177Do Targeted R&D Grants toward SMEs Increase Employment and Demand for High. . .

Previous attempts to circumvent the issues of appropriate matching and evaluation of effectiveness include Söderblom et al. (2015), Autio and Rannikko (2016), and Howell (2017). Their studies utilize detailed information on SMEs that applied for R&D support programs, using the outcomes of firms that were rejected in the final evaluation stage to control for selection effects. Söderblom et al. (2015) found that a Swedish R&D grant targeting growth-oriented startups increased the firms’ growth in terms of employees and sales. Further, the grant program made it easier for the firm to attract external financing and overcome “the liabilities of newness” (Stinchcombe, 1965), i.e., their struggle to develop routines, establish relationships with customers, and reach a more efficient scale of operation. Neither Autio and Rannikko (2016) nor Howell (2017) focused on employment growth, but their results provided indications that R&D grants toward growth-oriented SMEs spurred innovation and increased sales. The strategy to identify the link between receiving a grant and firm-level outcomes relies on the appropriateness of using a comparison group composed of firms that were rejected in the final round of the selection process. The relevance of this identification strategy depends on whether the outcome in the last stage of the decision-making process can be considered as approximately random. Otherwise, the results can be a consequence of an omitted variable that is correlated with the outcome variable and the probability of receiving an R&D grant. However, whether the probability of receiving a grant in the last stage of the decision-making process can be considered as random is highly questionable, since grant-issuing agencies often rely on elaborate ranking processes, usually conducting lengthy interviews to decide which firms will receive an R&D grant.

We estimate the effects of two Swedish government R&D grant programs administered by Vinnova, a Swedish government agency under the Ministry of Enterprise and Innovation. These programs seek to promote the development of new products and processes that could bring about new innovations and promote the long-run growth of SMEs. The instruction from the government explicitly states that Vinnova must account for the change in the number of employees that has taken place in SMEs to which they distributed support during the period 2006–2009. This is important as it articulates the effects on employment as a societal aim of these programs (Ministry of Industry, 2013). Note that these programs explicitly aim to promote new innovations and new knowledge, implying that they should have a particular effect on the recruitment of employees with high human capital. Consequently, we study the effectiveness of these grant programs by investigating whether they have a positive effect on the number of employees and the share of highly educated (our measure of high human capital) workers in the targeted firms.

To circumvent the methodological issues discussed above, we use Coarsened Exact Matching to construct a control group from the full population of Swedish firms. Our analyses are based on a unique longitudinal dataset on targeted R&D grant programs in Sweden, which is linked with matched employer–employee data from Statistics Sweden. We find no evidence that the government support programs had any positive and statistically significant effects on the number of employees brought into the targeted firms. This includes a lack of effect during the short-term S.-O. Daunfeldt et al.

178period in which the firm received grant support, as well as up to 5 years after the support program ended. We also investigate the effect of the government support programs on the demand for employees with higher human capital (as captured by the share of employees that have completed higher education) in the targeted SMEs. This is important because targeted R&D grants might encourage SMEs to recruit workers with higher human capital and higher salaries, rather than increasing their total number of employees. However, we find no robust evidence of any impact of the targeted R&D grant programs on the share of highly educated workers either.

The absence of any statistically significant effects is troublesome considering that targeted R&D grants require a positive impact that at least covers the administrative costs associated with these programs. Our results thus question the relevance of implementing government support programs targeting SMEs with high-growth potential. We believe that our findings challenge the more established norms that are widely accepted in terms of providing government grants to highly innovative firms. In illustrating the potential lack of impact of these grant programs, we hope to raise potential issues that relate to the selection mechanisms involved in such grant processes and consider alternative measures and outcomes of these grants.

2

Effects of Targeted R&D Grant Programs on Labor

Demand

There is an extensive literature evaluating different aspects of targeted R&D grant programs. Overall, this literature illustrates an equivocal state of affairs concerning