Readers who weren’t with me back in 2009 might not know why I chose the moniker “TheMoneyIllusion.”  From the beginning I realized that the term had multiple connotations, almost all of which dovetailed nicely with the content of my blog:

1.  In 1928 Irving Fisher wrote a whole book entitled “The Money Illusion,” full of delightful examples.  He was referring to the tendency of people to confuse real and nominal variables, or perhaps it would be more accurate to say people tend to see nominal variables as being in some sense “real.”  I still recall when we got no pay increase back in 2009, a year of minus 1% inflation.  My colleagues at Bentley greeted the anouncement very differently from when we got a 4% increase during a year of 3% inflation.  I see money illusion everywhere I look, so I didn’t need much convincing.

2.  Another form of money illusion is confusion about the nature of monetary policy.  The liquidity effect is actually just an epiphenomenon, and yet most people see it as not just a side effect, but rather as monetary policy itself.  And “most people” includes the Fed.  For this reason, policy is often perceived as being ineffective at the zero bound, unless perhaps longer term rates can be lowered via QE.  In fact, in the long run more money lowers the value of money (value in terms of the share of NGDP which can be bought with each dollar) for exactly the same reason that more apples lowers the value of apples.  (Admittedly the problem of expectations is more complicated in the money market.)

These two forms of money illusion have some interesting parallels.  In both cases the problem becomes much more severe at the zero rate bound.  For psychological reasons that are quite frankly irrational, both workers and central banks behave bizarrely at the zero bound.  Workers start being much more resistent to wage cuts necessary to restore labor market equilibrium.  And central banks become much more resistent to monetary policy steps needed to keep NGDP on track.  Their superstitious fears of the number zero (which after all is a foreign concept borrowed from the Islamic world) causes them to behave in ways that are both personally and socially destructive.

In terms of the previous post, the zero interest rate boundary contributes to the Fed’s failure to hit its NGDP target, causing NGDP instability.  Then the zero wage change boundary takes that NGDP instability and converts it into relative wage instability (unstable W/NGDP).

3.  There’s a third form of money illusion, to which economists are also subject.  Economists have just as much trouble as the Fed does in identifying the stance of monetary policy.  Because they think ultra-tight money is actually ultra-easy money, then are often not able to perceive the impact of very tight money.  This means that they find some other explanation for economic distress, such as financial instability.  Fisher said the business cycle is nothing more than the “dance of the dollar.”  Fisher defined monetary policy in terms of the price of money (1/P), not the rental cost of money (i.)  Because modern economists see tight money as high interest rates they confuse cause and effect, not understanding that financial distress is a symptom of a monetary policy-induced recession, not the cause.

I’ve followed Fisher’s approach, except that I believe the inverse of NGDP is a more useful indicator of the stance of monetary policy than the inverse of the price level.  One of my goals in the previous post is to get people to think about my message in a different way.  I’d rather people not think of me as calling for the Fed to stabilize NGDP (although obviously I am calling for that) but rather to see me as calling for the stabilization of its inverse—i.e. 1/NGDP.  Obviously those two goals are identical from a purely mathematical perspective.  So why do I prefer 1/NGDP?

I’ve often seen commenters, particularly those of the Austrian persuasion, complain that I favor some sort of central planning.  I’ve always found this claim to be rather bizarre, but haven’t been able to rebut it in a persuasive fashion.  If we talk about targeting 1/NGDP, then it will become clearer that all I want the Fed to do is stabilize the value of money.  Since they have a monopoly on the production of base money, is it really that unreasonable (even from a libertarian perspective) to ask them to at least try to stabilize the value of cash in terms of the share of NGDP that can be bought with each dollar?

This blog is about controlling the value of money, and letting the rest of the economy be determined by market forces.  I notice that critics of my blog often make the following claims:

1.  They claim the free market economy is so delicate and fragile that even the very small changes in the per capita money supply, or price level, or NGDP, during the 1920s somehow brought on major economic dislocation.

2.  Or, at the other extreme, they claim that even massive injections of money couldn’t possible be expected to create jobs.  Money just causes inflation, there are no real effects.

Oddly, many of my critics seem to hold both views at once.

And finally, you may wonder why I didn’t simply call the blog “Money Illusion,” which would sound more hip.  I seem to recall a scene in The Social Network where Justin Timberlake was discussing “The Facebook” with Zuckerberg.  Then he got up and left with a couple of starlets on his arms, and turned with one final piece of advice:  “By the way, drop the ‘the’, just Facebook.”

