New Empirical Evidence on the Positive Role of Market Leaders and IPRs Protection in R&D Investment

October 25, 2008

In a recent paper on "The Effect of Entry on R&D Investment of Leaders: Theory and Empirical Evidence" Dirk Czarnitzki, Federico Etro and Kornelius Kraft have provided new empirical evidence on the determinants of R&D investment and on the role of market leaders and of the protection of intellectual property rights. There is a long debate on the role of market leaders in investing in R&D. Following Arrow (1962), a popular view regards competitive pressure as being supportive for innovative activity and claims that incumbents tend to be less innovative than the outsiders. This article reconsiders this view both from a theoretical and an empirical perspective. Following the endogenous market structures approach, a theoretical model provides hypotheses on the incentives to invest in R&D for incumbent leaders and outsiders. It establishes the crucial role of entry pressure on the behavior of leaders and followers. In markets with endogenous entry each firm tends to invest less, but when the incumbents have a leadership position in the competition for the market, they tend to invest more than the average firm. Hence, the endogenous market structures approach obtains the exact opposite of the commonly held view associated with Arrow.

These predictions are tested through a statistical (Tobit) model for R&D intensity (RDINT). The empirical investigation is based on a unique dataset on the German manufacturing sector, the Mannheim Innovation Panel from 2005, which includes a wide number of firm level data and answers to a survey conducted by the Centre for European Economic Research (ZEW) with a special focus on innovation. Besides we summarize the number of firms per industry, the percentage of leaders (defined below) and the average R&D intensity. The Information and Communication Technology sector and the Instruments/Optics industry exibit the higher rate of investment in R&D. A novel aspect of this empirical approach is given by the fact that the same firms provide a subjective view on the key determinants of R&D intensity, the entry pressure and the leadership. Rather than determining arbitrarily the size and composition of a market, assigning a degree of entry intensity in a discretionary way, and assigning a status of leadership on the basis of predetermined variables, using the questionnaire of the Mannheim Innovation Panel allows the firms to identify the size of their main market, the existence of an endogenous threat of entry in the market and the identity of the leader in the market. The key variables identify the existence of endogenous entry pressure (ENTRY) and the status of market leaders (LEADER). Control variables include employment (EMP), capital intensity (KAPINT), a measure of the firms’ patent stock (PSTOCK), the Herfindahl index of concentration (HHI) and sector dummies.

The independence of the entry variable from the dependent variable, R&D intensity, is supported through an instrumental variable analysis and a number of exogeneity tests - instruments for entry are a default index, the minimum efficient scale (MES) and dummy variables on the importance of substitutability between products and advertising. Descriptive statistics for the main variables are reported in Table 1, which can be red as follows: the average firm invests almost 2.3 % of its sales in R&D, employs 300 workers with a capital endowment of 78 thousand Marks each, and 8% of the firms are leaders in their main market.

The main regressions for R&D intensity are reported in Tables 2 and 3 where we adopt different technical assumptions on the statistical properties of the relations under investigation. In Table 2 we propose the homoscedastic Tobit model and find that R&D investement decreases as the threat of entry increases. The leaders' investment does not differ from that of the outsiders. When we add the interaction term of leadership and entry threat, however, interesting differences occur. While the leader dummy is still insignificant, we now find that leaders who are faced by potential entry invest more than the outsiders. The results remain robust when we control for prior R&D using the patent stock. The patent stock is highly significant and positive, confirming that firms receiving stronger protection of IPRs through patents tend to invest more - alternatively, firms that (successfully) conducted R&D in the past will also invest more in the current period.

With respect to the other covariates, we find a positive and concave relation with employment, while capital intensity is positively significant in all models, and the Herfindahl index is always insignificant. Furthermore there are differences in R&D investment across industries. The industry dummies are always jointly different from zero in the regressions, and our results emphasize a high correlation of R&D spending with firms of the Information & Communication Technology.

As Table 3 shows, the assumption of homoscedasticity is rejected for all models (see Wald tests on heteroscedasticity), therefore we move to the heteroscedastic Tobit model. The industry and firm size dummies are always jointly significant. The main results are robust to the model modification. Leaders, in general, are still not differently investing in R&D than the outsiders, and R&D investment is negatively affected by the entry variable. Leaders that suffer from entry threat also invest more than outsiders in the heteroscedastic version. There are no dramatic changes in the estimates of the other covariates. The patent stock is still highly positively significant, confirming the positive relation between IPRs protection and investment in R&D. The estimated employment effect remains stable, however, the positive relationship between R&D and capital investment becomes statistically insignificant, once we correct for heteroscedasticity.

To sum up, our findings on entry are in line with the hypothesis that investment decreases with the strength of endogenous entry threats. Furthermore, incumbent leaders do not differ in their investment from other firms (LEADER is insignificant), unless they are threatened by endogenous entry. Then the negative investment effect is offset (see the positive sign of the interaction term LEADER×ENTRY). Thus, incumbents invest more than the outsiders under endogenous entry threat.

In line with the theoretical prediction of the endogenous market structures approach, the competitive pressure of the potential entry of other firms induces the market leaders to invest in R&D more than any other firm. This implies that we may have to change our way of looking at persistent market dominance: this may be the result of strong competitive pressure rather than of market power. Finally, the size of the firms and their patent stocks, proxy for the protection of IPRs, are positively related to R&D intensity. Therefore, markets with large incumbent leaders pressured by endogenous entry (not in the production process but in the innovative process) and whose IPRs are well protected may be more innovative than other markets.

 

Send us your feedback and comments about this article