Archive for January 2014


Updating prior beliefs

The New York Times has a piece discussing how two famous economists updated their beliefs on the efficacy of monetary policy in response to new information. First Narayana Kocherlakota:

“It’s a little embarrassing to say this, but you make a speech in August of 2010 and it inspires a whole quantity of work where people say, ‘This is what Kocherlakota says and we will now show in this paper that Kocherlakota was wrong,’ “ he said. “There’s a number of ways that people can react to that, and I reacted in the only way that a sensible person can, which is to update.”

In September 2012, shortly after the Fed announced a fresh expansion of its stimulus campaign, Mr. Kocherlakota returned to Michigan’s Upper Peninsula to announce that he had changed his mind. Contrary to his prediction, inflation was slowing. That meant the Fed had the opportunity and responsibility to do more.

Right now he’s the only one at the Fed arguing publicly that they should do more (according to the NYT.) That makes him my favorite government official.

Before discussing the second example in the NYT piece, a bit of background information.  Edward Prescott played a pivotal role in developing real business cycle theory in the late 1970s and early 1980s.  Prescott claimed that monetary shocks did not have significant effects on real output.  At about the same time both monetarist and Keynesian economists were arguing that if the Fed brought the rate of inflation down rapidly, unemployment would soar much higher in the short run.  The monetarists claimed unemployment would later fall back to the natural rate, even if inflation stayed low.  The Keynesians were less sure about that.

In 1981-82 the Volcker Fed did bring inflation down very sharply, and unemployment soared to 10.8%, the highest rate since 1940.  How did this event affect Prescott’s views?  This is from the same NYT column:

Edward C. Prescott, who won the Nobel in economic science in 2004 and is on the Minneapolis Fed’s research staff, said Mr. Kocherlakota was misjudging the Fed’s abilities. “It is an established scientific fact that monetary policy has had virtually no effect on output and employment in the U.S. since the formation of the Fed,” Professor Prescott, also on the faculty of Arizona State University, wrote in an email. Bond buying, he wrote, “is as effective in bringing prosperity as rain dancing is in bringing rain.”

Well that settles it.

Speaking of rain dancing, I used to think Kocherlakota was of Native American descent.  Not so:

Correction: January 27, 2014

An earlier version of this article misstated where Narayana Kocherlakota was raised. He was raised in Winnipeg, Manitoba, not Calgary, Alberta. It also misstated the age at which he received his doctorate. He was 23, not 24.

Because of an editing error, the article also misstated the origin of Mr. Kocherlakota’s name. His name is Indian, not American Indian.

HT:  Saturos, Michael Byrnes

A few short comments

1.  Many commenters were skeptical of my claim (in the previous post) that Robert Barro seemed dismissive of demand-side theories of the cycle.  Noah Smith linked to this 1989 paper by Barro, which is consistently skeptical of demand playing an important role in business cycles.  The impression created by the article is exactly the same impression I got from reading Barro’s 2011 WSJ piece. He didn’t completely rule out demand having some effect, been seemed very skeptical that it could have a significant effect.

2.  Several of my more reliable commenters thought this Econlog post gave useful insights into the way I look at NGDP targeting.  I can never tell which posts are interesting (I’m too close) but you might want to take a look.

3.  David Beckworth has a nice post on forward guidance, with a great title:

Forward Guidance is Hard. So is Navigating a Ship By Focusing on the Expected Path of its Rudder

Note that this is especially applicable to time dependent guidance.  State dependent guidance is better, albeit still has a problem with an interest rate rudder that locks up just when you need it most.

BTW, I see a lot of confusion in the press on forward guidance.  There are claims that the Fed has reneged on previous guidance.  That’s not quite right.  It you promises to visit California and then later promise to visit San Francisco, you have not reneged on your earlier promise.  The recent Fed decision to hold interest rates low at least until a considerable period after unemployment falls to 6.5%, is more specific that the promise to hold rates low at least until unemployment falls below 6.5%.  There is a problem with the current techniques for forward guidance, but it’s not about reneging on promises.  Rather the real problem is using unemployment and inflation as thresholds, when they should use NGDP.  And not doing level targeting.  In that case they would not have to constantly adjust the threshold as new information about the natural rate of unemployment came in.

Adding prose to Nick’s poetry

In recent years Nick Rowe has come up with a clever set of metaphors to describe a central puzzle of monetary economics—the fact that easy money both lowers and raises interest rates.  Here’s the punchline of his latest:

According to that theory: if the central bank wants to lower the real interest rate on paper money (because it thinks there’s a danger the real interest rate on paper money will rise above target unless it does something) it needs to lower the nominal interest rate on electronic money. The central bank calls this the “short run” part of its theory.

