It is conventional wisdom that printing more money causes inflation. This is why we are seeing so many warnings today of how Quantitative Easing I and II and the federal government’s deficit are about to lead to skyrocketing prices. The only problem is, it’s not true. That’s not how inflation works. Hence, this is yet another of the false alarms being raised (along with the need to balance the budget) that is preventing us from doing what we need to do to recover from the worse recession since the Great Depression.
Explaining inflation would be much simpler if not for the need to first spend so much time debunking the popular view. But, that’s the way it is. And so, let me start with the “money growth ==> inflation” view. This is based on the equation of exchange:
MV = Py
where M is equal to the supply of money, V the velocity of money (or the average number of times each dollar bill is spent), P the average price of goods and services, and y the total quantity of all goods and services sold during the time period in question. Thus, if there were 100 goods and services that sold for $10 each (on average), then that means a total of $1000-worth of transactions took place. Were there 200 one-dollar bills in this economy, then it must be that each was used 5 times (hence the “velocity” of money, or how fast they were spent again).
MV = Py
200 x 5 = 10 x 100
It is important to note here that the above is not the least bit controversial. No economist disagrees with the basic equation MV=Py. The arguments arise when additional assumptions are made regarding the nature of the individual variables. For example, this is what is assumed in the “money growth==>inflation” view:
M: That which is money is easily defined and identified and only the central bank can affect its supply, which it can do with autonomy and precision.
V: The velocity of money is related to people’s habits and the structure of the financial system. It is, therefore, relatively constant.
P: The economy is so competitive that neither firms nor workers are free to change what they charge for their goods and services without there having been a change in the underlying forces driving supply and demand in their market.
y: The economy automatically tends towards full employment and thus y (the existing volume of goods and services) is as large as it can be at any given moment (although it grows over time).
Now let’s go through an example, recalling the mathematical example from above:
MV = Py
200 x 5 = 10 x 100
Consider the assumptions made regarding each of the variables. P can’t change on it’s own, y is already as large as it can possibly be given current technology and resources, and V is constant. Only M can change in the short run and it must therefore logically be the starting point of any fluctuation we introduce. Furthermore, according to our assumptions, the central bank has the power to (for example) double the money supply at will. In Milton Friedman’s example from “The Optimum Quantity of Money,” a helicopter is used to accomplish this. Now what happens?
MV = Py
400 x 5 > 10 x 100
There is clearly a problem here which could be solved in one of three ways (assuming we don’t just lower M back to 200):
1) y could rise to 200, but of course it can’t because it’s already at its maximum;
2) V could fall to 2.5, but it is constant (something Friedman takes pains to emphasize in the original article); or
3) P could rise to 20.
It is of course the third that proponents of the “money growth ==> inflation” view say will occur.
MV = Py
400 x 5 = 20 x 100
Equality again!
Let me reemphasize why this is the only logical outcome. We have assumed that y and V are constant. Friedman says that y is constant at the level associated with the natural rate of unemployment, while V is indirectly related to agents’ demand for cash. When people want to hold more cash, V, the rate at which they spend cash, naturally falls, and vice versa. But, Friedman further specifies that V is relatively constant and so, therefore, is the demand for cash. Thus, when the central bank raised the supply of cash from 200 to 400, this meant that people were holding more cash than they wished to have in their portfolios. The Fed had created a situation in which the supply of money (newly raised) exceeded the demand (still at the original level). The result was that people, in the language of the “money growth==>inflation” view, rid themselves of excess money balances by spending that cash. They hoped to buy more goods and services but since, in aggregate, more did not exist, they only bid up their prices: money growth led to inflation.
This is this standard view. It makes for a great lecture in an intro or even intermediate macro class and I’ve done it many times (in fact, I just did it this week in my summer course). But the problem is that after the course is over, people only remember this:
increase M ==> increase P
What they don’t recollect are all the assumptions we made to get there! And not only are some questionable, they are downright inconsistent with other lectures we make in the very same class. Take for example y. One need only look out the window to see that it is not currently at the full-employment and therefore maximum level. Hence, given this scenario:
MV = Py
400 x 5 > 10 x 100
there is no reason that this could not lead to the rise in y shown below as those spending their “excess money balances” actually cause entrepreneurs to raise output to meet the new demand:
MV = Py
400 x 5 = 10 x 200
This is, of course, the goal of the government deficit spending that so many economically-ignorant people are trying to stop right now.
In addition, there is a great deal of evidence that the velocity of money IS NOT constant. As one would expect, it tends to decline in recessions when people do, in fact, want to hold more cash. Hence, if we assume that the central bank undertakes the above policy during such a period (as we see today), the final result might be this:
MV = Py
400 x 2.5 = 10 x 100
Or it could be some combination of a rise in y and a fall in V–this would make perfect economic sense. Notice how the process of making the initial assumptions of this approach more realistic is making it far from certain that a rise in M leads to a rise in P, particularly during an economic downturn.
But that’s not the worst of it. There is actually a much more fundamental problem with the “money growth ==> inflation” approach. Recall the original assumptions for M:
M: That which is money is easily defined and identified and only the central bank can affect its supply, which it can do with autonomy and precision.
What is “money” in a modern, credit-based financial system? Is it that stuff you carry in your pocket, the 1’s and 0’s of the electronic entries in your bank account, the available balance on your credit card, your checking account, your savings account? In practice, this question is so difficult to answer that economists actually offer several possible definitions, just in case! Suffice it to say that for present purposes, the idea that we can precisely identify the current “supply of money” in our economy is suspect. This by itself causes problems for operationalizing the above equation.
To make matters worse, the financial sector can create and destroy money without direct action by the central bank. Every time a loan is made, the supply of money increases. The bank is creating money out of thin air, with only a fraction of the total necessary to have already been in the vault as reserves. And when loans are repaid or there are defaults, the supply of money contracts. Hence, the private sector has a great deal of control over M.
