Archive for September 2012


Stable NGDP growth is a public good

We (almost) all benefit from low and stable NGDP growth.  But at an individual level we have no incentive to behave in such a way as to produce low and stable NGDP growth.  It’s a public good.

Tyler Cowen recently pointed out (correctly) that if we hold the money supply constant, the private sector could manufacture more NGDP growth by acting in such a way as to boost velocity.  That’s true.  Whether it would actually work depends on how the Fed responds to a rise in velocity.  That’s much more debatable, but I’m willing to entertain Tyler’s claim that we are in a place right now where expected NGDP growth is a bit less than the Fed would prefer.  (Obviously if true that would also imply a role for fiscal stimulus.)

But I’d like to put aside the monetary offset issue and focus on the public good problem.  It’s also true that it’s theoretically possible for everyone in the world to behave as if they are utilitarians, which (if they believe in global warming) would lead them to pretend that there is a carbon tax in effect, and base their consumption and production decisions on that assumption.  We typically assume that people are not that altruistic, which is why most economists favor a carbon tax.  (I am pretty sure that Tyler does as well.)  If we have a NGDP problem, the private sector could solve it by manufacturing more NGDP.  But they’d have no incentive to do so.

Clearly Tyler is aware of this public good problem, so he might be thinking about non-monetary policies that would encourage faster velocity, overcoming the public good problem.  The easiest way to solve the NGDP public good problem is to print more money (or even better, have the Fed boost V by setting a higher NGDP target.)  But there are other ways.  You could do fiscal stimulus.  You could replace the corporate income tax with a higher payroll tax on upper incomes, which would encourage more investment and boost V.  You could reduce burdensome regulations.

But these other ways of boosting (or “manufacturing”) NGDP are completely unrelated to the sticky-wage issue.  Perhaps I misread Tyler, but he seems to suggest that even if we need more NGDP, we might be better off thinking in terms of a lack of privately manufactured NGDP, not tight money.  But if (as Tyler seems to suggest) wage stickiness is not the problem right now, then having more NGDP wouldn’t help at all.  Rather we’d need supply-side policies that boosted RGDP, for any given level of NGDP.  So I find a puzzling “mixed message” in a post that talks about how the private sector can manufacture NGDP, and also expresses skepticism about the sticky-wage problem.

I may well have misread Tyler’s argument, so let me suggest how I would have re-framed the argument I think he was making:

1.  Suppose that the economy was subject to real shocks.

2.  Suppose nominal shocks had no real effects.

3.  Suppose the Fed successfully targeted inflation at 2%.

In that case movements in NGDP and RGDP would be perfectly correlated, even though there was no causal relation running from NGDP to RGDP.  NGDP shocks would not be as important as they seem, and sticky wages would not be the problem.

For quite some time I’ve suspected that my success in promoting NGDP targeting was partly undeserved.  The famous L-shaped NGDP graph for the last 10 years is a very powerful visual persuader.  Maybe too powerful.  So you might ask why I believe NGDP is so important, given that assumptions 1 through 3 make the hypothesis highly questionable.  Here’s why:

1.  During periods when the Fed wasn’t targeting inflation, we had lots of natural experiments with wild and crazy monetary shocks.  They seemed to have real effects.

2.  I put a lot of weight on the stylized fact that wages and prices are sticky, which means nominal shocks should have real effects (on theoretical grounds.)

3.  Asset markets behave in a way that seems consistent with the view that investors currently believe the real problem is (partly) nominal.  Equity investors presumably favor monetary stimulus right now because they (correctly) believe that wages are pretty sticky right now.

Tyler agrees with me on some of this, which is why he’s argued in the past that while monetary stimulus is not a panacea, it’s worth shooting for at least a bit higher NGDP growth.  We differ on the relative importance of real and nominal factors, and how long it takes for wages to adjust.

PS.  The private sector can also “manufacture” lower NGDP, if the Fed is passive.  For instance if the Fed had been passive (in terms of the monetary base) after the Soviet bloc broke up, then it’s very possible that the hoarding of dollars in that region would have led to a Great Depression in the US, comparable to the 1930s.

PPS.  I agree with David Beckworth and George Selgin that NGDP is a 100% nominal variable.  NGDP is “the real thing,” whereas P and Y are simply data points pulled out of the air by Washington bureaucrats.  There’s no good theory to back up their efforts, as it’s not even clear what “inflation” is supposed to measure.  (Suppose a $25,000 2012 Camry is just as good as a $25,000 Rolls-Royce sold in 1948.  Does that mean there has been no nominal increase in car prices since 1948?  What does “car price inflation” mean?) Tyler may have had in mind something like my example where despite the fact that NGDP and RGDP are highly correlated, there is no causal relationship.

