I have to say this article is a bit light on facts. For the curious, here is the actual press release from the Bureau of Labor Stats: https://www.bls.gov/news.

I have to say this article is a bit light on facts. For the curious, here is the actual press release from the Bureau of Labor Stats: https://www.bls.gov/news.release/prod2.nr0.htm https://www.bls.gov/news.release/prod2.nr0.htm On the face of it, this isn’t because of a decrease in hours worked because the productivity rate uses that in the denominator. Plus, hours worked actually increased. This is of course assuming that hours worked is accurate. If this number is actually higher than expected (for example, if workers are spending time commuting rather than working) then it will incorrectly underestimate productivity rates. I looked into how this is calculated: https://www.bls.gov/productivity/technical-notes/labor-productivity-hours-worked-method-using-ces-all-employee-hours-nov-2022.htm https://www.bls.gov/productivity/technical-notes/labor-produ... Importantly, see footnote 6 about salaried employees: > In the nonproduction-worker-to-production-worker hours ratio used in the current methodology, both the numerator and denominator are actual hours worked. However, to be conceptually consistent with the CES data, the denominator should be hours paid, which excludes off-the-clock hours worked by salaried production employees. That is, this ratio implicitly assumes that production workers are paid for all of the hours they work. However, according to the CPS, about 30 percent of production workers are paid a salary. Thus, the nonproduction-worker-to-production-worker hours ratio is biased downward and therefore does not capture all of the OTC hours worked by salaried nonproduction workers. Here, the BLS is explicitly saying they can’t capture all of the hours you work. Intuitively this is true. How could they? This is just my hypothesis. It would be great if there was some survey data to back it up regarding hours actually worked by salaried employees.