Tuesday, March 13, 2012

A less notorious lesson in Laffernomics

A few years ago, I bought a book co-authored by Arthur Laffer, Ph.D. and Stephen Moore  called, "Return to Prosperity."  Let's leave an appraisal of the book aside for the moment.  I have read the book and thought I might critically examine a few "nuggets".

The Phillips Curve has it exactly backward?

There is a point that is not involved in a more cohesive argument, so I will start with that.  On page 242, Professor Laffer writes, "Output growth--economic expansion--reduces inflation. Recession, contraction, and increased unemployment actually increase inflation.  Ironically, the Phillips Curve, as used by its modern adherents, has it exactly backward.  Putting people back to work doesn't cause inflation.  Period!"  That's very insistent, but does it make any sense?  The modern Phillips Curve links inflation with the output gap, instead of the unemployment rate.  If there is a lot of excess capacity, the logic goes, then there is a lot of extra supply compared with demand.  Under microeconomic models, this would imply a decline in the price of a good, which would imply deflation (decreasing prices).  The same models would imply that if demand increases without a corresponding increase in supply, inflation will occur (prices will rise), which is exactly the case of a negative output gap.  Let's look for what can be gleaned from actual economic performance:

Oh no, there doesn't seem to be any relationship between the two variables!  Actually, there does seem to be a lagged relationship.  The areas where output exceeds capacity, inflation follows, the areas where output is less than supply seems to show a decline in the rate of inflation.  Professor Laffer is wrong, wrong, wrong!

The Financial Crisis

Mostly in the next chapter of his book, Professor Laffer lays out an argument that we are in for high rates of inflation if the Federal Reserve does not change course.  Now, let's not dismiss this argument immediately just because it has not happened (some might say empirical rejection of his hypothesis).

Monetary Base, a Harbinger of Inflation?

Instead, follow his logic: he says that, "The Fed controls the monetary base 100 percent and does so by purchasing and selling assets in the open market" and "... the monetary base is essentially the reserves of the reserves of the system.  That's how the liquidity pyramid goes and the Fed controls that quantity 100 percent" (pg. 257, 251).  Mostly true, but there are shades of grey here.  I would posit that, generally speaking, the Federal Reserve (Fed) is able to control the monetary base (all the time, historically).  Nevertheless, it's problematic to say it has 100 percent control, since there has to be a counter-party to any transaction.  If the Fed wants to buy bonds, there have to be willing sellers; if the Fed wants to sell bonds, there have to be willing buyers.  So, 100 percent, no.

On page 257 Professor Laffer writes, "Monetarists believes excessive growth in the quantity of money (M1 for our purposes) leads to higher inflation and higher interest rates, while I believe excessive growth in the monetary base leads to higher inflation and higher interest rates."  This may seem muddy to most, but monetary base can be increased by adding to banks' reserves, without increasing credit or money in circulation.  To me, it seems quite clear that money has to be spent to cause inflation.  Here's a question that might make this clear:  How will money sitting in a bank vault increase the price of anything?  Remember, we're talking about not expanding credit, so interest rates are not lower, and no cash is put into circulation.  Professor, you are wrong.

Tight Monetary Policy Caused the Financial Crisis?

Then Professor Laffer begins weighing in on the financial crisis, which he calls a liquidity crisis.  He says, "The triggers of the liquidity crisis, the housing debacle, subprime mortgages and so on, didn't cause it" (pg. 255).  Okay, Professor, enlighten us.  "And when you combine this increase in the demand for money with the Fed's tight monetary policy over a long period of time, somethings going to happen that will trigger a financial collapse" (pg. 255).  He adds, "We have argued time and time again that the Fed should have increased the supply of money by increasing the monetary base.  The Fed waited until late 2008 and then it way overdid the increase" (pg. 255).  Okay, so the financial crisis was caused by tight monetary policy?  Other economists have been more likely to identify loose monetary policy leading to a housing bubble and, inevitably, its collapse, combined with speculation in complicated financial instruments as the core of the problem.  Maybe a look at a historical graph of the money supply measured in M2 would be illustrative?

