Research Findings

Should CEO pay rise with performance?


October 17, 2018

The median CEO pay in Standard and Poor’s 500 companies is about $10 million.  Many Americans think Chief Executive Officers (CEOs) are overpaid. Still, they underestimate how much CEOs are paid.

One recent study shows that many people in the U.S. support capping CEO pay at a maximum amount relative to the average worker. Not everyone thinks so, of course, but sizeable numbers of Democrats (66%), Republicans (52%) and Independents (64%) do.

While evidence of discontent with CEO pay continues to grow, we still know very little about why.

A challenge in the study of attitudes toward executive pay, as with the study of attitudes toward anything, is that people consider many factors in forming their attitudes and expressing their opinions. In the case of executive pay, these might include people’s perceptions of current pay levels, their perceptions about company performance, and their core values, among others.

My research examines one of the major determinants of attitudes toward pay: beliefs about whether and how rewards should rise with contributions.

A great body of work shows that people like to see rewards in the work setting distributed proportionally to relevant contributions, such as performance. This study tests two possible explanations of the widespread discontent toward executive pay: one possibility is that CEOs are seen as not contributing proportionally to their pay. Alternatively, such high pay itself may be objectionable.

If discontent is due to the perception that CEO pay is not proportionate to the their contribution, then we might see discontent diminished if people were convinced that CEOs contributed highly enough.  On the other hand, if the pay level itself is objectionable, then, even if people thought CEOs were highly contributing, they’d still be perceived as overpaid.

The experiment in this study examines whether fair pay for a fictitious CEO increases proportionally to the company’s changing performance.

Data for this study come from the Performance and Fair CEO Pay Study which is a survey experiment conducted in partnership with YouGov in 2013 with 989 respondents from across the United States. The survey measures respondents’ pay preferences for a fictitious CEO of a fictitious company, as the company’s performance goes from large losses to large profits. Each respondent is asked to react to seven levels of performance.

There’s one more piece to the experiment.

If people receive information that the average CEO is paid $5 million, we might expect them to gravitate toward this amount. This phenomenon called anchoring has been documented time and again in previous work.

Respondents are randomly assigned to four conditions: a control group where an industry average for CEO pay isn’t mentioned at all; a treatment condition where respondents are informed that the average CEO in the industry is paid $200,000; another where they are informed that the average CEO is paid $1 million, and one where the average CEO pay is $5 million.

How do participants’ perceptions of fair pay vary across the seven different levels of company performance and across each of these four reported levels of average CEO pay in the industry?

Respondents gravitate toward the industry averages mentioned in the experiment in deciding on the right amount of pay. However, they also make some adjustments upward or downward in relation to those anchors. When the fictitious CEO’s company is described as doing equally well as the average company in the industry, on average, respondents adjust the $200,000 suggested pay upward by 52.8%. Those given the $1 million average figure adjust it downward by 18.2%, and those given the $5 million average figure adjust it downward by 59.8%.

Respondents therefore also seem to hold their own beliefs about fair CEO pay.

How does fair pay vary with performance after those adjustments are made?

On average, fair pay rises linearly with performance. This finding indicates a strong commitment to the rule of proportionality of rewards and contributions, even at these high levels of pay.

However, responses also suggest that participants don’t think pay should rise as steeply with performance as it does in the real world. For example, for the group that is told that the average CEO is paid $5 million, median fair pay increases from $500,000 to $2 million as we go from large losses to large gains, never reaching the average pay suggested in the vignette.

Altogether the findings suggest a possibility to reconcile the widespread opposition to high CEO pay with the also widespread support for pay-for-performance. There might be a socially acceptable rate at which people, on average, believe pay should increase with performance. As one anonymous reviewer of my study nicely suggested, we might call this a social slope. This finding is especially of interest because most discussion of actual executive compensation is focused on how to tighten the connection between pay and performance.

The study also shows evidence that some predispositions, such as a belief in a free market ideology, have significant effects on these attitudes. Further examination of such predispositions is important, especially in understanding group differences in attitudes, and should be part of the future research agenda.



No Comments

Leave a Reply