What proxy advisors do

Plain English. Proxy advisors are research firms. Institutional investors — mutual funds, pension funds, index funds — hire them to analyze company ballots and recommend how to vote on director elections, executive-pay plans, mergers, charter amendments, shareholder proposals, and proposals to move the company’s state of incorporation. The recommendations can matter, especially when a vote is close, contested, or technical. But proxy advisors are not shareholders, do not sit on boards, and do not decide the legal effect of a vote. Their influence runs through investor voting behavior.³

Academic framing. Proxy-advisory firms are voting-information intermediaries in the shareholder-franchise system. Their effect on outcomes is an empirical question: recommendation direction, investor reliance, proposal type, issuer size, contestedness, and institutional-ownership mix all affect observed vote movement. The legal question is now live as well: federal proxy-solicitation rules, state disclosure mandates, First Amendment litigation, state consumer-protection pleadings, and issuer rebuttal filings increasingly treat proxy advice as a regulated governance channel.⁴

Institutional Shareholder Services (ISS) — wholly owned by Deutsche Börse Group through ISS STOXX (controlling stake 2021; the remaining 19.69% acquired from General Atlantic — signed Feb. 11, 2026, finalized end of March 2026, €1.15B; Deutsche Börse “now holds 100% of the shares in ISS STOXX”) via STOXX since February 2021 — and Glass Lewis — acquired by Peloton Capital Management and Stephen Smith in 2021 — together advise institutional investors on the substantial majority of U.S. proxy ballots.⁵ The classic empirical treatments (Choi, Fisch & Kahan, 59 Emory L.J. 869 (2010); Malenko & Shen, 29 Rev. Fin. Stud. 3394 (2016); Rose, 32 J. Corp. L. 887 (2007)) report advisor vote-effect ranges from low single digits to twenty-plus percentage points depending on ballot item, methodology, and identification strategy — SMU CGI cites each effect-size estimate to the specific study that supports it rather than collapsing them into a single headline number.⁶

1.1  Methodology terminology — advisor-following vote share, not “robovote”

SMU CGI uses the term advisor-following vote share — operationally, the share of institutional votes that aligns with the proxy advisor’s recommendation on a given ballot item, sensitive to ballot-question characteristics and identification strategy (see Malenko & Shen 2016’s regression-discontinuity approach) — in formal data products rather than the colloquial “robovote.” The replacement is methodological, not branding: robovote is colloquial and imprecise; advisor-following vote share is a ballot-item-level statistic that does not by itself entail causal inference. Causal inference requires design assumptions or instruments.⁷

02 The five-layer build map

Live: regulation & litigation — Building: empirical layers 01–04

The vertical organizes its coverage into five layers. Each layer is primary-source-backed and re-runnable. The Texas regulatory cycle accelerated the need to ship the legal layer first; Layer 05 is therefore live (first cycle May 2026; developments through Aug. 5, 2026 in § 5.3), ahead of the originally scheduled Q1 2027 launch of the empirical infrastructure. The four empirical layers continue to build toward Q1 2027 cadence and are reported here with explicit status indicators.⁸

03 The institutional landscape — ISS vs. Glass Lewis

Ownership — scope — methodology — sunset posture — active litigation

ISS and Glass Lewis differ in ownership, methodology posture, and litigation profile in ways that matter for the regulatory and empirical questions this vertical tracks. Figure 3 maps the differences across eight dimensions.⁹