Offshoring and Relative Labor Demand from a Task Perspective
ZEW Discussion Paper No. 13-067 // 2013Since trade is becoming more of a task-related phenomenon due to the increase in offshoring activities by firms, the pattern of effects across workers is changing as well. In particular, the skill-based characterization of differentially affected groups falls short of comprehensively capturing the effects. Offshoring occurs based on relocation cost for single occupations, which do not necessarily reflect skill intensities. This paper puts this claim to a test by analyzing industry level relative labor demand from a task perspective.
First, a model is developed to describe the mechanisms underlying the link between offshoring and shifts in employment and cost shares of tasks. This model explicitly treats occupations as fixed bundles of tasks which are, as a whole, subject to the offshoring decision. This is a crucial difference to skill-based approaches, in which individuals either supply high-skilled or low-skilled labor, with complete mobility across industries or stages of production. Furthermore, the model highlights the between-occupation dimension in task share changes in the economy. The model introduces a sorting mechanism of individuals across occupations, which relies on a trade off between occupations specific wages and individual abilities in terms of effort needed for the execution of certain tasks. In such a setting, an overall fall in offshoring costs can shift the cut-off occupation, which separates home from foreign production, such that the range of occupations at home shrinks. Crucially, this remaining range of occupations is characterized by a higher non-routine and interactive task content. Since workers whose jobs have gone offshore optimally re-sort into these occupations, average employment shares and cost shares of routine and non-interactive tasks fall in the home economy. Introducing occupations as bundles of tasks provides a new and consistent way of theoretically sound estimation of relative labor demand for tasks.
Second, the theoretical predictions are tested with industry level data from German manufacturing. A clear and robust result emerges. An increase in offshoring significantly reduces home country relative demand for routine and non-interactive tasks – in particular if this offshoring is directed towards non-OECD countries. This result is robust to various alterations and the use of instrumental variable methods. The additional value compared to skill related measures is demonstrated by the fact that this demand shift is also observed when controlling for the skill composition of industries. The main results show offshoring to account for up to 24% of the observed "task-upgrading" over the period 1998-2007.
Hogrefe, Jan (2013), Offshoring and Relative Labor Demand from a Task Perspective, ZEW Discussion Paper No. 13-067, Mannheim.