PPPS.  Commenters sometimes tell me that real world businessmen don’t care about NGDP.  In fact they care a lot.  I’m sure GM execs often sit around a table discussing how “the economy” will be next year, which actually means “how much consumers in aggregate will spend (in nominal terms) on cars next year.”  The execs already know pretty much what it will cost to manufacture cars next year, even in nominal terms.  So expected nominal spending on cars in 2013 is extremely important to GM execs when thinking about how much to invest in expanding their facilities.  Undoubtedly they break the forecasts down into unit sales and average prices, but both of those variables are very important to them.

Comments on Selgin and White

Lots of people, including both Austrians and market monetarists, use the concept of “monetary disequilibrium” to analyze the role of monetary policy shocks in the business cycle.  I generally stay away from that framework.  It’s not that monetary disequilibrium never occurs, but rather that I think it lasts for just a brief period.  The bigger problem is that equilibrium is re-established at a new and different NGDP level, which creates disequilibrium in the labor market.  So I see labor market disequilibrium as the essence of the business cycle.

Larry White recently had this to say:

Market Monetarists who have been celebrating the Fed’s recent announcement of open-ended monetary expansion (“QE3′) seem to believe that Sumner’s 2009 diagnosis still applies. But what is the evidence for believing that there is still, three years later, an unsatisfied excess demand for money? Today (September 2012 over September 2011) real income is growing at around 2% per year, and the price index (GDP deflator) is rising at around 2%. If the evidence for thinking that there is still an unsatisfied excess demand for money is simply that we’re having a weak recovery, then as Eli Dourado has pointed out, this is assuming what needs to be proved.

.  .  .

While saluting Sumner 2009, like Dourado I favor an alternative view of 2012: the weak recovery today has more to do with difficulties of real adjustment. The nominal-problems-only diagnosis ignores real malinvestments during the housing boom that have permanently lowered our potential real GDP path. It also ignores the possibility that the “natural” rate of unemployment has been hiked by the extension of unemployment benefits. And it ignores the depressing effect of increased regime uncertainty.

To prefer 5% to the current 4% nominal GDP growth going forward, and a fortiori to ask for a burst of money creation to get us back to the previous 5% bubble path, is to ask for chronically higher monetary expansion and inflation that will do more harm than good.

I don’t know of any theoretical models where steady 5% NGDP growth would create bubbles.  And even if bubbles did occur, I can’t imagine why they would represent a public policy problem if NGDP continued to grow at 5%.  I’d also point out that the US has experienced 3 major equity or residential real estate bubbles in periods of relatively low inflation and NGDP growth (1929, 2000, 2006) and zero major bubbles in periods with high inflation and NGDP growth (1968-81).  The 1987 stock market bubble was an intermediate case (which did zero harm, as NGDP continued growing after the bubble.)

However I do agree with some of Larry’s observations.  As I mentioned in my previous post, you’d expect the extended UI benefits to have raised the natural rate of unemployment (although I doubt it’s risen to 8.1%.”)  I also agree that the current situation is different from 2009. In 2009 I advocated going all the way back to the old trend line.  I currently favor going about 1/3 of the way back.  If we keep on the same track for a few more years I’ll through in the towel and advocate starting a new 5% trend line from where we are.  So my policy views have changed to reflect the changing nature of the crisis, and the fact that some wage adjustment has occurred.

In reply to David Beckworth, George Selgin makes the following claim:

All these appeals to different measures of the money stock as offering evidence as to the extent to which money is in short supply or has been exposed to demand shocks really are, or should be, considered quite beside the point in the MM and other nominal spending targeting frameworks. After all, nominal spending targeting makes sense precisely to the extent that fluctuations in nominal spending serve as a good indicator of money shortages or surpluses. So who cares what M2 or M3 or m$ or other still fancier M measures are of have been up to? If spending has remained stable, the presumption is that the economy has been getting all the liquidity or exchange media it needs, and that it is therefore not tending to ride up or down a short-run Philips curve. It is precisely because NGDP targeting and similar schemes dispense with the need to track particular monetary aggregates, or worry about the stability of demand for them, while still sticking to a nominal target, that they constitute an alternative and appealing alternative to conventional “monetarist” rules.

So, when it comes to demonstrating that money is or has been in short supply, a consistent Market Monetarist has no reason to appeal to the behavior of any measure of M. What matters is whether a plausible case can be made that spending is too low, or has been growing too slowly. That of course leads to thorny questions concerning the choice of an ideal growth rate and, what amounts in the short run to the same question, the fitting of a trend to past data with the aim of finding the once that would have been most conducive to the avoidance of booth booms and busts. This latter task, it seems to me, is not quite as simple a matter as some MM’s have made it out to be.

