Some new shortcomings of the AFL-CIO’s CEO-to-worker pay ratio

From Mark Perry.

"The AFL-CIO released its annual Executive Paywatch Report this week, based on CEO and worker pay in 2018. Here’s the press release, here’s AFL-CIO Secretary-Treasurer Liz Shuler remarks, and here’s more data. Like in the past, this year’s AFL-CIO’s report is based on some apples-to-oranges comparisons and additional statistical chicanery including:

1. The AFL-CIO reports an increase in the average U.S. rank-and-file worker’s pay over the past 10 years of slightly less than $8,000 or less than $800 per year. It also reports that the average CEO received a $500,000 raise last year compared to the average rank-and-file worker who received barely more than a $1,000 raise, bringing pay in 2018 to $39,888. 

Here’s the math behind those claims:

In 2008, the “Average Hourly Earnings of Production and Nonsupervisory Employees” in the private sector was $18.07 and the average work week for those workers was 34.3 resulting in annual pay of about $32,000 working 52 weeks per year (AFL-CIO’s assumption). In 2017, those workers earned $22.05 per hour and worked 33.7 hours per week for annual income of $38,640. Last year, those workers earned an average of $22.70 per hour, worker 33.7 hours on average and earned about $39,800 for the year, for an increase of about $1,140. Based on the approximate $8,000 difference in average annual income between 2008 and 2018 for rank-and-file workers, that translates to an average annual increase of less than $800. But it’s important to note that those are not full-time workers (35 hours per week or more), but part-time workers, or more accurately a mix of full-time and part-time workers whose average workweek is less than 35 hours per week.

To then compare the average annual pay increase of $800 (cash wages only) for mostly part-time workers to the average increase of $500,000 in total compensation for CEOs over the last decade is a misleading apples-to-oranges comparison. Deceptively, the AFL-CIO sometimes refers to “CEO pay” and other times to “CEO compensation” when comparing those figures to “average U.S. worker’s pay” that it says “highlights the continuing pay inequity between workers and CEOs.” But it’s statistical sleight-of-hand to compare the cash-only wages of mostly part-time workers to the total compensation of CEOs working 50-60 hour weeks, and yet the AFL-CIO makes this oranges-to-apples comparison every year in its Executive Paywatch Reports. 

2. The AFL-CIO changed its methodology this year and is using the new ratios of CEO compensation to median employee compensation for the S&P 500 (and other publicly traded) companies that is now mandated by the Dodd-Frank Wall Street Reform and Consumer Protection Act. You can view the AFL-CIO’s database of pay ratios for the S&P 500 companies here.

For example, the McDonald’s CEO Steve Easterbrook made $15.876 million last year and the median compensation for a McDonald’s employee was $7,473, resulting in a CEO-to-worker pay ratio of 2,124-to-1. The new required pay ratios compare CEO compensation to the median compensation of employees, so that’s an improvement over the AFL-CIO’s previous comparisons of CEO compensation to the cash wages of employees (excluding fringe benefits), but there’s a new issue. S&P 500 companies like McDonald’s operate globally with workers around the world, and so McDonald’s (and other companies) compares the median compensation for global workers to the compensation of the company’s CEO.

According to McDonald’s most recent proxy statement, it considered all full-time, part-time, seasonal and temporary workers employed on October 1, 2017 in restaurants across the globe and reported that “Our median employee for 2018 (a part time restaurant crew employee located in Hungary) had 2018 total compensation of $7,473.” Similarly, Gap, Inc. has a whopping pay ratio of 3,566-to-1 because its median employee compensation is only $5,831 because it employs a lot of part-time workers globally at its stores in low-wage countries like Mexico, Pakistan, Russia, Philippines, Thailand, India, Ukraine, Peru, etc.

So while the new pay ratios overcome the flaw of comparing employee cash wages to total CEO compensation  the new ratios include the compensation of many workers outside of the US, which will significantly inflate pay ratios by including the compensation of workers in low wage countries around the world working for US-based S&P 500 companies.

3. But here is perhaps the most significant statistical legerdemain employed by the AFL-CIO this year. It takes the reported pay ratios of the S&P 500 companies (available here), which range from a high of 3,566-to-1 for The Gap to a low of 6-to-1 for Copart, Inc., and then calculates the average CEO-to-worker pay ratio of 287-to-1 for 2018 (see chart above). Note that the new 2018 ratio is much lower than the 2017 CEO-to-worker pay ratio of 361-to-1 reported by the AFL-CIO using its old methodology, a decrease of 20.5% in just one year. Overall, the pay ratio decreased significantly because it’s more accurate to: a) do a company-by-company comparison of CEO compensation to the compensation of workers at the same company and b) compare CEO compensation to employee compensation. But that decrease didn’t stop the AFL-CIO’s secretary-treasurer from reporting that:

Not surprisingly, the CEO-to-worker pay ratio remains high: 287 to 1. I’ll repeat that: 287 to 1. Meaning the average CEO earns 287 times what an average employee earns.

Conveniently, no mention of the fact that the pay ratio was 20.5% lower than the previous year.

Most importantly though, it’s always more accurate to use the median than the mean to find a “typical” value in a distribution, especially when  there are outliers on the high end like in this case. Therefore, it would be more accurate to calculate the median pay ratio for the S&P 500 companies instead of the mean pay ratio. And when you do that, the CEO-to-worker pay ratio for 2018 falls to 169-to-1 (see chart above).

Bottom Line: It’s an improvement in the AFL-CIO’s methodology to use the new required company pay ratios based on median employee compensation to calculate its CEO-to-worker pay ratio, but the union federation is still reporting an inflated statistic that now includes global workers and is also based on the flawed use of mean instead of median to calculate the typical pay ratio for the typical worker. For S&P 500 companies last year, the median (typical) CEO-to-worker compensation ratio was only 169-to-1, not the 287-to-1 ratio reported by the AFL-CIO. And as I posted previously:

Q: The average CEO of an S&P 500 company received $14.5 million last year, and as a group those 500 CEOs received about $7.25 billion in total compensation in 2018. If the AFL-CIO could wave a magic wand and confiscate that entire amount and redistribute $7.25 billion to the current 104.3 million rank-and-file workers, how much would each one get?

A: An annual increase in pay of $69.50 for each rank-and-file worker before taxes, or about $1.34 more per week and about only 3 cents per hour. In other words, complete confiscation and redistribution of S&P 500 CEO compensation would make almost no difference for the average rank-and-file worker."