Persistence of Innovation in Dutch Manufacturing: Is It Spurious?

Abstract

This paper studies the persistence of innovation and the dynamics of innovation output in Dutch manufacturing using firm data from three waves of the Community Innovation Surveys (CIS), pertaining to the periods 1994-1996, 1996-1998, and 1998-2000. We estimate by maximum likelihood a dynamic panel data type 2 tobit model accounting for individual effects and handling the initial conditions problem. We find that there is no evidence of true persistence in achieving technological product or process innovations, while past shares of innovative sales condition, albeit to a small extent, current shares of innovative sales.

What determines the innovation capability of firm founders?

“…Our main variables describing the founder characteristics have no variation over time within a firm. Hence, these variables would be wiped out when applying a fixed-effects estimator and we would not be able to identify the effects of the founder characteristics (see Raymond et al, 2010 for a detailed discussion on this issue). To deal with unobserved firm heterogeneity Innovation capability of firm founders Table 2 Estimates of innovative activity (random-effects probit) See Table 1 for the variable definitions; to estimate marginal effects, we fixed the group specific intercept at 0, but otherwise averaged the marginal effects over the other explanatory variables; SEs are in brackets under the coefficients.…”

Does history matter for the relationship between R&D, innovation, and productivity?

“…Mañez et al (2009) study the persistence in the firm R&D status, i.e., in the decision to engage in R&D activities, while Peters (2009) analyzes whether firms innovate persistently, defining an innovator as a firm which exhibits positive innovation expenditure in a given year. In contrast to these studies, Duguet and Monjon (2004) and Raymond et al (2010) examine the persistence in innovation outputs, although, as they use CIS data, their indicators as to whether a firm has introduced an innovation are related to a 3-year period, which could induce an artificial persistence due to overlapping time periods and double counting (Peters, 2009). However, Raymond et al (2010) find that there is only true persistence of innovation in high-technology industries.…”

“…In contrast to these studies, Duguet and Monjon (2004) and Raymond et al (2010) examine the persistence in innovation outputs, although, as they use CIS data, their indicators as to whether a firm has introduced an innovation are related to a 3-year period, which could induce an artificial persistence due to overlapping time periods and double counting (Peters, 2009). However, Raymond et al (2010) find that there is only true persistence of innovation in high-technology industries. For low-technology industries, past process and product innovations and past shares of innovative sales do not affect current process and product innovations and innovative sales 2 .…”

Persistence of Cooperation on Innovation: Econometric Evidence from Panel Micro Data

“…Peters (2009) and Raymond et al (2009) found support for persistence of both innovation inputs and outputs. Raymond et al (2010) using a novel method to treat the initial conditions problem only confi rmed true persistence in the propensity of fi rms to innovate in the high-tech sector, while other results came out inconclusive. Earlier studies by Van Leeuwen (2002), Duguet and Monjon (2004) and Rogers (2004) found support for the persistence thesis but because of data constraints could not control for unobserved individual heterogeneity.…”