The impact of investments in R&D on firm performance: Applications of quantile regression with instrumental variables

Abstract

This study analyzes the impact of R&D in firms’ sales levels, taking a sample of 2,000 firms. Using the technique of instrumental variables in the quantile regression, the coefficient of elasticity in R&D demonstrated to be considerably higher compared to the traditional method (Koenker and Bassett estimator) and a trend reversal with rising quantile. The estimator of the quantile regression with instrumental variables reveals the risks of bias and inconsistency when the exogeneity assumption is violated. Unlike the results presented in the recent literature, the implications of this study show that firms with higher sales (higher quantis of conditional distribution) obtain higher returns on the use of R&D investments.