The Lucas critique has received deficient attention in labour demand studies. In this paper, we examine the empirical importance of the critique in light of labour demand in Norwegian manufacturing. The modelling takes place within multivariate cointegration analysis to develop a conditional equilibrium correction model of labour demand. To detect the applicability of the Lucas critique, the role of expectations and measures of model constancy are examined with variable-addition tests and recursive methods. The labour demand model is found reasonably stable despite major policy regime shifts over the past decades, supporting the conclusion that the Lucas critique lacks force.