R&D, Innovation and The Business Cycle

Abstract

Technological Innovation is a major driver of long-run economic development. It increases wealth generation capacity, productivity, and rewards innovators with temporary monopoly profits. Part of the innovation effort, investments in R&D are uncertain and long-run oriented, and absent market frictions, should not be impacted by current economic conditions. This thesis investigates whether firms’ R&D expenses respond to changes in short-run macroeconomic activity in the G7 countries, while considering the endogenous relation between innovation and economic growth. Empirical studies attribute the procyclical behavior of R&D, especially in the U.S, to financial constraints that drive firms away from the optimal allocation of resources. Moreover, substantial evidence indicates that managers manipulate discretionary expenses to meet their current earnings targets, compromising other dimensions of the business, such as future growth. We study whether macroeconomic conditions affect firm-level R&D investments beyond what can be explained by financial factors and growth opportunities. Our results show that contemporaneous macroeconomic conditions, as well as access to financing, are positively associated with firm-level R&D intensity, most notably in young and high-tech firms. We find a positive association with the 1-year lagged economic conditions in the U.S and Canada, but not in the other countries of the G7. Recessions are especially significant in explaining cuts in R&D investments in all countries. Therefore, R&D seems to be pro-cyclical regardless of financial constraints, suggesting that the latter relation with earnings management activity might be complimentary, rather than mutually exclusive. We address the endogenous nature of innovation and economic growth by estimating R&D and GDP in a panel Vector Auto Regression (PVAR) and find no evidence that reverse causality impacts our results. Furthermore, although aggregate level and the weighted average change in R&D of our sample are reasonably correlated, we do not find a relevant long-run relation between firm-level R&D intensity and economic activity, which calls for further investigation.