The contribution of interregional and inter-field knowledge spillovers to regional Smart Specialisation

ABSTRACT

Industrial policy based on Smart Specialisation emphasizes the exploitation of industrial linkages based on technological rather than intermediate product linkages. This paper develops microlevel analysis using the Organisation for Economic Co-operation and Development’s (OECD) patent citation database on the intensity of technical relatedness depending on the cited and citing technological fields. The results are used to estimate the patterns of interregional knowledge spillovers in the European Union and explain spatial differences in patent growth. The approach to identifying spillovers provides an improved toolbox to guide implementation of Smart Specialisation policies.

KEYWORDS:

• Smart Specialisation Strategy
• relatedness
• interregional knowledge spillover
• regional development
• distance decay

JEL:

• O31
• O32
• R11
• R58

ACKNOWLEDGEMENTS

The author is grateful to Fushu Luan, Shunfeng Song, Yanmin Yang and Anthony M. Yezer for advice; as well as for the constructive comments and suggestions received from two anonymous reviewers and the editors.

DATA AVAILABILITY

The data that support the findings of this study are openly available as the appendix in ‘figshare’ (see https://doi.org/10.6084/m9.figshare.15163968.v5). These data were derived from the following resources available in the public domain: the OECD’s Citations database, July 2020 (see https://www.oecd.org/sti/inno/intellectual-property-statistics-and-analysis.htm).

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the author.

Notes

1. The central strategy is accompanied by four macro-regional strategies aimed at fostering interregional linkages and cooperation: European Union Strategy for the Baltic Sea Region (EUSBSR), EU Strategy for the Danube Region (EUSDR), EU Strategy for the Adriatic and Ionian Region (EUSAIR) and EU Strategy for the Alpine Region (EUSALP).

2. Harmonized Commodity Description and Coding Systems (HS), International Standard Industrial Classification (ISIC) and Standard International Trade Classification (SITC).

3. International Patent Classification (IPC), United States Patent Classification (USPC) and Cooperative Patent Classification (CPC).

4. For access to the data, see https://www.oecd.org/sti/inno/intellectual-property-statistics-and-analysis.htm/.

5. The priority year is the submission year of first application, and it serves as the best approximation of the year in which the innovation activity takes place.

6. Because the number of citations received increases with time, comparison of patents applications submitted in different years can be difficult without controlling for the rightward truncation in potential citing patents.

7. The distance between inventors in the same NUTS-3 region is given by: ${\displaystyle \frac{128}{45\pi }\sqrt{S/\pi },}$where S is the area of the NUTS-3 region, following Murata et al. (2014).

8. The regression model fails to converge when the sample size is small. For instance, there are 320 NUTS-2 regions in the dataset. That means there are 640 fixed-effect dummies in the regression, which impose a restriction on the subsample size of each cited class.

9. Two outliers, IPC classes B04 (${\text{β}}^0$= −1.52, p-value = 0.948) and C05 (${\text{β}}^0$= 42.35, p-value = 0), are omitted from Figure 3 for better layout. C05 is the only class with a significantly positive coefficient. Because of the abnormal magnitude and its counterintuitively positive value, it is to be studied as a special case in the future.

10. Though further investigation of the abnormally positive cases is interesting, this paper considers these five pairs, all in cited class C05, as counter-intuitive outliers that should be omitted from analysis.

 

Additional information

Funding

This paper was supported by the National Natural Science Foundation of China [grant number 72103099].