The EPI's numbers are wrong/misleading for a few reasons.
- The productivity and compensation graphs are often inflation-adjusted using incomparable deflators. Productivity is measured using a GDP inflator (which includes investments) whereas compensation is measured using a CPI deflator (which only includes consumer products). CPI is historically higher[1] which means compensation is artificially being deflated more than productivity.
- The most well-known graph from the EPI[2] shows productivity for all workers but shows compensation for only 80% of workers, with the provided reason being to exclude high earners like management and executives. Which, sure I guess, but surely the two metrics should be kept consistent.
If you correct for these errors you end up finding a productivity/compensation gap that is much smaller than the EPI claims, though notably not zero. The EPI themselves have a figure that includes all workers and uses matching deflators ([3], figure C, "Real producer average hourly consumption") so perhaps some of the blame here is on readers.
I think this BLS paper on the subject[4] is well worth a read. An interesting result they found is that the industry with the greatest productivity/compensation gap is...computing. Probably not the jobs people usually imagine when they think of stagnating pay.
- The productivity and compensation graphs are often inflation-adjusted using incomparable deflators. Productivity is measured using a GDP inflator (which includes investments) whereas compensation is measured using a CPI deflator (which only includes consumer products). CPI is historically higher[1] which means compensation is artificially being deflated more than productivity.
- The most well-known graph from the EPI[2] shows productivity for all workers but shows compensation for only 80% of workers, with the provided reason being to exclude high earners like management and executives. Which, sure I guess, but surely the two metrics should be kept consistent.
If you correct for these errors you end up finding a productivity/compensation gap that is much smaller than the EPI claims, though notably not zero. The EPI themselves have a figure that includes all workers and uses matching deflators ([3], figure C, "Real producer average hourly consumption") so perhaps some of the blame here is on readers.
I think this BLS paper on the subject[4] is well worth a read. An interesting result they found is that the industry with the greatest productivity/compensation gap is...computing. Probably not the jobs people usually imagine when they think of stagnating pay.
1. https://www.bls.gov/opub/mlr/2016/article/comparing-the-cpi-...
2. https://www.epi.org/productivity-pay-gap/
3. https://www.epi.org/publication/understanding-the-historic-d...
4. https://www.bls.gov/opub/btn/volume-6/pdf/understanding-the-...