When I started out I tried to buy “MoneyIllusion,” but it cost $1800.  So then I checked “TheMoneyIllusion” and it was $12.  Keep in mind that I’m not as rich as either Timberlake or Zuckerberg.  If it had also been $1800, you’d be reading something like “TheseMoneyIllusions,” or God forbid “TheMoneyAllusion” (yes, I’m that cheap) so count your blessings.

PS.  This article discusses the actual origins of zero.  It was invented multiple times, and as usual not by the people who got credit (which were the Arabs.)  Interestingly, however, it appears to have been first invented in what is now an Arab country (Iraq.)  Didn’t the Iraqis also invent writing, law, money, literature, the calendar, and pretty much everything else?

PPS.  Ambrose Evans-Pritchard once called me the “eminence grise” of market monetarism.  I’m going to return the favor, calling him the poet laureate of market monetarism.

HT:  Lars Christensen

Macro is basically composed of three fields:

1.  Long run RGDP growth.

2.  Nominal macro variables in the long run (P, NGDP, i, E, etc)

3.  Business cycles.

Money is all of part two, and half of part three.  That means monetary economics is half of macro.  It’s also far and away the more interesting half.  The real side is a long, boring and random list of factors that affect RGDP (capital, labor, technology, good governance, natural disasters, etc.  Monetary economics is an interesting, counterintuitive and elegant structure that fits together wonderfully, using concepts like money neutrality and superneutrality to generate the QTM, the Fisher Effect, PPP and lots of other models filled with beautiful symmetries.

A work study student helped me create a couple of graphs that form my “model” of the macroeconomy.  I’ll use them to explain a few issues.  First a model of NGDP determination:

This is similar to the regular model of money supply and demand in Mankiw’s textbook, except that the value of money is defined as 1/NGDP, not 1/P.  I.e., the value of money is defined as the share of NGDP that can be bought with a single dollar.  As with the traditional model, the demand curve is a rectangular hyperbola. In this case the area of a rectangle under any given point on the demand curve is M* (1/NGDP), or M/NGDP, not M/P as on the traditional graph.  Thus a given demand for money is defined as a given “Cambridge k” (which is the inverse of velocity.)

Market monetarists differ from traditional monetarists in that we see monetary policy as shifting both money supply and money demand.  As in traditional monetarism, any one-time and permanent increase in the monetary base increases NGDP in proportion.    That would be like a shift from point A to point C.  (Note that 1/NGDP falls at point C, hence NGDP rises.)  On the other hand setting a higher NGDP growth target would raise NGDP by reducing the demand for base money–shifting the demand curve to the left.  Ditto for a lower IOR.

If NGDP is slow to adjust to a permanent rise in the base, interest rates may fall in the short run (the liquidity effect) and you initially go to point B as the lower interest rates cause the demand for money to rise.  Then over time NGDP starts rising, and you gradually move from point B to point C.  Actually that’s not quite right, as even in the very short run NGDP will respond somewhat to monetary stimulus.  After all, some prices (like food, metals and crude oil) are highly flexible.  So the short run equilibrium is actually slightly below point B.  NGDP rises immediately, by a small amount.

If the monetary stimulus is not expected to be permanent then you will see interest rates fall and you would go to point B in the short run.  That’s what happened in Japan between 2003 and 2006.  Then in 2006 the MB was reduced by 20% and Japan went back toward point A.

OK, so that’s my model for explaining one third of macro; inflation, NGDP, nominal interest rates, nominal exchange rates, etc.  It’s a monetary model.  Then we add the assumption that nominal wages are sticky in the short run:

This one is kind of tricky to explain.  I’ve assumed the long run supply of labor is perfectly inelastic, but nothing of importance would change if you made it slightly positively sloped, or even backward bending.  The key assumption here is that, unlike in the money market, we are usually not at equilibrium.  Let’s start with the case where the Fed reduces NGDP unexpectedly, causing W/NGDP to rise unexpectedly to point B.  We assume that employment is demand-determined when we are above equilibrium.  Lots of people want to work at that wage, but employers determine how many actually get jobs.  The gap between supply and demand is cyclical unemployment, which is currently about 2.6% in the US.  The other 5.6% is the natural rate of unemployment, and occurs even if we are at an equilibrium relative wage rate.  In terms of hours, unemployment is even higher, as many who want to work full time are working part time, or so discouraged they leave the labor force.

