One issue that sometimes comes up in inequality debates is the idea that cash wages are a misleading indicator because the median family also enjoys employer-provided health benefits. Lane Kenworthy notes that we can address this by looking at CBO data on income, which does include the value of benefits. The black line shows us the "old" naive comparison and illustrates that GDP per capita started diverging from median household income.
The red lines give us the "new" more sophisticated comparison. They don't allow us to look at median household income, but they do let us look at the household income of the third quartile, which is similar. And they show the same basic pattern. Average (i.e., mean) household income has closely tracked GDP per capita except with some added volatility that by eyeball is due to capital gains. And household income in the middle quartile has fallen far behind.
That said, while painting a very similar picture in terms of the basic divergence story I do think it's worth noting that the red and black lines tell a very different story about the past decade. According to the naive view, not only did we just have a very nasty recession but the aughts as a whole were a time of falling living standards. When you put the red line into it, the aughts look much better. They simply turn out to have been a decade in which approximately 100 percent of the value of the middle quartile's increased compensation went to in-kind health care services. That may or may not have been a good social allocation of resources, but it represents real spending by employers on employee compensation, which presumably isn't something firms were doing just to be nice. Workers were doing things of value to firms and were in turn extracting money for their efforts. The really atrocious time for median compensation came way back in the '80s.