8  Final Considerations

What was predicted as a tangible possibility before embarking on the analysis turned out to be true: the lack of a precise data-set codebook and of clear theories that could guide configurational choices have weighted a lot in the impossibility of concluding an analysis like the QCA. In this case, the algorithmic modelling approach have been proved to be the most functional in terms of predictiveness (the machine learning algorithm run by the ESGeo preliminary analysis did give reliable results), but the ESG scenario is still in need of methodological and theorical formulations that can become the backbone of the standardization process of ESG ratings around the globe.

I would not rule out the Qualitative Comparative Analysis although the halt in the analysis registered in this thesis, the final chapters give a sense of how much can be done in this direction and the various paths that one could follow to finally achieve an unpacking of an ESG rating score through a research approach that enables theory making.