Activist investors often think that the classification of boards abets sloth, protecting directors from shareholder input. Yet boards understandably value the durability and continuity of multi-year terms, which give them the bargaining power to pursue long-term plans. Companies often try to navigate these differences peaceably. For example, Phillips 66’s board publicly supported a de‑classification proposal back in 2015. But the proposal failed then – and four more times after – despite winning as much as 99 percent of the votes actually cast. Why? Phillips 66’s certificate of incorporation (“charter”) requires an 80 percent outstanding‑share vote to amend the charter or bylaws. Routine annual meetings never hit that turnout, so reform stalled.
2025 may bring a different result. Activist Elliott Investment Management has waged a proxy fight against Phillips 66, hoping the heat of battle will coax more than 80 percent of shares to vote. But Elliott has a contingency plan: to push for a board policy requiring every director to tender an annual resignation letter. The policy would, in effect, de‑classify the board without touching the charter – and therefore without the super‑majority vote.
Governance experts have called Elliott’s plan “innovative” and “creative.” It is certainly creative; it is also, for three distinct reasons, illegal.
The Three Pillars Protecting Classified Boards
To appreciate the impermissibility of Elliott’s plan, consider the current legal authority for creating and protecting the company’s staggered board. Three different sources reinforce the status quo.
- Statutes. DGCL Sections 141(d) & (k) expressly authorize staggered boards and protect directors from removal without cause between elections. A forced early exit guts the statutory bargain. If directors could be required to step down early, shareholders would never need to wait out the three-year cycle.
- Fundamental Documents. Phillips 66’s charter divides directors into three classes, “each elected for a term expiring at the third succeeding annual meeting,” and allows mid‑term removal only for cause. Its by-laws do the same, doubling down on the DGCL structure.
- Fiduciary duty. Directors must implement, not sabotage, the governing instruments of the corporation. A board that coerced mid‑term resignations would breach its duty of loyalty by nullifying shareholder‑approved charter provisions. Directors who dislike classification may work to amend the charter; they may not ignore it.
Because Elliott’s proposal would make every director resign after one year without cause, it collides with all three legal authorities.
Elliott’s Proposal
Here is Elliott’s proposal, which contemplates requiring directors to serve less than their three-year terms:
RESOLVED, that stockholders request that the Board adopt an annual election policy for directors, requiring each incumbent director (including directors with terms not set to expire at the next annual meeting) to deliver to the Board a letter of resignation effective at the next annual meeting of stockholders, each year prior to the nomination of director candidates for election at the annual meeting.
Styled as “non‑binding,” a shareholder proposal is improper if it calls for illegal action – as this one does. It would compel directors serving three‑year terms to leave after one year, forcing annual elections for the entire slate. To the degree that Elliott’s policy is adopted and effective, Phillips would operate as other declassified companies do, without going through the process designated for declassification .
This change contravenes DGCL Section 141(k) and the company’s fundamental documents. If the board adopts this non-binding proposal, its members may have breached their fiduciary duty to effectuate, or at least not undermine, the governing law and corporate documents.
Consider Airgas
This is not the first time parties have tried to skirt legal authority and impose declassification in slightly different language. In Airgas, Inc. v. Air Products and Chemicals, Inc., (Del. 2010) an Airgas shareholder sought to move the company’s annual meeting eight months forward, effectively shortening staggered terms. The Delaware Supreme Court held the plan invalid because it contravened the statutory protection of a directors’ full three-year-term. Elliott’s policy is more aggressive: It would shorten every term, every year. If eight months was too much in Airgas, twenty-four months is a fortiori impermissible.
The Airgas plan was carried forth through a bylaw amendment, and the Supreme Court likewise upheld the longstanding principle that bylaws cannot undo what the charter seeks to achieve. The same reason applies even more strongly to Elliott’s plan for Phillips. It seeks to use a lower authority (a mere “policy”) to undermine higher authority (both the by-laws and the charter).
The Airgas court rejected arguments that the shareholder’s plan was, in some narrow sense, compatible with the law and fundamental documents. Instead, the plan was invalid because it served “to frustrate the plan and purpose behind . . . [Airgas’s] staggered terms and [ ] it is incompatible with the pertinent language of the statute and the [Charter].” The plan did not mention staggered boards or early termination, but it nevertheless “amounted to a de facto removal without cause of those directors.” The same is true of Elliott’s plan for Phillips, which would de facto eliminate directors’ three-year terms without any showing of cause and frustrate the charter’s system of classification.