I completely agree about the redundancy of money data in the MM model.  NGDP is a sufficient statistic for any problems with monetary policy.  And I agree that finding the optimal trend line for NGDP is a non-trivial exercise.  All I would add is that showing the folly of the actual 2007-2012 path of NGDP is a trivial exercise.  I am pretty sure that George agrees with me on that point.

A pragmatic approach to monetary policy

In his recent post, Eli Dourado raises a number of interesting points:

The empirical point is summed up in the graph below. NGDP grew around 5 percent per year until around 2008, and then it fell, and then it grew at around 5 percent””or slightly less””per year again beginning in mid 2009. These facts are well known, but I bring them up here because they do constrain the kind of stories we can tell about the economy. Any story you tell has to contain a one-time shock that ended years ago, and it has to be consistent with NGDP that has grown at about the same rate over the last 3 years as it did before the shock arrived.

To be precise, NGDP has grown at a 4.1% rate over the last three years. Normally NGDP would grow faster than trend during a recovery, which means the Fed added insult to injury.  If NGDP had grown at 5%, then RGDP would have also grown faster, and unemployment would be lower than 8.1%.  But I accept Dourado’s broader point—you’d have expected wages to have mostly adjusted by now, and thus you’d have expected more progress on unemployment, even with a 4.1% NGDP growth rate. But the fact of the matter is that nominal wage growth has not slowed in tandem with NGDP growth.  The facts are incontrovertible:

Take a close look at that graph (courtesy of George Selgin.)  That’s not just roughly what the sticky wage theory would have predicted, it’s EXACTLY what the sticky wage theory would have predicted.  The ratio of NGDP to nominal wages ratio of nominal wages to NGDP soared in 2008-09, and unemployment soared from 5% to 10%.  Then nominal wage growth slowed modestly, and this slightly reduced the ratio of wages to NGDP.  And of course unemployment fell slightly, from 10% to 8.1%.  The theory fits the data perfectly.

[The post initially got the W/NGDP ratio backwards.]

That’s leaves Dourado with just one valid complaint; market monetarists haven’t explained why nominal wage growth fell only modestly, despite high unemployment.  But here’s the problem, the brightest minds in macro have been unable to find plausible microfoundations for the standard macro model.  I’ve offered several suggestions:

1.  The zero lower bound produced by money illusion, combined with the fact that wages are still rising in healthy sectors, or sectors shielded from market forces.

2.  The 40% boost in the minimum wage at the beginning of the recession.

3.  The unusually long extension of UI benefits, which made our labor market more “European.”

4.  A long term trend toward a growing share of national income going to capital, which makes the need for wage reduction even greater.  This also helps explain why corporate profits are doing well in the recession.

But even I don’t think these are completely persuasive.  So I try to take a pragmatic view of all this.  If the best minds in the profession can’t come up with plausible microfoundations, then how am I supposed to do so?  Why should I even try?  I can’t even figure out why the NFL says that wasn’t an interception last night!

Given our limited knowledge, what sort of advice should we give our monetary policymakers?  Let’s break the problem down into pieces:

A policy of NGDPLT is a nice safe choice when there’s lots of uncertainty about which macro model is correct.  I gather that Dourado agrees on this point.

That means the entire debate is on where to start the new trend line, if the Fed adopts that sort of plan.  Some favor going all the way back to the pre-2008 trend line, which would call for a period of rapid NGDP growth.  Critics say that would risk another destabilizing boom.

Others say start the new trend line right here, and aim for 5% NGDP growth going forward.  To that group I say that under current policy we are likely to fall a bit short of 5% NGDP growth.  So if that’s your position then you should not only favor QE3, you should be calling for QE4.

I believe the balance of factors suggests that the safest choice is to go about 1/3 of the way back to the old trend line, and then level off at 5%.  That’s based on the following pragmatic judgments:

1.  I fully accept that the natural rate of unemployment might have risen to 8.1%, as it has in France, Spain, and Italy.  But it’s by no means clear this has occurred.  In the US the natural rate has probably been in the 4.5% to 6.5% for over 70 years.  It’s never been close to 8.1%.  So I regard 8.1% as very unlikely.

2.  If it did rise to 8.1%, the most likely explanation is that the policies I mentioned above (minimum wage increase, extended UI, etc) caused the increase.  But monetary stimulus would help on both fronts; reducing the real minimum wage (which never would have been passed had Congress know how little NGDP growth we were going to get) and also causing Congress to reduce the maximum UI benefit more quickly, as they do after every previous recovery from a recession.

