Abstract
This paper investigates whether firms innovate persistently or discontinuously over time using an innovation panel data set on German manufacturing and service firms for the period 1994–2002. It turns out that innovation behaviour is permanent at the firm level to a very large extent. Using a dynamic random effects discrete choice model and a new estimator recently proposed by Wooldridge (2005), I further shed some light on the driving forces for this phenomenon. The econometric results show that past innovation experience is an important determinant for manufacturing as well as for service sector firms, and hence confirm the hypothesis of true state dependence. In addition, the results highlight the important role of knowledge provided by skilled employees and unobserved individual heterogeneity in explaining the persistence of innovation.
Similar content being viewed by others
Notes
Both studies applied a cross-sectional probit approach including a dummy variable for whether the firm was an innovator in the previous period as an explanatory variable.
Detailed information on the surveys, the construction of the panel data sets and their representativeness can be found in Peters (2008). By and large, both panels still reflect total-sample distributional characteristics quite well and don’t give any obvious cause for selectivity concerns.
As an example, in the 2001 survey a firm is defined as an innovator if it has introduced a new product or production technology in the period 1998–2000, in the 2002 survey this indicator is related to 1999–2001.
Instead of \(\bar{x}_{i},\) the original estimator used x i = (x i1, …, x iT ) but time-averages are allowed to reduce the number of explanatory variables. Using x i instead of \(\bar{x}_{i}\) leaves the results nearly unaltered.
A fixed effects (FE) model would be preferable because it assumes that μ i is random but leaves its distribution unspecified. However, no general transformation is known how to eliminate μ i in dynamic FE binary choice models. For the dynamic FE logit model, Honoré and Kyriazidou (2000) proposed a semiparametric estimator, which is, however, extremely data demanding and cannot be used here.
Note that attrition is here mainly due to the voluntary character of the survey and not due to innovation.
To the best of my knowledge there is no test on exogeneity available for this type of model. In a static RE probit model, Wooldridge (2002) suggested to add the lead of the variable and to test on its significance. In the static RE probit model only PUBLIC is not strictly exogenous on the 1% and SIZE on the 10% level.
References
Aghion, P., & Howitt, P. (1992). A model of growth through creative destruction. Econometrica, 60(2), 323–351.
Bartelsman, E., & Doms, M. (2000). Understanding productivity: Lessons from longitudinal microdata. Journal of Economic Literature, 38(3), 569–594.
Cefis, E. (2003). Is there persistence in innovative activities? International Journal of Industrial Organization, 21(4), 489–515.
Cefis, E., & Orsenigo, L. (2001). The persistence of innovative activities. A cross-countries and cross-sectors comparative analysis. Research Policy, 30(7), 1139–1158.
Cohen, W., & Levinthal, D. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128–158.
Crépon, B., Duguet, E., & Mairesse, J. (1998). Research innovation and productivity: An econometric analysis at the firm level. Economics of Innovation and New Technology, 7(2), 115–158.
Duguet, E., & Monjon, S. (2004). Is innovation persistent at the firm level? An econometric examination comparing the propensity score and regression methods. Cahiers de la Maison des Sciences Economiques v04075, Paris.
Flaig, G., & Stadler, M. (1994). Success breeds success. The dynamics of the innovation process. Empirical Economics, 19(1), 55–68.
Geroski, P. A., van Reenen J., & Walters, C. F. (1997). How persistently do firms innovate? Research Policy, 26(1), 33–48.
Heckman, J. J. (1981). Statistical Models for Discrete Panel Data. In: C. Manski, D. McFadden (Eds.), Structural analysis of discrete data with econometric applications (pp. 114–178) Cambridge, MA.
Honoré, B., & Kyriazidou, E. (2000). Panel data discrete choice models with lagged dependent variables. Econometrica, 68(4), 839–874.
König, H., Laisney, F., Lechner, M., & Pohlmeier, W. (1994). On the dynamics of process innovative activity. In: K. Oppenländer, G. Poser (Eds.), The explanatory power of business cycle surveys (pp. 243–262). Avebury.
Malerba, F., & Orsenigo, L. (1999). Technological entry, exit and survival: An empirical analysis of patent data. Research Policy, 28(6), 643–660.
Manez Castillejo, J., Rochina Barrachina, M., Sanchis Llopis, A., & Sanchis Llopis, J. (2004). A dynamic approach to the decision to invest in R&D: The role of sunk costs, mimeo.
Mansfield, E. (1968). Industrial research and technological innovation: An econometric analysis, New York.
Nelson, R., & Winter, S. (1982). An evolutionary theory of economic change, Cambridge.
OECD, & Eurostat (1997). Oslo manual, proposed guidelines for collecting and interpreting technological innovation data, (2nd edn.). Paris.
Peters, B. (2008). Innovation and firm performance: An empirical investigation for German firms, ZEW Economic Studies, Vol. 38, Heidelberg.
Phillips, A. (1971). Technology and market structure: A study of the aircraft industry. Lexington.
Raymond, W., Mohnen, P., Palm, F., & Schim van der Loeff, S. (2006). Persistence of innovation in Dutch manufacturing: Is it spurious? UNU-Merit Working Paper 2006–011, Maastricht.
Reinganum, J. F. (1983). Uncertain innovation and the persistence of monopoly. American Economic Review, 73(4), 741–748.
Rogers, M. (2004). Networks, firm size and innovation. Small Business Economics, 22(2), 141–153.
Romer, P. (1990). Endogenous technological change. Journal of Political Economy, 98(5), 71–102.
Schmookler, J. (1966). Invention and economic growth. Cambridge, MA.
Sutton, J. (1991). Sunk costs and market structure. Cambridge, MA.
Wooldridge, J. (2002). Econometric analysis of cross section and panel data. Cambridge, MA.
Wooldridge, J. (2005). Simple solutions to the initial conditions problem in dynamic nonlinear panel data models with unobserved heterogeneity. Journal of Applied Econometrics, 20(1), 39–54.
Acknowledgements
The author gratefully acknowledges the comments of Martin Biewen, Francois Laisney, Ulrich Kaiser, Jacques Mairesse, Pierre Mohnen, Christian Rammer, Jeffrey M. Wooldridge and the anonymous referees. I would also like to thank all participants in the ZEW conference on Innovation and Patenting (Mannheim, 2005), the Conference on Innovations and Intellectual Property Values of the Applied Econometrics Association (Paris, 2005), the IIOC (Boston, 2006), the DRUID Summer Conference (Copenhagen, 2006), the International J. A. Schumpeter Society Conference (Nice, 2006) and the ESEM (Vienna, 2006) for valuable comments on previous versions. Any errors remain those of the author.
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
Rights and permissions
About this article
Cite this article
Peters, B. Persistence of innovation: stylised facts and panel data evidence. J Technol Transf 34, 226–243 (2009). https://doi.org/10.1007/s10961-007-9072-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10961-007-9072-9
Keywords
- Innovation
- Persistence
- State dependence
- Unobserved heterogeneity
- Dynamic random effects panel probit model