The ECGI blog is kindly supported by

The ECGI blog is kindly supported by International Board Foundation

Does sustainable investing work? A Literature Review

Reading time:

13 minutes

Date published:

Review of the Evidence

In this note we set out the core academic references that we drew on to develop the views set out in our series of articles on the Three-Stage Rocket in sustainable investing (see article links below). This has been written for a practitioner rather than academic audience. We hope that the note provides a structured guide and set of references for practitioners who want to find their way into the extensive academic literature on the topic of the real-world impact of sustainable investing approaches. 

Harald Walkate and Tom Gosling, April 2024

Given the difficulty for practitioners of navigating academic research, we start by providing some commentary on some issues relating to interpretation of the academic evidence, and the approach we have taken in our review.

Interpreting evidence

Busy practitioners can face challenges in interpreting academic research papers. It is not uncommon for such papers to extend to 50 pages and they may include detailed analysis, statistical tests, and methodological choices that are tough for practitioners to evaluate. This is, in part, why we undertook this work, providing our own summary evaluation of the literature in order to provide a structured pathway for practitioners to access the literature in increasing levels of detail as they so wish. However, this does leave practitioners with the challenge of how to evaluate our evaluation! To assist with this, we share some comments below on issues faced when interpreting academic evidence, which we have sought to take into account in our own work. 

  1. Reliability of research

Academic evidence is not necessarily reliable. Working papers may contain errors that are subsequently corrected. Even reviewed and published papers may contain errors that slip through the net and are only subsequently discovered. Results may arise from specific choices in relation to data processing or methodology which vanish when other reasonable choices are made. Professor Alex Edmans at London Business School has provided an excellent guide for journalists and writers on how they should think about evaluating the quality and reliability of research[1]. We have taken Professor Edmans’ guidance into account when considering how to evaluate and weight the significance of different research findings.

  1. Economic versus statistical significance

Academic papers, especially in the finance literature, place significant weight on the concept of statistical significance. When a relationship is found between two factors – for example, investor engagement and subsequent changes in corporate action or performance – then statistical significance tests provide a guide as to whether we should consider the relationship to be real rather than random. However, such tests say nothing directly about the economic significance of the relationship. For example, a hypothetical study may find a statistically significant relationship between collaborative climate engagements by investors and subsequent emissions reductions by firms. But if the reduction in emissions is a one-off non-repeated reduction of 2%, for example, there is a question about how economically significant the result is in the context of the challenge of climate change, which requires at least high single digit, and probably double digit, reductions in emissions annually to contain global warming within tolerable levels. Often the economic significance has to be interpreted from regression tables or, even if disclosed in the paper, needs to be translated out of statistical terminology. This distinction between statistical and economic significance is discussed more in Section 5 of Gosling (2024). That paper finds that many results found in the literature on sustainable investing have low economic significance, even if they may be statistically robust. This distinction is often lost in the marketing or public discussion of such papers, which is frequently simplified into binary statements such as “research shows that engagement works”.  

  1. Stage 1, 2, and 3 impacts

One of the motivations behind our Three-Stage Rocket analogy was the common misunderstanding we perceive in how academic research on sustainable investing impact is interpreted. Frequently, impacts that are identified are, at best, Stage 1 (some impact on companies), with Stage 2 impacts (change in company impact on the real world) highly questionable, even before we consider Stage 3 impacts (contributions to systemic change). For example, in studies on engagement, “successful results” often amount to commitments by companies relating to disclosure (e.g. TCFD or GHG emissions) or future target setting (e.g. net zero or SBTi). Such results are frequently talked of as being “impacts” of sustainable investing. However, such claims ignore the question of whether such commitments, even if delivered, actually result in any real-world change either at the Stage 2 level, or whether any Stage 2 impact that is delivered is simply undone at Stage 3 (e.g. coal assets moving from public to private ownership). In our view, whether accidentally or by design, simplistic presentation of academic findings often results in impact being over claimed when the research frequently shows, at best, impact at the level of Stage 1.