3.  David Glasner showed that equity markets are clearly rooting for higher inflation.  That was true in the 1930s, but generally is not true during periods when the economy is close to the natural rate.

4.  Ten year bonds yield 1.7%, suggesting we are much more likely to err on the side of excessively slow NGDP growth, a la Japan, than excessively rapid growth.

5.  We know that nominal wages are very sticky, and that unemployment closely tracks the ratio of wages to NGDP.  It is POSSIBLE that faster NGDP growth would merely lead to higher wages, and no gain in jobs, but how likely does that seem in this sort of labor market?

6.  Level targeting calls for a catch-up period, and NGDP growth over the last 4 years has been the slowest since Herbert Hoover was President, about 2% per year.  Admittedly the case for “catch-up” gets weaker as time goes by (a point on which I agree with Dourado) but I don’t think that the catch up argument has completely gone away.

That’s why I end up favoring going 1/3 of the way back to the old trend line.  I balance Dourado’s very good arguments for caution, with what I see as very powerful arguments that the economy is still depressed by a lack of demand.

Ultimately it’s a judgment call, and I’d like to reiterate that in the long run the issues we agree on, (NGDP targeting) are far more important than the issues where we disagree (where to start the trend line.)

Regarding the Berger paper mentioned by Dourado, I don’t have time tonight to take a close look, but here are a few initial reactions:

The previous two recessions saw small drops in NGDP growth, and slower than normal recoveries.  So there’s really no big mystery to explain, except to the extent that productivity behaved abnormally.  I’ll accept Berger’s claim that it did, but it really doesn’t explain much, at least in this recession.  The unemployment rate is not surprisingly high, it’s surprisingly low given the sub-3% RGDP growth since unemployment peaked at 10%.  Indeed it’s not clear why unemployment fell at all.  (I’ve had posts called “our job-filled non-recovery.”) And in the 2001 recession unemployment peaked at 6.3%, which probably isn’t much above the natural rate.  I wasn’t complaining about tight money in 2001, it’s the recent recession that is the outlier.

So yes, things change over time and cycles today are different from the 1950s, just as they were different from the 1920s.  But I fail to see how sudden drop in NGDP and RGDP could not be quickly reversed.  It’s a mistake to think in terms of firms rehiring the workers they laid off.  My understanding is that this doesn’t occur all that often.  In a normal year where there is a net job creation of 1 million, you might have 29 million jobs lost and 30 million jobs gained.  Most laid off workers go on to different firms.  With that massive churn, it’s not at all difficult to create a couple million more net jobs if the economy starts from a depressed condition and the Fed makes sure we have rapid NGDP growth, as in 1983-84.

P.S.  Eli also had this to say:

I assume that when he says “cutting-edge” he is not referring to the papers cited in Ryan’s post, since those are both from the 1980s.

Well at least he’s more polite than other bloggers.  For a dinosaur like me who went to grad school in the 1970s, anything after 1980 is “cutting-edge research.”  Remember that Steven Wright joke; “About 20 years ago . . . no wait, it was just last week.”  That’s me—in reverse.

PPS.  Ramesh Ponnuru has one of the best pieces on NGDP targeting that I’ve ever seen by a journalist.

The Price of (limited) Success

The recent success of market monetarism has attracted a lot of attention.  People who attract a lot of praise (especially excessive praise) will naturally attract their fair share of critics, particularly when their credentials are  . . .  questionable.  Still, I suppose it beats the alternative.

There’s a natural human tendency to want to boil down someone’s message, to make it easier to visualize what makes a group distinctive. I do think it is fair to say that market monetarists are obsessed with NGDP (at least I think it’s fair in my case, I can’t speak for others.)  But I don’t think this statement by Eli Dourado fairly characterizes the market monetarist position (again, speaking for myself):

I would like to see a greater emphasis in the blogosphere on understanding stylized facts about recessions, a greater willingness to explore micro phenomena (even if we are not using fully microfounded models), and more macro-ecumenicism. No one school of macro has it all figured out, and that includes market monetarism. There is enough ambiguity in our current situation that reasonable people can disagree about what is going on. But I don’t think that reasonable people can be totally certain that all we need is more nominal stimulus.

I’m not totally certain that my brain is not in a vat right now.

