The goal of this dissertation is to facilitate the assessment of impacts from sustainable measures and projects with an emphasis on impact reporting for Green, Social or Sustainability Bonds in the Sustainable Finance market. It does so by providing analysts with the means to develop, depict, formulate, and assess a causal hypothesis between an intervention and its subsequent effects in an impact-chain, represented by desired environmental (E), social (S) or governance (G) changes. This is achieved by developing a methodology for so-called ESG Logic Models or ESG-LM, that combine heuristic Theories-of-Change with propositional logic and Bayesian Reasoning.

Three research questions are investigated and responded to. Research Question 1 asks how such Theories-of-Change can be developed for any type of ESG-related issue and how the different process steps in a causal chain can be classified, hierarchised, and prioritised regarding their efficacy towards overarching sustainability goals and their plausibility. Research Question 2 studies (a) the means by which the analyst or any other interested third party might be warranted in believing the causal claims from an ESG-LM, and (b) how an ESG-LM can be improved if this credence is low. Research Question 3 then looks at the reporting of impacts themselves regarding indicator selection, indicator assessment and indicator quantification as well as the provision of information on the contributions and attributions by different actors.

The dissertation draws on a variety of theories and adapts existing methods to achieve that. It operationalises concepts from empirical Sustainable Finance research and already existing impact assessment methodologies. It adapts scholarly and practitioner approaches for theory-based evaluation and applies a qualitative social science perspective towards theory-building and evaluation, while some of the assessment tools in the dissertation are grounded in Logic, Set Theory and Bayesian Epistemology. Examples for such tools include rules for the Attribution by actors, heuristics for the abduction of plausible outcome pathways, or a four-stage Argument and Decision-Tree to assess the credibility of ESG-LM claims (based on Bayes Theorem).

My assessment of the entire methodology is positive overall, as it provides solutions to each of the three research areas. Limitations of the approach, and thus opportunities for further research, are the additional expertise and time required by analysts compared to the existing, and somewhat more pragmatic, solutions in the current market. However, this is outweighed in my opinion by the ability of the framework to strongly mitigate impact washing by actors in the financial markets as well as biases by analysts. Its overall methodology also provides opportunities for new research angles in the area of sustainability indicators and assessments