October 12 2010 - Regardless of economic conditions, there is one constant in the workplace: CEO salaries keep going UP. A
study published earlier this year by Thomas A. DiPrete, Greg Eirich and Matthew Pittinsky puts much of the blame on the practice of compensation
Standard practice in American corporations, benchmarking is a process used by compensation committees
to compare peer groups of executives at similar organizations in order to establish a "fair" market salary for their CEOs. However,
it seems that each year a few CEOs "leapfrog" their peers by getting large increases in compensation that have little
to do with the financial performance of their organizations. Their unjustified raises are used by other firms in subsequent benchmarks. Gradually, the study
concludes, executive pay is ratcheted up for everyone through a 'contagion effect.' According to DiPrete:
"We show that rising CEO pay is not simply a function of what individual companies do, but is influenced by the behavior of leapfroggers at
Using procedures laid out in reward management handbooks the researchers reconstructed likely peer groups for
CEOs listed in Standard and Poor's annual compensation surveys. Then they looked for evidence of leapfrogging in those likely peer groups over
time. They found that leapfrogging could explain around half the overall increase in CEO pay from 1992 to 2006.
Previously, the debate on the causes of ever-increasing CEO pay had mainly fallen into two groups:
- CEOs were overpaid because of failures in corporate governance at individual organizations
- CEOs were paid what they deserved on the basis of profits they delivered to shareholders in a "superstar" job market
The researchers say that their findings broaden the debate as unjustified increases
for a few CEOs can lead to "legitimate" salary increases for others. They write that 'the linkages among firms produced by the benchmarking process
guarantee that firm-level governance failure becomes a factor in the environment of other firms.'
After the Enron and WorldCom scandals inf the early 2000s, the Securities and Exchange Commission changed its rules to require corporations to
disclose benchmarking information. The research team state that they are 'now using the new data mandated by the SEC to better understand how the network structure
of peer groups affects executive pay setting in American corporations.'
"Whether the SEC regulatory change reduces the ratcheting effect of leapfrogging on CEO pay - creating more transparency about who is in the
peer group and at what level the company is benchmarking - is an important question for future research" says
Powerful CEOs pay employees more
Research on perceptions of
fairness in executive pay and how CEO over- or underpayment cascades down to
lower organizational levels was published in the September/October 2006 issue of Organization Science. The study looked at data from over
120 firms over a five-year period. Authors James Wade, professor at Rutgers University, Charles O'Reilly, professor at Stanford's Graduate School
of Business, and Timothy Pollock at Pennsylvania State University's Smeal College of Business, used techniques of operations research (OR) in their analysis.
Key findings include:
- CEO overpayment has higher costs than previously realized. It has been argued that even if a CEO is overpaid, a large company can easily absorb the cost. However, the study found that CEO pay has direct consequences for compensation at lower employee levels. The effects of CEO overpayment cascade down to subordinates at diminishing degrees. For example, where one CEO was overpaid by 64 per cent, individuals at Level 2 (chief operating officer, chief finance officer, etc.) were overpaid by 26 per cent, while individuals at Level 5 (division general managers) were overpaid by 12 per cent. The cumulative effect of this systemic overpayment impacts on overall organizational performance and shareholder value.
Charles O'Reilly commented:
"Given the large sums paid to some senior executives, the total cost for
overpayment could be a big number - and, in some cases, significantly affect shareholder returns."
- CEO pay impacts subordinate turnover. The study found that CEOs serve as a key reference point for employees in determining whether their own pay is fair. If the CEO is overpaid, subordinates are more likely to leave. The turnover effect becomes more pronounced the farther away you get from CEO level. Even if an employee is overpaid relative to the market, they will have a greater propensity to leave if their CEO is overpaid by a larger percentage than they are.
James Wade said:
"CEO compensation impacts employee retention more than we realized. Our research found that CEO overpayment is related to turnover, which can have important long-term consequences. It is quite possible that those most likely to leave because of perceived unfairness are precisely those employees coming up in the organization that would eventually rise to the top management team level."
- CEO underpayment also cascades. The study found that CEO underpayment tends to get cascaded through an organization, with multiplying effects. If the CEO is underpaid more than you are, you are less likely to leave, but if the CEO is underpaid less than you are, you are more likely to leave. As James Wade put it, "underpaying a CEO could reduce turnover if subordinates are underpaid less than the CEO is underpaid."
- Notions of fairness are powerful. The study found that CEOs tend to be concerned with the perception of fairness. If the CEO is paid generously, they will typically use their influence to pay others generously as well. And, if they are seen as being underpaid, that will also have an effect.
Timothy Pollock commented:
"Our research shows evidence that CEOs are concerned with fairness, and that they are likelier to share rewards than they are to share burdens. But this can be expensive and has the potential to hurt a firm's bench strength if the rewards aren't fully shared."
- Powerful CEOs pay employees more. CEOs who also serve as chairman of the board tend to pay their employees more. This effect is more pronounced at higher levels, but diminishes at lowers levels. The effect disappeared at Level 5 (division general managers), but was strong at Levels 2 - 4 (the top management team through the junior vice president ranks).