations. There are nominal rigidites, but they are rationally based and institutionally dependent. Change policy and institutions, and rigidities change. Reinforce them by policy, and they get worse.

[Pietro M.]

I do want to go to a strip club with t-bills and see what happens, it would be a nice economic experiment.

[Pietro M.]

A note on the stripper example, which is one of the reasons why Boettke is right in saying that "economics is one of the sexiest things in the world".

Cochrane is of course reasoning in some never-never land frictionless Walrasian model. This is the worse model we can manage to have to understand what money is about, no doubt.

However, what he says can be forced to make sense even in an Austrian setting.

Firms gain a profit out of investment because of relative price differentials. Adjusted for risk, this differentials need be compared with the cost of credit, which is greatly influenced by the Fed, both in the short run (no one denies it) and in the long run (next year's short-term rate will mostly likely be close to zero, too, and the same applies on longer horizons if the economy doesn't recovery: the expected cost of credit is thus almost zero in the long run).

Firms' net profits are thus given by price differentials, minus risks, minus Fed's rates, with some allowance for eventual interest rate spikes in the future). This is what drives investment.

In the Walrasian world, relative prices immediately adjust to neutralize the effects of the Fed nominal targeting. In the real world, non-neutral effects are persistent.

When the target rate is zero, there is nothing the Fed can do to foster additional investments, money is not neutral, but it is ineffective: it can't perpetuate a boom. This is no longer in the Fed's power.

On empirical grounds, it is evident that even in these conditions the Fed is capable of preventing economic adjustement by loan evergreening and moral hazard. The reasons for this are not very clear to me, as it would be much easier to claim that once interest rates are zero the economy worked as if the Fed didn't exist. It just ain't so.

[Mario Rizzo]

It is important to note that Bernanke's (and others) focus on the (un)likelihood of inflation is symptomatic of the problems with his analysis. As Hayek noted many times, distortions can occur and multiply without there being inflation in the conventional sense. But this consideration is quite a pain-in-the-neck for the simple aggregate demand approach.

In this framework, what is to be done if you want to boost demand to reduce unemployment and yet avoid distortions that will come to haunt you later? The easiest --but incorrect -- thing to do is to ignore (for the time being) the distortions. And so the Fed-induced cycle continues.

What this suggests is that attempts to synthesize the "Keynesian" demand management approach with a more sophisticated bubble-distortion-malinvestment analysis will suffer from intractable problems.

[Current]

I wouldn't go as far in the quasi-monetarist direction as Bill Woolsey goes. But, would go a little way.

We live in a world where there is long-run gradual price inflation. I'd argue that people take that into account in their decision making, though not perfectly. Even if there isn't an announced price level trend target, NGDP target, or inflation target, in practice in the past a gradual rise is what we've got. I'd argue that when planning people consider this as the default scenario.

If a central banker announces he will aim to hit an inflation target like this then what he's saying is "I will fulfill your normal expectations". I think this is important because investment depends upon a reasonably predictable price level year-after-year. He could just as easily say "from now on the inflation target is 0%". Indeed that would be better in the long run. But, the point is that he must ensure that the market isn't uncertain about the future.

Clearing this uncertainty will probably do much more than QE2 itself can achieve.

Pledging to use a new inflation target of 0% in the long run wouldn't necessarily be good for employment in the short run. In the short run as Bill and Pietro have pointed out we will have nominal price stickiness. I don't believe that all agents can gather good enough expectations for that to not be a problem. Especially because of the Hayekian argument for price-stickiness: the time taken for price signals to propagate through the economy to final goods prices.

[Barkley Rosser]

I agree that QEII is not likely to stimulate the economy very much (nor is it likely to be very inflationary either), but Cochrane is one of the last people who should be consulted on what to do about our current situation. He is one of the leading advocates of the "all the unemployment is voluntary" school.

So, supposedly one day back in 2008, a bunch of people just woke up with their intertemporal elasticity of substitution shocked, shocked I tell you! They just up and quit work and have not been looking for any since.

[Troy Camplin]

Has anyone done any work on relative price stickiness -- which network topologies increase it, which ones decrease it? (Which would speak to those policies which affect network topology.)

[Bill Woolsey]

O'Driscoll:

What do you want to know about relative prices? I don't think the purpose of quantitive easing should be to create some pattern of relative prices. If relative prices need to adjust, let them adjust in an environment of stable growth of money expenditures. The notion that nominal quantities are unsustainable is absurd. I don't imagine that any particular constellation of relative prices will be maintained for long.

[Rosetta Stone Portuguese]

oh my!! i love you so much!!

Thank you so much.