I very much enjoyed Edwin Hu, Nadya Malenko, and Jonathon Zytnick’s new paper, Custom Proxy Voting Advice.   They find that most institutional investors who buy proxy voting advice from ISS and Glass Lewis don’t use their benchmark recommendations, but instead create a tailored set of preferences and get recommendations that are based on those preferences.  Then, in particular cases, they may depart from those recs and vote another way – which in fact appears to happen quite a bit for shareholders who use customized recommendations, because, the authors speculate, the customized recommendations free up attention from less contentious votes, and permit shareholders to focus on the more contentious ones.

The point is important because, first, it may mean that headlines like “ISS recommends XXX” may be less meaningful than we think, because the benchmark recommendation may not be what many clients receive.  And second, these findings continue to demonstrate the folly of the perennial corporate complaints that proxy advisors have too much power and/or shareholders “robovote” in response to proxy advisor recommendations.   The real complaint is that shareholders have too much power and too many preferences, and if that’s the problem – well, management should take it up with them.

The final thing to note is that much of the differential comes, unsurprisingly, environmental/social proposals.  Which makes me want to draw attention to this paper by Roni Michaely, Guillem Ordonez-Calafi, and Silvina Rubio, Mutual Funds’ Strategic Voting on Environmental and Social Issues.  They find that ESG-themed mutual funds within larger mainstream families engage in a subtle form of greenwashing, whereby funds within larger families tend to vote for the E/S proposals when the proposals are very likely to pass, or very likely not to pass – and they deviate and vote with the family for the closer votes.  So they vote E/S more than regular-themed funds, but only when those votes won’t make a difference in outcome.  Which is consistent, I think, with Hu, Malenko, and Zytnick’s findings regarding how institutions use custom proxy voting advice, and deviate from it.