To be precise, for "production" and "nonsupervisory" workers. Doesn't say anything about the size of this population, and how representative it is of the general earnings potential of the US as a whole.
This pool of workers includes those in manufacturing and construction jobs, as well as all “nonsupervisory” workers in service industries such health care or fast food. The group accounts for about four-fifths of the privately employed workers in America, according to BLS.
Production and non-supervisory is most people
It's probably less "most people" than it was in 1973, though.
Why would that be the case? Has there been a significant spike in farm workers or managers per capita?
Well, I was assuming that people like engineers and programmers weren't "production". There are more of them now than there were in 1973, and fewer factory workers.
Yeah, for service industries, it counts anyone not in a supervisory role, so it'd include them as well. :)
Maybe, but I can't tell by looking. It appears to be roughly the same to me. But my subjective assessment is meaningless. We'd need data.