Multiple Voting Rights Are Arriving in France

The French Parliament recently adopted an Act introducing for the first time multiple voting rights in France, anticipating the forthcoming adoption of the EU Multiple Voting Rights Directive (MVRD). While the MVRD mandates the introduction of multiple voting rights in Member States’ national laws, it leaves them substantial flexibility to restrict the conditions under which these rights can be issued. In particular, the MVRD allows Member States to provide for sunset clauses capping the number of voting rights per share, as well as the maximum time duration during which these rights can be granted, and only requires them to allow the issuance of multiple voting rights at the time of the ‘admission to trading of shares’ (as opposed to companies whose shares are already listed).

France made the choice of capping the number of voting rights per ‘preference share’ to 25 (as opposed to one voting right per share for ordinary shares) when the shares are listed on a multilateral trading facility (MTF). When the shares are listed on a regulated market, the issuer is free to issue preference shares with an unlimited number of voting rights. The new regime also provides for a mandatory sunset clause capping the duration of multiple voting rights to ten years, with a possible extension for five more years that can be decided through a majority vote of ordinary shareholders (that is, excluding multiple voting right holders). Finally, companies can only issue preference shares with multiple voting rights at the time of the initial admission of their shares to trading, but not if their shares are already listed.

These main restrictions are complemented by various ancillary rules. For instance, preference shares must be issued to the benefit of ‘named persons’ and thus cannot be transferred unless and until they are converted into ordinary shares. Furthermore, multiple voting rights cannot be exercised for voting certain resolutions at the general meeting, such as those appointing auditors and modifying the company’s bylaws.

This reform represents a major increase in terms of French listed companies’ flexibility to design their corporate governance structure. Until now, these companies could only issue preference shares with double voting rights as well as non-voting shares representing up to 25% of their capital. It also represents a better alternative to previous suggestions to further limit the maximal number of voting rights that could be issued per preference share. Nevertheless, these limitations on shareholders’ freedom to issue multiple voting rights hardly seem justifiable as a matter of public policy. We show it by focusing on the restriction of preference shares with multiple voting rights to companies initiating the admission of their shares to trading and the mandatory sunset clause.

1. The Restriction of Multiple Voting Rights to the Context of Initial Admissions to Trading

One of the main interests of multiple voting rights is that they allow the founders of successful startups to retain control over their company beyond its IPO, even after they have been strongly diluted by new investors. Founders who already benefited from multiple voting rights at the pre-IPO stage (something made possible by French law) may only accept to launch an IPO subject to keeping these rights post-IPO, while (venture capital) investors may count on an IPO to obtain a favourable exit from their investment. Thus, allowing startups the possibility to issue multiple voting rights post-IPO is likely to encourage startups to launch IPOs, to founders’ and investors’ mutual benefit. From this perspective, the MVRD did well by requiring Member States to authorise companies to issue preference shares with multiple voting rights at the time of the initial admission of their shares to trading.

While authorising the issuance of these shares by companies that are already listed is arguably less relevant, one may wonder why France decided to prevent them from doing so. This issuance could have been allowed subject to a vote of the company’s general meeting excluding shareholders benefiting from these shares. This mechanism would have provided shareholders of listed companies with an additional tool to shape the governance of their company at no obvious cost.

The Report from the Finance Commission of the National Assembly on the Act implicitly justifies this choice by stating that the Act aims to encourage the access of companies to public markets, which is also the main concern of the MVRD. In line with this objective, the MRVD only requires Member States to allow the issuance of preference shares with multiple voting rights by companies at the IPO stage (art 4). However, the MVRD also explicitly provides that ‘Member States may introduce or maintain in force national provisions that allow companies to adopt multiple-vote share structures in situations not covered by this Directive’ (art 3). This possibility seems to have been ignored by the drafters of the Act.

2. The Mandatory Ten-Year Sunset Clause

The Report on the Act justifies the enactment of a mandatory ten-year sunset clause, renewable once for five years, by the fact that ‘multiple voting rights are not designed as a way for certain shareholders to consolidate their power over a listed company at low cost and for all time’.

This choice is disputable. To be sure, the reason why multiple voting rights were initially issued to the benefit of a specific person (typically, the company’s founder) may fade over time (typically, because the founder’s vision becomes less relevant than managerial skills, as the company grows). But if the shareholders consider that, for whatever reason, multiple voting rights should be issued for a longer period, why prevent them from doing so, absent any meaningful externality? While the issuance of multiple voting rights for a very long period of time may eventually turn out to be a bad idea, shareholders are ex ante ideally placed to decide what is best for their company. Justifying the mandatory sunset clause by considerations of good governance, as the Report from the Finance Commission does, paternalistically assumes that shareholders cannot decide for themselves on their company’s best governance structure.

Admittedly, startup founders with strong bargaining power may, in theory, ‘twist the arm’ of their investors, by extracting private benefits of control through the issuance of preference shares with multiple voting rights for an excessively long period of time whose corporate governance cost for other shareholders (in terms of voting power) supersedes these benefits. However, nothing suggests that such a disequilibrium between investors’ and founders’ bargaining power exists in practice. At any rate, this consideration does not seem to have been accounted for in the elaboration of the Act.

The risk associated with long-lasting multiple voting rights also seems exaggerated in that shareholders may very well accompany their issuance with various safeguards—eg by preventing their transferability, providing for their automatic cancellation when a specific event occurs, or issuing them to the benefit of a legal entity with appropriate statutory mechanisms rather than to a physical person.

Furthermore, the ten-year mandatory sunset clause could prevent the implementation of original governance mechanisms that have clear benefits as a matter of public policy. A company and its shareholders may for instance decide to pursue the interests of some constituency (other than its shareholders) or humanity at large rather than profits. This company may be willing to issue one preference share with (almost) 100% of the company’s voting rights to a foundation whose mission is to foster the company’s peculiar mission, with ordinary shareholders receiving all of the company’s profits without having a say on the conduct of its business. This type of governance structure is for instance likely to become crucially relevant in the artificial intelligence sector, where key players—such as OpenAI and Anthropic—claim to put the interests of humanity above profits.

Lastly, ordinary shareholders’ ability to vote on an extension of the initial duration of multiple voting rights, but only once and for a maximum of five years, is somewhat puzzling. Indeed, the new regime already assumes that shareholders should be able to decide the issuance of multiple voting rights for up to ten years. In this context, there does not seem to be any clear justification for preventing minority shareholders from extending multiple voting rights for another ten years after the initial period has expired, and to do so more than one time.

In conclusion, the introduction of a mandatory sunset clause capping the maximum duration of multiple voting rights may bring unnecessary rigidity into the shaping of companies’ governance structure, and thereby prevent the adoption of governance structures benefiting both companies and the public at large. The impossibility for companies whose shares are already listed to issue preference shares with multiple voting rights also seems hard to justify as a matter of public policy. However, the decision to allow the issuance of an unlimited number of voting rights per preference share on regulated markets, and up to 25 voting rights per share on MTFs, is both courageous and welcome and deserves to be recognized.

Olivier Gossner is a Research Director at CNRS-CREST, a Professor of Economics at Ecole Polytechnique and a Professor of Mathematics at the London School of Economics.

Paul Oudin is a DPhil in Law candidate at the University of Oxford and a practising lawyer at Vermeille & Co.