Counterfactual Impact Evaluation of Public Funding of Innovation, Investment and R&D

This report uses data from Efige and from Bureau Van Dijk’s Amadeus and Orbis to estimate the effect of funding from the EU and national programmes on firms’ employment, sales, added value, productivity and innovativeness. It also looks at the impact of subsidies to investment and R&D (irrespective of the source of funding) on the same variables. In the first part of the report we use only the Efige dataset (covering the years 2007- 2009) and we look at (contemporaneous) correlation between public support (from national and EU sources) and product and process innovation. Our results indicate that national and EU funding are equally important in stimulating product innovation. However, EU funding has a higher correlation with process innovation. We also find a positive correlation between public support to private R&D and product innovation (but no significant correlation between the former and process innovation). On the other hand, public support to private investment (including ICT capital) is positively associated with process innovation but not with product innovation. In the second part of the report we perform a proper counterfactual analysis, where we merge the Efige dataset with the Bureau Van Dijk’s Amadeus (years 2001-2012) and Orbis (2006-2012) databases. This allows us to test whether firms funded between years 2007 and 2009 have a significantly different economic performance (measured in terms of employment, sales, and value added) in the years 2009-2012, while controlling for firms characteristics measured prior to 2007 (i.e. in the pre-treatment period). Our results indicate that receiving public support from national funds generates positive increments in employment, sales and added value, compared to the counterfactual status of the absence of public intervention. We do not find evidence that EU funds have additional impacts on employment, sales or value added (relative to firms receiving only national funding or no funding). This result is most likely due to the small sample size of firms receiving EU funds, which does not allow us to precisely estimate the impact of EU funding alone or in conjunction with national funding. It is also likely to depend upon the features of EU funding, which is geared towards research that produces results over a longer time horizon than the one observable in our data. We also find that generic support to firm-level investment projects has positive impacts on employment and added value. However, no statistically significant impacts are estimated for subsidies which support R&D expenditures exclusively (possibly due to the nature of R&D support policies, which often require more time to yield noticeable impacts on general firm-level performance).