International trade, technological shocks and spillovers in the labour market: a GVAR analysis of the US manufacturing sector

We empirically analyse the response of labour market related variables in the US manufacturing sector to various shocks, notably to trade openness and technology, as well as examining spillovers from industry-specific labour market shocks. The econometric approach involves an application of the recently developed Global Vector Autoregression methodology of Dees et al. (2007) to 12 manufacturing industries over the period 1977-2003. The framework allows us to analyse the response of a standard set of labour-market related variables (employment, real compensation, productivity and capital stock) to exogenous factors (a sector-specific measure of trade openness, a common technology and oil price shock), along with industry spillovers using specific measures of manufacturing-wide variables for each sector. Generalized impulse responses indicate that increased trade openness negatively affects real compensation, has negligible employment effects and leads to higher labour productivity. These impacts, however, are relatively weaker than those induced by technology shocks, with the latter positively and significantly affecting both real compensation and employment. There is also evidence of positive spillovers across industries from sector-specific employment and productivity shocks. Impact elasticities suggest strong intra-sectoral linkages for employment and capital stock formation, contrasting with weak linkages for what concerns real compensation and productivity.