Danone Made Faber's Company a 'Mission' Enterprise. It Didn't Make Him Harder to Fire.

Nine months before Danone's board fired him, Emmanuel Faber stood in front of the company's shareholders and told them they had just toppled a statue of Milton Friedman.² They had voted, 99.4% in favor, to rewrite Danone's bylaws around a "raison d'être" and a mission committee empowered to check the board's decisions against it.¹ It was the kind of line that gets you a standing ovation and a case study. It was also, as a description of what had actually changed inside French corporate law, wrong in a way nobody in the room had reason to check.

The board that removed Faber on March 15, 2021 didn't need shareholders, didn't need a vote, and didn't need to explain itself beyond a two-paragraph press release.¹⁰ It needed a majority of a board that numbered around a dozen directors, and an afternoon.¹⁶ The statue Friedman's admirers actually rely on, it turns out, was never in the room Faber was toppling things in.

The "entreprise à mission" status Danone adopted comes from France's 2019 loi PACTE, which added a new paragraph to Article 1835 of the Code civil letting companies write a "raison d'être" into their bylaws, and created a separate status in Articles L.210-10 through L.210-12 of the Code de commerce for companies that also commit to specific social and environmental objectives, monitored by a mission committee and audited by an independent third party.¹⁴ Danone met all five conditions and became the first listed company to do it.¹

Read the statute and the appeal is obvious: a board that has bound itself, in its own constitution, to something other than the share price has a rhetorical shield against exactly the kind of investor who thinks the share price is the whole point. That is the sales pitch, and it is not dishonest. It is just describing a different part of the building than the part an activist fund actually knocks on.

Articles L.210-10 through L.210-12 govern what the company must report and who gets to challenge the roadmap against it. They say nothing about who elects the chairman, who can remove him, or on what notice. That power lives in Articles L.225-47 and L.225-55, in an entirely different section of the same code, and the loi PACTE didn't amend either of them. Danone's shareholders had, with genuine and near-unanimous enthusiasm, rewritten the company's stated purpose. Nobody had rewritten the paragraph that decides who gets to run it. A mission statement is a description of what the board must weigh. It was never going to be a description of who gets to sit on it.

By November 2020, Danone had reported its first sales decline in years, its stock had fallen 29% on the year against a much softer drop at Nestlé and Unilever, and the board approved cutting up to 2,000 jobs, a quarter of them at headquarters, chasing roughly $1.2 billion in savings under a plan called Local First.⁴⁵ A company that had just told the market it existed to serve more than shareholders was now explaining to a quarter of its head office why they no longer had a job there. Nobody enjoyed pointing that out more than the people about to make the next move.

Bluebell Capital Partners, a small London activist fund whose Danone position never crossed the 5% disclosure threshold, sent the board a letter that same month asking it to separate the chairman and CEO roles and replace Faber, then made the letter public on January 18, 2021.⁶ Artisan Partners, a considerably larger long-only manager that had built a stake above 3% and become Danone's third-largest shareholder, wrote its own letter in February and met board members directly, arguing that "on almost every measure, Danone's performance has lagged."⁷ Together the two funds held less than 6% of the company.⁸ Neither had filed for a shareholders' meeting. Neither needed to.

On March 1, 2021, the board met and announced its answer: unanimous support for Faber, full confidence in his operational leadership, and a plan, proposed by Faber himself, to separate the chairman and CEO roles "in the near future" once a new CEO was found. Jean-Michel Severino was named Lead Independent Director with immediate effect.⁹ Read on its own, this is a board closing ranks around its chief executive. Read against what happened thirteen days later, it looks more like a board buying two weeks to find out whether closing ranks was actually possible.

It wasn't. On March 14, the board met again, and on March 15 announced that Faber would step down as chairman and CEO immediately. Gilles Schnepp, a director for less than a year, became non-executive chairman on the spot; Véronique Penchienati-Bosetta and Shane Grant were named interim co-CEOs while a search went ahead.¹⁰ "Unanimous support" turned out to have a shelf life of thirteen days, which is a useful number to remember the next time a board issues one.

Here is the mechanism that did the work, stated exactly the way the code states it. Article L.225-47 of the Code de commerce gives the board of directors the power to elect its own chairman from among its members, and to revoke him at any time: "Le conseil d'administration peut le révoquer à tout moment." The same article adds a sentence worth reading twice: "Toute disposition contraire est réputée non écrite." Any provision to the contrary is deemed unwritten.¹²

That sentence is doing more than describing a default rule that a company could contract around if it cared enough. Ad nutum revocability of the chairman is a rule of French public policy, and a company cannot amend its way out of it, whether the clause it tries to use talks about job security, a supermajority, or a company's sacred and board-approved raison d'être. If Danone's mission bylaws had somehow included a sentence promising Faber he couldn't be removed without cause, that sentence would have been void the moment it was written, not merely unenforceable in a hard case. The statute doesn't ask what the company meant. It just erases the clause.

Article L.225-55 covers the directeur général, and it carries the same "at any time" language for the CEO seat, with one further detail that founders who like consolidating titles should sit with for a moment: revocation of a CEO without just cause can normally give rise to damages, "sauf lorsque le directeur général assume les fonctions de président du conseil d'administration." Except when the CEO also holds the chairman's seat.¹³ Faber held both, having added the chairmanship to his existing CEO title in December 2017.¹⁷ The board's own March 1 proposal, to split the roles Faber had spent a little over three years combining, would have given him more legal protection going forward than he had on the day it was announced. The board acted on March 15, while the split was still just a proposal, and the one damages claim French law hands an ordinarily-revoked chief executive never came into play. Whether that timing was deliberate or simply how these things fall apart is not something the public record settles, and it doesn't need to for the lesson to hold: combining the two most powerful seats in a French boardroom is often described as consolidating control. Under this particular article, it is also the arrangement with the thinnest exit terms.

