The Operationalization of Smart Specialization in Estonia

The smart specialization concept has become a central policy-prioritization framework in the European Union (EU) aimed at promoting growth through innovation. This framework emphasizes the importance of contextual specificity in understanding different comparative advantages and development trajectories of EU regions. Instead of top-down industrial policy approach focusing on specific sectors and picking winners, smart specialization emphasizes bottom-up process of entrepreneurial discovery in finding domains for specialization.

Although, the smart specialization has been extensively and intensively covered in the academic literature and various EU guidelines and policy papers, the operationalization of concept remains a challenge on the ground. The purpose of this paper is to trace the process of smart specialization operationalization in Estonia. It does so by reviewing key Estonian policy documents and reports on smart specialization in the context of academic and policy literature.

It argues that policy reports used for selecting smart specialization domains in Estonia suffer from a weak conceptualization of smart specialization. More extensive and intensive conceptualization would have contributed to a better public understanding of smart specialization as well as more focused operationalization of smart specialization in empirical work.  The measures used for identifying key domains oversimplified the meaning of smart specialization. A survey was conducted among  experts with question whether a domain has strong basis and potential for entrepreneurship and science. However, smart specialization emphasizes that innovation does not need to be science-based but it can be knowledge-based. Such oversimplified interpretation of smart specialization unnecessarily narrowed down choices for experts.

Third, the methodological choices made are unclear and poorly documented. A conference, interviews and surveys for selecting three smart specialization domains were used in Estonia. However, it has been poorly documented who were interviewed, who were experts and who participated in the conference. The selection criteria for surveys and interviews as well as different analytical methods used are not clarified. Questions are not made available in the reports. All of this reduces measurement validity.

The problems highlighted above reduce the reliability and trustfulness of analytical work conducted for selecting smart specialization domains. Hence, the last part of assessment contributes to the policy debate on smart specialization by offering a way for operationalization of concept. On the basis of literature, it is suggested that smart specialization can be divided into four sub-concepts: embeddedness, relatedness, connectivity and diversity. The sub-concepts can be operationalized by using different quantitative indicators which are highlighted in the paper.  

The last part assesses possible policy mix by reviewing outcomes of previous policies as well as suggestions from the EU guidelines. If theoretical logic of smart specialization suggests finding a few domains for specialization, then public policy making in practice suggests selecting a wide range of domains as this increases the flexibility. If the entrepreneurial process of discovery would lead finding a new domain, then transaction costs for replacing existing domains are too high. De-centralized entrepreneurial process of discovery and centralized process of policy-making create tensions in the implementation of smart specialization.