تصمیمات بلندمدت در برابر تصمیمات کوتاه مدت: تحقیق و توسعه و ساختار بازار در شرکت های اسپانیایی

I present new econometric evidence on the relation between market structure and R&D using data on Spanish firms. I adopt a different approach from previous studies by distinguishing between long-run and short-run decisions of firms regarding R&D. I assume that the long-run or strategic decision is whether to conduct R&D or not, and the short-run choice is how much to invest once the firm decides to be innovative. I argue that market structure affects long-run R&D decisions but does not affect short-run ones. A Heckman-type selection model is used to test such a relation. The results are robust to several specifications and measures of monopoly power.

The relation between R&D and market concentration has been widely studied in the Industrial Organization literature, perhaps because of the contradictory evidence found when using different data, variables, and econometric models. For this reason, a strong case should be made before embarking in a new study of R&D and market structure. In this paper I present new evidence on such relation using a panel of Spanish manufacturing companies. Several factors make the effort worthwhile. I adopt a different approach from previous studies by distinguishing between long-run and short-run decisions of firms regarding R&D. I assume that the long-run or strategic decision is whether to conduct R&D or not and the short-run choice is how much to invest once the firm decides to be innovative. I argue that market structure affects long-run R&D decisions but does not affect short-run ones. A Heckman-type selection model is used to test such a relation. The advantage of this model compared to the usual OLS or Tobit models is that it allows us to gain information from the fact that the decision of whether to innovate or not and how much to expend on R&D once the firm decides to innovate are different but, depend, to a certain extent on related market and firm characteristics. Most of the previous literature in which a significant relation is found to exist between market power and innovation activities focuses on a particular measure of market power. I include up to five different measures and show that including just one does not fully capture market power and could be misleading. Finally, our rich database makes it possible to control at the same time for many of the characteristics of companies and markets that previous studies have found to be important, including technological opportunities, appropriability and demand conditions, and industry characteristics. In a paper that uses several measures of market power Geroski (1990) finds no evidence of the effect of these variables on innovation. My results show that concentration and other measures of monopoly power have a significant effect on the decision to innovate but not on the intensity of R&D. This result is robust to several specifications and measures of monopoly power. In addition, my results provide insight on several policy issues: more monopoly power is associated with more firms choosing to innovate, but on the other hand, innovative firms could, as a result, innovate less. If the decision to innovate depends in part on characteristics of the firm that do not affect R&D intensity, as is the case in our study, then, policy measures affecting those parameters would lead to more firms undertaking R&D without affecting R&D intensity of already innovative firms. The remaining sections of the paper are organized as follows. In Section 2, I review the relevant literature. In Section 3, I present the econometric methods. In Section 4, I discuss the data and measurement issues. I offer the main results in Section 5. I devote Section 6 to extensions and robustness checks. Finally, Section 7 concludes.

Theoretical models predict that more monopoly power has both positive and negative effects on R&D. Depending on model assumptions, one or the other effect is stronger. Empirical papers provide evidence on which of the two effects dominates. Many papers find no evidence of a significant effect after controlling for industry characteristics, appropriability, demand conditions and technological opportunities. I reach the opposite conclusion by focusing on the decision of firms of whether to perform R&D activities or not. I argue that the decision of whether to do R&D or not and the decision of how much to spend on R&D are related and depend on similar factors. When firms decide to perform R&D, they make a careful evaluation of what are the returns to be expected from the different possible levels of R&D spending given expected market structure. Due to the long-run nature of innovation activities, that involve uncertainty over the invention and high fixed costs, the decision on how much to spend on R&D is not very likely to experience big changes with short-run changes on market structure. Therefore, I use a two-step procedure in which in the first equation I estimate the effect of different market structure variables on the probability of a firm performing R&D activities and in the second equation I estimate the effect of market structure on R&D intensity taking into account the fact that we only observe R&D of firms that decide to undertake R&D. When controls for industry and market characteristics and technological opportunities are taken into account, R&D intensity is not affected by monopoly power, but the probability of a firm being innovative increases with it. The results have interesting policy implications. Policy measures that increase market power of already innovative firms are not likely to increase overall innovative effort. However, changes in market structure can create incentives for new companies to perform R&D activities. This conclusion comes from the fact that market structure variables have different effect in the decision to undertake R&D than on how much to spend. While the debate on the effect of market structure on R&D is far from being solved, the results presented here suggest that more attention should be paid at the effect of market structure on the decision of firms to perform R&D or not. A more careful look to which policy channels can create more incentives for new innovators without distorting the activities of already innovative firms is also suggested by the results.