To be or not to be innovative: an exercise in measurement

Working Paper: NBER ID: w8644

Authors: Jacques Mairesse; Pierre Mohnen

Abstract: In this paper, we put forward the idea of an innovation accounting framework and consider two main indicators based on it: expected innovation and innovativeness. The framework is the analogue of the standard framework of economic growth accounting, with innovativeness being a parallel notion to that of (total factor) productivity. We provide an illustration of the idea using data from the European Community Innovation Surveys (CIS1 and CIS2) and measuring innovation by the share of firm innovative sales. We adopt a generalized tobit model of the propensity and intensity of innovation as our accounting framework. We first apply the framework to a comparison of the innovation performance of French manufacturing industries, while also checking the robustness of our estimates to the use of micro- aggregated firm data provided by Eurostat versus the original individual firm data. We also provide an overview of the results of a larger comparison of innovation across seven European countries.

Keywords: Innovation Accounting; Expected Innovation; Innovativeness; Tobit Model

JEL Codes: C35; L60; O31; O33


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
expected share of innovative sales (O35)predicted based on characteristics of firms (D22)
characteristics of firms (D21)expected share of innovative sales (O35)
firm size (L25)expected share of innovative sales (O35)
intensity of R&D efforts (O32)expected share of innovative sales (O35)
economic environment (P42)expected share of innovative sales (O35)
observed share of innovative sales (O36)innovativness (O35)
expected share of innovative sales (O35)innovativness (O35)
innovativness (O35)total factor productivity (D24)

Back to index