Multiple causation and the measurement of unemployment

The establishment of appropriate policy measures for fighting unemployment has always been difficult since causes of unemployment are hard to identify. This paper analyses an approach used mainly in the 1960s and 1970s in economics, in which classification is used as a way to deal with such a complex, multiple causal phenomenon like unemployment. The method is based on decomposing unemployment into classes of unemployment and the measurement of each of these classes by reference to stable, measurable macroeconomic relationships like the Phillips curve and the Beveridge curve. In this way economists were able to ‘diagnose’ unemployment and make policy recommendations for fighting unemployment without making explicit reference to the underlying singular causes of unemployment.