But, according to that same theory: if the central bank’s target for the real interest rate on paper money were lower, the nominal interest rate on electronic money would need to be higher. The central bank calls this the “long run” part of its theory.

Very few people understand the central bank’s theory. Even some very good monetary economists don’t get the short run part, and think that if the central bank wants to lower the real interest rate on paper money, all it needs to do is raise the nominal interest rate on electronic money.

Just in case you haven’t figured it out yet: this is not an imaginary world. It’s the real world. But it definitely is weird. Only a finance theorist could have dreamed up a world this weird.

This is a bit unfair to Nick, like quoting just the last stanza of a poem.  Please read the whole thing. Here I’d like to talk a bit more about the money/interest rate puzzle—what drives it, and why it’s so important.

Let’s start with a simple flexible price world, where the central bank decides to increase the trend growth rate of zero interest fiat money.  Most monetary models would predict superneutrality, which means that the inflation rate and the NGDP growth rate and the nominal interest rate would rise in proportion to the rise in the money supply growth rate.  We will call this assumption A.  I’m going to try to amend this example with two other assumptions, to see if they can cause this result to “flip,” so that easy money makes nominal rates fall.  Neither will work.

B.  Now assume that prices are sticky.  What happens if the money supply growth rate increases? Most likely rates will still rise over time.  They might even rise immediately.  This is something like what occurred during the “Great Inflation.” And nominal rates rose during that episode.  Trend rates by definition involve long periods of time, and prices are flexible over long periods of time.

C.  Now go back to the flexible price assumption, but change the example from an increase in the growth rate of M to a one time increase in the money supply.  With flexible prices money should be neutral even in the short run (technically you also need flexible debt payments, i.e. indexed mortgages.)  If money is neutral then P and NGDP immediately rise in proportion to the rise in money, and nominal interest rates are unchanged.  This is what happens during a “currency reform.”

So we still have not reversed the results.  We still don’t have a clear and unambiguous example of easy money lowering nominal rates.  Sticky prices didn’t do the job, nor did switching to a one-time change in M.  But now let’s see what happens if we try both assumption B and assumption C at the same time:

D.  Assume a sticky price world and a one-time rise in M.  Now we have short run disequilibrium in the sense that the supply of money exceeds the demand for money, and since prices are sticky the price level has not risen enough to restore equilibrium.  But nominal interest rates can immediately fall, and this lowers the opportunity cost of holding cash and electronic reserves.  So we do reach an equilibrium in one sense, nominal rates adjust to equilibrate the supply and demand for money. Then in the long run prices adjust and nominal rates return to their original level.  Money is neutral in the long run.  We have a more comprehensive macroeconomic equilibrium in the long run, whereas we only reach a monetary equilibrium with the adjustments in nominal rates.

Thus to reverse the Kocherlakota/Williamson result that inflation and nominal rates move in the same direction, we need not one but two assumptions—a onetime rise in M and sticky prices. That’s why it seems like such a perplexing puzzle, the resolution is not simple.  And I think the reason Nick and the rest of us MMs keep harping on this issue is that it seems really important.  Consider a few facts:

1.  The declines in NGDP during 1929-33 and 2008-09 seemed to be lose-lose events.  Almost everyone lost money, from the poor to the banksters.  Hard to believe that special interest politics would have led Fed officials to intentionally engineer such catastrophes.

2.  Easier money would have prevented or at least greatly moderated these two disasters.

3.  One reason that money wasn’t made easier is that almost everyone (wrongly) assumed money was already incredibly easy.  If nominal rates had been 8% during 1930 or 2008, the Fed would have cut them sharply.

4.  Their assumption was wrong because very few people understand Nick’s nice little parable.

As so we must suffer.

PS.  Nick needs to collect all his little stories into a book.  I’m imagining the monetary equivalent of Italo Calvino’s Invisible Cities.  Something like “Imaginary Economies.”

Nick Rowe puts it all in perspective

The world is full of brilliant macroeconomists who fail to see the big picture.  Very few of them are able to step back and ask the sort of basic questions that Nick Rowe likes to pose:

Is the macroeconomic importance of finance, likewise, merely an artefact of a particular monetary policy?

What lesson should we learn from the recent recession? Is it that macroeconomists should pay more attention to finance? Or is it that we should change monetary policy?

