The effects of business innovation modes on eco-innovation: where do environmental benefits materialize?

ABSTRACT

This study investigates the efficacy of various innovation modes in fostering eco-innovation (EI). EI refers to innovations that provide environmental benefits both within a firm and during the use or consumption of products by end users. The study focuses on the impact of cooperation with science, technology, and innovation (STI) partners, as well as learning by doing, using, and interacting (DUI) partners on EIs. Utilizing data from the Basque Innovation Survey spanning 2018–2020, we employ a logit function to estimate the average treatment effect of the treated (ATT) resulting from cooperation with different partners. Specifically, collaboration with research centres, consultants, and suppliers yields significant benefits for EI within firms. Furthermore, all partnerships positively influence EI with environmental benefits during end-user consumption, albeit to varying degrees of relevance. These results offer valuable insights into cooperation-oriented strategies and provide recommendations for policies aimed at fostering EI, thereby mitigating the environmental impact of firms’ operations and products.

KEYWORDS:

• Eco-innovation
• cooperation
• STI mode
• DUI mode

Acknowledgements

The authors are grateful for the work of the editor and the valuable advice of the reviewers.

The authors would like to thank Prof. Juan José Gibaja for his comments during the development of the paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Section C.5 in the BIS 2020 asks questions regarding EIs and the decision factors driving their introduction.

2 Using a 3-point Likert scale (no; yes, but insignificant; yes, significantly), firms selected from a set of typologies of environmental benefits that they might have introduced.

3 It must be noted that these are self-reported environmental benefits, not objectively measured via environmental indicators; therefore, there is a potential bias given the subjectivity of the respondent.

4 For a detailed discussion of the advantages of this approach over others, such as instrumental variables, see Parrilli, Balavac, and Radicic (2020).

5 We conducted a robustness analysis using alternative matching algorithms: full matching and nearest neighbour with exact matching on firm size. These models are not reported in the paper because of space limitations, but they are available in the supplementary material. See Appendix A for the visual balance assessment for the entropy balancing technique.

6 For technical details, see Hainmueller (2012).