Seriously, I consider myself a moderate supply-sider, which puts me to the right of most economists.   That means I certainly don’t think “more NGDP” is all we need.   We have structural problems. To get back to where we were before the recession we’d also need to:

1.  Reduce the minimum wage rate, which was raised by over 40% right on the eve of the recession.

2.  Cut the maximum unemployment insurance benefit back to 26 weeks, from a peak of 99 weeks (and now 79 weeks.)

Other reforms would also help to reduce unemployment, such as legalizing self-employment in job categories that currently have occupational licensing laws.  I must have discussed the first two 100 times in my blog, if not more.  I would add that higher NGDP would reduce the real minimum wage, and would encourage Congress to lower the maximum unemployment insurance benefit, so even “structural problems” can be partly ameliorated by demand increases.  I frequently talk about how supply and demand are “entangled.”  From my point of view it’s the rest of the profession that is simplistic, treating AS and AD as if they were independent, like S&D.

I would add that under NGDP targeting it would make no difference whether unemployment is structural or not—policy would be exactly the same in either case.  That’s why it’s so important to get a rule-based, non-discretionary policy regime.  As long as we are having this debate, then monetary policy has failed.  Eli Dourado raises some other good arguments, which I’ll address in a separate post.

Noal Smith has a satire on the various internet schools of thought, including market monetarists:

Favorite blog: The Money Illusion

Favorite dead economist: No one. The spot is being reserved for Scott Sumner, along with thousands of life-sized terra cotta grad students.

Will appear in response to posts regarding: Monetary policy, macroeconomics, any word containing the letters “NGDP”

Craziest idea: Pegging the monetary policy of the world’s leading nation to an obscure and highly illiquid futures market

Special attack: NGDP-style kung fu

Secret weakness: Supply shocks

Notes: Having now effectively swayed the Federal Reserve and won essentially all of the Econ Blogosphere to their way of thinking, the Market Monetarists can hardly be classified as “trolls” any longer…

I know that it’s a great honor to be satirized by Noah Smith, but at the risk of seeming churlish I’m going to get all defensive here.  To be effective, satire must exaggerate certain recognizable characteristics. Let’s take the easiest case first, supply-shocks.  Supply shocks are the great weakness of inflation targeting, and what has drawn great minds like Bennett McCallum to NGDP targeting.  If it’s a secret weakness, the secret has been incredibly well kept.

And what can we make of the “highly illiquid futures market” comment? The central bank would promise to buy and sell unlimited quantities of NGDP futures at zero transactions costs at the target NGDP level.  That makes my proposed NGDP futures market the most liquid asset market in all of world history.  If people find Noah Smith’s market monetarism satire to be funny, it just shows how much work we have to do to educate the public as to what NGDP targeting is all about.

On the other hand Noah Smith’s barbs aimed at the other internet schools of thought are all completely accurate and hilariously funny. Highly recommended.  Well done Noah!  🙂

HT:  Tyler Cowen.

PS.  Given my current pace I’ll be dead soon enough, so start preparing that terra cotta army.

How far we’ve come

Here’s the respected Wolfgang Münchau, columnist at the Financial Times:

This is a debate about nominal income targeting, where a central bank no longer stabilises the inflation rate directly but focuses instead on stabilising nominal gross domestic product. You can think of nominal GDP as the sum of real GDP and inflation. If real growth falls, the central bank would thus have to drive up inflation. Conversely, if real growth rises, the central bank would have to bear down on inflation much harder than it would do under the pure inflation targeting regime used by central banks such as the ECB.

.   .   .

It was probably inevitable that this long-lingering clash of economic cultures has finally come out into the open. As Europe’s recession gets worse, the ECB will soon have to adopt similar measures to those used by the Fed. It may even end up adopting a nominal income target, explicitly or implicitly, for the simple reason that the crisis in the eurozone is ultimately insoluble without annual nominal growth of at least 5 per cent. And I cannot see the Bundesbank support for any of it.

Nearly 4 years ago I and a few others started arguing that the crisis was being misdiagnosed by the economics establishment.  Market monetarists argued that while banks did make some foolish loans, the deeper cause of the global financial crisis was a lack of nominal income; the resource that people, businesses, and governments have available for repaying nominal debts.  The collapse in nominal income in the US and Europe was larger than anything since the 1930s, and hence most pundits and policymakers had no experience dealing with this situation.

It’s amazing how far we’ve come in less than 4 years.  NGDP is the hottest idea in macroeconomics, and the lack of NGDP growth is seen as the root cause of the financial crisis.  I recall in 2009 how commenters were incredulous when I said “NGDP,” not reckless banking, was behind the financial crisis.  They thought I was crazy.  Now it’s practically conventional wisdom.  This means that all those “policy implications of 2008” from both the left (who blamed the banks) and the right (who blamed the regulators) will have to be completely re-thought.  It’s about time.

HT:  JimP

PS.  I finally figured out how to type an umlaut—a sign of my respect for Mr. Münchau.