Okay, there seems to be a little decline in the rate of M2 money growth midway through 2003 going forward.  I'm not sure there's anything here that looks like tight monetary policy.  Maybe the Prime Interest Rate would be instructive?

Wow, he's right!  That's some tight monetary policy!  It must have really hurt the economy!

Okay, he's wrong.  the economy was doing fine during that period.  Why would he want easy monetary policy if Real GDP was growing fairly well already?  (Refer to my previous post where I demonstrate that the CBO thinks the U.S. economy was at its output capacity prior to the crisis.)

Return to 70s-Style Inflation?

Professor Laffer says that he believes the U.S. is headed for 70s style inflation if the Fed does not change its course.  He writes, "The period I was writing about then [in "Global Money Growth and Inflation", WSJ] was not all that different from today with the Federal Reserve headed by Ben Bernanke" (pg. 229) and "To date what's happened is potentially far more inflationary than were the monetary policies of the 1970s, when the prime interest rate peaked at 21.5 percent and inflation peaked in the low double digits" (pg. 259).  That's scary.  He explains, "Reduced demand for money combined with rapid growth in money is a surefire recipe for inflation.  The higher interest rates themselves will also further reduce demand for money, thereby exacerbating inflationary pressures" (pg. 259).  This makes no sense!  A shortfall in the demand of money would imply declining interest rates to bring demand and supply to a new equilibrium.  Keep in mind, the reason interest rates are low by this analysis are by the decline in demand for money.  If that is true, then how can inflation be a problem?  Maybe he's talking about a relative decline in the demand for money (compared with goods and services).  In that case, demand would be pushing prices up rather soon.  But why would there be an apparent shortfall in aggregate demand and so much unemployment if everyone is rushing out with such money as they have to purchase goods and services?  Wait a minute...  The private savings rate has increased!

There is a real problem with the clarity of Professor Laffer's writing.  Frankly, there seems to be no way to make his assertion make sense.  The fact is, the current situation resembles the paradox of thrift from the Old Keynesian liquidity trap.

Professor Laffer also says, "The expansion of money, given an increase in the monetary base, is inevitable, and will ultimately result in higher inflation and interest rates" (pg. 258).  He recommends, "[The Fed] should contract the monetary base back to where it otherwise would have been, plus a slight increase geared toward economic expansion.  Absent this major contraction in the monetary base, the Fed should increase reserve requirements on member banks to absorb excess reserves" (pg. 259).  How would one assess Professor Laffer's predictions about policy?  The policy has clearly not changed course to tighten monetary policy to the extent that he recommends, but there is no real evidence of rising inflation.  There was a short period of elevated commodity price growth, but it has since subsided, with no major contraction of the monetary base.


Does Professor Arthur Laffer really believe what he says?  He explains what would occur should the Federal Reserve have changed course, "This would put the Fed in direct competition with the Treasury's planned issuance of about $2 trillion worth of bonds over the coming twelve months.  Failed auctions would become the norm and bond prices would tumble, reflecting a massive oversupply of government bonds" (pg. 260).  Okay, lets leave aside that he's saying that the government would be insolvent (because nobody would be willing to buy its bonds for any price).  He's saying bond prices would decline, which seems reasonable.  But if bond prices decline, then interest rates would be higher!  If bond prices are higher because the Fed buys U.S. treasuries instead of selling them, then it leads to lower interest rates.  How does the Fed buy bonds?  By increasing the monetary base (with purchases of bonds with new money).  Increasing the monetary base decreases interest rates!!!  It's clear that he understands the process and I doubt he has confusion about the relationship between bond prices and interest rates.  He's lying!

A dash of frosting on the cake: "In addition, a rapid contraction of the monetary base as I propose would cause a contraction in bank lending, or at best limited expansion" (pg. 260). He admits that his "solution" would only make problems worse!


  1. Not trying to address everything in the post, or the book (which I have not read), but a few points stood out to me.

    A lot of what Laffer has to say sounds like a mix between old-school monetarism and the market monetarism of Scott Sumner, Nick Rowe, Lars Christensen, and many others. This school emphasizes the importance of using monetary policy to maintain nominal GDP on a stable path.

