Unemployment and growth dynamics : theory and OECD evidence

• Main Author: Balakrishnan, Ravi
• Published: London School of Economics and Political Science (University of London) 1999
• Subjects: 331.13
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.645527

We study unemployment and growth dynamics. A search theoretic approach, augmented by exogenous and endogenous growth considerations, is used. We apply a variety of macro-econometric tools, across OECD countries, namely: structural vector autoregression (SVAR) analysis; simulations; frequency domain analysis and panel data regression analysis, to test out a variety of hypotheses drawn from the theoretical literature. First, we look at unemployment dynamics, using a search model and an SVAR methodology, and discover that the European Community and the US have faced similar shocks, mainly aggregate ones, but have reacted very differently. The exception is Spain, where most of the unemployment dynamics have been driven by reallocation. Overall, this implies that EEC economies might be 'dynamically sclerotic' when compared to the US, though simulations do not confirm this result. Next, we re-examine the link between growth and unemployment. Using: frequency domain analysis, panel data regression analysis and looking at cross correlations, we find that the interactions are weak, with at best a marginally significant negative effect of growth on unemployment. This is consistent with theories that predict capitalization effects dominate creative destruction, in the effect of growth on unemployment. It is not consistent with theories that imply a strong effect of unemployment on growth, through: loss of skills; learning by doing; cleansing effects and savings effects. Even when the capitalization effect is significant, it is not very large. A 1% increase in steady state growth would only reduce equilibrium unemployment by 1%. Finally, we look at the links between growth, R&D and job flows, as an alternative way to isolate creative destruction effects. We find that creative destruction mechanisms are only important for US, i.e. for a country on the technological frontier.