Nick Rowe continues to provide interesting feedback.  This is the tail end of a much longer comment he left at the end of my Richard Rorty post from a few days back:

Here’s another one, that’s both fun, and deadly serious in current contexts:

Horizontal axis: belief that monetary policy is interest rate policy.
Vertical axis: truth that monetary policy is interest rate policy.
Here we have a positive feedback system. There are probably two (stable) equilibria: one in which monetary policy *is* interest rates, and we can get stuck in liquidity traps; and a second in which it isn’t, and we don’t get stuck. Social construction of reality, again, which is just an extreme form of a convention, which not only affects how we behave, but how we see the world. It defines the “social facts” of what central banks “do”.

Question: if this (above) view is correct, what exactly is it that Scott Sumner is *doing* (on this blog)? What *is* he? Not a policy advocate, in the standard sense.

And, if this view is correct, forget about trivia like cutting the overnight rate by 0.25%. Banning all public mention of the overnight rate would be a more effective policy!

I think I may have jumped the shark. Never mind.

He is being too modest; this raises all sorts of interesting issues.  Yes, other economists have discussed somewhat similar ideas.  I seem to recall Krugman used the “Peter Pan” analogy at one point.  Something to the effect of “if we think monetary stimulus can work in a liquidity trap, then it can work.”  But Nick’s comment raises some deeper issues.

In another post one commenter contested my view that investors were paying a lot of attention to Fed policy in October 2008, the period when I alleged that policy was too tight.  He argued that investors were focusing on falling GDP.  I replied that expectations of falling NGDP is exactly what I mean by tight money.  At this point you might say “but for the Fed to have been able to prevent this, they would have had to have been able to change NGDP growth expectations with their policy tools.  Unless investors see the connection, they won’t be able to do that.”

Normally it would be easy for me to swat away that objection, as stock prices often rise or fall 2% or 3% minutes after the 2:15 Fed policy announcements.  So clearly Fed policy has a powerful effect on expectations.  And I think my argument even applies to early October 2008, when the Fed funds target was still 2%.  But what about later, when there was a widespread sense that rates could be cut no further.  Was the Fed no longer able to affect market expectations of NGDP growth?

That is where Nick’s observations come in.  Nick and I have both argued that there are a variety of policy levers by which the Fed can impact nominal spending.  But suppose everyone else thinks the fed funds rate is the only possible tool?  Does that make the Fed powerless?  I am going to argue that the answer is no.  But first suppose that the answer were yes.  Then Nick is arguing that my blog (and by implication blogs like his and Bill Woolsey’s as well) are providing a valuable public service.  We are like those slightly ridiculous 60s hippies that were “trying to blow people’s minds, to get them to see reality in a new way.”  And if people can see there are other ways out of the liquidity trap, then they will be more optimistic that the Fed can avoid the worst case of another Great Depression.

[BTW, my blog started in February 2009, and caught on in March 2009.  Just saying . . . ]

Well this is all very flattering, but I don’t think it is that simple.  It’s up to the Fed to decide what sort of policy tools they want to use.  In the early 1980s the Fed announced it was targeting M1, and almost immediately the various Wall Street markets started reacting strongly to the unanticipated part of the money supply announcement.  When they stopped a few years later, the markets stopped reacting to M1 announcements.  In 1933 FDR announced he was using the price of gold as the level to move prices higher, and the markets immediately started responding to changes in the price of gold.  It’s really all up to the Fed.  Whatever tools they choose to use, the public will follow.  In my view the tool should have been NGDP futures prices.  If they had done so, the public would have followed these prices as closely as they now follow changes in the fed funds rate.

So I don’t buy versions of the Peter Pan argument that make success seem very iffy, very contingent on luck, on whether the well-intentioned Fed is able to convince to the narrow-minded public of their wisdom.  That’s getting the problem backward; the public knows exactly what is going on, it is the Fed that needs to wake up and “have its mind blown.”