  1. Individual papers versus the literature as a whole

There are very few academic papers that provide results that, by themselves, provide definitive conclusions that can be applied to real world practice. This is because each academic study involves detailed analysis of a particular issue, based on a particular set of data, in a particular setting, using a particular research design. Often the results are not widely generalisable. Therefore, interpreting any single paper as showing that X, Y, or Z sustainable investing practice “works” is fraught with problems. Instead, each paper should be viewed as a single tile in a mosaic, which, taken as a whole, can reveal the picture painted by the literature in its entirety. This is why our approach has been to take a broad review of the research literature in informing our conclusions, rather than relying on one or two studies. 

  1. The inevitable need to interpret

Continuing the mosaic analogy, even a complete review of the academic literature results in a minority of the tiles being definitively and accurately placed. These tiles give a sense of the whole picture but there will always be substantial gaps and uncertainties. Therefore, any broadly applicable real-world conclusions drawn from even a body of academic literature will inevitably require significant application of interpretative judgement on which reasonable people could certainly disagree. Moreover, such judgement is prone to error and, particularly, confirmation bias. It will never be entirely true that “research shows that…”

Our approach

In undertaking this review, we have sought to take as balanced and wide-ranging an approach as possible. But ultimately, as outlined above, our conclusions set out in three short articles are our interpretation of the literature. And in coming to that interpretation, our judgement is subject to exactly the limitations outlined in point 5 above. As will be clear from our writing, we are somewhat towards the skeptical end of opinion on what sustainable investing can realistically achieve. We like to believe that dispassionate review of the evidence has led us to that conclusion. But even if that is true, it would be arrogant of us to assume that we are immune from confirmation bias, which may now be keeping those views in place. 

This is why, as well as providing our summary articles (see article links below), we have chosen to provide this guide to our thinking and process, in the hope that transparency builds confidence in our views. Moreover, anyone who chooses can look at the same evidence we have looked at and then determine whether they reach the same conclusions. 

As a further self-check mechanism, we have also submitted our articles, and this literature review summary, to two academics in the field of sustainable finance, who have themselves written literature review papers: Falko Paetzold at the University of Zurich and Julian Kölbel at the University of St. Gallen. They kindly agreed to review drafts of these articles and have provided us with invaluable input. Noting that reasonable people can disagree on how the academic evidence is interpreted and on which papers are most relevant, we have not asked these academics to say they agree with any of our specific choices or conclusions. Rather we have asked them to identify any critical research papers we have missed or to provide their feedback on whether we have made a reasonable interpretation of the evidence. 

We thank these reviewers for the feedback generously provided, which we have sought to incorporate. We did not agree with all the points made, and so for the avoidance of doubt, we cannot claim that they have approved every one of our statements. Ultimately, our views are our own. But we do feel confident, as a result of this review, that there are at least no glaring errors or omissions.

We now turn to the specifics of the academic papers that we reviewed for the articles.

Academic review papers

Our starting point was to consider recently published academic review papers. We drew on several such papers, because different review authors have different perspectives and scopes. Moreover, while there was significant overlap in papers referenced by, and interpretations of the evidence expressed in, the four papers considered, this overlap was far from complete. In this way we sought to take the broadest possible view of the evidence. The papers are:

  • P Matos (2020), ‘ESG and Responsible Institutional Investing Around the World: A Critical Review’, CFA Institute Research Foundation Literature Reviews, May 2020, ISBN 978-1-944960-97-1, available at: https://ssrn.com/abstract=3668998 or http://dx.doi.org/10.2139/ssrn.3668998

  • J Kölbel, F Heeb, F Paetzold, and T Busch (2020), ‘Can Sustainable Investing Save the World? Reviewing the Mechanisms of Investor Impact’ Organisation and Environment 33(4), 554, available at: https://doi.org/10.1177/1086026620919202 

  • E Marti, M Fuchs, M R DesJardine, R Slager, and J-P Gond (2023), ‘The Impact of Sustainable Investing: A Multidisciplinary Review’, Journal of Management Studies, available at: https://doi.org/10.1111/joms.12957

In addition, we drew on the review of sustainable investing impact mechanisms contained in Section 5 (and the Appendix) of the working paper below:

The focus of these four papers is covered briefly below. 

