The Cyclical Relationship Between Real Wages and Employment

Abstract

<p>The standard textbook Keynesian macro model predicts counter-cyclical movements of real wages when cycles are assumed to be caused by aggregate demand shocks. However, most of the empirical work seemed to suggest that real wages actually moved procyclically. Rather than abandon any crucial assumptions in the "micro base" of the model (i.e. perfect competition, diminishing returns, and variable coefficients of production ) , neoclassical theorists either modified the standard model in other ways, or "refined and clarified" the nature of the data required to test this model. While the "statistical clarifications" are for the most part well grounded theoretically, the theoretical modifications were generally found to contain undesirable features. In response to this, two models were developed which could explain- procyclical real wage movements and which did not contain any undesirable features.</p> <p>The empirical part of the thesis involved testing four different hypotheses about the determination of employment, taking into account the "statistical refinements" suggested in the literature. We found that the data rejected both of the models developed in the theoretical part of the thesis. In addition, we found that the best specification of the labour market was "employment equals the minimum of demand and supply", and since most of the observations were to be found on the demand curve, a simple fitting of a labour demand curve would do as an approximation. Frictional unemployment and vacancies were found to be important in the determination of employment , but were not easily modelled. None of these modifications, neither the "theoretical" nor the "statistical", had any significant effect on the coefficient relating real wages and employment. Though this was consistently significantly negative, it was too small~in absolute size, causing us to reject the Cobb-Douglas specification of the production function. Finally, using time series methods, a causal relation was found running from real wages to employment when wages were deflated by the wholesale price index. However, this relationship disappeared when wages were deflated by industry selling prices.</p>