Government support for SME innovation: Does it work?
Until 2014, there was no pan-European policy tool designed to directly support the innovative efforts of individual SMEs. EU innovation policies were typically focused on cooperative R&D projects bringing together science and businesses to promote cross-border technological innovation.
In 2014, the Executive Agency for Small and Medium-Sized Enterprises (EASME) rolled out an SME Instrument to provide innovation support to SMEs. With a budget of around €3 billion over 2014–2020, its goal was the selection and funding of companies with the most innovative ideas and highest growth potential.
Applicants competed for grants that could range from €0.5 to €2.5 million. Fundable R&D activities included prototyping, testing, design, performance evaluation, monitoring, demonstration, piloting, validation for market duplication, scaling up, and application development.
A recently published study evaluates the impact of these R&D grants. In each competition, firms were ranked by independent external experts. Grant winners were then selected based on the available budget. From a methodological point of view, this feature allows the authors to identify the effects of grants based on a regression discontinuity (RD) design. The idea is to compare projects close to the cut imposed by the budget threshold, namely projects that just made the cut and projects that just failed to make the cut.
The authors estimate that grants trigger sizable impacts on a wide range of firm-level outcomes. The figure below illustrates the RD design and the impact estimate in one specific dimension: citations to patents by grant applicants. Panel A depicts pre-grant data, whereas Panel B depicts post-grant data. In each panel, firms are ordered left to right based on rank (higher rank to the right). The vertical line corresponds to the divider between firms that made the cut (right) and firms that did not (left).
Consider first Panel A. As can be seen, higher-ranked firms were already (slightly) more innovative than lower-ranked firms. That said, there is no discernable discontinuity at zero, that is, at the grant selection threshold. This is consistent with the hypothesis that comparing threshold neighbors (i.e., those that made the cut with those that did not make the cut) provides a quasi-natural experiment of the causal effects of receiving a grant.
Consider now Panel B, corresponding to post-grant data. Differently from Panel A, we observe a clear discontinuity at zero. The difference between the left and the right levels at zero then provides a good estimate of the impact of receiving a grant. Specifically, the authors estimate an increase within the range of 15% to 31% depending on the bandwidth employed.
The authors also show that the effects are higher for younger and smaller firms, as well as for firms in countries with lower credit availability. These results are consistent with the explanation that the positive effect of grants works through the alleviation of firm financing constraints.
Additional evidence can be obtained based on the feature that firms that deserve funding according to experts’ evaluation, but do not obtain it only because of budget constraints, receive the so-called “Seal of Excellence” (SOE). The authors show that most of the effect comes from actually receiving a grant, thus concluding that the certification effect is less important than the actual financing effect.
Pietro Santoleri, Andrea Mina, Alberto Di Minin, Irene Martelli
The Causal Effects of R&D Grants: Evidence from a Regression Discontinuity
The Review of Economics and Statistics (2024) 106 (6): 1495–1510.