In the past, the lesson we learned is that we should change monetary policy. When fluctuations in the demand and supply of gold became macroeconomically important, we dropped the gold standard. When fluctuations in the trade balance became macroeconomically important, we dropped fixed exchange rates. When fluctuations in the demand for money became macroeconomically important, we dropped money growth targeting. When fluctuations in finance became macroeconomically important, what should we do? Why should this time be different?

Now, the analogy is not exact. The Bank of Canada does not peg a nominal interest rate, except for 6-week periods. It adjusts that nominal interest rate peg eight times a year, as needed, to try to peg the CPI inflation rate. But interest rates, in particular the gap between market rates of interest and their corresponding “natural” rates of interest, play a key role in monetary policy tactics. And the monetary policy strategy, targeting 2% inflation, can be seen as pegging the real rate of interest on currency at minus 2%, which is the interest rate differential between holding currency and holding the CPI basket of goods. Current monetary policy uses one interest rate differential as an instrument to target a second interest rate differential.Current monetary policy, in both tactics and strategy, is all about interest rate differentials. And the theory of interest rate differentials is finance. A financial crisis is a big sudden change in interest rate differentials.

It doesn’t have to be this way.

In the past, we changed monetary policy to make the demand for gold, the trade balance, and the demand for money, macroeconomically unimportant.

We should now change monetary policy to make interest rate differentials macroeconomically unimportant.

Excellent.  I’d like to propose NGDP futures targeting as an alternative to the failed interest rate targeting regime.

PS.  In the comment section Vaidas raises some interesting objections to futures targeting, but I doubt any of them would be of economic significance. However it’s hard to be sure, as we have no empirical evidence for this sort of system, at least that I know of. Which is why we need to create the market!! Andy Harless says a NGDPLT target of central bank internal forecasts could do about as well, and I suspect he’s right.

PPS.  I started teaching today (after being on sabbatical last year) and thus am suddenly much busier. Blogging will probably be slower, but if it seems to dry up you can always check over at Econlog, where I will continue to guest blog.

Further thoughts on Robert Barro

Here’s an early indication of where Robert Barro is going in the 2011 article Paul Krugman criticized:

The overall prediction from regular economics is that an expansion of transfers, such as food stamps, decreases employment and, hence, gross domestic product (GDP). In regular economics, the central ideas involve incentives as the drivers of economic activity. Additional transfers to people with earnings below designated levels motivate less work effort by reducing the reward from working.

So that suggests that “regular economics” is based on incentives, like “supply-side” economics, or new classical, or RBC, or whatever.  So then I keep reading, waiting for the “to be sure, demand shocks also matter.”  But I never find it.  Nor do I find any clear statements to the effect that demand shocks don’t matter.  But what I do find over and over again is the sort of rhetoric used by economists who think that demand shocks don’t matter.

In fairness to Barro, the Keynesians are partly to blame here.  He’s wrong about AS/AD in my view, but he’s right that aggregate demand is an unfortunate term.  The AS/AD model only makes sense if you assume the AD curve is nominal GDP.  It has nothing to do with “demand.”  I don’t doubt that Barro would accept the notion that certain types of government expenditures would boost measured GDP (albeit without boosting welfare), so it would be possible to find something Barro wrote that suggests “demand matters.”  But in my view the two key issues here are whether fiscal stimulus in the form of tax cuts or transfers boosts NGDP, and the completely separate question of whether nominal shocks impact RGDP.

These two issues are often confused.  The first is about monetary offset, or crowding out, or Ricardian equivalence.  The second is about sticky wages and prices. In a long opinion piece Barro could have separated the two vastly different concepts with a single sentence, or even a phrase within a sentence, and yet he doesn’t do so.  So I can’t tell whether Krugman is right, but the way it’s written left me with the exact same impression as Krugman, as far as whether Barro thinks nominal shocks matter.

Tyler Cowen quoted from a 1995 paper where Barro was critical of the AS/AD model, but certainly understood the basic concept of aggregate demand.

If you read the paper, you will see three things.  First, Barro is fully aware of “AD-like” phenomena and does not reject that notion.  Second, Barro seems to prefer the IS-LM model to AS-AD, albeit with some caveats about possible false predictions of IS-LM and also noting in footnote two that he prefers his own presentation in his 1993 text.

I’m not sure what “AD-like phenomena” means to Tyler, but I interpret it as nominal shocks. Obviously Barro understands the concept of nominal shocks or monetary shocks or whatever you want to call them.  The question is whether he thinks they have important real effects.

