This paper has two related objectives. The first is to scrutinize the case for an Independent European System of Central Banks (hereafter IESCB), which is rejected on the grounds that it would worsen the performance of the real economy. The second objective is to propose an alternative institutional arrangement more conducive to the achievement of high levels of effective demand and thus of employment.
We start by briefly summarizing a number of interrelated theoretical propositions that underpin the orthodox approach to unemployment and the impact of the financial system on unemployment. These propositions, which have been influential on the IESCB proposals, start with the classical dichotomy whereby there is a separation betWeen the real and the monetary sides of the economy, with the (equilibrium) level of unemployment and output determined on the supply side of the economy and the level of prices (and hence the rate of inflation) set by the rate of expansion of the money supply. Hence the level of unemployment is not considered to be influenced by the level of aggregate demand; Say's law is, in effect, operating and inflation is viewed as a monetary phenomenon in a causal sense. The monetary and financial systems are viewed as stable and do not have any significant impact on the real side of the economy. Savings and investment are seen to be brought into balance by movements in the rate of interest, with the rate of inflation essentially determined by the rate of growth of the money supply.
In much current debate, these propositions are reflected in the concept of the nonaccelerating inflation rate of unemployment (NAIRU), which is clearly a supply-side equilibrium position and which can be seen as a recent expression of the classical dichotomy. There is the strong suggestion that the equilibrium position is unique. The unique equilibrium and the complete separation of the supply and demand sides are not inevitable features of the type of models from which the NAIRU is generated. For example, Manning (1992) provides a model of price and wage setting that has two equilibrium positions and Sawyer (1982) provides a model with multiple equilibria and a lack of a clear separation between demand and supply sides.
The NAIRU is a construct of economic theorizing that may (or may not) be helpful in thinking about the economy. But there is no way in which it can be demonstrated actually to exist in the real world, and hence using estimates of the NAIRU involves a leap of faith to accept the existence of some (unique) NAIRU. Estimates of the NAIRU are typically derived from price- and wage-setting equations, and the theoretical construct is imposed on the estimation of those equations. The price and wage relationships that interact to determine the NAIRU have proved unstable and subject to shifts, which makes any estimate of the NAIRU also unstable. It is argued here that the estimates of the NAIRU (or related concepts) that have been produced have shown a strong tendency to move in sympathy with actual levels of unemployment.(1) Further, the estimates are sensitive to the precise form of equations estimated. Setterfield, Gordon, and Osberg (1992) "suggest that estimates of the NAIRU [for Canada] are extremely sensitive to model specification, the definition of variables and the sample period used. [Further] . . . the final range of all NAIRU estimates . . . is about 5.5 percentage points. Indeed, the size of this range is so great that it covers virtually the entire range of male unemployment rates in Canada since 1956" (p. 134). The Directorate-General for Economic and Financial Affairs of the European Commission concluded that the concept of the NAIRU is "unusable operationally" because "empirical studies on both sides of the Atlantic have shown that large variations in NAIRU may be caused by apparently small differences in sample, retained explanatory variables and analytical formulation. …