ESTIMATING THE INNOVATION FUNCTION FROM PATENT NUMBERS - GMM ON COUNTPANEL-DATA

The purpose of this paper is to estimate the patent equation, an empir ical counterpart to the 'knowledge-production function'. Innovation ou tput is measured through the number of European patent applications an d the input by research capital, in a panel of French manufacturing fi rms. Estimating the innovation function raises specific issues related to count data. Using the framework of models with multiplicative erro rs, we explore and test for various specifications: correlated fixed e ffects, serial correlations, and weak exogeneity. We also present a fi rst extension to lagged dependent variables. (C) 1997 by John Wiley & Sons, Ltd.