Beyond input & output additionality of public R&D support: behavioural additionality & implications for industrial strategy

Having identified a gap in the literature of innovation policy I would like to explore further the behavioural additionality of public R&D support programmes. This project will involve collection of survey data on firms' innovation behaviour. Given that well-established practices on collecting behavioural additionality data do not exist, the survey design is rather critical in the success of the PhD project. Firms that have received R&D support & firms that have not received but applied for R&D support will consist the units of analysis. Data on public R&D support will entail both R&D tax credits and R&D subsidies, which would enable identification of potentially differential behavioural additionality effects across the two measures. This has never been explored so far. These effects can be explored within European economies for which rich data on innovation are available; namely, Flanders, Germany, Spain, UK and Turkey (if possible through OIVs). More specifically data will be collected on: R&D tax credit/subsidy; Private R&D spending & innovation outputs of firms; Behavioural additionality indicators; Firm characteristics such as number of employees, sectoral affiliation, year of establishment, spending on R&D, turnover etc which can all serve as control variables. In this PhD project, an attempt to build upon & extend the literature on the determinants of R&D behavioural additionality will be made. Both quantitative and qualitative analysis will take place. From a quantitative perspective, parametric and non-parametric micro-econometric approaches will be used to test the hypotheses developed in the theoretical part of the PhD project. Propensity Score Matching (PSM) will be implemented as a means to evaluate the behavioural additionality of public R&D support. This evaluation method will address the potential endogeneity arising from the selection of firms in the R&D support schemes. The matching will take place conditional upon pre-specified firm characteristics(eg industry, size, innovation intensity etc) & some comparison behavioural indicators could be the size of the projects, the willingness to cooperation and willingness to engage in risky R&D. However, a serious weakness of matching approaches is that these methods account only for observable and not for unobservable influences. To this end, & given that longitudinal data may be difficult to obtain (to adopt a Difference-in-Differences estimation), we will also use appropriate instruments to estimate the corresponding regressions. All these types of additionalities are expected to contribute to higher levels of R&D & innovation which, in turn, can lead to higher productivity & eventually higher growth rates in the economy. This PhD project will inform the policy maker of the potential need for restructuring existing STI policy instruments &/or the development of new STI & industrial policies able to also accommodate the behavioural effects stemming from such R&D policy measures. Further, given that R&D subsidies can be more targeted to specific sectors while R&D tax credits are of a more horizontal nature (they apply equally to all eligible firms), perhaps a vertical innovation & industrial policy allowing for sectorally-differentiated R&D policies would be desirable. Such policies would be determined by the potentially various degrees of input, output & behavioural additionality realised across different sectors of the economy.