HETEROGENEOUS AGENTS AND SPILLOVERS IN INNOVATION PROCESSES

This work is a collection of papers on innovation, a broad theme that covers several but interconnected issues. Innovation is one of the main determinants of firms’ performance in advanced economies and it is also an important driver of growth ((Romer 1990, Aghion and Howitt 1992, Acemoglu 2002, Jones 2002). Firms often invest huge amounts of resources in R&D to improve their production technology (process innovation, aiming to a cost reduction), create new products and increase the quality of the existing ones (product innovation); in this way, firms aim to increase their market shares and profits. Sometimes R&D investment is necessary to enter the market or simply not to exit. Firms’ innovative activities contribute to make the surrounding economic system more competitive and stimulate further investment and innovation. At the same time,consumers enjoy the benefits accrued by lower prices, better quality and more variety. Despite of the beneficial effects they bring to the society, innovative activities are often associated with a non competitive market structure (harmful for consumers) and with externalities, meaning that innovation is strictly related to situations that bring to market failures and to not socially desirable outcomes (in terms of prices, quantities produced and R&D effort). These considerations often justify the request of government intervention in terms of subsidies to R&D and patent protection. The level of innovation of an economic system is affected by the interaction of several agents and also influenced by external factors. Understanding how the different forces at work interact and which factors enhance or prevent innovation is of primary importance from the point of view of policy makers; aware of the possible incentives and obstacles to innovation, they can develop policies aimed to create an environment that favor investment and growth. In this work I investigate the role of agents’ heterogeneity in innovation processes, an issue that has received attention in the literature on innovation only quite recently, tough its importance has been recognized for a long time. I treat this topic from both a theoretical and empirical point of view, though applied to different subjects. In the theoretical part of the work (Chapter 2 and Chapter 3), I focus on a particular aspect of the issues related to innovation and market failure, namely the spillover externality problem and the use of R&D cooperation agreements (RJV) as a way to enhance innovation and lead the level of investment closer to its socially optimal level. The positive externality is caused by non-complete appropriability of the results of R&D activity and is responsible for R&D underinvestment. Since d’Aspremont and Jacquemin (1988) and Kamien et al. (1992), the positive effect of R&D cooperation agreements in presence of high spillovers has been widely analyzed. However, most of the previous works only consider symmetric firms, discarding the potential relevant impact of asymmetries on firms’ decisions. The empirical literature has emphasized the role of asymmetries in terms of gains from cooperation, that in turn affect decisions about RJV membership (Kogut 1991, Röller, Tombak and Siebert 1998), but these issues have been scarcely taken into account in the theoretical literature, with few exceptions (Baerenss 1999, Atallah 2005). So, in this part of the work I try to fill this gap in the literature taking into account firms’ heterogeneity. Firms can be heterogenous in many respects: efficiency level, type of technology, experience, market size etc. Here, firms’ heterogeneity regards the efficiency of R&D effort. In the empirical part, I study a quite different aspect of innovation, namely the link between immigration and innovation. Owing to the size that the phenomenon of immigration has assumed in the advanced countries in the last decades, immigration has been recently at the centre of the political and economic debate. Economists have studied extensively the potential impact of immigration on a variety of economic and social indicators of host countries, such as natives’ wages (Borjas 2003; 2005, Ottaviano and Peri 2012) and employment opportunities (Pischke and Velling 1997, Card 2001; 2005), firm productivity (Peri 2012), trade creation (Gould 1994, Rauch and Trinidade 2002, Peri and Requena-Silvente 2010) and crime (Bell et al. 2010, Bianchi et al. 2012), just to take a few examples. Until very recently the effect of immigration on innovation and technical change was instead much less studied. Although new evidence is progressively accumulating, it remains nonetheless mostly limited to the impact of skilled immigration in the U.S (Hunt and Gauthier-Loiselle 2010, Stuen et al. 2012, Lewis 2011, Peri 2012).Immigration can affect local innovation in several ways. First of all, immigration entails an inflow of foreign population into a region, and produces changes (i) in the size of the population; (ii) in the average skill level of the population; (iii) in the age structure of the population. All these variables have been recognized to be powerful predictor of innovation. Immigration has also a direct effect on innovation through cultural diversity (spillovers may arise from complementary abilities and different backgrounds, with a positive effect in the production of new ideas). At the same time, greater difficulties in communication and reduction of social capital can act as obstacles to innovation and growth (these negative