"I have yet to see the case that investing through the stock market has anywhere near as much impact as donating to effective nonprofits." - where do you believe the post claimed otherwise?

The same logic also naturally applies to ESG funds. If a manager is known to outperform because of their superior use of ESG information then, on average, you can expect their fees to rise to reflect this and to neutralize the benefits to incoming investors to the fund.

Great to see this discussion happening! As a practitioner and researcher who has been thinking about these issues for a while I am still highly uncertain about what conclusions we should make about impact investing (II). So the more thoughtful discussion on this the better.

I have some general comments on this space & EA-bias and also some specific follow-up questions.

In the SRI/ESG/II space there is so much variety that a lot of the time people are talking past each other because they aren't talking about the same thing. So, one way for those who are interested in this debate to contribute to it would be to define one very specific impact investing strategy for analysis at a time. This post by Max and Gabe is a good start in this regard as it is focused on shareholder advocacy. John and Hauke's simple analysis of Acumen vs Donate Now vs Invest to Give is also a good start. Their example would be really useful if it was improved to fully represent each strategy (e.g. as John and Hauke are careful to mention they do not account for 'that [Acumen] could continue to reinvest the profits in socially impactful businesses in perpetuity', nor 'unaccounted for advantages to donating now, such as diminishing returns and compounding social benefits') as well as the risks and uncertainties involved.

Another tricky thing with this space is the varying quality of the arguments and literature. It was a strong choice to base this post on shareholder engagement as studies of this have some of the most compelling results and have been published in the best journals (e.g. Strickland et al., Becht et al., and Dimson et al.). Studies focused on the links between environmental performance, financial performance and investment performance typically have weaker data, they often apply weaker methods, and they generally don't get accepted into the best journals. John and Hauke mention some common short comings of studies of screened funds. More generally, my guess would be that many ESG performance papers get desk rejected by top journals based just on unclear logic in their abstracts. For example, not only does a correlation between ESG and CFP not prove causation, it also doesn't necessarily tell us anything about the link between ESG and investment returns (because CFP != stock returns).

The EA community itself also sometimes makes overly simplistic arguments when discussing this topic. My perception is that a very common mistake is always appealing to a somehow infinite supply of 'socially neutral' investors. I was pleased to see that John and Hauke's report pretty much avoided this (that is, they were careful not to say that buying and selling stocks has no price impact). Heinkel, Krau and Zechner (2001) (https://www.jstor.org/stable/2676219) analyze a simple model for what happens if you correctly model the market as being finite. They find that for realistic model parameters you need about 20% of investors to commit to 'green investing' in order to induce some polluting firms to switch. And they estimated that at most 10% of investors were 'green' at the time. The lesson from this basic model is that it makes complete sense in theory that groups of investors can influence corporate behaviour - the groups just need to be large enough. Another, point is that the threshold number of investors depends on the cost to each firm of buying greener tech. This suggests thinking of other (possibly complementary) ways of inducing the reform you want to see like investing in (or grant funding) R&D to make the switching cost lower.

@Max, @Gabe, some questions in line with my points above to make sure I understand what you wrote:

How would you precisely define the Shareholder Advocacy Impact Investing strategy that you discuss in your post? For example, is the idea that by selecting and investing in a stock, and then using my shares to support an advocacy campaign, that I will marginally increase the success probability of the campaign, and that the success will generate abnormal returns that will reward me for the time involved and give me more ammo to make my next advocacy investment? And that this works out to possibly be a better investment of my time and money than other impact opportunities?

Because if there is no investment reward, why not pursue advocacy by some other possibly more effective means? Or, if there is no investment reward but shareholder advocacy is somehow much more effective than other forms of advocacy, what is the minimum number of shares I (or a like-minded group) need to hold to achieve this level of effectiveness? How much does my impact scale with a greater holding?