None of this is exotic or improperly used. It is the default architecture of a French unitary board, applied exactly as written, to a company whose own bylaws happened to contain a paragraph about purpose that the activists never once had to cite.

It is worth sitting with how small the winning coalition actually was. Bluebell's stake was never disclosed because it never had to be. Artisan's was a little over 3%. Together, under 6% of a company with tens of billions of euros in market value.⁷⁸ No proxy statement went out. No extraordinary general meeting was called. Nobody counted a single share on a ballot, because the decision that mattered was never going to be made by shareholders in the first place.

What Bluebell and Artisan were actually running was a persuasion campaign aimed at a board of about a dozen directors, using public letters, media coverage, and direct meetings to make continued support for Faber look, to each of those directors individually, like the more exposed position. That is a fundamentally different exercise from winning a shareholder vote, and it has a fundamentally different cost structure: you don't need to build a majority of the register, you need to build a majority of a room, and a room is a much smaller and much more persuadable body than a share count. A defense built to withstand the first kind of contest tells you nothing about your exposure to the second, and Danone's mission-company architecture, whatever else it accomplished, was built entirely for the first kind.

By 9:05 a.m. on March 15, Danone's shares were up 4.2%,⁵ part of a move that Reuters reported reaching as high as 4.9% during the session,¹¹ on no news except that the company's champion of stakeholder capitalism had just lost his job. If the market believed Danone's 2020 bet, that a legally embedded social purpose would compound into long-run value the way a share buyback never could, the correct reaction to losing the person who built that bet was a selloff. It got a rally instead.

That is not proof the mission strategy was wrong, and it is not proof Faber was a bad operator; the stock's own two years of underperformance against Nestlé and Unilever is a more direct piece of evidence on that question than a single morning's move.⁵ But a same-day 4% pop is a specific, recorded number, and it says the market's near-term read on "purpose-led governance protects against short-termism" was that removing the person most associated with it was worth buying into. Danone's own investors, the ones the raison d'être was written for, priced Faber's departure as good news before lunch.

If you are a founder or a director looking at a mission statement, a benefit-corporation election, or any ESG-flavored bylaw amendment and wondering whether it buys you room to operate against short-term pressure, the Danone sequence gives you a concrete checklist rather than a vibe.

• Find the chapter, not the clause. A purpose provision usually lives in the articles governing corporate object and reporting. Board election and removal usually live somewhere else entirely. Read both, and check whether the purpose language actually amends the removal provision, because in most jurisdictions it structurally cannot.
• Ask what a public-policy rule voids. In France, ad nutum revocability of the chairman is not a default you can draft around with a strongly worded bylaw; any contrary clause is deemed unwritten. Check whether your own jurisdiction treats board removal power the same way. If it does, no purpose clause, however sincere, changes who can fire you or how fast.
• Price what combining titles actually costs you. Holding chairman and CEO together concentrates authority day to day. Under French law it also strips out the one damages claim an ordinarily-revoked CEO would otherwise have. Before you consolidate two seats into one, find out whether your jurisdiction's version of that trade exists, and whether it's worth what it saves you in board meetings.
• Count the room, not the register. A campaign aimed at your board doesn't need to win a shareholder vote, and building takeover defenses calibrated to shareholder votes won't tell you anything about your exposure to a campaign aimed at a dozen directors instead. If your governance planning only models proxy contests, it is modeling the wrong contest.
• Separate "cannot be removed" from "removal has a price." You typically cannot contract out of at-will board removal. You often can contract for what happens financially if it's exercised: notice periods, severance, or change-of-control terms sitting in an employment or service agreement rather than the bylaws. That is a private negotiation, not a statutory shield, and it is worth having before you need it.

This describes French mechanics specifically; board removal power, damages for wrongful termination, and the legal effect of purpose or benefit-corporation provisions vary by jurisdiction and by a company's own instruments. Confirm the equivalent rules where your company is actually incorporated with qualified counsel before you rely on any of this.

It did what it was built to do: the mission committee kept reviewing Danone's roadmap against its stated social and environmental objectives and reported on it to shareholders, through the crisis and after.¹⁵ What it did not do, because it was never given the power to do it, was cast a vote, block a board resolution, or slow down the timeline between the March 1 concession and the March 15 removal. Reporting oversight and removal power are different grants of authority, and Danone's bylaws only expanded the first one.

Not against the removal itself; ad nutum revocability of a French board chairman is void-any-contrary-clause by design, which is precisely what makes it durable. What he could have negotiated, separately, is what removal costs the company financially: a notice period, a severance formula, or change-of-control terms written into an employment contract rather than the bylaws. That kind of protection is available in most jurisdictions. It just isn't the kind a purpose clause provides, and it isn't the kind Danone's public disclosures describe him as having secured.

The details differ, but the shape of the gap is the same question to ask. A Delaware public benefit corporation statute changes what directors must balance when they act; it is not drafted as a change to who can remove a director or an officer, which remains governed by the ordinary provisions of the DGCL and the company's own charter and bylaws. Before assuming a benefit-corporation election insulates management from a board or shareholder move, confirm what the statute actually amends, the same way you would for a French mission company.

Faber's real mistake wasn't writing a purpose into Danone's bylaws. It was letting the size and sincerity of that vote, 99.4%, stand in for a defense he never actually built. A board that can revoke its own chairman at any time, under a rule no bylaw can override, was always going to be the forum where his job got decided, whatever else the company had put in writing about what it existed for. If you're a founder writing your own purpose clause, benefit-corporation election, or mission statement this year, write it because you believe it. Just don't file it in the part of your brain marked "control," because your board's lawyers already know it doesn't live there.