If monetary policy pushes NGDP above expectations (from when labor contracts were signed), then we have an overheating economy.  In this case the actual level of excess hours worked is not necessarily equal to the gap between supply and demand, as there may be a labor shortage at point C.  In that case the level of hours worked would lie somewhere between the supply and demand curve, but the qualitative results would be the same—too much employment.  In a richer and more realistic model employment would depend on more than just unexpected NGDP shocks; because of money illusion workers may be reluctant to accept nominal wage cuts.

One problem with the preceding model is that it implies that equilibrium is “best” and that the public suffers if there is either too much employment or too little.  And yet it doesn’t seem that way in the real world.  It seems like boom periods are much better than recession periods.  And they are!  That’s because this simple model abstracts from all sorts of real world policy distortions.  The vast majority of government intervention tends to discourage employment (minimum wages, marginal tax rates, welfare, unemployment comp., occupational licensing laws, etc, etc.)  Thus when we are on the long run labor supply curve the level of hours worked is actually far too low, even in the US.  It’s even worse in Europe, which explains why Europeans are less happy than Americans (see if that gets some comments.)

So there’s a reason why booms feel like good times, even though the simple model says we are working too much.  The stickiness of wages temporarily pushes us toward the socially optimal level of hours worked.  We produce more goodies, and collectively we consume more as well.  We are happier.  Alas, we can’t stay there, as wages adjust and we go back to the long run equilibrium.  Even worse, attempts to use monetary policy to create booms tends to end up creating more business cycles, which makes people less happy.  Better to do as the Aussies do and throw Schumpeter in the trash can.  End business cycles with stable NGDP growth, or at least level targeting if there are unavoidable short term blips like 2009.

I agree with Matt Yglesias about 99% of the time on monetary policy.  The one area I slightly disagree is that I think he’s too inclined to view the Great Moderation as being a period of excessively tight monetary policy.  That might be slightly true, as inflation and NGDP growth did decline very gradually during that period.  But the decline was so slight it doesn’t seem to me that it mattered much, until 2008.  I wonder if he isn’t relying on his intuition that it’s better to err on the side of higher inflation.  In the short run that’s true, but only because of all the other policy distortions.  This is actually a very counterintuitive point, and is similar to a policy debate I occasionally have with Karl Smith.  I insist that there is no first order income effect for tax changes, and hence that only the substitution effect matters.  That means taxes unambiguously reduce hours worked.  It’s really hard to see at the individual level, because the average American feels that higher taxes make him worse off, even as a first order effect.  But that can’t be true, as the tax money is not just destroyed.  Similarly, if wage stickiness is making you work harder than you’d prefer, it seems like it would be making you worse off.  But what you don’t see is that the average person working harder also gives more tax money to the government, and that at least some of this money comes back to you (you the average American) in the form of Social Security, better schools for your kids, or better roads.

Booms feel good.  It’s a nice example of the fallacy of composition.

PS.  Mark Sadowski sent me the following graph of unemployment (red) and nominal wages divided by per capita NGDP:

It breaks my heart to think about all the garbage we force our students to learn in macro.  Is inflation good or bad for growth?  They leave the course not having a clue.  If only we’d teach a simple monetary model of NGDP determination, and then a simple W/NGDP model of business cycles, using the sticky wage assumption.  They might actually be able to understand that model.  Of course if our politicians understood the model then they never would have allowed the Fed and ECB to create the Great Recession.  No more macro.  Economics would become a one semester course, and we’d loose lots of teaching jobs.

PS.  Matt Yglesias has a great post on why non-sticky wage models of unemployment (such as a loss of wealth) are a dead end.  They can’t explain why people would work less:

It is both true that we are not as wealthy as we thought we were and that there’s a lot of joblessness in the United States, but I struggle to grasp a model in which the former causes the latter. Imagine a reverse situation. A town full of working class people sees its unemployment rate suddenly shoot up from 11 percent to 27 percent. Concurrently, it turns out that the town’s residents were much wealthier than they thought they were””each one of them actually had a check for $1 million sitting in their pockets. We might say it’s pretty clear what’s happened here. These folks are wealthier than they thought they were so they raised their reserve wage. But then suppose it turns out the checks were fraudulent and they all bounce. The reserve wage should fall and joblessness should decline. That it seems to me is the supply-side story about the relationship between wealth and employment.