But given where we are, stuck with a Fed that can’t think beyond interest rates, what would be the best type of Fed chairman?  In the 1980s and 1990s, when most academics thought the “time inconsistency problem” biased us toward excessively fast NGDP growth, there was a view that we needed a sober, conservative central banker.  Maybe even someone a bit heartless.  Someone who could hold down inflation expectations.  Of course macroeconomists are exactly like the generals who are always fighting the last war, and this was all an attempt to build models out of a single observation—the 1970s.  In retrospect the models were wrong.  So if we are stuck with interest rate targeting, what sort of central banker do we need?

During most times a sober, responsible figure like Volcker/Greenspan/Bernanke is fine.  They can use their Taylor Rules to try to keep NGDP growing at a low and stable rate.  But the minute interest rates hit zero (or even close to zero) there should be a provision that the entire FOMC is replaced with a new FOMC.  And who should be on the new FOMC?  Sit down before you read this . . .

I’m thinking of people like Jim Cramer and Larry Kudlow.  When I say “like them,” I don’t mean people with similar views on monetary policy, I mean people with similar personalities.  People who are optimistic about America, strongly pro-growth, people that are not passive, but rather want to make things happen.  You get the idea.  The point is to instill the public with a feeling that good times are just around the corner, and that they will pull out all the stops to make it happen.  The last thing we need right now is a central banker with shoulders hunched over, who solemnly intones about the looooong difficult period America has ahead if it.  Who keeps talking about how we need to tighten our belts, and get used to a lower standard of living.  Come to think of it, I’d like to nominate my commenter JimP to the FOMC.

Back to reality, I have two serious points in this post.  One is that Nick has correctly pointed to the key role played by beliefs in the transmission of monetary policy.  Academic economists have an enormous responsibility to make people understand that liquidity traps do not limit the effectiveness of monetary policy.  Unfortunately among the people who need to be educated are Fed Presidents like Janet Yellen, and what Krugman calls the “economic analysts” who advise them.  But my second point is that this “expectations” issue should not be misunderstood.  The previous examples of monetary policy at the zero rate bound (US in the 1930s, Japan more recently) make it crystal clear that (contrary to Krugman) there is no such thing as an expectations trap.  No central bank has ever tried but failed to reflate.  Central bankers set the agenda, it is up to them to provide leadership, if they wish to do so.  But right now the leaders of the Fed, ECB, and BOJ, have absolutely no desire to raise NGDP growth expectations.  And that’s a shame.

PS.  Regarding Nick’s proposal for no more discussion of the short term policy rate; my dream is a macroeconomics with no discussion of either interest rates or inflation.  Just aggregate hours worked for the real (cyclical) variable, and NGDP for the nominal variable.  Monetary policy targets NGDP expectations rather than the short term rate.

I’ve already discussed John Taylor’s appearance on the Big Think.  Today I’d like to do two more posts delving more deeply into Taylor’s responses and discuss some of the problems that I see in his approach to monetary economics.  I asked Taylor whether money was too tight in late 2008.  Here is what Bill Woolsey had to say about Taylor’s response to my question:

what is most striking about Taylor’s response is that “monetary stimulus” is simply identified as lowering the Fed’s target for the federal funds rate.

More troubling, however, is the notion that the targeted level of the federal funds rate was fine, but it was rather uncertainty generated by the financial policy — the bailouts — that caused the problem.

.   .   .

This view is wrongheaded. “The” natural interest rate is the interest rate that coordinates saving and investment–given the expectations that households and firms actually have. It is the level of the interest rate that makes total real expenditures equal the productive capacity of the economy–again, given the expectations that households and firms actually have.

If uncertainty about government action raises saving or reduces investment, the natural interest rate is lower. The role of any market price, including the market interest rate, is to coordinate given actual market conditions, not hypothetical ideal conditions.

