The missing link between innovation and performance in Brazilian firms: a panel data approach

This study aims to verify if there is a positive relationship between innovation and productivity and/or profit in Brazil using a recursive model in line with the Crépon, Duguet and Mairesse (CDM) model. Using a rich combination of three databases, this paper considers a sample of more than 10,000 Brazilian industrial firms and the period 2001–2008. Besides using R&D expenditure as a measure of innovation input, this study also tests the technical-scientific personnel stock as a more appropriate measure of innovative effort in emerging countries. This variable considers the tacit knowledge intrinsic to the worker and corrects a possible measurement error bias in the R&D expenditure variable. The empirical strategy uses a reduced form of the CDM model in a panel data structure to provide consistent estimates as it controls for selectivity, simultaneity and endogeneity due to unobserved firm effects. There are still few studies that apply the CDM model for panel data, especially regarding developing countries. The main results suggest that technical-scientific workers positively affect the firms’ probability to innovate while R&D expenditure has no effect. The results also highlight the absence of the effect of innovation on productivity and profit, suggesting a missing link between innovative efforts and Brazilian firms’ performance.