Martin Andersson: Claudio Petti, Lauretta Rubini, and Silvia Prodetti present an interesting contribution that puts the influence of innovation-support policies in China at center stage. Analyzing the influence of innovation-support policy is clearly warranted, not least in China—where innovation, science, and technology are high on the policy agenda. The authors note that the country recently developed a new long-term plan (SciTech Guideline), which focuses on stimulating technology-led innovation and growth. It appears to be an ambitious plan comprising significant financial resources, as well as a variety of policy measures. At the same time, our current understanding of if (or when) innovation-support policy is effective is in fact rather limited. Many governments have different types of policies in place; some are general and span all types of firms and industries, whereas others are targeted toward certain types of industries and/or firms. There is thus a clear need for more systematic analyses of when innovation-support policy actually works that can guide the design and focus of policy initiatives.

Petti, Rubini and Prodetti (henceforth PRP) study one specific type of policy (i.e., government technology-related financial support to firms) and test whether it has any influence on firms’ profits. They use a Chinese firm-level survey data set comprising 405 small- and medium-sized enterprises (SMEs) operating in high-technology sectors, which include information on the amount of financial government support each firm has received. PRP design a so-called integrative moderated moderation model to test whether firms that receive financial innovation support from the Chinese government perform better in terms of profits than other firms.

What is refreshing in PRP's analysis that they acknowledge that the influence of innovation-support policy may depend on the nature of firms’ own R&D investments. Their argument is that the influence that government support has on firms’ performance depends on the nature of firms’ own R&D expenditure. To this end, they make a distinction between informal and formal R&D, where informal R&D is more oriented towards incremental or process innovation. The influence of government support is then expected to be greater in firms that invest more heavily in formal rather than informal R&D. They find some support for their conjecture. Innovation-support policy appears to positively moderate the relationship between R&D investments and performance, but this positive moderation effect diminishes when the firm invests more heavily in informal R&D.

I regard PRPs contribution as an exploratory and empirical contribution that provides additional insights into the complex interplay between innovation-support policy and firm performance. The arguments supporting the idea that the nature of firms’ own R&D investments (informal or formal) seem plausible, but there is no deductive reasoning or conceptual framework presented that lead to clearly articulated hypotheses. Although there is indeed no established theory of innovation in firms readily available to build upon, the literature on the role of firms’ resource bases and absorptive capacity could be used to discuss the role that firms’ own “assets” and investments play in moderating the influence of innovation support on their performance in more depth. For example, is it the nature of R&D investments that is the central factor of importance, or is that rather a proxy for some other type of underlying firm-level resource or asset?

A particular question that arises when reading PRP's contribution concerns the choice of studying firms’ profits to evaluate the effectiveness of innovation support. One line of reasoning suggests that policy is warranted when there are significant social returns to innovation and R&D. From this perspective, policy should support R&D and innovation in firms whose R&D/innovation activities generate the largest social returns. It is indeed established in the research literature that R&D is associated with spillovers. In a recent survey of the literature on R&D and productivity at the firm-level, Wieser (2005) shows that R&D leads to the accruement of spillover benefits by other firms and that estimated elasticities or rates of return of R&D spillovers are in most cases significant and positive. These findings are the typical argument in favor of R&D support to individual firms because it is the R&D activity that gives rise to spillovers and social returns. It is also one reason why most of the literature analyzing the influence of R&D support focuses on whether the policy does lead to more R&D investments in firms rather than simply substitute investment that would have been financed by the firms themselves without support. Against this background, I would have liked to see a bit more discussion of choosing firms’ own profits as the “outcome variable” in focus when analyzing effectiveness of innovation-support policy. Particularly because PRP discuss the issue of policy targeting and directed support at some length in the section on policy implications.

The moderated moderation model is an interesting empirical tool that seems well suited to study the often complex interplay between firm-level outcomes and various investments. It is unclear, however, whether the model allows for causal interpretation. There is always a risk that unobserved firm-level heterogeneity and confounding factors work in the background. Although it is certainly the case that researchers rarely have sufficient data to account for many relevant factors, I would have appreciated a discussion of what potentially relevant control variables are and what it may mean that they are not part of the data set. On an empirical note, it would also have been interesting to see a specification where the authors not only have the sheer amount of financial government innovation support, but instead innovation support as a fraction of total spending on innovation and R&D. Such a specification could yield additional insights as it would inform how large the government support is in relation to firms’ own investments.

In summary, I think PRP present an interesting set of empirical results on the interplays between R&D investments, innovation support, and firms’ performance. The results definitely stimulate the reader to reflect on when and how firm-level innovation support works.


Research and Development Productivity and Spillovers: Empirical Evidence at the Firm Level
Journal of Economic Surveys