The phrase, “hypothetical ideal conditions” got me thinking about the Fed’s fateful decision not to ease policy on September 16, 2008, and how the Taylor Rule may have contributed to this mistake.  Recall that the Taylor Rule is backward-looking; policy reflects lagged data on inflation and output gaps.  As of September 2008 the lagged inflation data was fairly high.  There was a small output gap, but the recession was not yet very severe.  So if the Fed was using a backward-looking Taylor Rule, then there may have been no need to ease policy.  These are the “ideal conditions” Woolsey alluded to.

Now let’s think about how policymakers could respond to a financial crisis.  From a backward-looking Taylor Rule perspective, the financial crisis is not something you’d want to keep in mind when setting policy; after all, its effects had not yet shown up in the price and output data.  At the same time, it was clearly a problem that could have a deflationary effect in the future, as it was associated with an increased demand for liquidity and a lower Wicksellian natural rate of interest.  So what can a Taylor Rule economist suggest?  From their perspective, the only solution is to deal directly with the banking problem.  It doesn’t show up in their interest rate equation, so it becomes a sort of “non-monetary factor’ which is reducing aggregate demand.  That is, a problem that cannot be addressed through ordinary monetary stimulus.  Thus the financial crisis seems like it is the “root cause” of our problems, precisely because the Fed doesn’t have a good way of using monetary policy to respond to the crisis.

From the perspective of a forward-looking NGDP targeter, a financial crisis is just one more factor that can lead to velocity shocks, and hence is something that should be offset with a change in the monetary base (or fed funds target) sufficient to keep expected NGDP growth on target (say around 5%.)  That’s not to say that the government might not want to do something about the banking crisis, after all it could have undesirable real effects even in a nominal economy expected to grow at 5%.  But those real effects would be much smaller than you might imagine, partly because much of the output collapse that has been attributed to banking problems is actually due falling NGDP, and partly because if NGDP was expected to grow at 5% then the banking problems would have only been a small fraction of what we actually observed.

Taylor’s mistake is nothing new, many prominent economists such as Irving Fisher, Herbert Simon and Milton Friedman once toyed with the idea of “100% money,” which is a banking system where bank deposits are 100% backed by reserves.  Under this system a banking crisis would not reduce the money multiplier.  Why were these free market-types led to such draconian regulatory proposals?  If you believe monetary policy should target the money supply, then you’d view a banking crisis as something that could shock the multiplier, and hence aggregate demand.  And if you didn’t want to give the Fed a lot of authority to try to fine tune the economy, then you’d do whatever seemed necessary in order to prevent banking crises from disturbing the monetary multiplier, so that the Fed could implement a simple, nondiscretionary policy rule for the monetary base.  You’d even be willing to pass a law banning fractional reserve banking.

Today most people think that 100% banking is a bad idea and that the Fed should offset shocks to the money multiplier by adjusting the monetary base.  I predict that we will eventually see the backward-looking Taylor Rule in the same light.  Instead of blindly following this sort of rule during a financial crisis, we should follow Svensson’s advice and set the money supply or the fed funds target at a level that is expected to hit the central bank’s policy objective.

BTW,  Fisher didn’t favor a money supply rule.  I presume he proposed the 100% banking idea because he feared the impact of excess reserve hoarding in a world where the monetary base was still constrained by the size of central bank gold stocks.

The Big Think has just posted an interview with John Taylor, and he provides some very interesting answers to my questions on monetary policy.  Before I try to systematically demolish his response, let me first say that I agree with John Taylor on most things, including most of his answers to questions raised by others:

1.  Fiscal stimulus was not very effective, and led to a counterproductive increase in the national debt.

2.  The Fed should focus on monetary policy, rather than supporting the housing market.

3.  The big mistake was not the failure to bail out Lehman, but rather policy errors a few weeks later.

4.  The State of California needs to curb the growth rate of spending.

5.  The Fed should not focus on asset bubbles.

So we agree on most things.  But with all due respect I think his attachment to a backward-looking Taylor Rule has led him into all sorts of blind alleys.

