Real earnings and business cycles: new evidence

Abstract

In the time domain, the observed cyclical behavior of the real wage hides a range of economic influences that give rise to cycles of differing lengths and strengths. This may serve to produce a distorted picture of wage cyclicality. Here, we employ frequency domain methods that allow us to assess the relative contribution of cyclical frequency bands on real wage earnings. Earnings are decomposed into standard and overtime components. We also distinguish between consumption and production wages. Frequency domain analysis is carried out in relation to wages alone and to wages in relation to output and employment cycles. Our univariate analysis suggests that, in general, the dominant cycle followed by output, employment, real consumer and producer wages and their components is 5–7 years. Consistent with previous findings reported in the macro-level literature, our bi-variate results show that the various measures of the wage are generally not linked to the employment cycle. However, and in sharp contrast with previous macro-level studies we find strong procyclical links between the consumer wage and its overtime components and the output cycle, especially at the 5–7 years frequency. Observed real wages are not constant over the cycle, but neither do they exhibit consistent pro- or counter-cyclical movements. This suggests that any attempt to assign systematic real wage movements a central role in an explanation of business cycles is doomed to failure. (lucas 1977)