We rarely get to conduct scientific experiments in economics, but for the last 3+ years The Federal Reserve has unintentionally conducted a test of an economic theory called the “Quantity Theory of Money” (QTM). QTM makes some very specific predictions – predictions that Ron Paul, conservatives on Wall Street, and others have been repeating a lot.  Unfortunately for them, QTM has failed the test.

First, some background on the theory. The QTM is and has been one of the foundations of both monetarist thought and Austrian economic thought.  In it’s base form, it’s based on an accounting identity that must, by definition, be true.  The notation sometimes varies, but the quantity theory of money is based on a definition called the equation of exchange.  This equation goes like this:

M times V = P times Q
where:
M: Money supply
V:  velocity of money, or the number of times the average dollar changes hands and is spent during the same time period as Q is measured.
P:    price level
Q:   real GDP (sometimes real National Income, Y, is used – same thing essentially)

So what does the equation say?  If you look at the left hand, M x V, you get a representation of the total spending in the economy.  It’s how much money was in circulation times the number of times that money was spent.  The right side, P x Q, gives us the value of nominal GDP.  It’s the value of all the real stuff we bought (Q, real GDP) times the Price level P which translates it into today’s prices.  Put the two sides together and you’ve got total nominal spending in money terms must be the same as the total value of the things we bought.  Duh.  Of course it is.  It’s an identity.   It’s the macro equivalent of saying that if I spend $5 each time (M) on 7 trips to the grocery (V) to buy 70 apples (Q) at $0.50 each (P), then I will spend $35 on $35 worth of stuff.

As an identity definition it’s not really very interesting.  It’s when economists begin to use it as a model of future outcomes that problems arise.

The typical way QTM is used, and the simple way folks like Ron Paul and a lot of folks who are upset at Federal Reserve efforts to stimulate the economy, is by thinking of what happens when M, the money supply, is suddenly increased.  The thinking is that an increase in M must result in an increase in P in order to keep the equation balanced.  This is the foundation of modern inflationary fears in the last  few years.  Typically folks using the QTM this way don’t say things like “an increase in M must lead to an increase in P”.  They say things like “The Fed is printing money like mad and that’s going to lead to inflation”  since an increase in the price level, P, is how we measure inflation.

But there’s actually four terms in this equation.  Any of them can change.  That’s where assumptions come in.  The advocates of QTM, whom I’ll call “inflation-phobes” for the moment since they’re always fearful of inflation, make some strong assumptions.  They assume three big things.  First, they assume that the velocity, V, is constant.  In other words, according to them, you and I always spend our money at exactly the same rate. Suppose I spend my whole paycheck every two weeks now.  They assume that I’ll always spend my whole paycheck every two weeks no matter how big or small that check is or whether I’m suddenly fearful of losing my job next month.  The evidence for the constant velocity assumption is weak.  You be the judge using the St.Louis Fed data:

To me, that doesn’t look constant.  If V is constant, then any increase in M also increases spending, MxV.  But if V isn’t constant, then an increase in M can be offset by simultaneous slowdown in velocity.

The next assumption is that real GDP is always at capacity.  In other words, there is no unused capacity in the economy such as unemployed workers or empty office buildings or factories running only 1 shift when they can run 2.  This is assumption is essential to the inflation-phobes because it means that Q can’t be increased.  This is necessary to their desired outcome because it would imply that the only way for PxQ to rise to meet an increase in MxV is by having P increase.  I won’t go to the trouble of showing data and graph to prove that Q isn’t at capacity.  If you have doubts, see last week’s update on employment and GDP.

There’s one more unstated assumption by the inflation-phobes.  They assume that any increase in base money, which is primarily the bank reserves The Fed makes available to commercial banks, will necessarily translate into M1, money in circulation among the public.  This too is a bad assumption.  There times, like the last 3 years, when commercial banks don’t want to or can’t lend.  In times like this bank reserves just sit there on the books safely tucked away from any kind of productive economic activity or spending.  The Fed can push reserves onto the banks’ books, but it can’t turn those reserves into loans or spending by customers. Another weakness in this assumption is the idea that even if the money ends up in private hands it will be used for spending on goods and services. Instead, what we’ve seen is that much of what little lending the big banks have been doing has been to finance financial market trading and speculation – things like oil futures.  Speculating in oil futures isn’t the same thing as actually refining and selling oil.  The speculation doesn’t create jobs and isn’t part of the circular flow. Production is.

So we’re back to the question of testing the QTM theory.  The QTM theory of money as inflation-phobes express it, says that increases in base money (M) necessarily must result in inflation (increases in P) at some time in the near future.

In 2008 and 2009 The Federal Reserve expanded bank reserves greatly.  It expanded the monetary base dramatically.  The Fed invented a variety of new names and methods for doing it, although almost all of them involved The Fed buying some kind of bond, security or financial asset.  If the QTM theory and the hard-money inflation-phobes are right, there should have been a dramatic increase in inflation.  They predicted it.  Again and again.  It simply hasn’t happened.  Paul Krugman put together an nice little graph showing the failure of the QTM theory:

The thing is, of course, that the past three years — the post-Lehman era during which the Fed presided over a tripling of the monetary base — have been an excellent test of that model, which has failed with flying colors. Here are the data — I’ve included commodity prices (IMF index) as well as consumer prices for the people who believe that the BLS is hiding true inflation (which it isn’t):

A couple of notes: for the commodity prices it matters which month you start, because they dropped sharply between August and September 2008. I use the IMF index for convenience– easy to download. (Thomson Reuters I use when I just want to snatch a picture from Bloomberg). But none of this should matter: when you triple the monetary base, the resulting inflation shouldn’t be something that depends on the fine details — unless the model is completely wrong.

So, we’ve had a test, a pretty substantial test of the Quantity Theory of Money and the assumption that any increase in monetary base must lead inevitably to an increase in prices and inflation.  The theory has failed.  It should be put to rest.  Milton Friedman, a man as responsible as any other for pusing QTM, once famously claimed that “inflation is anywhere and everywhere always a monetary phenomenon”.  He was clearly wrong, there’s more involved than just base money growth.