The Determinants of Equilibrium Unemployment in Britain

A lot of energy is currently being expended in trying to understand what has been happening to unemployment in Britain and the rest of the developed world.' In this paper, we look at this question in the British context by trying to confront some possible explanations with the data. Since the data we use are quarterly over rather a short period we are setting the explanations a very hard test. Most explanations tend to be rather casual and the theories underpinning them ill developed. Thus, for example, the so called 'real wage' argument is based on the idea that real wages affect labour demand and that the equilibrium real wage bears some relatively stable relationship to output per head. But how stable is not clear. Thus whether or not we can expect a well defined, stable ceteris paribus dynamic relationship between real product wages as a proportion of output per head and unemployment is equally unclear. So if such a relationship does not show up in a quarterly regression model this may not be that serious for the theory. Nevertheless we must look for the relationship, for how else are we ever to discriminate between a plethora of conflicting views? The primary emphasis in this paper is on the secular factors affecting unemployment rather than the cyclical variables. Consequently we shall not be concerned with developing a coherent theory of cyclical fluctuations and in the empirical work we shall control for the cycle simply by making use of those exogenous or predetermined variables which other work in this field suggests are important.2 The remainder of the paper is set out as follows. The next section is concerned with specifying a model which generates a definition of equilibrium unemployment. The outline of this model is a pair of equations determining the flows into and out of unemployment and these are specified in Section II which looks closely at the possible variables involved. Section III then deals with the results and presents some general conclusions.