But perhaps the real nail in the coffin of the “money growth==>inflation” view is this: the phenomenon that Milton Friedman identifies as key to the whole process, i.e., the excess of the money supply over money demand, cannot happen in real life. The irony here is that something else we already cover in the intro macro class makes this evident. How is it that the Federal Reserve increases the money supply? Remember that Friedman used a helicopter–indeed, he had to, for there was no other way to make the example work. This wasn’t just a simplifying device, it was critical, for it allowed the central bank to raise the money supply despite the wishes of the public. However, that can’t happen in the real world because the actual mechanisms available are Fed purchases of government debt from the public, Fed loans to banks through the discount window, or Fed adjustment of reserve requirements so that the banks can make more loans from the same volume of deposits. All of these can raise M, but, not a single solitary one of them can occur without the conscious and voluntary cooperation of a private sector agent. You cannot force anyone to sell a Treasury Bill in exchange for new cash; you cannot force a private bank to accept a loan from the Fed; and private banks cannot force their customers to accept loans. Supplying money is like supplying haircuts: you can’t do it unless a corresponding demand exists.
The bottom line is that the “money growth==>inflation” view makes perfect sense in some alternate universe where all those assumptions regarding the variables DO hold, but not here, not today, not in the United States of America in 2011. That’s not how it works. It’s a damn shame, I know, because it’s so simple and intuitively appealing and it would make controlling inflation really simple. But, if we are to develop useful policies then we need a model better suited to the way the modern financial system works.
There’s no reason to throw the baby out with the bath water, so let’s retain the equation. However, we need new assumptions with respect to M, V, P, and y:
M: A precise definition and identification of money is elusive in a modern, credit-money economy, and its volume can change either with or without direct central bank intervention. In addition, the monetary authority cannot raise the supply of money without the cooperation of the private sector. Because central banks almost always target interest rates (the price of holding cash) rather than the quantity of money, they tend to simply accommodate demands from banks. When private banks communicate that they need more reserves for loans and offer government debt to the Fed, the Fed buys it. It’s the private sector that is in the driver’s seat in this respect, not the central bank. The central bank’s impact is indirect and heavily dependent on what the rest of the economy is willing to do (which is, incidentally, why all the QE and QE II money is just sitting in bank vaults).
V: The velocity of money is, indeed, related to people’s behavior and the structure of the financial system, but there are discernable patterns. It is not constant even over the short run.
P: While it is true that factors like production bottlenecks can be a source of price movements, the economy is not so competitive that there are not firms or workers who find themselves able to manipulate the prices and wages they charge. The most important inflationary episode in recent history was the direct result of a cartel, i.e., OPEC, flexing its muscle. Asset price bubbles can also cause price increases (as they are now). The key here, however, is that P CAN be the initiating factor– in fact, it has to be, since M can’t.
y: The economy can and does come to rest at less than full employment. Hence, while it is possible for y to be at its maximum, it most certainly does not have to be.
A number of scenarios can be described based on this more realistic alternative and it would be nice to go through each. Unfortunately, as I suggested above, the big problem with this topic is that it takes to long just to reject the popular view! So, I’ll avoid the temptation to write a book here and offer just a quick example (maybe a future post can go over some other interesting possibilities).
As already mentioned, the most important inflationary episode in post- WWII history was that during the 1970s and early 1980s. From 1968 through 1972, consumer price inflation averaged 4.6%. Over the next ten years it was 7.5%. What happened? What caused this sudden and dramatic acceleration in prices? Did the Fed accidentally print too much money? As already explained, that can’t happen–you simply can’t raise the money supply above the demand. M did rise, however, and largely proportionally to the increase in P. This is a much more realistic story of those events.
As the price of oil skyrocketed, so costs of production rose for many, many US businesses. Because there is a lag between purchasing inputs and selling output, most firms have to borrow money (working capital) to bridge the gap. As the ripple effect of the OPEC price increases moved throughout the economy, the demand for cash by these businesses rose. Quite reasonably, private banks and the Fed did what they could to accommodate. These were fair requests on the part of US entrepreneurs. Loans were extended and government debt sold by the private sector to the central bank. This raised the supply of money. Therefore, the rising prices led to an increase in the supply of money and not the other way around. QE, QE II, and the federal government deficit cannot by themselves cause inflation.
And this is how it really works, at least until the Fed starts using helicopters for monetary policy.
Originally published in Forbes on May 14, 2011 and recently taken down. Republished here with permission from the author.
Talking about demand for money seems strange. There will always be demand for money that can be spent to buy other stuff; only its purchasing power fluctuates. So where does that leave the idea that the Fed can’t create demand for money where it doesn’t exist?
Say the government wants to buy 50 more fighter jets. Boeing wants $1 billion a piece for them. But Congress has already spent all them money the government actually has. So the Fed creates $50 billion in new money, which is injected into the economy via payments to Boeing. This inflates the money supply with only the consent of one company, the direct beneficiary (although not as much of a beneficiary as government, since they still have to supply the jets; the governments effectively gets the jets for nothing, as the $50 billion cost of the planes is spread out among all holders of dollars in the world, each of which is now worth a bit less since there is now $50 billion extra cash in the economy chasing the same amount of goods and services that would otherwise be chased by an amount of money lower by the sum of $50 billion.
Why wouldn’t government engaging in deficit spending by inflating the money supply in this manner tend to cause price inflation?
Unless external, non-financial, factors prohibit it, industry will always provide whatever goods necessary to absorb whatever amount of money is available. That is the fundamental proposition of capitalism. Without that proposition, capitalism is jut warlordism without the swords.