CEO Pay, Performance, and Value Sharing, a paper by academics at the Stanford Business School, discusses the disconnect between the perceptions of CEO pay among directors (who set CEO pay) and the public (who ultimately pay it). According to the paper, two 2016 surveys, by the Rock Center for Corporate Governance and by Heidrick & Struggles, showed that 65% of directors believe that the level of CEO pay is “not a problem,” while 70% of the public believe that it is a problem. These two groups also differ on how to address the issue: 62% of the public believe there should be a maximum amount that a CEO should be paid relative to the average worker, regardless of the company and its performance, while 84% of corporate directors believe there should be no maximum. Almost half of the public surveyed believe that the government should intervene to address this problem – including through measures such as substantial tax increases, strict limits on absolute and relative pay levels, required increases in the proportion of performance-based compensation and elimination of stock options and equity-based awards – while 98% of directors strongly oppose governmental intervention (which, since they don’t think there’s any problem, only makes sense). These discrepancies in points of view between the public and directors, the paper concludes, are bound to perpetuate the controversy over CEO compensation, creating a serious PR challenge as boards try to justify CEO comp levels.

But the disconnect is not limited to the one between the public and directors. Directors also disagree with CEOs about the most appropriate methods of determining CEO comp. For example, while both groups might agree that the extent of value creation is an important element of the equation, they tend to differ on the appropriate method of measuring value creation. According to the paper, directors think total shareholder return (TSR) should be used to measure value creation, while CEOs tend to look instead to profitability measures (such as operating income and free cash flow), which they are more likely to be able to directly influence. In reality, the paper observes, both measures are subject to outside forces, such as broad-market trends, behavioral sentiment or cyclicality. Which of these measures, the paper asks, is more accurate, and, when measuring performance, are there ways that the board can control for fluctuations in the market and general economy? Would a more effective way of demonstrating “pay for performance” be to “calculate the relation between compensation realized by a CEO over a designated period and value creation during that period…. Would the results of this analysis assuage the controversy over CEO pay or exacerbate it?”

What’s more, measuring the contribution of a CEO to the performance of the company can be problematic. According to the paper, research has yielded mixed results, with measures of CEO contribution to the variance in company performance ranging from 3.9% to 29.2%. These disparities are further compounded by the differences among companies, particularly the distinction between those with short product development or life cycles –where CEO impact can be immediate – compared to those with a long cycle, where near-term impact is unlikely. Surprisingly, survey data shows that directors attribute 40% of a company’s overall performance directly to the conduct of the CEO. Perhaps, the paper speculates, this particular perspective of directors might explain, at least in part, “why current CEO pay levels are as high as they are.”

About the only area where directors and public seem to have relatively similar perspectives is on the economic value-sharing arrangement, that is, the percentage of TSR that is awarded to the CEO. According to survey data presented in the paper, shareholders believe it is fair to award the CEO 1% of TSR, while directors would view 2% as fair. Should companies make the economic value-sharing arrangement explicit, the paper asks?