استراتژی های شرکت های فرانسوی برای محافظت از مالکیت معنوی خودشان

In attempting to protect their innovations, firms can choose from a range of mechanisms, which may be either non-statutory (trade secrets, design complexity, and lead-time advantage over competitors) or statutory (patent, design registration, trademark, copyright). Yet, little is known about how firms do actually make their choices from among these different appropriability mechanisms. The aim of this paper is to determine how French firms’ use of intellectual property protection mechanisms relates to the type of innovation, the characteristics of the market sector in which they operate, the firms’ characteristics, and their human resources strategies. Our empirical model draws on four French databases covering the period 2001–2004. Our results show that the choice of a means of protection matters and emerges out of a complex strategy. Our results also reveal that the different statutory and non-statutory means of protection are complementary within their own categories but hardly so between categories.

The protection of intellectual property (IP) is a much debated and crucial question in economics. Patenting seems to be the most obvious protection mechanism. However, Scherer et al. (1959) and large-scale industry surveys carried out over recent decades (Levin et al., 1987, Cohen et al., 1996 and Cohen et al., 2000) show that firms improve the conditions for appropriating the returns on their innovations via different channels. These include patenting but also lead-time advantage over competitors and, moving quickly down the learning curve, secrecy, exploiting their reputation and implementing sales and services efforts (Mendonca et al., 2004). The economic literature seeks to identify which channel is the most efficient. It appears that the ranking of these strategies is not constant and varies with the sector in which the firm operates and the nature of the innovations protected. However, few studies have investigated the strategies which lead firms to use one or another of these mechanisms and the potential complementarities among them, as a single firm may use more than one means of protection. In particular, the studies oppose patenting to secrecy while ignoring the other formal or informal means of protection (Arundel, 2001). We note two main exceptions Amara et al. (2008) and Hanel (2008), although the latter fails to take into account the complementarities between the means. Our study is original in three ways: first we are able to compare the use of seven means of protection: (i) secrecy, (ii) complexity of design, (iii) lead-time advantage (iv) patents, (v) design registration, (vi) trademarks, and (vii) copyright. Second, we test their complementarities by using a multivariate probit model. Finally, we include four types of explanatory variables that are not included together in previous studies. Thus, we test the impact of (a) the type of innovation (product, production method, logistics, support activity), (b) the market sector (demand pull, technology push, number of competitors, R&D cooperation), (c) the firm's characteristics (size, market share, R&D, group membership), and (d) the firm's human resources strategies (employee loyalty, fringe benefits, wages to reduce job mobility). We analyse the use of protection mechanisms by innovative French firms. The study uses the fourth Community Innovation Survey (CIS4) data to provide empirical estimates of the propensity to choose one protection mechanism rather than another. The R&D database provides information on the stock of firms’ research expenditure in 2001. The 2001 Annual Survey of Firms (EAE) provides us with information about the individual characteristics of firms such as size and sector. It also enables us to build indicators as to the number of competitors and market share. The DADS database compiles the annual statements of corporate data (Déclarations Annuelles des Données Sociales, DADS). This database contains information on the total amount of wages and fringe benefits the firm gives to its employees. It thus gives us information about the strategy chosen to secure employee loyalty in 2001. Our results show that the choice of a means of protection matters and arises out of a complex strategy. Thus, non-statutory means of IP protection are mainly chosen by firms that cooperate in R&D and have intensive R&D activities driven by the technology-push process. These firms also develop a human resources strategy to reduce job mobility. The differentiation between secrecy, design complexity and lead-time advantage is based on the type of innovation (e.g. production method vs. support) and the size of the firm. Statutory means of IP protection are more likely to be chosen by firms with a high market share and engaging in external R&D. Our results also reveal that the different non-statutory means of protection are complementary, as are statutory means, within their own categories. However, there is little complementarity between the two categories. Thus, patenting and design registration are independent of secrecy. The remainder of the paper is organized into five sections. Section 2 discusses the range of choice of protection mechanisms for intellectual property. Section 3 describes the model. Section 4 presents the databases and variables. Section 5 reports and discusses the results. Finally, Section 6 concludes.

