Print
E-MailEdwin Hu is an Associate Professor of Law at the University of Virginia School of Law, Nadya Malenko is a Professor of Finance and Wargo Family Faculty Fellow at the Boston College Carroll School of Management, and Jonathon Zytnick is an Associate Professor of Law at the Georgetown University Law Center. This post is based on their recent article, forthcoming in the Georgetown Law Journal.
There has never been more interest in restraining proxy advisors. President Trump has issued an executive order devoted to “[p]rotecting American [i]nvestors” from them. Elon Musk has called them “corporate terrorists.” Two House committees are investigating them. The SEC has adopted two sweeping, and entirely conflicting, regulatory regimes, now subject to clashing rulings from three circuit courts. Texas has passed a law that could expose proxy advisors to a wave of lawsuits, and at least thirteen other states have proposed similar legislation.
Yet many of the proposals advanced in this debate rest on flawed economic foundations. In a new article, Other People’s Votes, we offer an economic roadmap for understanding proxy advice and its role in shareholder voting, in the hope that policymakers will target legitimate sources of concern rather than continue to propose counterproductive fixes.
Institutional investors vote on consequential questions: who sits on the board, whether executive pay is excessive, whether a merger should proceed. They must do so across tens of thousands of proposals each season. But the incentive structure of shareholder voting systematically discourages informed participation. An investor who becomes informed bears concentrated costs, while the benefits of any individual vote are diffuse, probabilistic, and shared pro rata with everyone else. The predictable result is free-riding and underinvestment in monitoring.
Proxy advisors emerged as an economic response to this problem. By pooling demand across many investors, firms like Institutional Shareholder Services (ISS) and Glass Lewis exploit economies of scale to lower the cost of participating in governance and to generate more informed voting than most investors would undertake on their own. The industry’s central features—standardized recommendations, automated execution, and concentration—are consequences of the underlying economics of voting at scale.
But the same forces that make proxy advice valuable also weaken incentives for precision and oversight. They limit how much firm-specific analysis investors will pay for, and they reduce investors’ incentives to monitor, verify, or supplement the advice they receive.
Critics have this weakened oversight in mind when they charge that proxy advisors induce “robo-voting”—that investors mechanically rubber-stamp off-the-shelf recommendations down the ballot. We situate this behavior within the broader context of investor decision-making. Drawing on a growing empirical literature, we develop a three-stage framework that shows most investors take a more active role than critics presume:
- Selecting an advisor. Investors choose whether to use a proxy advisor and, if so, which one. The ability to switch creates at least some competitive discipline.
- Customizing policies. Investors depart from benchmark recommendations in favor of policies tailored to their own preferences. Proxy advice is a two-way street: investors help construct the advice they receive, rather than passively accepting an advisor’s judgment.
- Allocating attention. Investors direct scarce attention to the proposals that need it most—larger portfolio positions, closer contests, higher-stakes decisions—and review those individually rather than simply accepting a recommendation.
While some investors robo-vote without tailoring preferences or devoting attention to specific proposals, most use automated execution as a tool to implement deliberate ex ante choices at scale while conserving attention for the votes that matter.
This more active picture is not, however, a flawless one. Under-investment in informed voting is inherent to the system; proxy advisors did not create it, and no regulation can legislate it away. The question is how to channel proxy advisor influence toward better aggregation of investor information and preferences. We propose two complementary reforms grounded in investors’ limited attention budgets.
The first is a system of attention flags: proxy advisors would identify recommendations that involve material trade-offs, firm-specific context, or high-stakes consequences, redirecting investor attention to where additional scrutiny has the greatest expected value and enriching what is currently a largely binary recommendation. Such alerts are not foreign to the industry; many customers already receive analogous flags today. But these services are opt-in, taken up chiefly by the more engaged investors who least need the prompt. Our proposal inverts the default. A naive baseline rule would flag material ballot items for all clients, so that the investors most prone to inattention receive the alert automatically, while those with the capacity to do better are encouraged to tailor the flags to their own priorities and portfolios. A default of this kind reaches the investors who would not ask for it, while leaving sophisticated investors free to adjust it to their needs.
The second is a fiduciary floor: minimum obligations governing proxy voting that prohibit pure robo-voting and require some meaningful human judgment where it is warranted. This floor matters most now because of the rise of AI proxy advice. Advances in artificial intelligence make it cheap to convert raw information into voting recommendations algorithmically, and cheaper still to skip the human review that should accompany them. Because deploying a purely data-driven system costs little while pairing it with oversight costs a great deal, competitive pressure points toward a race to the bottom in human involvement. Yet firm-specific, soft-information signals still require human judgment, and the latent biases of large language models, trained on vast and opaque datasets, are largely unknowable. A fiduciary floor would welcome greater algorithmic customization while ensuring these tools supplement rather than replace judgment. The two reforms work in tandem: heightened duties incentivize targeted engagement, while attention flags lower the cost of that engagement, mark the proposals where unsupervised automation is least defensible, and create a clearer record for assessing compliance with investors’ voting obligations.
Finally, we use our economic framework to evaluate the leading regulatory approaches—asymmetric burdens on negative recommendations (Texas SB 2337), issuer standing to sue, mandatory issuer pre-publication review, restrictions on auto-submission, antitrust action against the duopoly, and efforts to shoehorn proxy advice into existing solicitation or beneficial-ownership regimes. Several would function less as neutral regulation than as indirect efforts to drive proxy advisors from the market or to suppress recommendations opposed to management.
Because institutional investors face strong incentives to conserve attention, weakening proxy advice is unlikely to be offset by greater investor engagement; more probably, it would shift voting toward deference to management, worsening the underinvestment in oversight that the reforms are meant to address. Proxy advisors are not the source of shareholder voting’s tensions; they are an institutional response to them. Effective regulation should start from that premise.
The complete article, Other People’s Votes: The Law and Economics of Proxy Advice, is available here.
