Thomas I. Palley
Dept. of Economics
New School for Social Research
New York, NY 10003
Revised July 1993 I Introduction. Within orthodox monetary macroeconomics the determination of the money supply is widely regarded as unproblematic. Recently, Post Keynesian economists have sought to re-open this issue, arguing for a re-focusing of attention away from the money multiplier toward the role of bank lending in this process. The current paper presents three competing models of the money supply process which illuminate some of the issues in this debate. The first model, labelled the "pure portfolio approach", corresponds to the orthodox description of the money supply process. The second model, labelled "the pure loan demand approach", corresponds to the Post Keynesian
"accommodationist" view of endogenous money. The third model, labelled the "mixed portfolioloan demand approach", corresponds to the Post Keynesian "structuralist" view of endogenous money (see Pollin, 1991a). This third model is very much in the spirit of the earlier "New View" developed by Gurley and
Shaw (1960), and Tobin (1969) in the 1960's. However, the model explicitly focuses on the money supply implications of the banking system's response to expansionary shifts of loan demand. The earlier New View theorists emphasized asset substitutabilities, and focused on changes in asset prices. This was consistent with their interest in the monetary transmission mechanism, but they took the money supply to be exogenous. Post Keynesians focus on the implications of asset substitutabilities for the money supply, and the capacity of the banking system to underwrite economic activity. The critical difference between the "pure loan demand" and "mixed portfolio-loan demand" models concerns the significance ascribed to the private initiatives of banks in accomodating increases in loan demand. In the pure loan demand model, accomodation depends exclusively on the stance of the monetary authority, and its willingness to meet the reserve pressures generated by increased bank lending. However, in the mixed model accomodation depends on both the stance of the monetary authority, and the private initiatives of banks. These initiatives are independent of the monetary authority, and are therefore suggestive of the structurally endogenous nature of "finance capital". II Three competing models of the money supply.
A. The portfolio choice money multiplier model. We begin with the orthodox money multiplier model given by +
(1) Hs = NBR + max [ 0 , BR(i - id) ] - +
(2) Dd = D(i, y) - +
(3) Cd = C(i, y) + +
(4) Td = T(i, y)
(5) Rd = k1Dd + k2Td - +
(6) Ed = E(i, id) (7) Hd = Rd + Cd + Ed
(8) Hs = Hd
(9) M = Cd + Dd where Hs = supply of base NBR = non-borrowed reserves BR = borrowed reserves id = discount rate Dd = demand for checkable (demand) deposits Cd = demand for currency Td = demand for time deposits/bank cerificates of deposit Rd = required reserves Ed = demand for excess reserves Hd = demand for base i = nominal interest rate y = nominal income k1 = required reserve ratio for demand deposits k2 = required reserve ratio for time deposits M = M1 money supply
Signs above functional arguments represent signs of partial derivatives. Equation (1) describes the base supply function, which consists of non-borrowed and borrowed reserves. The level of discount window borrowing is a positive function of the gap between market interest rates and the discount rate. Equations (2) and (3) describe the demands