Skip to main content

Advertisement

Log in

Embeddedness and local patterns of innovation: evidence from Chinese prefectural cities

  • Regular Article
  • Published:
Journal of Evolutionary Economics Aims and scope Submit manuscript

Abstract

The diffusion of innovative activities has been very fast in China since the mid-1990s. The literature nonetheless suggests that internationally-relevant innovation may have delayed gaining embeddedness in some places, depending on the strategy it was “seeded”. This paper posits that different degrees of embeddedness are linked with different local patterns of innovation and investigates these linkages across Chinese prefectural cities. Four research hypotheses are stated, one for each indicator identified in the literature to investigate technological catching up. The empirical exercise is set as an ordered logistic regression of data rearranged from the OECD Patent Databases for the period 1981–2009. The results show that embeddedness is positively linked with innovation that increasingly relies on its own local past and negatively linked with innovative activities more concentrated across patent owners. The evidence of a nexus with originality and technology cycle time is less clear and requires appropriate investigation in future research. At the state of the art, the main hint is that embeddedness is gained where the knowledge paths increase in complexity.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2

Similar content being viewed by others

Notes

  1. The literature also identifies types of RIS based on the characteristics of nodes and ties (among others: Cooke 2004; Asheim and Coenen 2005; Zukauskaite 2018). Their discussion is beyond the scope of this paper, however. What concerns us here is the extent, not the modes, that innovative activities can be said to be anchored into a place.

  2. Compound Annual Growth Rate. The counts here consider domestic applications for creations and inventions from China’s National Bureau of Statistics, China Statistical Yearbooks. Editions referred for data collection are 1996, 1998 to 2003 and 2005 to 2011, available at the National Bureau of Statistics of China, Annual Data: http://www.stats.gov.cn/english/Statisticaldata/AnnualData. Population data are from the United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospect 2017: https://esa.un.org/unpd/wpp. Hong Kong, Macao and Taiwan are not included for statistical consistency. Data extracted on May 21, 2018.

  3. Compound Annual Growth Rate. Patent counts refer to the priority date and applicant location based on fractional counts from the OECD.Stat: http://stats.oecd.org. Population data are from the United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospect 2017: https://esa.un.org/unpd/wpp. Hong Kong, Macao and Taiwan are not included for statistical consistency. Data extracted on May 21, 2018.

  4. Shares are computed on fractional counts of patent applications to the EPO by priority date and applicant location. Data are collected from the OECD.Stat: http://stats.oecd.org. Hong Kong, Macao and Taiwan are not included for statistical consistency. Data extracted on May 21, 2018.

  5. Shares are computed on domestic applications for creations and inventions from China’s National Bureau of Statistics, China Statistical Yearbooks available at the National Bureau of Statistics of China, Annual Data: http://www.stats.gov.cn/english/Statisticaldata/AnnualData. Hong Kong, Macao and Taiwan are not included for statistical consistency. Data extracted on May 21, 2018.

  6. Data are collected from the OECD.Stat: http://stats.oecd.org. Hong Kong, Macao and Taiwan are not included for statistical consistency. Data extracted on May 21, 2018.

  7. By definition, fi is never less than ai or bi (fiai,bii) and the rescaled gaps (gi/fi) are consequently ranging between –1 and + 1 (− 1 ≤ gi/fi ≤ + 1).

  8. Data on population in prefectural cities (pop) are from China Data On Line, City statistics, http://chinadataonline.org/member/city/. Data extracted on February 18, 2016.

  9. An additional rank 5, that is, a prevalence of d/f alone, may result in theory from grouping. It is hard (not impossible), however, to find out that innovators largely agglomerate in a place while their activities are mostly performed elsewhere. The results presented in Prodi et al. (2018) actually report this hypothetical rank 5 as an empty group.

  10. The ordered logistic regression is based on the assumption of proportional odds or parallel regressions. More precisely, the changes in regressors xit are assumed to produce the very same increase in the odds that the response variable crosses any threshold θy (Agresti 2013). Given that the dependent variable takes four possible outcomes ωy = 1, 2, 3, 4, there are three implied thresholds to be considered (θy = ωy − 1). As an example, one standard-deviation increase in the concentration indicator Hit is expected to reduce about 87% the probability that the response variable \(y^{*}_{it}\) moves, let’s say, from the lowest class (Yit = 1) to the three-class block above (Yit > 1) as much as from the lowest two-class block (Yit = 1, 2) to the highest two-class block (Yit = 3, 4). A Brant test based on χ2(2) statistics (p-values between brackets) confirms that the assumption is met by each individual regressor in M3: Oit 0.79 (0.674); Lit 2.43 (0.296); Hit 8.40 (0.103); Tit 1.18 (0.554); Oit × Tit 1.08 (0.582). Accordingly, the null hypothesis of proportional odds cannot be rejected.

  11. Regional dummies are based on the three initiatives of coordinated development launched by China’s national government since the late-1990s to support reducing economic gaps across regions (Li and Wu 2012).

  12. Data for control variables are from China Data On Line, City statistics, http://chinadataonline.org/member/city/. Data extracted on February 18, 2016.

References

Download references

Acknowledgements

The authors are truly thankful to the editors and the two anonymous reviewers whose comments have helped pave a path towards more consistent evidence and results.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Federico Frattini.

Ethics declarations

Conflict of interests

The authors declare that they have no conflict of interest.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Prodi, G., Nicolli, F. & Frattini, F. Embeddedness and local patterns of innovation: evidence from Chinese prefectural cities. J Evol Econ 30, 1219–1242 (2020). https://doi.org/10.1007/s00191-020-00667-7

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00191-020-00667-7

Keywords

JEL Classification

Navigation