Scott Sumner  

Everything's endogenous

Open Borders Comes to Illinois... Gross Evidence on Immigrant Vo...
Your debutante just knows what you need

But I know what you want -- Bob Dylan

Over at MoneyIllusion I did a post---partly tongue in cheek---pointing out that if Milton Friedman were alive today he might be forced to change his mind as to the proper definition of "money." When he was alive he preferred the M2 money supply, but I suggested that in the 21st century the quantity of coins seemed to better meet his (empirical) criterion for money.

As I expected some took it more seriously than it was intended, but some of the complaints were quite revealing. Consider this representative comment:

The chief problem with using coinage as any sort of indicator of monetary policy is that people decide how many coins to hold, not the Fed or any other monetary authority.
To paraphrase Mr. Dylan, the public knows how many coins they want, but the Fed determines how many they need.

I certainly understand the sense in which the coin stock could be viewed as "endogenous", i.e. not directly controlled by the central bank. But there's another sense in which it is no different from any other monetary indicator/target. When the Fed increases the monetary base it leads to a hot potato effect, which causes the following to happen:

1. The public wants to hold more M2, because with a bigger NGDP they feel they need to hold more M2.
2. The public wants to pay more for foreign exchange, because with a higher price level they need to pay more for foreign exchange.
3. The public wants to lend money at a lower interest rate (perhaps), because with larger real cash balances they need to offer lower rates to lend money.
4. The public wants to hold more coins, because with a larger NGDP they need to hold more coins to do their shopping.

The Fed does not directly control M2, exchange rates, market interest rates or the stock of coins. Those variables reflect the decisions of non-Fed actors. But Fed policy can push people to behave in such a way that those variables change. Thus it's not crazy to talk about the Fed targeting M2 (as Friedman preferred) or exchange rates (as Mundell often prefers) or interest rates (as Woodford prefers) or NGDP futures prices (as I prefer) or even the quantity of coins, as I half-jokingly considered.

If women consistently wore shorter skirts during booms than recessions (as someone once claimed) then it would not be crazy to talk about the Fed targeting average skirt length.

There's lesson here. People prefer to think about macro issues in concrete terms. I go to the bank, and I decide how many coins I will hold. I decide how much cash I will hold. Banks decide how much reserves they will hold. But guess what, the Fed decides the total monetary base. If they double the base, then you and the banks will WANT to hold a lot more coins, paper money and bank reserves, whether or not you currently think you will want to hold more. Goods, services, and asset prices will change in such a way that you actually want to hold more base money.

As individuals we control how much base money we hold. In aggregate, the Fed controls the total. The adjustment in goods, services, and asset prices necessary to reconcile those two sets of wishes is what lies at the very heart of monetary policy. It's the hot potato effect. If you don't really get this idea at an intuitive level, you'll end up spouting one fallacy after another. You'll focus on "concrete steppes", missing the big picture. You'll focus on credit, when you need to be focused on monetary policy. You'll focus on finance, when you need to be thinking in macroeconomic terms.

Comments and Sharing

COMMENTS (2 to date)
Shayne Cook writes:


In your last paragraph you state,

"As individuals we control how much base money we hold." [emphasis mine]

Could you provide a definition of the "hold" term that can be consistently applied as you are referring to here and elsewhere?.

I.e., Held/hold, as in "withheld/hoarded" and thereby not part of the macroeconomic money circulation calculus ...


Held/hold, as in "not spent for new goods/services by the money's actual owner" but still available for use by other macroeconomic entities (as in savings, other components of M2, etc.) ...


Held/hold, as in just "owned"?

Robert Schadler writes:

Seems to me this points to a much larger problem: modern social science has been leery of positing causation -- even though that i a (and possibly the) main reason why people want to know things. A correlation is very often what "research' concludes -- that two things occur together. The deep-seated human desire to understand tries to smuggle something toward a cause, some kind of inference.
A correlation is a tautology. When a and b occur in some kind of relationship .... they occur in some kind of relationship, assuming all other things are the same. The amount of coinage that the Fed minted ... is the amount of money the Fed minted. The amount of the coins that the public wants to keep ... is the amount of coins the public wants to keep.

Comments for this entry have been closed
Return to top