Exactly.

Matt Yglesias responds to Michael Mandel’s suggestion that we should try to eliminate the trade deficit in manufactured goods, even if it results in slightly higher inflation:

Mandel thinks this would be a reasonable price to pay to rebuild America’s manufacturing base and get us back into the neighborhood of full employment. And I agree. But I would note that when you try to think clearly about what he’s saying you see that this idea””like virtually all possible roads back to full employment””crucially depends on the existence of a stimulative monetary policy paradigm. Mandel is assuming that lots of the new manufacturing employment would represent a net increase in employment, and not just displace workers from existing jobs. But that’s assuming the Federal Reserve would allow the inflation rate to rise by 0.2 percentage points per year each and every year for ten years. There’s no indication that the present Federal Reserve would in fact do this. Instead, consistent with their policy of capping price increases as a two percent annual rate, monetary policy would force net job creation to stay on a low trajectory.

Paul Krugman likes to talk about how he’s been right about almost everything, and his conservative opponents have been wrong about almost everything.  While that’s certainly overstating things, I do believe he’s been right about the need for more demand-side stimulus, and that government policies since 2008 were not likely to lead to higher inflation and interest rates.

Let’s assume Krugman was correct.  Why wouldn’t the left gradually win the vigorous debate among intellectuals?  Why wouldn’t the gradual accumulation of facts tend to discredit one model and support the other?  Isn’t the intellectual debate in the blogosphere a sort of Darwinian struggle?

Krugman’s answer is that the right is dishonest.  The smarter people on the right don’t believe what they are saying; rather they’ve been corrupted by the promise of plum jobs in DC, or money from the Koch family.  But that doesn’t seem plausible to me.  It doesn’t explain why I became a right-winger when young, or why I’ve morphed into being a pragmatic libertarian (out of step with both the GOP and the Koch family.)  And when I talk to other right wing intellectuals they seem a lot like me.  So unless I’m incredibly naive, I think there is an honest debate taking place.  If so, why doesn’t one side gradually win?

My guess is that both sides are right, but about different things.  The left is right about two big things, the value of social insurance and the need for a government policy that stabilizes aggregate demand.  Right wing intellectuals (not necessarily the GOP!) are right that prosperity depends on a set of “conservative” ideas like property rights, prudence, thrift, hard work, ownership, personal responsibility, and the various government policies that encourage those things.  In other words the left is right about the demand side of the economy and also the steeply diminishing marginal utility of consumption, and the right is right about the supply-side of the economy.  Of course I’m oversimplifying, the left has made some contributions to the supply-side, and the right has a better understanding of the potency of monetary stimulus than the left.  But as broad generalizations, they explain why both sides keep seeing confirmation of their model in the news they read—they are looking at a different set of issues:

1.  The left sees the folly of a monetary regime that allows AD to collapse, as in Greece and Latvia.  The right sees that the state willing to do more “responsible” supply-side policies (Latvia) does better than the one which is more statist (Greece.)

2.  The right sees sound economic policies and low taxes producing very rich countries (Singapore, Hong Kong, Switzerland) whereas the left sees fairly rich and more egalitarian models in places like Sweden and Denmark.  If I’m right that both sides are partly correct, then you’d expect the most successful countries on Earth to embody ideas from both sides of the spectrum, and they do.

Another point I’d make is that whereas left wing pundits are right about some issues, and right wing pundits are right about other issues, the markets are always right.  Now that’s a pretty bold assertion, so let me qualify it.  I don’t mean the market have perfect foresight, and can predict the future.  And I am restricting this claim to policies about efficiency, supply-side and demand-side policies, not egalitarian policies.  My claim is that the markets are left-wing on the need for adequate AD, and right-wing on the need for sound pro-growth supply-side policies.  The markets believe that each side of the ideological debate has a sort of blind spot—the left underestimates the importance of incentives, and the right underestimates the damage done by demand shortfalls.  Here’s Arnold Kling discussing the recent moves to save the eurozone:

I believe that for the crisis to end, two things have to happen: the defaults by insolvent governments must be formalized, so that creditors know exactly how far to mark down the value of their holdings; insolvent banks must be resolved, as Kapoor defines resolution.

The new agreement accomplishes neither of these.  The financial markets were pleased.     As is often the case, I am baffled by the markets.

During the 1930s the slow motion collapse of the gold standard was extraordinarily damaging to AD.  The markets know this, which is why oil soared $6 and equities also rallied on the positive eurozone news.  (This post was written a week ago.)

