Case study research on innovation systems: paradox, dialectical analysis and resolution

This paper deals with a largely unnoticed methodological paradox concerning the scientific status of case study research on innovation systems (ISs). While case study research constitutes the methodological catalyst to the genesis and establishment of the ISs approach as one of the most widely used theoretical and policy-relevant perspectives on innovation in the social sciences, many ISs scholars believe that this type of research can not infer causality and generality. To heighten our understanding of, as well as resolve, such an enduring methodological tension, this paper utilises the dialectical method, particularly the analytical triad of thesis, antithesis and synthesis. It shows that the paradox is attributable to the absence of a compelling antithesis to the deductive thesis, wherein (and grounded in the hypothetico-deductive model of science) case study research on ISs is incapable of drawing causal, generalisable knowledge. In line with a growing number of critical realist studies on innovation, the paper employs the realist mode of scientific explanation known as retroduction as a means by which to articulate the retroductive antithesis, whereby case study research on ISs is necessary to learn about the general aspects of causality, and especially causal mechanisms. The dialectical tension between the two antagonistic theses culminates into an original methodological perspective– the detroductive synthesis, wherein the ability of case study research on ISs to study causality and generality depends mainly upon the model of science that, either explicitly or implicitly, informs the analysis. Overall, the present paper not only resolves the case study paradox but also enables innovation researchers to reimagine and conduct case study research in a stand-alone, paradox-free, causal explanatory and generalisable way. Thought-provoking implications for the peer-review process and innovation policy evaluation are also discussed.