Modelling labour market transitions in Australia

Abstract

This thesis investigates issues relating to two aspects of unemployment that has gained importance in recent years. The first aspect concerns the growing numbers of hidden unemployed in the economy. The implications of this phenomenon for economic analysis is demonstrated by its impact on the macroeconomic analysis of the unemployment-vacancy relationship. A simple stock-flow model is used to predict the effect of hidden unemployment on the UV loci. A systematic approach to estimation of the UV curve is adopted and the empirical results confirm the a priori predictions of this model. The second aspect concerns the state dependence interpretation of unemployment behaviour. A search-theoretic framework is adopted to provide the basis for specification of the models to be estimated. The determinants of transition from unemployment of Australian youths are investigated at the micro level using an Australian labour market longitudinal data set. Survival modelling techniques are applied to estimation of the hazard functions. Structural estimation of the job search model which involves the estimation of the reservation wage function is also undertaken. Positive duration dependence in unemployment is implicated by the results from these analyses. A novel approach to the modelling of recurrent unemployment is also introduced in this thesis. These models, based on the stochastic counting processes theory, allow the pattern of occurrence of employment events and duration of successive unemployment spells to be determined. Evidence of positive occurrence dependence are found from this analysis. Finally, some issues relating to targeted employment susbsidies are also investigated. The costs of this policy intervention are compared with the alternative of providing only unemployment benefits. It is shown that the nature of duration dependence is vitally important for the economic justification of the subsidy policy.