The aim of this article was to further our knowledge of firms’ strategies with respect to IP protection. Contrary to most previous studies, we were able to consider seven means of protection: secrecy, complexity of design and lead-time advantage over competitors, which are non-statutory mechanisms and patents, design registration, trademarks, and copyright, which are IP rights. We used a multivariate probit model to take into account the correlation between the equations, as firms may simultaneously use different methods of protection as sets of complementary methods. Our estimations simultaneously tested different types of variables to explain the strategy of IP protection, which, to the best of our knowledge, had not been done before. Indeed, we tested the impact of the different types of innovation, market sector characteristics, firms’ characteristics and also of the human resources strategy to limit job mobility. The main result of our study is that the choice of IP protection method can be explained by the type of innovation, the strategy of gaining employees’ loyalty, the size and market share of the firm, its R&D activities, its membership of a group and its sector. Moreover, our results reveal that while the different non-statutory means of protection are complementary so are IPRs. However, the strategies are more similar for the former than for the latter. The IPR strategies are little correlated with the strategies of non-statutory mechanisms. Thus, patenting and design registration are independent of secrecy. Trademark registration is a substitute for every non-statutory means of protection. Non-statutory means of IP protection are mainly chosen by firms from the intermediate goods sector that cooperate in R&D and have intensive R&D activities driven by the technology-push process. These firms also develop a human resources strategy to retain their employees. The differentiation between secrecy, design complexity and lead-time advantage is based on the nature of the innovation (production methods vs. support) and the size of the firm. IPR protection is more likely to be chosen by firms with a high market share and engaging in external R&D. In particular, it appears that large firms with high market shares, belonging to a group and having intensive R&D activities are more likely to use patents. By contrast, the intensity of internal R&D activities is not a factor in choosing one of the other three IPRs. Thus the choice between the different means of protection is partly explained by the organization and the intensity of the R&D. The fringe benefits are quite an important factor in the IPR strategy but further research is required to test alternative specifications such as other measures of compensation or the wage dispersion within a firm if data are available. The difference of choice between patent and secrecy is based on the nature of the innovation and the membership of a group, which is consistent with the results of previous studies (Arundel and Kabla, 1998). Results for technology push suggest that developing advanced technologies requires IP management that strongly considers and uses non formal appropriation methods. It is also important that managers understand what non formal forms of appropriation are and what the present legal environment suggests for their rights and responsibilities in protecting their innovations. Hence, our results show that the choice of a means of protection matters and arises out of a complex strategy. This study encourages further comparison of the different IP protection strategies. It also calls for new theoretical models that go beyond the opposition between patents and secrecy and take into account the other types of protection. Nevertheless, our work also suffers from several limitations. Compared with other studies, we were unable to measure the intensity of the use of each form of IP protection, which could have helped in identifying the value of each of them. Moreover, some of our explanatory variables, such as subsidies or cooperation, are dummy variables. It would have been interesting to identify the level of public funding or the nature of the different partners. Furthermore, as we already noted, it would be very interesting to distinguish innovation in services from innovations in goods and to identify at least, radical and incremental innovation, to obtain a proxy for the value of innovation. Further research is needed therefore. It would be useful to improve the characterization of our explanatory factors. For instance, we should take into account other human resource strategies such as bonuses, the level of qualification and wage dispersion. Moreover, it would be worthwhile having panel data to analyse the evolution of the firm's IP strategy. Furthermore, as we now have more knowledge of IP strategies, further research should test the impact of the different IP strategies on the performances of firms. Finally, it would be very interesting to conduct the analysis at the innovation level rather than at the firm level or at least to divide the sample into firms that innovate in a single direction of innovation and those that innovate in multiple directions.11 This approach might reduce the correlation between the different types of innovation. Even if our study does not indicate whether some means of IP protection are more efficient than others, and so fails to identify the forms to be encouraged, it can contribute to the debate about the public policy on IP. First, the diversity of the means of IP protection seems to match different situations. This study then confirms the need for choice in the use of IP protection. Second, the results lead us to wonder whether access to the different IP protection mechanisms is identical for every firm. Patenting is used more by large firms which engage in intensive R&D. Is patenting an activity for a specific organization of R&D or do small firms suffer from limitations in making patent applications? Do firms without R&D have difficulties patenting their inventions? Depending on the answers to these questions, public policy might facilitate patent registration for some kinds of firms. Furthermore, the use of IP protection mechanisms as a performance indicator should be revisited through our results. Among means of IP protection, patents are almost the only form to be used to evaluate the activities of research and innovation, at the firm, territory or country level. However, our results show that firms use a range of means of protection. Using patenting only reduces the scope of innovative activities: patents identify the activities of one type of firm, and one type of innovation. While innovation protected by non-statutory means is always difficult to identify, the use of copyright and registration of design should be generalized in the studies which aim to measure innovation. Finally, innovation policy, at least in France, mainly tends to encourage R&D activities. Our results show that some types of protections are not characterized by the level of R&D expenditure. In this context, in order to improve innovation capabilities, policy makers should think about proposing other types of public aid to involve a larger number of different firms.