The impact of offshoring on employment in Belgium
Worries about massive job losses have prompted a fast-growing literature on offshoring and its impact on employment in advanced economies. This paper examines the situation for Belgium. The offshoring intensity is computed as the share of imported intermediate inputs in output, based on a series of constant price supply- and-use tables for 1995-2003. Estimations of static and dynamic industry-level labour demand equations augmented by offshoring intensities, do not reveal a significant impact of either materials or business services offshoring on total employment for Belgium between 1995 and 2003, but this does not preclude a differential impact by skill-level.
The fears raised by offshoring used to be focused on manufacturing, but now they also extend to certain kinds of services, mainly business services that have become increasingly tradable and thereby subject to offshoring. In this paper, we present evidence on both materials and services offshoring for Belgium. The offshoring intensities are computed for the period 1995-2003 as imported intermediate inputs, drawn from constant price supply-and-use tables, divided by output. It does not come as a surprise that the levels of the offshoring intensities are very different for materials and business services: they are high for the former and still relatively low for the latter, even in 2003. In terms of the growth rates, the intensity of materials offshoring stagnates, whereas the intensity of business services offshoring is on the rise. Offshoring to high-wage and low-wage countries is also addressed through a split-up according to the origin of the imported intermediates. This is done by combining detailed trade data by country of origin with the data on imported intermediate inputs. The split-up shows that for materials it is offshoring to high-wage countries that drags the growth rate down. The highest growth rates for both materials and services can be observed for offshoring to Central and Eastern European countries.
To determine the impact of offshoring on employment in Belgium, standard neoclassical labour demand equations have been defined and augmented by the offshoring intensities. Both static and dynamic versions of these equations have been estimated for 58 manufacturing industries and 35 service industries. They fail to reveal a substantial impact of the offshoring intensities on total employment in these industries. In all specifications, the employment impact is found to be either non-significant or very small. This is true for both materials and service offshoring. These results are in line with previous findings in the literature. The main conclusion to be drawn is that these results show that, at the industry level, offshoring did not massively depress employment over 1995-2003. We believe that this is consistent with the view that materials offshoring is mature and no longer rising strongly, while services offshoring still remained at low levels in Belgium during this period despite substantial growth rates. Moreover, our estimation results are in line with the argument put forward by many observers that the number of jobs lost because of offshoring remains small compared to total job turnover in an economy. Nonetheless, these are results at the industry level and they do not mean that there are no jobs lost due to offshoring. These results may indeed hide disparities in demand for different skill categories, which is influenced by offshoring, as well as differences in trends at the level of the firms. Both these issues deserve to be carefully examined in future research for Belgium.