effects are more likely to arise in presence of low skilled immigrants). Finally, immigrants flows affect firms’ choices concerning technology adoption and investment in physical capital, according to the change in the average skill level they cause in the population. So, in this part of the work, heterogeneity concerns the greater cultural ‘diversity’and the changes in the average skill level of the population induced by large immigration flows. The thesis has the following structure: Chapeter 1 provides an overview of the way in which the main issues related to innovative activity have been treated in the theoretical literature. Starting from earlier works on innovation, mainly focused on the value attached to innovation in monopolistic and competitive markets, it develops analising the two main fields in the literature on innovation: patent race and spillover externality. The part related to spillovers and R&D cooperation is treated in a more extensive way, since the theoretical models I present in Chapter 2 and Chapter 3 belong to this strand of the literature. This chapter contains also a review of the past literature on incomplete information in R&D models. In Chapter 2, I extend standard models on R&D competition vs R&D cooperation in a context of non-complete appropriability of the results of R&D activity. In a Cournot duopoly model with R&D investment stage and spillovers, I introduce asymmetries in R&D productivity between firms that may engage in R&D cooperation. Also, with the introduction of a further stage, I analyze the incentive to cooperate in R&D by forming a RJV. While the existent literature focuses on the comparison between two scenarios exogenously given, I endogenize the formation process and show that, when spillovers are high, due to firms’ asymmetries, RJV is not formed for most of the parameters’ values and does not fulfill the aim of stimulating innovation. This contributes to explain the relatively low diffusion of cooperative agreement in R&D and supports some empirical findings about and the determinants of RJVs formation. Also, in line with the theoretical literature, I find that, when spillovers are low, R&D cooperation reduces the total level of investment; in this case allowing this kind of agreements can be harmful, since in some regions of parameters both symmetric and asymmetric firms have incentive to cooperate in order to avoid investment. Chapter 3 presents a further extension of the model discussed in Chapter 2, namely the introduction of the incomplete information assumption. Here, I investigate the role of R&D cooperation agreements (RJVs) in a context of incomplete information with asymmetric firms, where firms, in addition to set the optimal R&D investment under two regimes (R&D competition and RJV), have also to take decisions about RJV membership. Some interesting results arise from this extended model. (i) When firms compete in R&D, incomplete information about rival’s R&D productivity leads to inefficient investment choices in some regions of parameters; in particular, when firms are actually symmetric, asymmetric information further reduces the investment, with respect to the complete information setting. (ii) A signaling role of cooperation agreements emerges, in addition to the already recognized role in reducing the inefficiencies arising from free riding problem. Revealing its willingness to participate, the efficient firm to signal its type, thus increasing the investment level (innovation enhancing effect) and improving total welfare. (iii) When firms are asymmetric, for most of the parameters’ values, RJV is not formed and does not fulfill the role of stimulating innovation. Chapter 4 (joint with Massimiliano Bratti1) investigates the causal effect of foreign immigration on innovation (patents’ applications) in Italian provinces. We provide evidence for a country which was exposed to a very fast and large wave of immigrations during the 2000s, using a very small geographical scale of analysis (NUTS-3 regions), which enables us to better control for differences in institutional and socio-economic factors which are difficult to observe but which may simultaneously contribute to both attracting new immigrants and increase the innovation potential of a region. Moreover, unlike most papers in the literature which only considered the effect of skilled immigration, (i) we first focus on the general impact of immigration, and then (ii) separately look at the effects of low-educated and high-educated immigrants on innovation. Using instrumental variables’ estimation (and instruments based on immigrant enclaves), we find that the overall stock of immigrants has a significant negative effect on innovation of Italian provinces: rising the share of immigrants by one percent point (p.p.) decreases patenting by 0.064 percent. However, distinguishing the effect between low and highskilled migrants shows that the aggregate negative effect is driven by the prevalence in Italy of low-educated immigrants. In fact, our estimates suggest that an increase of 1 p.p. in the share of low skilled foreign migrants on the population induces a reduction in patents’ applications per 1000 inhabitants in a range between 0.094 and 0.186 percent, according to the method used to classify immigrants by skill level. Instead, presumably due to the extremely low presence of high skilled immigrants in Italy and to the underutilization of their competencies, the impact of high skilled immigrants on innovation is positive, but cannot be precisely estimated.