    His argument about the Phillips Curve, for example, works out exactly if you assume "constant monetary policy" in the market monetarist sense of 5% nominal GDP growth. If real GDP growth increases by 1%, but nominal GDP growth remains 5%, then inflation has to fall by 1%, which is precisely the story Laffer is telling. The historical data simply indicates from this view that monetary policy has been magnificently mis-managed.

    Or take the argument that tight money caused the recession. Old-school monetarists would have seen M1 or M2 as the indicator of monetary policy (and would view it this way whether the Fed is explicitly and intentionally targeting a monetary aggregate or not), but market monetarists view nominal GDP as the indicator of monetary policy. When NGDP growth falls 8% below trend, this is viewed as an unprecedented level of tightening.

    The argument about interest rates and money demand goes back to the quantity equation, M x V = NGDP. M is a measure of money supply, and 1 / V is a way of describing money demand (in terms of the fraction of income people hold as money, M / NGDP). if velocity is a function of interest rates V(i), we might expect people to hold a larger fraction of income as money when rates are low, reducing V, and a smaller fraction when rates are high, raising V. From this angle, it does make sense to say that high interest rates reduce 1 / V, meaning they increase V for a given level of M, putting upward pressure on NGDP.

    The peculiar part, and I think the part where Laffer seems to separate from the market monetarists, is the idea that raising NGDP is a bad thing! For the market monetarists, raising NGDP *is* raising aggregate demand, and given how far below trend we still are, we would see a lot of this increased nominal growth take the form of increased *real* growth before inflation really could become a major problem.

    So as near as I can tell from what you've excerpted, Laffer's model is a strange mix of old-school monetarist emphasis on monetary aggregates, especially the base; market monetarism, especially the role of nominal GDP; and a very, very steep SRAS curve. I don't think he's right, especially on the last point. But it's not really incoherent, either. For further reading on this sort of model I recommend Scott Sumner's blog www.themoneyillusion.com.

  2. I really don't read that as his argument. If you read the chapters in question, I doubt you'd come away with that impression, either. There might be a slight confusion because of the use of term "money demand" perhaps in different contexts. It seems to me, from what I've read from his book, he is either deeply confused (like about the relationship between bond prices and interest rates) or he is dishonest, but I think there are enough glimmers of clarity, at least in the last two paragraphs of a Chapter, which is what I quote in my last section, that I lean toward dishonesty. I am always glad to have discourse and even disagreements, as long as we aren't ad hominem (and, technically, I'm not, even though I'm somewhat rude about Professor Laffer, I suppose).

  3. As I said, I haven't read it, so I could be quite off. But I think there are a number of people on the right who, contrary to the interpretations of Krugman and DeLong, *do* have a model. It looks too close to an RBC model for me to find it credible, and as you point out I think it makes wrong predictions, but having a bad model with inappropriately strong priors isn't the same as having no model and lying about it. I still believe the philosopher's rule of charity is the best guide to discourse among academics.

  4. Well, I think Professor Laffer might very well have a model but he doesn't necessarily really believe it (I hope he's smart enough not to). As for empirical hypothesis testing, I wanted to hold off on that a little bit, and, if you notice, my graphs were all ending late in 2010, which maybe is a little bit later than when Mr. Laffer wrote his book, but most of the data to which I point is earlier. Yes, I'm fairly confident his model (such as it is) doesn't work, but I was hoping to demonstrate that there are very confused elements in his model. He does not present the model in a very clear and consistent manner and I suppose there might be something I'm misinterpreting, although I've tried to examine things from at least a few angles, where possible. If I misunderstood some Monetarist or Market Monetarist insights, it'd be because I generally am less familiar with those models, at this time, except for my readings of Professor David Glasner's blog and a few dusty textbooks which include some discussions of Milton Friedman (but don't treat his model very comprehensively). I did lay off on his arguments about purchasing power parity, since I don't consider that very unusual and not really empirically rejected.