Economist Mark Perry writes in Carpe Diem When we consider all US CEOs and all US workers, the ‘CEO-to-worker pay ratio’ falls from 331:1 to below 4:1


The AFL-CIO is comparing: a) the average salary of a small sample (350) of the highest paid US CEOs, out of a total CEO population in 2013 of 248,760 CEOs, according to BLS data here, and b) the average worker pay for production and nonsupervisory workers, which represents only 8.5 million factory workers out of a total of 136.3 million payroll employees nationwide. In other words, the AFL-CIO’s reported “CEO-to-worker pay ratio” of 331:1 is calculated by ignoring 99.9% of all US CEOs and 93.8% of all US workers. A more accurate description would be to call it a ratio of the pay for 350 of the highest-paid US CEOs to the pay of only 6.2% of the American labor force, or a ratio of an unrepresentative, infinitesimally small, and statistically insignificant group of CEOs to a small minority and unrepresentative group of US factory workers. It’s a completely bogus and meaningless comparison.

The top chart above shows a more statistically valid comparison of CEO pay to average worker in the US pay by considering: a) the average annual pay of all US CEOs in every year from 2002 to 2013 (data here) and b) the average annual pay of all US workers in a comprehensive, national BLS dataset that includes workers in 22 major occupational groups, 94 minor occupational groups, 458 broad occupations, and 821 detailed occupations (132.6 million workers for 2013). Based on those data, the average CEO earned $178,400 last year, the average worker earned $46,440, and the “CEO-to-worker pay ratio” was 3.84:1, and that’s a LOT different from the AFL-CIO’s ratio of 331:1 by a factor of more than 86 times! Call it a “statistical falsehood-to-truth ratio” of 86:1 for the AFL-CIO’s exaggerated, bogus ratio. The chart also shows that the real CEO-to-worker pay ratio has not been increasing as is frequently reported, but instead has been remarkably constant over the last 12 years, averaging 3.8:1 in a tight range between a maximum of 3.89:1 in 2004 and a minimum of 3.69:1 in both 2005 and 2006. The ratio of 3.84:1 in the most recent year (2013) was actually slightly lower than the ratios in 2004 (3.89:1) and in all years between 2009 and 2012.

Likewise, the bottom chart displays a more statistically valid comparison of average CEO pay to the annual pay of a full-time minimum wage worker. In 2013, a full-time minimum wage worker earned $14,500, and therefore the CEO-to-minimum-wage-worker pay ratio was only 12.3:1 compared to the grossly inflated 774:1 ratio reported by the AFL-CIO. That’s a “statistical falsehood-to-truth ratio” of 63:1 for the AFL-CIO’s exaggerated ratio. Because of the recent increases in the minimum wage between 2007-2009, the CEO-to-minimum-wage-worker pay ratio in recent years has been lower than the most recent 12-year average of 12.76:1.


They cherry pick data to compare a subgroup of the wealthiest CEOs with a subgroup of the lowest paid production workers.  This is statistical malfeasance of such a high order that any student of statistics can see right through it.  Is this a matter of incompetence or fraud?

This gross misuse of such data is becoming both typical and dangerous.  Is the media too ignorant of basic statistics to see such obvious flaws or are they so monolithically indoctrinated that they refuse to consider that the complete information may shake their poorly founded beliefs about their reality?

A simple Google search will reveal that there are far more articles available that reinforce this falsehood than there are to correct or clarify it.  This is also true in the discussion of inequality,  AGW, and other social issues that require clear information to make correct policies.

“A lie gets halfway around the world before the truth has a chance to get its pants on.” Winston Churchill