Working Paper

The Mismeasure of Mammon: Uses and Abuses of Executive Pay Data

Download paper

Report to the Institute for New Economic Thinking on the statistical measurement and policy implications of the compensation of the highest- paid U.S. corporate executives

On April 7, 2016, the Wall Street Journal ran an article headlined “CEO pay shrank most since financial crisis,” while on May 27, 2016, a similar New York Times story declared “Top CEO pay fell – yes, fell – in 2015.” Unfortunately, both the Journal and the Times mismeasured the actual take-home pay of each and every one of these CEOs in 2014 and 2015. The reason for this mismeasure is that both articles relied on “fair value” estimates of the stock-based pay of these CEOs as reported in the Summary Compensation Table of the definitive proxy statement (Form DEF 14A) that each publicly listed company files annually with the U.S. Securities and Exchange Commission (SEC). Yet the very same proxy statements also report the actual realized gains of these CEOs in the Option Exercises and Stock Vested Table. It is the realized gains on stock-based pay, not fair-value estimates, that enter into the total compensation that a CEO actually takes home and reports as income in his or her income-tax return.

Moreover, including actual realized gains instead of estimated fair value of stock– based pay in the measure of total executive compensation can make a big difference. In 2014 average total compensation of the 500 highest-paid executives named on corporate proxy statements based on actual realized gains was $34.3 million, with 81 percent coming from stock-based pay. But average total compensation of the 500 highest paid based on estimated fair value was $19.3 million, with 62 percent attributable to stock- based pay. The excess of total actual realized-gains compensation over total estimated fair-value compensation was greatest in those years when the stock market was booming.

Why would the Wall Street Journal and the New York Times report estimates of executive pay when they could be reporting the CEOs’ actual pay? In this paper, we answer this question by explaining the origins of the “fair value” estimates of stock-based pay and how the obsession with these estimates by the SEC, relying on the business-run Financial Accounting Standards Board (FASB), has relegated to statistical obscurity executives’ readily available, accurate, and actual realized gains from stock-based pay. We use Standard & Poor’s ExecuComp database to document that a) stock-based pay, in the forms of realized gains from stock options and stock awards, dominates both the size of and the changes over time in the total compensation of the highest-paid senior executives; and b) the fair-value estimates of stock-based pay tend to understate, often substantially, the realized gains from stock-based pay that these executives actually receive.

An irony is that even critics of excessive executive pay, most notably the AFL-CIO on its Executive Paywatch website, use the fair-value estimates when the actual CEO compensation numbers would reveal a much larger ratio of CEO pay to the earnings of the average worker. Indeed, as we discuss in the conclusion to this paper, as mandated by the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010, this mismeasure of executive pay has become institutionalized in U.S. government policy in the SEC’s Pay Ratio Disclosure Rule, which beginning in 2017 requires every company to publish the ratio of CEO to median-worker pay. Under this rule, the SEC requires companies to use the fair-value measure of CEO pay. The Pay Ratio Disclosure Rule is supposed to provide the public with a company-level indicator of income inequality. Instead it will tend to underestimate inequality, substituting fictitious estimates for actual known amounts of income that CEOs put into their bank accounts and declare in their income-tax returns.