Modelling aggregate demand for labour: A critique of Lewis and MacDonald

In a recent contribution to this journal, Lewis and MacDonald (2002) argue that Australian literature on aggregate demand for labour is permeated with misunderstandings and, as a result, existing empirical work has been misinterpreted. The objective of the present note is to argue that the interpretation of existing empirical studies, to the extent that they are based on a CES production technology, is broadly correct. We demonstrate that Lewis and McDonald have no basis for estimating anything more than the partial elasticity of labour demand, holding output constant, because their single-equation estimation does not identify the elasticity of demand for output.