Tests of the Solow efficiency wage model using Australian aggregate industry and macro economic time series data

摘要

This thesis assesses the efficiency wage hypothesis using Australian industry and macroeconomic time series data by focusing on two questions: whether paying an abovemarket clearing wage called the efficiency wage raises industry output and productivity,and if such a payment causes unemployment at the macro economic level. The wage productivityor wage-output nexus is investigated using three techniques; namely adecomposition procedure used by Huang, Halam, Orazem, and Paterno (1998), aninstrumental variable estimation method, and the Solow residuals approach. Further, anexamination of macro economic unemployment involves developing an aggregateunemployment equation, where the Solow (1979) model is used to derive a testablehypothesis. The Solow model argues that effort, which is a function of the wage, entersthe production function when the real wage is rigid. By introducing profit maximisingbehaviour and making further economic assumptions, the Solow condition that the effortelasticity with respect to the wage is one can be derived. The theoretical framework ofSolow is useful as specifying a production function allows the possibility of aggregatedata being used to assess the wage-productivity prediction. The Solow condition is alsouseful because it provides the basis for constructing a testable hypothesis using anunemployment equation. Solow’s theoretical framework and the Solow condition doesnot rely on the economic assumptions of the shirking, labour turnover, sociological andadverse selection [micro economic] efficiency wage models. Therefore, the innovationof this thesis is to treat the efficiency wage hypothesis as an imperfectly competitivemodel of the labour market using applied macro economic methods. Previous Australianmacro economic literature in the 1970s and 1980s have argued that the wage is eitherharmful to employment prospects (ie unemployment is classical), or that factors such asconsumption and investment are more important (ie unemployment is Keynesian). One ofthe aims of the thesis is to use the empirical analysis to suggest that neither of thesepropositions is entirely correct. Rather, an intermediate position is arrived at by arguingthat there is some empirical evidence in Australian industry and macro economic timeseries data to suggest that the wage plays a dual function: both as a small source ofproductivity and also a minor cause of involuntary unemployment.