Introduction

Public subsidies to private R&D constitute in most countries one of the main instruments for supporting innovation. The most common argument justifying these programs is based on the presence of market failures, which would lead to insufficient incentives or finance for firms to innovate. The presence of market failures is believed to be particularly important for small and medium-sized enterprises (SMEs) (Peneder 2008). This has greatly encouraged public support in this area through supranational (EU), national and regional programs, with these three levels often interplaying within a multilevel governance framework (OECD 2011). Although programs in favor of private R&D have been implemented for a long time, the literature devoted to evaluating the effects of these programs has a much shorter history. However, during the last decade or so, this literature has benefitted from the methodological developments achieved in the econometrics of program evaluation (Imbens and Wooldridge 2009; Imbens and Rubin 2015). As a result, the number of works trying to establish whether subsidies are effective has increased rapidly, using both general survey data and ad hoc program data. Although findings are often mixed, some recent surveys have highlighted that the majority of works support the effectiveness of R&D subsidies (García-Quevedo 2004; Zúñiga-Vicente et al. 2014; Caloffi et al 2016). Small-business programs, which are very common, but are perhaps more likely to be implemented at the regional policy-making level (Blanes and Busom 2004), have received only limited attention in the applied microeconomic literature, although there are some notable exceptions. One of these is Wallsten (2000): focusing on the well known SBIR program in the US, the author finds that subsidies had no effects on R&D expenditure. On the other hand, more recently, both Lee and Cin (2010) and Czarnitzki and Lopes-Bento (2013) find that small-business R&D programs, respectively in South Korea and Belgium, had relevant effects on some major input indicators, such as R&D investment and/or R&D employees. Other contributions, although not specifically focusing on a small-business program, account for possible effects arising for SMEs. For example, both Lach (2002) and Bronzini and Iachini (2014) find that subsidies had a positive effect on the R&D expenditures only of small firms, while González et al.(2005) and González and Pazó (۲۰۰۸) report relevant R&D-inducement effects for smaller non-R&D-performing firms. Other works consider a wider range of outcome variables: for instance, Merito et al. (2010) find that subsidies shifted the employment structure of SMEs towards more skilled workers and raised their employment levels, Bronzini and Piselli (2016) report positive effects on their probability of applying for patents, while Criscuolo et al (2012) report positive effects on employment and investments limited to smaller firms. Most of these studies have focused on effects that are contemporaneous to subsidy receipt (Zúñiga-Vicente et al. 2014). We instead focus on what happens to firms shortly after the completion of the subsidized projects, in order to establish whether the subsidy has encouraged non-transitory R&D, enhanced the propensity to intellectual property protection or to collaborative R&D with other firms or research centers. As will be explained later in this article, these are aspects of major importance with small-business programs, which often aim to promote behaviors that SMEs are reluctant to adopt.