Tuesday, December 2, 2014

That Krazy Keynesian Multiplier Again

We recently got some more questions about the Keynesian money multiplier, developed by Baron Kahn in the 1930s, and embedded in the monetary and fiscal policy pretty much throughout the world ever since.  That is in spite of the fact that there are some glaring inconsistencies and contradictions in Keynesian multiplier theory — as well as most of the fundamental assumptions in Keynesian theory in general, for that matter.

The enquirer got a bit of the picture, mostly by reading pages 77-84 of the Just Third Way Edition of Dr. Harold G. Moulton’s 1935 classic, The Formation of Capital.  There were, however, a few misconceptions that needed clearing up — such as the belief that binary economics rejects the multiplier effect.

No, binary economists reject the Keynesian analysis of the multiplier effect, which is something entirely different.  Neither Moulton nor we are saying that there is no multiplier effect.  What we and Moulton point out is that Kahn made a huge error in his analysis: it double counts money by including checks accepted for deposit as reserves in the accepting bank, rather than as a claim on the reserves of the banks on which the checks are drawn.

Checks are not, however, reserves.  They are orders to deliver reserves.  Take the scenario where Guy 1 (not CESJ’s “Fulton Sheen Guy” from Iowa, who appeared last week on Hungry for More to talk about stuff) takes $1,000 and deposits it in Bank A, and Bank A takes $900 of that money and makes a loan to Guy 2.

This is where it gets complicated, at least a little.  When Guy 2 draws a $900 check on Bank A and deposits it in Bank B, he is not giving $900 to Bank B.  Instead, Guy 2 is giving Bank B the right to demand that Bank A deliver $900 when the check Guy 2 deposited is presented by Bank B to Bank A for collection.  Bank A transfers $900 to Bank B, and the amount of reserves in the system stays the same: $1000.  The only difference is that where Bank A had $1,000 and Bank B had $-0-, Bank A now has $100, and Bank B now has $900.

That is what happens when loans are made out of reserves, as is the case with banks of deposit.  The Keynesian analysis, however, mixes in features of issue and discount banking with its assumption that only banks of deposit exist.  Viz. (which is short for “videlicet,” meaning “namely” or “that is to say,” or “as follows”), Keynesian (or Kahnian) multiplier theory assumes that the only loanable funds (even the whole of the money supply) are in the form of the reserve currency, and confuses the different forms of money, some of which qualify as reserve currency, while others do not.

In other words, all reserve currency is money, but not all money is reserve currency.

Thus, a commercial bank (a type of bank of issue and of discount) can create one form of money that is not the reserve currency by accepting (discounting) a contract called a bill of exchange, and issuing a contract called a promissory note.  In that case, a loan is not made out of the bank’s store of existing reserve currency (past savings), but consists of new money in the form of newly issued banknotes (very rare these days) or newly created demand deposits, neither of which constitute the reserve currency into which all other forms of money should be convertible.

When a commercial bank issues promissory notes to discount bills of exchange, everything else being equal, the bank’s reserves do not change, but the money supply increases.  Given zero transactions demand for cash (everything done by check), and offsetting balances (the amount of checks drawn on other banks accepted equals the amount of checks presented for payment at other banks through the clearinghouse), the commercial bank’s cash reserves will stay exactly the same.  A commercial bank’s cash reserves (under U.S. law, reserves must be in the form of vault cash or demand deposits at the Federal Reserve) are only useful for “cashing checks” (i.e., converting one form of money into the reserve currency), or to balance the difference between claims presented for payment, and claims accepted for collection.

When making loans out of existing reserves, the amount of money in the system necessarily stays the same.  When creating money by accepting bills of exchange and issuing promissory notes, with existing reserves set aside as a sort of insurance pool that is only tapped if loans go bad or someone demands conversion of his or her form of money into the reserve currency, the amount of money in the system can be increased simply by banks accepting bills and issuing notes.  A reserve requirement puts a brake on this by limiting the amount of bills that can be accepted to a multiple of existing reserves, e.g., a 20% reserve requirement for a commercial bank with $1 million in the reserve currency in its vault would limit the bank to accepting $5 million worth of bills and issuing notes to that amount.

With a central bank and modern communications, the fractional reserve requirement is essentially meaningless.  All bills accepted can be converted instantly to the reserve currency by rediscounting them at the central bank, and all new demand deposits can be backed 100% by vault cash or demand deposits at the central bank.


No comments: