Hillary Clinton has decided to make income inequality, especially the compensation of CEOs, a major focal point of her campaign. Last week she announced her campaign for President and has criticized the pay gap between higher executives and average workers.
Striking a populist note, Clinton, who announced on Sunday she was running for president in 2016, said American families were still facing financial hardship at a time “when the average CEO makes about 300 times what the average worker makes.”
This statistic of “300 times compensation” has been around for awhile. A quick Google search links to an ABC News article citing a fall 2009 study.
A study last fall by the Institute for Policy Studies, a liberal Washington D.C. research group, found that CEOs in the country’s S&P 500 companies make, on average, 319 times more than the average American worker.
Notice that the study analyzed only the S&P 500 companies. This is only a small subset of many of the largest companies in the United States. There are literally millions of companies not counted because they are privately owned, and so not reporting figures in the stock market, and thousands more not listed because they don’t fit the Standard & Poor’s unique and fairly arbitrary criteria to be on the 500 list.
The problem with the “300 times average worker” statistic is that it’s poorly defined and highly misleading. Who is the “average” worker? All American’s employed? The average employee at that company? The media articles and pundits won’t tell you. If Ms. Clinton is going to make a push for income equality, shouldn’t she have some idea of what these statistics actually mean?
While this author couldn’t find the original IPS analysis, a 2013 study by the Economic Policy Institute also purports a 300-ish-to-1 ratio for CEO pay to average workers, and in this case they identify “Private-sector production/nonsupervisory workers” in their chart.
But here’s the problem with the statistics they’re quoting:
Average CEO compensation was $15.2 million in 2013, using a comprehensive measure of CEO pay that covers CEOs of the top 350 U.S. firms and includes the value of stock options exercised in a given year, up 2.8 percent since 2012 and 21.7 percent since 2010.1
They gave away the goat right in the first bullet point describing their data: They’re comparing the 350 biggest-name CEOs (sans Mark Zuckerberg, who apparently earns more money than God) to the average of all American workers in the industries covered by these 350 firms, to get a 295.9-to-1 ratio of CEO compensation. That’s like comparing the average size of the 350 largest trees in Sequoia National Forest to the average size of every plant on in California. Of course the ratio will look huge, because the sample set is a statistical outlier.
But wait: Wasn’t it made clear earlier that there are millions of private sector companies in the US? And thousands of them trade on the stock market? Why didn’t they analyze all private sector CEOs? In fact, it would be very simple to do this, because the Bureau of Labor Statistics tracks the data. So why didn’t they?
There’s a very simple answer to that: The “statistical analysis” wouldn’t back up the political talking point.
The May 2014 BLS statistics linked above shows the average CEO earned $180,700 in 2013, less than four times what the average American worker earns, at $47,230. The average For-Profit Private Sector CEO earned a bit more than that. The top industry to work in is apparently “Pipeline Transportation of Natural Gas,” where the average of 40 CEOs is $250,030.
Let’s go with $250,030, since that’s the top industry. Even at this level, the average CEO only earns about 5.3 times what the average worker earns. That’s about 60 times less than the media-quoted average, and less than the oft-touted ratio reported for Japanese companies of 10-to-1. One might ask Ms. Clinton if she’s familiar with the BLS data if one were intellectually curious. That would, of course, require the news media to be intellectually curious.
Let’s go a step deeper: Payscale has compiled a list of the top 100 publicly traded companies in 2013, their CEO non-stock compensation (that is, their cash salary, which is a bit different from EPI the study above), and their average employee’s compensation. On this list, only two companies had CEOs earning more than 300 times their employees: CVS Caremark (422-to-1) and Goodyear Tire & Rubber (323-to-1). In fact, the average of these 100 companies, compared to their own employees, was 85.6-to-1. And that’s leaving out the CEOs of Sears and Google, who earn $1 a year. In fact, one need only work a quarter of the way down the list to reach the 100-to-1 ratio.
So what does all of this mean? It means we as Americans have to be careful to understand the statistical data we’re viewing and on making policy decisions based on it. The oft-quoted 300-to-1 ratio has become all but gospel in the media. Few pundits, prognosticators, or politicians even dare to question it, let alone critique it.
Again, it’s not that the statistics are bad: The top 350 CEOs actually do earn total compensation of 300 times the average American worker’s salary. The idea that these CEOs represent the “average” of American Chief Executives, or that by comparison to other countries, the US is a tremendous statistical outlier, is bald-faced lie.
Ms. Clinton famously earned $200,000 to $300,000 per speaking engagement before she declared her campaign for the Presidency. That means in one event, she earned more than the average CEO did in 2013. One wonders what she thinks about that pay ratio?
1. Stock option compensation is often a large cash-dollar number. 100,000 shares at $35 per share is $3.5 million, which is a lot of money. What usually isn’t quoted in these compensation figures is the cost of the option: That $35 stock option has to be bought first. That price may be $10 per share, or it may be $30 per share. At $10 per share option price, the option is a $2.5 million pre-tax paycheck, less any fees to the bank and/or broker. At $30 per share, it’s only worth $500,000.