Consider this ambiguous remark from the 2011 article:

How can it be right? Where was the market failure that allowed the government to improve things just by borrowing money and giving it to people? Keynes, in his “General Theory” (1936), was not so good at explaining why this worked, and subsequent generations of Keynesian economists (including my own youthful efforts) have not been more successful.

Now consider the following from Barro’s 1995 AS/AD paper:

Barro and Grossman [1971; 1976, Ch. 2] showed that the IS-LM model is a useful representation when nominal prices and wages are sticky at excessive levels . . .

His “youthful efforts” that he now seems to reject include the 1971 and 1976 papers with Grossman, don’t they?  The ones that argued IS/LM was useful when prices were sticky.

For people who believe AD is important, the “market failure” is easy to find–sticky wages and prices. And I recall that that assumption is included in Barro and Grossman’s work. But Barro now says that effort no longer seems successful.

Here’s how Barro ends the 2011 piece:

There are two ways to view Keynesian stimulus through transfer programs. It’s either a divine miracle””where one gets back more than one puts in””or else it’s the macroeconomic equivalent of bloodletting. Obviously, I lean toward the latter position, but I am still hoping for more empirical evidence.

That’s a disappointing conclusion.  Not because he’s wrong, but rather because if he’s right it’s written in a misleading way.  For instance, suppose Barro thought boosting NGDP through money creation could create jobs, but not via deficit spending (because of Ricardian equivalence/monetary offset/crowding out, etc.)  In that case he’d be correct, but I’d argue the conclusion is very misleading.  It’s written to appeal to right wing readers of the WSJ who also would consider the notion that printing money could create jobs to be a “divine miracle.”  Now it’s possible Barro is just throwing red meat to the WSJ readers, and that his actual views are far more nuanced and sensible. But then he ought to be prodded until he clearly spells ought what those nuanced views actually are.

When I wrote my previous post I was defending Krugman against the charge that he was unfair to Barro on the unemployment insurance issue.  Now Tyler Cowen criticizes Krugman for assuming that Barro doesn’t think AD matters.  If I was in Krugman’s shoes I would have been more careful, as I don’t know that Krugman expressed Barro’s views correctly.  Tyler might be right.  But I will say that my reading of the article suggests that Barro was trying to give his readers the impression that incentive effects determine employment, not demand-side effects.  Maybe he thinks both do.  My hunch is that he thinks sudden nominal shocks have a modest effect at certain times, but explain essentially none of the current unemployment rate.  In that case Krugman’s charge was a bit too extreme, but not far off.  I suppose I sympathized with Krugman a bit because I found the Barro article to be so frustratingly vague.

Let’s get away from the specifics here and look at the picture more broadly.  Here’s what I see over and over again:

1.  Conservative economists who are contemptuous of “Keynesian economics” in general and fiscal stimulus in particular.  Fine, so am I.

2.  Conservative economists who have much less to say about monetary stimulus, but seem quite skeptical.

3.  A blurring of the distinction between policy not boosting NGDP, and NGDP failing to boost RGDP (radically different ideas.)

4.  If conservatives really believe “the real problem is real” then they shouldn’t be contemptuous of “Keynesian economics,” they should be contemptuous of “the entire Keynesian/Milton Friedman strand of economics.”  After all, Friedman also thought nominal shocks had long lasting effects on unemployment. But they never add Friedman’s name.

5.  If conservatives want to discriminate between Keynesian fiscal views and Friedman’s monetarist views, then they ought to advocate monetary stimulus after they dismiss fiscal stimulus as bloodletting.  But they never seem to do that.  (In other words they should become market monetarists.)

6.  In general, when a group of people shy away from an issue, it makes me think they aren’t entirely confident that their views hold up to empirical scrutiny.  Maybe that’s why Barro is so forceful on fiscal (when he has studies to back him up) but silent on whether more nominal spending would help right now (where the empirical evidence for rejecting monetary stimulus is far weaker in my view.)

Tyler suggests we ought to be able to find out what Barro thinks by looking at his 40 year track record of published work.  I’m a big fan of his work; I once used his textbook for a grad class in macro.  I’ve read the 1995 article.  He’s brilliant.  But I’ve found that almost all economists seem to have suddenly abandoned what they believed in 2007, and adopted a radically new economic model in 2009.  New Keynesians became old Keynesians.  Monetarists became Austrians.  How do I know Barro is any different from the others?  Recall that the standard view in 2007 was that the Fed drove nominal spending and that the zero bound was not a barrier to monetary stimulus.  This was mainstream economics in 2007.  And now it’s been abandoned for no apparent reason other than that if it were true it would imply the economics profession caused the Great Recession of 2008-09.