We've just about come to the end of the 2011 proxy season, the first to feature widespread “say-on-pay” votes. While the vast majority of compensation plans have been approved, remember the old proverb that failure teaches more than success. So what can we learn from the 21 companies (as of mid-May; the numbers have been rising since) that had a majority of shareowners cast “no” votes on executive compensation plans?

To begin, performance matters. In 2010, the Standard & Poor's 500 returned 12.8 percent (excluding dividends). Yet, companies that tripped over the say-on-pay hurdle returned an average of -4.7 percent, with the median return at -5.4 percent.  Of the 21 companies that failed say-on-pay, 16 trailed the S&P 500 and 12 had negative returns. Under-performance, then, is highly correlated to a negative say-on-pay vote.

But a declining share price isn't enough on its own; thousands of companies under-performed in 2010, after all. What made shareowners turn their thumbs down at these 21 and not at others? Examine the rejected pay plans, and you find numerous compensation no-nos: a lack of alignment between compensation and total shareowner return (TSR); tax gross-ups; single-trigger change-in-control provisions; and high dilution rates. Not every single one of the 21 had every single one of those blemishes, of course. But if you were to take a group portrait, they would be prominent.

The question is: Why would the boards of those 21 companies wave known red flags in front of a newly empowered electorate? We considered four hypotheses.

First, did the compensation committees believe that the triggers were necessary for the company, even if they were known to be controversial? There's no evidence of that, either in the Securities and Exchange Commission filings we reviewed, or in the published reports following votes. We interviewed executives at, and consultants to, a number of the companies, and didn't find any evidence that compensation members considered the triggers to be indispensible.

Second, might these 21 companies have smaller market capitalizations and fewer resources to inform their boards about current compensation practices? Not according to the data. The median market capitalization of the 21 is $1.3 billion, just slightly above the median $1.1 billion median market capitalization of the Russell 3000 index.

Third, could there be a correlation to the compensation consultants used? Of the 21 companies, four tapped Towers Watson, four employed no consultant, two used Frederic Cook & Co., and two used Hewitt. No other consultancies were hired by more than one company. Given Towers' Watson's prominence in the field, those numbers appear normal. So we ruled out any pattern in a chosen adviser.

Say-on-pay appears to be working as an engagement tool. Investors report that companies that have lost votes have promised that changes will come, and that corporate representatives will communicate with their owners this summer and fall to redesign compensation plans.

One finding, however, does suggest a systemic issue. According to the National Association of Corporate Directors, the median tenure for all public company directors in the United States is less than 7.5 years. Yet the median board tenure for the compensation committee chairs at the companies with failed say-on-pay votes is approximately eight years, and the mean is 10 years. (One company recently emerged from bankruptcy; we omitted it from the analysis as the board is composed primarily of new, post-restructuring directors.)

Put another way: Statistically, the average chair of the compensation committees at those companies with failed say-on-pay votes should already be retired from the board. And the figures probably understate the issue, since two of the long-tenured chairs—one with 12 years, the other 17—have been with their respective companies since the initial public offerings. One compensation committee chair has served on his company's board for an eye-popping 39 years!

Why does that matter? Shouldn't the most experienced directors chair this important committee? That's certainly one viewpoint, but two less benign possibilities may partly explain why the plans at these companies seem out of line.

The first is “client capture.” Organizational theorists describe client capture as the development of long-term, close relationships in the regulatory world, that allow the regulated to gain influence over those who are supposed to regulate them. The result is that the regulators try to please the regulated, rather than guard against abuse.

That analogy isn't perfect. Directors are not regulators; one of their critical functions is to counsel and advise management, not simply to oversee and govern it. But that dual role might facilitate client capture over time, as the close working relationship encourages a desire to please. That is why Britain's Combined Code, that country's corporate governance standard, suggests that directors who have more than nine years of tenure should no longer be considered independent.

Behavioral science suggests another reason for concern. Long-tenured directors may “anchor” their vision of what should be to what has been. As a result, they may resist change because they see that change as an implicit repudiation of how they have operated in the past. From their point of view, there is no reason to shift just because the world's expectations have changed. Alternately, they may be blinded by habit, and so remain unaware of the need to modernize compensation standards. From the perspective of institutional investors, either explanation is problematic: The compensation committee is either resistant to change, or has been asleep.

Investors made those views clear through their “no” votes at these 21 companies. In a series of private meetings with some of the largest investors in the country, we've heard remarkable consistency about what will happen next.

First, the good news: Say-on-pay appears to be working as an engagement tool. Investors report that companies that have lost votes have promised that changes will come and that corporate representatives will communicate with their owners this summer and fall to redesign compensation plans. But we also heard another clear message from investors. They intend next year to vote against directors serving on compensation committees at companies that lost say-on-pay votes but still refuse to make adequate changes. For companies with majority voting, directors serving on pay panels could be at risk of being ousted if they don't respond.

Don't Make the Same Mistakes

Like it or not, the advent of the Compensation Discussion and Analysis, together with say-on-pay, means compensation for the top named executive officers is no longer just an internal matter. So what is a compensation committee to do? Let's look at the micro-level first.

In the immediate future, you have a checklist of decisions that are relatively straightforward: Does the company really believe it needs tax gross-ups or single-trigger change-in-control provisions? If the answer is yes, the compensation committee is going to have to find new ways to persuade owners. Other questions are trickier: How do you align pay with performance, when performance is usually measured by TSR, which can be influenced by all types of market moves unrelated to the fundamental condition of the company? But in the end, a checklist is a mechanical reminder not to make the same mistakes twice.

Even more important is to consider the responsibilities of the comp committee, so it can anticipate changes rather than react to them after damage is done. We maintain that the time is right for a new dynamic among the compensation committee, management, outside consultants, and shareowners. Compensation committee chairs, for instance, need to play a far more assertive leadership role in understanding investor expectations, crafting credible policies, and explaining them persuasively to a broad constituency. And boards need to make sure committee membership is refreshed and not simply left untouched for years.

Luckily, we have a well-tested model that all companies and board directors already know: the audit committee. Audit committees must also negotiate the four-way relationship among board, management, shareowners, and an independent expert. Audit committees today are, as a rule, professional in their approach; they stay up-to-date on changing regulations and disclosure requirements, and they supervise the company's public statements relating to its accounting policies and financial condition. In other words, audit committees do for financial accounts and the Management Discussion and Analysis everything that the compensation committee must do for executive pay and the Compensation Discussion and Analysis.

The demands upon the compensation committee are new, and practice has not yet caught up with the new reality. But that's exactly the point. The world has changed. The compensation committee needs to change along with it.