Kling is also right that Europe is not addressing the fundamental problems.  They are just kicking the can down the road.  But markets know that can kicking can be very important, if it staves off disaster until a more durable solution can be developed.

This perspective doesn’t just apply to economic policy, but to social science more generally.  The right is correct that culture helps explain why some groups are more productive than others.  The left is correct that those on the right are often too resistant to the sort of progressive cultural change that comes from increasingly polyglot cosmopolitan societies.  In criminal justice you can build similar dynamic around deterrence vs. civil liberties.

Because both the left right embody perspectives that at least to some extent “work” (I’m a pragmatist so I can’t say “are true”) they may lose a battle here or there, but can never lose the ideological war of the social sciences.

David Glasner has a post that sharply criticizes Allan Meltzer’s recent WSJ op ed. Like Glasner, I greatly respect Meltzer’s work on monetary economics. But I’m afraid that David is correct in taking Meltzer to task for the editorial–it’s very disappointing. (However I wish he hadn’t used the term ‘corrupt,’ which is obviously not the issue here.)

David points out that Meltzer equates ultra-low interest rates with easy money, whereas Milton Friedman frequently pointed out that it actually means money has been tight. And he’s also correct in suggesting that rejecting the low rates = easy money equation is virtually the definition of being a monetarist. But I’d like to focus on a different part of Meltzer’s column:

Consider also how, in the summer of 2010, the Fed allowed itself to be spooked by cries about a double-dip recession and deflation. It added $600 billion to banks’ reserves by buying up federal Treasurys and mortgage-backed securities. Today, $500 billion of those reserves remain on bank balance sheets, and most of the rest of the dollars are held by foreign central banks. Not much help to the U.S. economy. By early autumn 2010, it had become clear that fears of a double-dip recession and deflation were just short-term hysteria.

Glasner criticizes the economics here, but this is also a basic failure of logic. If someone says “I’m really concerned about X happening, and therefore I will do Y to prevent it.” And then they do Y, and X does not happen, then the policy has been a success, not a failure. I see people make this logical error all the time, but I’m surprised to see it made by someone as bright as Meltzer.

And it’s even worse.  Back in 2009 Paul Krugman and I were highly critical of Meltzer’s claim that Fed policies would lead to high inflation. Here’s something I wrote in 2009:

One of the things I find most frustrating about this crisis is the way my intellectual allies on the right keep shooting themselves in the foot, and thus unwittingly tend to discredit their otherwise defensible ideologies.  More specifically they continue to warn of high inflation, which is about the last thing we need to worry about right now.  In my previous post I tried to provide a psychological theory for these bad forecasts.  But whatever the reason, when these predictions don’t come true it will (unjustly) tend to discredit both the good and bad parts of their theoretical apparatus.

A few weeks ago it was Allan Meltzer, now it is Arthur Laffer.

Obviously Krugman and I were right and they were wrong. But that’s not what concerns me now. If the field of economics is to have any credibility at all, then when two contrasting hypotheses are clearly tested by a policy experiment, the loser needs to admit that they lost. In mid-2010 the core rate of inflation had fallen to 0.6%.  At that point the Fed sent out signals that it was going to do QE2. The Fed claimed that QE2 would prevent deflation, but not result in high inflation. Meltzer and many others on the right warned of high inflation.

The Fed policy signals in late 2010 clearly boosted TIPS spreads, and thus we moved closer to the Fed’s inflation target. Yet both actual and expected inflation have remained low. Bernanke was unambiguously proved right—we avoided deflation without suffering from high inflation—and Meltzer was wrong. And the bond market says inflation is likely to remain low over the next 30 years.

It seems to me that old-style monetarism has reached a dead end. In the op-ed Meltzer sarcastically dismisses the stock market’s obvious preference for monetary stimulus in the period since AD crashed. I’d expect that sort of dismissal of important market information to come from the left, not from a school of thought that supposedly believes markets are much more efficient than government. When the markets disagree with the monetarists, the markets are usually right. When the markets disagree with the Keynesians, the markets are usually right. When the markets disagree with me, the markets are usually right. When the markets disagree with you (the reader), the markets are usually right. When the markets disagree with God . . . then it’s probably about 50-50.

What will replace old-style monetarism? Will it be Williamson’s “new monetarism,” or will it be “market monetarism?”