The practice of allowing shareholders to cast non-binding say-on-pay votes has spread quickly and broadly throughout the world. It seemed that investors would finally get the opportunity to express their dissatisfaction with outrageous or ill-conceived compensation packages.

The practice was, at first, voluntary, with companies having the option of submitting their compensation policies to a vote. As the number of volunteers remained small, though, investors submitted proposals for requiring companies to carry out the non-binding votes.

In some jurisdictions like the United States, non-binding say-on-pay votes were made mandatory. In Canada, say-on-pay votes are not required, but 80 percent of the largest companies have adopted the practice voluntarily or as a result of pressure from investors.

Now that say-on-pay has been around a few years, what does research tell us about its effectiveness? Academic studies provide a mixed view at best. It seems that say-on-pay has led to more dialogue between the company and large shareholders but has not stopped the rise in executive compensation.

Some findings reveal disturbing and unintended consequences. For instance, studies suggest that shareholders base their votes on the performance of a company’s stock rather than on an analysis of the firm’s compensation policies and practices. If company shares do better than those of its peers, almost any compensation package will be approved. This perverse result tends to increase the pressure on management to focus on short-term stock performance, sometimes through decisions that may negatively affect future performance.

This is not surprising, though. It has become far harder to read and understand the particulars of executive compensation. Indeed, for the 50 largest (by market cap) companies on the Toronto Stock Exchange in 2015 that were also listed back in 2000, the median number of pages needed to describe their executives’ compensation rose from six in 2000 to 34 in 2015, with some compensation descriptions consuming as many as 66 pages. Investors holding shares in hundreds of different firms face a formidable task. The simplest approach is to vote according to the stock’s performance or, more likely, to rely on the recommendations of proxy advisory firms, which also base their “advice” in part on relative stock market performance.

Thus, 66 percent of corporate directors do not agree that say-on-pay resulted in a “right-sizing” of CEO compensation. Yet 83 percent of directors very much agree or somewhat agree that say-on-pay increased the influence of proxy advisors, according to a 2016 PwC and Cleary Gottlieb survey: Boards, shareholders, and executive pay.

Boards of directors, compensation committees, and their consultants have come to realize that it is wiser and safer to toe the line and put forth pay packages that will pass muster with proxy advisory firms. The result has been a remarkable standardization of compensation, a sort of “copy and paste” approach across publicly listed companies.

Thus, most CEO pay packages are linked to the same metrics, whether the companies operate in manufacturing, retailing, banking, mining, energy, pharmaceuticals, or services. For the companies on the S&P/TSX 60 index, the so-called long term compensation for their CEOs in 2015 was based on total shareholder return (TSR) or the earnings per share (EPS) growth in 85 percent of cases. The proxy advisory firm ISS has been promoting these measures as the best way to connect compensation to performance.

In spite of, or perhaps because of, the limited usefulness of non-binding say-on-pay votes, various parties are promoting a binding shareholder vote on pay. That is, shareholder rejection  of pay packages or policies would force the board to change and resubmit them to a shareholder vote. Promoters of this measure are a bit hazy on the details and particulars, but the notion is alluring to many investors and government policy makers.

Indeed, the UK has already adopted a form of binding say-on-pay, and France recently enacted a compulsory and binding say-on-pay for French-listed companies. Shareholders of each UK public company vote every three years on the remuneration policy of the company, and the results are binding. Rejection of the policy requires the company to continue to operate according to the previous remuneration policy or to call a general meeting and present a new remuneration policy to shareholders for approval. UK shareholders  also vote yearly and in a non-binding way to approve the total pay (a single figure) awarded to company executives.

The French government is seeking to adopt a system of shareholder votes similar to that of the UK. The enabling act is now stalled in the French senate, but some version akin to the UK’s should emerge soon.

Should Canada or the United States go the way of the UK and France? A binding vote on executive compensation raises many technical issues: Given the complexity of current compensation programs, what are shareholders voting on, and what does a negative vote really mean? In case of a negative vote, will the company carry on with its current policies, which may be worse than the proposed and rejected policies?

At a more fundamental level, the setting of pay policies should be the preserve of the board, as Canadian corporate law clearly states. When egregious pay packages are given to executives, a say-on-pay vote, compulsory or not, binding or not, will always be much less effective than a majority of votes against the election of members of the compensation committee. But that calls upon large investment funds to show fortitude and cohesiveness in the few instances of unwarranted compensation which occur every year.

That is as it should be, notwithstanding the indirect benefits claimed for the practice of say-on-pay.

This post comes to us from Yvan Allaire, the Executive Chair of the Institute for Governance of Private and Public Organizations, and Francois Dauphin, IGOPP’s Director of Research. It is based in part on a recent formal Policy Paper adopted by the Institute’s Board or Directors.