Browsing by Author "Louhivuori, Valtter"

Innovations can be seen as an engine of long-term economic growth. Firms conduct research and development (R&D) activities to create new production technology, methods or products in order to rival their competitors. In addition to benefiting the inventor, new innovations have considerable positive externalities through knowledge spillovers. However, the socially optimal level of innovations may not be achieved, because firms can underinvest in R&D if they are not compensated for the positive externalities produced by their R&D activities. Public R&D programs aim to encourage innovation by compensating firms for the positive externalities that they produce. Finland’s recent public efforts on fostering innovation have been globally high by many indicators. Nevertheless, the effectiveness of these efforts has been relatively little scrutinised. This thesis studies the effectiveness of Finnish R&D program in fostering innovation outputs at the firm level. Firm-level patent statistics are used as a proxy for the innovativeness of a firm. A major contribution of this thesis is the comprehensive database that has been constructed and employed for the analysis. The database includes firm-level innovative characteristics for all the Finnish firms during a ten-year sample period, altogether covering more than two million observations for over 400 000 firms. Most of the studies on the effectiveness of the Finnish R&D program rely on the assumption that the researcher has full information on the relevant innovative characteristics that affect a firm’s program eligibility. This thesis addresses the program selectivity concern by employing an instrumental variable approach that exploits regional variation in public R&D funding stemming from the European Regional Development Fund (ERDF) aid regulations. The estimates suggest that when the program selection bias is neglected, program participation is associated with around 10 percentage point increase in patenting probability among active patentees, whereas for all firms, the increase in patenting probability is only around 0.1 percentage points. However, the instrumental variable estimates do not confirm any significant causal effect of R&D program on patenting. This thesis highlights the importance of accounting for the selection bias induced by the R&D program selection criteria. The public R&D agency is found to select firms strongly based on the same characteristics that are highly associated with innovation within firms. Therefore, it is important to ask if some of the supported firms might have had conducted their R&D projects even in the absence of the public support. Analysing the R&D program’s selection criteria plays a major role in scrutinising the effectiveness of public R&D subsidies and in the further development of public innovation policies.