Taylor responded to two of my questions.  I will quote in full, breaking occasionally to interject my own views:

Question: Is it better to have price level targeting or inflation targeting in a liquidity trap? (Scott Sumner, The Money Illusion)

John Taylor: I think the distinction between price level targeting and inflation targeting, which has had a huge amount of attention in the academic literature and has been referred to in the past, such as Ben Bernanke before he became Chairman. I think those distinctions in practice get blurred by the realities, quite frankly. It of course would be good to have a price level targeting other things equal that would sometimes require deflation, however, and a lot of evidence that that’s not such a good thing to have; even occasionally.

So, I’ve always been of the view that an inflation target makes more sense, it’s easier for people to understand, we’ve had a lot of practice with it. And while there is this problem of lack of a drift, the drift of the price levels sometimes called base drift, overall seems to me if we were able to keep the inflation rate at a very low level, then we would effectively get similar results to price level targeting.

What do you think?  It doesn’t seem to me that he addresses any of the reasons Woodford puts forth for price level targeting in a liquidity trap.  Just to review, Woodford argues that with an inflation target, monetary policy has no traction in a liquidity trap.  Assume the inflation target is 2%, but also suppose inflation keeps coming in under 2%.  How does the Fed react?  Under inflation targeting the only answer is “try the same thing again, and hope for better luck next time.”  With price level targeting, if you undershoot the 2% price level growth trajectory the target inflation rate automatically rises.  If nominal rates are stuck at zero, sub-target inflation automatically lowers the real rate.  Does Taylor’s response address Woodford’s argument?  Maybe I missed something, but it seems to me that Taylor’s answer referred to the opposite situation, where inflation has exceeded the target.  Next question . . .

Question: Should the Fed have been more aggressive with monetary stimulus in September-November 2008? (Scott Sumner, The Money Illusion)

John Taylor: I think the Fed cut interest rates during that period just about right. I think it was basically coming down, it could have been a little faster at that point, and of course, GDP did fall tremendously. But I think if they had kept the interest rates along the same path it would have been fine.

Now stop right there.  I didn’t ask whether the Fed should have had lower interest rates, I asked whether their policy should have been more stimulative.  Of course Joan Robinson would say that nominal interest rates and monetary policy are one and the same.  Joan Robinson would say that monetary policy was not easy during the German hyperinflation, after all nominal rates weren’t very low.  Am I being unfair to John Taylor?  In one sense yes.  The inventor of the Taylor Principle clearly understands the distinction between nominal and real interest rates as well as any person alive.  But then when I asked about the stance of monetary policy, why did Taylor not criticize the Fed for dramatically raising real rates between July and November 2008?  Why did he merely talk about changes in nominal interest rates, as if they provided meaningful information about changes in the stance of monetary policy?

A few months ago I had a series of posts basically calling out the entire monetary economics establishment.  I accused them of throwing around terms like ‘easy money’ and ‘tight money’ without assigning any coherent meaning to the terms.  Taylor’s answer is a perfect illustration of what I was complaining about.  I still await an answer as to what these terms mean.  I might have to offer a $200 reward as a cheap stunt, just to generate some interest.  How hard can it be to define the stance of monetary policy?  How hard can it be to explain what we mean by ‘easy money’ and ‘tight money?’  Apparently it’s pretty hard, because no one has yet been able to answer my question.

Taylor continues . . .

The problems I see at that period was the tremendous amount of uncertainty added by these new effects, if you like, the quantitative ease, and sometimes I’ve called somewhat **** industrial policy to where the actions went well beyond the usual interest rates. And I don’t think they were appropriate. We’re still studying them. Some of them were inappropriate. I just finished a study on mortgage interest rates.

I agree that we wasted precious time focusing on non-monetary issues during September-October 2008.  Taylor continues . . .

So, the question about whether the policy could have been even easier going into the panic, it is certainly something to examine, but I think basically, it’s about all it could do at that point. The problems I saw were not so much a monetary policy at that stage, but the really ad hoc chaotic interventions that occurred around the time of the rollout of the TARP, not clear kind of actions with respect to what happened to Lehman Brothers. So, a lot of surprises and ultimately a lot of panic. And I think that panic was induced as I’ve argued in other places by government actions. I think some of the actions of the Fed after the panic began were constructive to be sure. But I don’t see that in terms of a faster reduction in interest rates.

