Oxford University Press's
Academic Insights for the Thinking World

Voting and Corporate Governance:
Having a Say

Chris Mallin is Professor of Corporate Governance and Finance & Director of the Centre for Corporate Governance Research at the University of Birmingham. She is the author of Corporate Governance and she blogs with fellow OUP author Bob Tricker at Corporate Governance. The below post is an adapted version of one found on that blog, and is about the important of votes in corporate governance. Her previous OUP posts can be found here and here.


Voting ones’ shares is seen as one of the main tools of corporate governance. In recent times, votes have been cast against adoption of the annual report and accounts, against the appointment, or re-appointment, of certain directors, and against certain proposed strategies. Votes can also be used via the ‘say on pay’ to signal displeasure at executive remuneration packages. Although the ‘say on pay’ is an advisory vote, it may nonetheless be quite effective at making boards think twice about the proposed pay packages for executive directors.

However companies do not always take as much notice of the votes cast as one would like. For example, the recent annual general meeting of Marks and Spencer is a case in point as regards the use of voting as a (vociferous) voice. Andrea Felsted and Samantha Pearson in their article ‘M&S chief defiant amid revolt by investors’ highlight that nearly 38% of votes cast backed a resolution seeking the appointment of an independent chairman within the next year. Sir Stuart Rose, who has been the centre of much criticism since taking on the roles of both chairman and chief executive, did not seem overly bothered by the investors’ views on this matter. There was also much shareholder dissent over the re-election of the chairman of the remuneration committee and over the adoption of the remuneration committee report.

Withheld votes
Whilst the importance of the vote is universally accepted, let us consider what happens in the UK when a vote is withheld. A withheld vote may signal that an investor has reservations about a resolution, or it may be a stronger expression that an investor is unhappy about a resolution, whilst falling short of actually voting against the resolution. However when the ‘vote withheld box’ is ticked on proxy forms in the UK, the withheld votes are not counted as part of the votes cast. For example, after its annual general meeting in May 2009, Shell published the voting results on its website. On Resolution 1 : Adoption of Annual Report & Accounts, there were: ‘votes for’ 3,301,631,965, ‘ votes against’ 3,394,595, and ‘votes withheld’ 16,026,721. However when indicating the percentage split of the votes, ‘votes for’ are shown as 99.90% and ‘votes against’ as 0.10%. The votes withheld were nearly 5 times that of the votes against but nowhere are they reflected in the percentage totals of votes cast. Similarly, on Resolution 4 : Re-appointment of Lord Kerr of Kinlochard as a Director of the Company, there were ‘votes for’ 3,161,974,849, ‘votes against’ 77,443,311, and ‘votes withheld’ 77,876,289. The percentage allocation indicated 97.61% ‘votes for’ and 2.39% ‘votes against’. The ‘votes withheld’ which again exceeded the ‘votes against’ were not reflected at all in the percentage totals. It should be said that Shell does clearly state that “a ‘vote withheld’ is not a vote under English Law and is not counted in the calculation of the proportion of the votes ‘for’ and ‘against’ a resolution.”

The Combined Code on Corporate Governance (2008) under Code provision D.2.1, states that ‘For each resolution, proxy appointment forms should provide shareholders with the option to direct their proxy to vote either for or against the resolution or to withhold their vote. The proxy form and any announcement of the results of a vote should make it clear that a ’vote withheld’ is not a vote in law and will not be counted in the calculation of the proportion of the votes for and against the resolution. However it’s interesting to note that a decade ago, the Report of the Committee of Inquiry into UK Vote Execution (1999), published by the National Association of Pension Funds, stated that whilst it was initially attracted to the idea of adding a third box (being an ‘abstention’ or ‘vote withheld’ box), it then decided that there were several arguments against the inclusion of such a third box. Firstly it might provide investors with an ‘easy option’ so that rather than voting against, they withheld their votes; and secondly since withheld votes are not counted, and have no legal effect, then it could drive down the level of recorded votes.

However as we have seen, the Combined Code does advocate the inclusion of a ‘vote withheld’ box on the proxy form. Therefore, it could be that in practice the addition of a third box which allows a withheld vote but which is not counted, may lead to the understatement of the level of dissatisfaction with some resolutions. Given that institutional investors are coming under more and more pressure to be seen to be active owners of shares, it may be that a ‘vote withheld’ will increasingly become seen as sitting on the fence, rather than taking a decision to vote against. In the US, it would seem that abstentions do have a legal effect under a majority voting system. For example, in a director election if there were more votes withheld than were voted for the candidate, then the candidate would not be elected, hence the abstentions (votes withheld) would have a legal effect.

Broker votes
Turning to US issues, the SEC has recently made some important changes to proxy voting. Weil, Gotshal and Manges (2009) report that ‘the SEC approved a change to NYSE Rule 452, eliminating broker discretionary voting of uninstructed shares in uncontested director elections, which will have the effect of reducing the number of votes cast in favor of the board’s nominees in the election of directors and strengthen the influence of institutional investors and activist shareholders.’

Blank votes
However James McRitchie has brought to our attention the problem of blank proxy votes which go to management. He highlights that fact that the broker vote issue that the SEC took care of is ‘where retail shareowners don’t submit a proxy (or voter information form) at all. When that happens, the broker or bank can vote within 10 days of the meeting. The “blank vote” issue arises when the shareowner votes at least one item on their proxy but leaves some other items blank……..[the voting] platform for institutional investors doesn’t allow submission of blank votes, [but the] platform for retail holders does and the SEC allows them to fill in the blanks as instructed by brokers and banks (always with management)’. Furthermore he states that ‘As shareowners who believe in democracy, we have filed suggested amendments to take away that discretionary authority to change blank votes, or non-votes, as they might be termed. We believe that when voting fields are left blank on the proxy by the shareowner, they should be counted as abstentions.’

Clearly the area of voting is a complex one and changes are being brought in over time to remove barriers to voting and to help ensure that votes are cast in a way which fairly reflects the owners’ intentions. A decade ago it would have seemed highly unlikely that many institutional shareholders would publish their voting levels in individual companies and on individual resolutions but many institutional shareholders now do this. In the US a number of institutional shareholders have gone a stage further and disclose their voting intentions prior to a company’s AGM. Hopefully institutional shareholders in other countries will adopt this approach in future.

Read More in…

Recent Comments

  1. Thanks for mentioning our petition to the SEC on “Blank Votes.” Readers who agree with us can voice their opinion by sending an e-mail to rule-comments@sec.gov with File 4-583 in the subject line. See submitted comments at http://www.sec.gov/comments/4-583/4-583.shtml

Leave a Comment

Your email address will not be published. Required fields are marked *