Choose your battles

MV = PQ

Divide both sides by Q.

(MV)/Q = P

In English: (Money times Velocity) per unit of output equals price.

Less rigorously: The price level depends on the quantity of money and how frequently it is spent.

And that is still pretty damn rigorous.

The price level depends on the quantity of money and how frequently it is spent.

I don't have a problem with that.

Some people say "but money is endogenous" -- money is created within the economy, by the people who engage in economic activity. Okay. So maybe those people are disagreeing with one detail of something Milton Friedman said. But that doesn't mean that "The price level depends on the quantity of money and how frequently it is spent" falls apart.

They wish it did, maybe. But it doesn't. Prices still move with the Q of M. And with hiccups and lags.

 

The price level depends on the quantity of money and how frequently it is spent.

"But you have not defined money!"

Exactly! It is not a problem that money is here undefined. It is a solution. For example, borrowed money works just as well as paycheck money when it comes to influencing prices. So we can include credit (which is money that we have to pay interest on) along with money (which we don't pay interest on), and call it all "money".

And the other thing, V, the Velocity of money -- If we consider Velocity to be a property of money, then we can drop V from the equation and simplify things. If we have M, we'll spend it. Maybe we spend it faster; maybe we spend it slower; it depends on conditions and what we choose to do. But I don't have a problem saying that both borrowing and velocity are included in my concept of "money".

So we can tidy up the "(MV)/Q = P" thing by saying that the M implies the V. So the "V" drops out of the equation, and then we don't need the parentheses because we have simplified things. And the slash in the equation means "per unit of" Q -- per unit of quantity: "Each".

We are left with only

M = P

But I don't want to say Money equals Price. I'd rather say Money determines Prices. Or again, the price level depends on the quantity of money (where our definition must say money is inclusive and flexible, because the economy changes over time). Or better: Money and Prices move together and, perhaps with some exceptions, Money determines Prices. Trying to get everybody on the same page here.

What we are left with, I think, is something that Milton Friedman could have accepted. I don't know, really, if he would have. But I can accept "Money and Prices move together" as derived above, and still accept Friedman's explanation in Money Mischief and elsewhere about the quantity of money being the main cause of inflation. I can even accept the "always and everywhere" part.

 

I worry that some people might object to my reasoning because they disagree with Milton Friedman. But those people may be working backwards from their conclusion that "Friedman is wrong" and making arguments that they hope will prove him wrong. 

I have no patience with that. You don't begin with your conclusion and work backwards from there. 

Some of those people might say it is true that the quantity of money and the price level are related, but point out that this is only because economic activity (buying stuff) can change the quantity of money. Again, they are still trying to prove Friedman wrong. I say, chicken or the egg? I say it doesn't matter which comes first, the economic activity or the quantity of money. If the deal affects the quantity of money, then the deal cannot be completed without the quantity of money being affected.

(Note that I am not saying anything here about policy.)

If you still want to disagree with Milton Friedman, there are much better things to disagree with than his "always and everywhere" view. Take his graphs, for example. The "money relative to output" graphs in Money Mischief and Free to Choose. Friedman says that money relative to output increases on a path that is extraordinarily similar to the path of the price level, "always and everywhere". He presents the graphs as evidence that his claim is correct.

I first looked into this in the 1970s because my textbook (the 1975 McConnell) said the monetarists and the Keynesians disagreed. The disagreement was over some detail regarding the velocity of money, as I recall. But the textbook only said they disagreed, and left it hanging like that. That was unacceptable. That was when I went to the library and discovered the Historical Statistics and Statistical Abstract.

Friedman said the path of the price level and the path of "money relative to output" (MRTO) are similar. The price level goes up, and up. 

I used M1 for money, because M1 is money that circulates, and it is spending that impacts prices.

I used GDP -- but I guess it must have been GNP back then, in the 1970s -- as my measure of output. I'll call it GDP, here. I made a MRTO graph, M1 relative to GDP, and stared at it. It went down, down, down. It didn't go up and up.

 

Years went by.

In Friedman's Money Mischief, in the chapter with the graphs, in a footnote, Friedman identifies the data he used to make the graphs. He used "real" output. Inflation-adjusted output. He divides the quantity of money by a set of numbers that have inflation stripped out of them.

He divides numbers that do have inflation in them by numbers that don't have inflation in them. That's the reason his MRTO goes up like prices: there is inflation in the numerator and not in the denominator. Friedman's calculation is like dividing Money by GDP at actual prices, and then multiplying the result by the price level!

Of course his MRTO looks like the price level. He multiplies the price level into it!

Friedman's graph shows MRTO going up on a path similar to inflation because he multiplies the MRTO by the price level. Only he doesn't do it on the up-and-up. He takes the inflation out of the output number and calls that number "real" output. Then he takes the quantity of money (which is worth less and less over time because the money numbers have inflation in them), and divides by "real" output.

Sounds good, doesn't it? The output is real, the graph must be good. But Friedman's arithmetic is devious and deceptive. You want to complain about Friedman? Complain about that.

I start out by saying I accept the argument that Money determines Prices. I do. But I cannot accept Friedman's graphs. His arithmetic is bad, outrageously bad, and his graphs are definitely deceptive. 

Bad arithmetic does not make good evidence. Far as I'm concerned, yes, money and prices move together. But Friedman offers no evidence of it. His graphs are garbage.

Did he do it on purpose? Oh, I couldn't know that. You might think he had to know the graphs were bad. 

I sometimes think he had to know -- but of course I cannot be sure. I can say that I first came across the MRTO calculation around 1976-77, and I did not figure out why his MRTO went up-and-up until after Money Mischief was published in the 1990s. Took me 20 years to figure it out. And I was a math major.

Hey! The mind works at its own pace.

You want to criticize Friedman, do it for his graphs. Don't do it because the quantity of money and the level of  prices tend to move together.