Why the “Voluntary Resignation” Defense Fails
Professor Jonathan Macey argues that because directors may resign at will, it is no problem to require resignations:
[D]irectors are free to resign their board positions at any time, and nothing in the Phillips charter or bylaws possible can be construed as preventing directors from voluntarily offering to resign. In other words, Elliott is not proposing that directors be compelled to resign.
I find this argument confusing and unpersuasive. Professor Macey seems to accept that there would be a problem if the proposal compelled resignation; he just concludes that Elliott is not proposing that directors be compelled to resign. Yet Elliott proposes “requiring each incumbent director . . . to deliver . . . a letter of resignation.” Is the distinction that Elliott would “require” resignation rather than “compel” resignation? But how could that distinction matter, and what would it even mean? The proposed policy is simply that the company try to stop directors from serving their full terms. Both that goal and the result are invalid and will remain so until the company’s governing documents are amended.
And why are we discussing directors’ freedom to resign? Of course directors are always free to resign their board position. They are also free to drop dead. Does that mean the company can require them to do so at the end of the year? The resignation requirement does not frustrate the charter because the charter forbids resignation, it frustrates the charter because the charter establishes a classified board and a defined procedure for amendment – and the proposal purports to impose declassification without satisfying that procedure.
Focusing on the presumed voluntariness of resignation elides the central point: The proposed plan would require resignation letters. Conditioning renomination (and practical ability to serve) on automatic resignation may be indistinguishable from statutory removal. Delaware courts look to substance, not labels; calling the mechanism “voluntary” won’t mask its coercive effect.
The Charter‑Erosion Problem
The distinction matters because, even if most shareholders want de‑classification, Elliott’s workaround could allow policy to neuter almost any charter or statutory requirement.
The result would sometimes shift power from the board. DGCL 141(a) gives the board authority over the business and affairs of the corporation. That is why shareholders cannot propose resolutions requiring the board to take certain business actions. But why not a shareholder-proposed board policy requiring directors to take a specified action or else resign?
Other times, policy might increase board power, undermining protections and assurances otherwise assured by the fundamental corporate documents. Suppose XYZ Corp’s charter grants nomination rights to one class of investor. Could the company adopt a policy requiring the board to ignore the shareholders’ nominations, or to take any frustrating actions just short of that? Such a policy would vitiate negotiated investor protections.
The predictable next step would be a cottage industry of bespoke policies designed to sidestep super‑majority or quorum protections. Charter provisions would risk becoming mere suggestions, not the “contracts among the shareholders” Delaware insists they are.
Worse, the policy terms would quickly grow even more extreme: If annual resignation letters are valid requirements, why not withhold compensation or D&O insurance from directors who refuse to sign? Unchecked creativity would hollow the corporate contract sub rosa.
Gaming the Proxy Guiderails
Why is the proposal on Phillips 66’s own proxy card at all? Normally, shareholder proposals appear on the company proxy card only after they pass through Exchange Act Rule 14a-8’s substantive filters. Shareholders never see proposals that violate state law. When investors see Elliott’s proposal on the company’s card, they may assume that the proposal is valid and appropriate. Otherwise, the company would have excluded it.
The trick is that Elliott did not advance its proposal through the usual channels. Instead, Elliott included its proposal within its own proxy solicitation. Starting the proposal with its own card exempts Elliott from the requirements of Rule 14a-8. But once Elliott gets its proposal onto the agenda for the annual meeting, Exchange Act Rule 14a-4(a)(3) and practical realities mean that Phillips must print the proposal, too. Investors will thus see the proposal in the same “real estate” as an SEC‑vetted 14a‑8 item even though it never faced those guardrails. That is not illegal, but it erodes federal proxy protections.
Conclusion
Delaware corporate law tolerates many creative corporate‑governance devices, but not those that override a statute, a company’s charter, and fiduciary duties all at once. When super‑majority hurdles genuinely impede necessary corporate action, the answer may be even more concerted proxy solicitation, as Elliott now undertakes. Other times, the answer may be a statutory amendment. Section 242(d) was passed in 2023, allowing smaller majorities to approve new stock, after AMC was unable to muster high levels of participation from otherwise enthusiastic retail investors. The answer is not an end-run around the charter.
Innovation has its place; a violation of the corporate contract does not. Elliott’s resignation policy is a stimulating classroom hypothetical, not a lawful path to de‑classification.
This post comes to us from Professor Andrew Verstein at UCLA School of Law.