Matos covers the role institutional investors have played in governance, including the role of blockholders and activist investors. The paper also reviews the evidence on whether ESG performance is beneficial for corporate performance and investment returns. Although Matos pays less attention to the core focus of our own articles – the role of investors in creating environmental and social impact – useful discussion of this topic is included, and the paper helpfully situates this in the broader discussion about the role of ESG. Matos’ stance is perhaps best summarised in the paper’s conclusion which describes the literature on ESG investing as having a “healthy dose of skepticism” with much more that remains to be explored. This includes evidence of the impact of ESG investing on positive societal outcomes such as the SDGs, where little evidence is found in top-ranked finance journals.

Kölbel et al focus very much on whether sustainable investing approaches have real-world impact, rather than on their impact on corporate performance and returns, and so directly relates to the content of our articles. They analyse the literature under engagement, capital allocation, and indirect impacts (which includes things like stigmatisation, benchmarking, and demonstration effects). They also set out a model of investor impact which, positioning the company as the “intermediary” between investors and real-world impacts, has some alignment with our Three-Stage Rocket model. Overall we would describe their conclusion as one of cautious and limited optimism about the impact of sustainable investing approaches. Engagement is found to be better evidenced than capital allocation, with little evidence supporting indirect effects. The authors conclude that “our results suggest that the current practice of SI has only a modest investor impact, and call for the development of investor impact metrics that reflect the contribution of SI to societal goals.”

Marti et al provide a useful complement to the prior papers by explicitly broadening the frame of the review to include key elements of the social science literature outside finance. They organise their review under three headings of portfolio screening, shareholder engagement, and field building, which are very aligned with the three headings used by Kölbel et al, and under which they build a taxonomy of, in total, 15 potential impact mechanisms, which they use to organise the literature. They also develop the idea of ‘field building’, which involves investors influencing the system in which actors make decisions on sustainability. Field building includes activities like public advocacy, influencing the views of other investors, changing social norms, and developing voluntary standards. The tone of this review is overall more positive on the potential for sustainable investment to have impact than the other reviews. However, the evidence provided adds little additional support for this from the existing literature, and so the paper sets out a broad research agenda for the future. The authors postulate that sustainable investing is most likely to contribute to real-world improvements in ES outcomes by acting as a distributed process in which direct and indirect impacts reinforce and recommend a research agenda accordingly.

Gosling provides an analysis of arguments relating to universal ownership theory rather than expressly providing a comprehensive literature review. However, his review of research on investor impact channels in Section 5 of the paper, plus the Appendix, provides a useful complement to the three academic review papers outlined above. In particular, he investigates the economic as well as the statistical significance of the findings of the papers underlying, in particular, the capital allocation and engagement headings, which contributes to our analysis of the Stage 1, 2, and 3 impacts in our Three-Stage Rocket framework. Overall, he finds that while there is evidence for impact of sustainable investing strategies, the evidence is mixed, the real-world impacts rather small, and their robustness to stage 3 offsetting impacts from competitive or consumer responses uncertain. Moreover, even in relation to the best-evidenced channel – engagement - he finds that the successes identified in the literature often relate to items that are either low cost for companies to implement or win-win for companies and shareholders. He therefore questions the extent to which sustainable investing approaches are likely to bring about real-world change on anything like the scale required.

Additional papers

Where papers that we considered are referenced in the four review papers outlined above, we do not cite them separately in this note, because for practitioners, at whom this work is directed, the literature is most readily accessed via those review papers in the first instance.

However, in addition to studying these review papers and the references contained therein, we also identified a number of papers, both published and unpublished, relevant to the issues we are considering and which were not referenced in the review papers outlined above. Many of these papers were recommended to us as relevant by leading academics in the field with whom we interact. A number are recent working papers. For completeness, we list these papers in the ‘Additional Bibliography’ below. The fact that we have separately listed these papers purely reflects their absence from the review papers cited above and should not of itself be interpreted as us ascribing any greater or lesser importance to them.

Additional Bibliography

-------------------

By Harald Walkate and Tom Gosling

Read all three articles in the series

Article 1 – Introducing the three stage rocket analogy

Article 2 – Launch and reaching the earth’s lower atmosphere

Article 3 – Reaching orbit?

ECGI Blog: Call for Views: Does sustainable investing work?

Project Page

 

The ECGI does not, consistent with its constitutional purpose, have a view or opinion. If you wish to respond to this article, you can submit a blog article or 'letter to the editor' by clicking here.

This article features in the ECGI blog collection Responsible Investment

Related Blogs

Top