This paper examines the application of count data models to firm level
panel data on technological innovations. The model we propose exhibit
s dynamic feedback and unobserved heterogeneity. We develop a fixed ef
fects estimator that generalises the standard Poisson and negative bin
omial models allowing for dynamic feedback through both the firm's sto
ck of knowledge and its product market power. By using the long pre-sa
mple history of innovation information this ''entry stock'' estimator
is shown to control for correlated fixed effects and is compared with
an alternative nonlinear GMM estimator. We find evidence of history de
pendence in innovation activity although variables reflecting the comp
any's economic environment are also found to play a major role.