For a household size of 2.6, we wind up with 3,846 households. If there's 3,846 households with a median income of $100, that makes our total economy $384,600.
Now let's assume enough of those families divorce such that there are now only 2.5 people per household. We wind up with 4,000 households. We still have the same total economy of $384,600 but split across more households (4000) that results in a median household income of $96. That's a 4% drop in median income.
Now of course in real life the change in household size isn't just people getting divorced (or never marrying). It's also reduce birth rate. But I hope you can see how a small change in single people makes a relatively big change in median income.
Real per capita GDP rose by around 25% over this period. What accounts for the other 21%?
I'm not making any claims that the gains have been perfectly proportional across all segments of the population. Only that the average worker is really better off today than they were before, and that the median wage charts don't give us very helpful information to gauge how much better/worse off they are.
These new additions to the labor pool don't just add another worker/job at the lower end affecting the median chart, they provide downward pressure on everyone's wages because they compete for those jobs. These workers are productive, so GDP goes up, but there is downward pressure on all wages. That is a big part of the story of the increasing wage/wealth disparity between the rich and everyone else. It's not the only factor, but it's a significant one.
I don't understand why you bring up women entering the workforce again. As I mentioned, that's wrong twice: the proportion of women in the workforce is flat over this timeframe, and additional women in the workforce would push median household income up, not down.
So we're left with immigrants. Thus my original question way up there: is it all down to underpaid immigrants, then?
What is the point of mentioning that decreasing household size could make the number misleading, when household size didn’t decrease by a significant amount? What is the point of repeatedly bringing up workforce participation by women?
It’s fine to say that a measure might be misleading because of other factors. But when those factors don’t actually make it misleading for the example being discussed, I don’t really get the point.
It’s also weird that you focus entirely on the possibility that everyone is better off. Isn’t it also possible that everyone is worse off while the median household income remains flat? Smells of ideological bias.