What’s spurious, what’s real? Measuring the productivity impacts of ICT at the firm-level

In order to assess the productivity effects of information and communication technologies (ICT), regressions based on cross–sectional firm–level data may yield unreliable results for the commonly employed production function framework. In this paper, various estimation biases and econometric strategies to overcome their sources are discussed. The effects are illustrated on the basis of a representative set of panel data for German service firms covering the period 1994 to 1999. The application of a suited SYS–GMM estimator yields evidence for significant productivity effects of ICT. However, these are substantially smaller than those suggested by cross–section estimates.I would like to thank Irene Bertschek, François Laisney, Georg Licht, Werner Smolny, Kevin Stiroh, Elke Wolf, Thomas Zwick and two anonymous referees for their helpful comments and the MIP–team for providing me with the data. This paper was written as part of the research project ‘‘Productivity and Spillover Effects from ICT as a General Purpose Technology’’ commissioned by the Landesstiftung Baden–Württemberg foundation.Final version received: September 2002/Final version accepted: April 2004