I absolutely agree that the government was creating a lot of panic.  If I was running a highly leveraged commercial or investment bank, and I saw NGDP growth expectations rapidly plunging into negative territory, and I saw the Fed making no effort to adopt a more stimulative policy, but instead adopt an interest on reserves program to prevent reserve injections from having any effect, well then I’d panic too.  But those aren’t the actions that Taylor thinks created the panic.  Rather it was Bernanke and Paulson’s admittedly clumsy efforts to save the banking system.  The reason we both felt that late September and early October was important was that it was during this period that asset markets showed panic.  And we even agree that it was related to Fed actions.  But Taylor (and Cochrane) seem to think it was what they said that was causing panic, whereas I think it was the fact that they got distracted from monetary stimulus.  Throughout history there are many examples of tight money policies severely depressing the stock market.  I simply don’t think it is plausible that a few clumsy regulatory moves that were quickly reversed under political pressure could have severely depressed equity prices.  There’s no precedent for that.  That’s not how the stock market works.  It is not that sensitive.  Oddly, on this point I agree with Paul Krugman (as does Tyler Cowen.)

Taylor continues . . .

In fact, if I could just add. One thing I think people should recognize is that while the federal funds rate target of the FOMC came down gradually in the fall of 2008, the actual rate came down quite a bit more rapidly. In fact, each of the FOMC decisions effectively ratified what the rate was already at. And that rate came down so rapidly because of the large expansion in the reserves. So, it’s hard to see how measured in terms of interest rates policy could be much easier in the October, November, December period.

The fact that the Fed was scrambling to catch up to market rates is one indication of just how far behind the curve it was in the fall of 2008.  But the last sentence was a big disappointment.  Taylor previously indicated that policy was fairly expansionary in late 2008, and ended his answer by suggesting that there wasn’t much more the Fed could have done.  In contrast, I think monetary policy was highly contractionary, and that the Fed could have and should have been much more stimulative.

So how does the Taylor Rule hold up in the crisis of 2008?  The first question to be answered is which version of the Taylor Rule?  I seem to recall that Taylor likes to cite the “Marshall McLuhan” scene in the old Woody Allen movie, so I’ll assume the Taylor Rule is whatever the hell John Taylor says it is.  Here’s how I see things:

1.  In October 2008 I was running around calling for a much more expansionary monetary policy.

2.  John Taylor seems satisfied with the stance of monetary policy in October 2008.

3.  In 2009 NGDP fell at the fastest rate since 1938.

In retrospect, would you say I was right in calling for more monetary stimulus, or was Taylor right?

I do think the Taylor Rule works reasonably well when you are not in a liquidity trap, although even in that case I think NGDP futures targeting is superior.  But if there is one thing we have learned in the past two years it is that the Taylor Rule provides absolutely no guidance to policymakers when interest rates approach zero.  Indeed I believe it quite likely that the “panic” Taylor refers to in early October 2008, which was very real, was caused precisely by the fear that the Fed would adhere to the Taylor Rule, and would not adopt the sort of policies that Bernanke had recommended the Japanese try once their rates hit zero.

PS.  If you haven’t read my earlier posts on monetary policy, I should clarify a few things.  I know full well that Taylor and others understand the distinction between real and nominal rates.  But they insist on continuing to discuss the stance of monetary policy with reference to nominal rates.  As long as they keep doing so, I will keep needling them with my insulting Joan Robinson comparisons.

I am not saying that real interest rates are a good indicator of monetary policy.  Rather that’s what others often say when I point to the flaws with nominal rates.  My point is “OK, if real rates are the right indicator, then why no criticism of the Fed’s extraordinarily contractionary policy during July to November 2008, when 5-year real rates jumped from 0.5% to 4.2%?”

In my view the only meaningful indicator of monetary policy is the expected growth in the policy target variable, which I think should be NGDP.