The impact of technology on the demand for heterogeneous labor is controversely discussed throughout the literature. New technology which is said to favor high skilled labor and to substitute low skilled labor is often considered as the main reason for the decline in relative demand for low skilled labor. While most analyses focus on manufacturing industries, this paper presents empirical evidence that technological skill bias is also present for business related services, an increasingly important sector in the German economy. Cross-sectional data from an innovation survey and panel data from a quarterly business survey in the service sector are used in the empirical investigation. The data allow to directly distinguish among five different skill groups. The micro-level data also allow the analysis of shifting employment patterns for a single economic unit. Ordered probit models are utilized to study the determinants of skill shifts in business related services. It turns out that investment in information and communication technologies is a complement of university graduates and a substitute for workers with completed vocational training. New capital goods are substitutive to low unskilled labor whereas the demand for technically skilled labor remains unaffected by investment decisions. A puzzling finding is that labor costs do not play an important role in the demand for university graduates, technically skilled and unskilled labor but turn out to be highly significant for skilled labor. Also, expected foreign competition has a significantly positive effect on the demand for both university graduates and unskilled labor alike while present foreign competition is only positively significant in the demand for university graduates.
skill-biased technological change capital-skill complementarity panel data service sector ordered probit model