| Abstract: | A number of recent articles have attempted to restore the use of a simple measure of the money supply as an indicator of future price levels and to re-establish a causal link from money to prices. Most notably Hallman, Porter and Small (HPS) (1989a), (1989b) originated the approach using US data and Hannah and James(1989) have applied it to the UK The approach broadens the traditional idea of a constant velocity of money by introducing the notion of V* and Q*, the long-run value of velocity and income. These are then used to define P from the traditional quantity theory of money as the long-run equilibrium price level. The analysis then proceeds to estimate a standard Error Correction Model (ECM) for price determination with the levels effect given by (P-P*)t-1. The conclusion drawn is that 'a measure of money that determines the long-run future level of prices is useful in determining the proper monetary policy for attaining price stability. We have shown, through the construction of P*, that M2 can serve as this determinant for the price level' (Hallman, Porter and Small (1982a) p. 23).
We argue in this paper that the P* approach is flawed. It is certainly more complex than traditional monetarist approaches but the fundamental questions of causality are in no way either affected or resolved. The P* analysis is a variant on more conventional cointegration analysis (Engle and Granger (1987), Johansen (1988), Hall (1989)) and we argue that the Johansen framework allows us to address the question in a formal and more complete way. When this approach is applied to the US data used by HPS, we find that while the P* relationship does indeed represent a cointegrating one, it does not have a causal link with prices but rather the causality runs from prices to money - this result conforms well to the work of Hendry and Ericsson (1990) or Hall, Henry and Wilcox (1990), which use this form of relationship to model the demand for money. |
