I think that broad aggregates of wages and productivity may not be the most informative way of looking at these trends. Instead, I want to give a perspective based on Robert Fogel's view, which is that the share of consumption going toward food and durable goods has been trending down while the share of consumption going toward health care, education, and leisure is going up.
The aggregate earnings of males seem to have fallen because there are fewer males working full time. This may be a substitution effect. As the cost of food and durable goods falls, what are you going to do? You only consume more health care services if you think you are sick and that the doctor can do something for you. So you either consume more education or more leisure. The way I look at it, a lot of males are choosing more leisure. Maybe they would prefer to work 2000 hours a year if they had a really high after-tax wage. But in any case, they can get by on more sporadic employment.
Suppose we were living in the alternate universe where these males have technical training in computers, finance, or some other highly marketable skills. Would total employment and output be much higher? Instead, perhaps the distribution of leisure would be different. It would be less concentrated among high-school educated males and more widespread among females and highly-educated males.
In the universe as it exists, those with the right skills can afford to live expensive lifestyles. What do these lifestyles consist of? For many, the expensive lifestyle consists of sending their children to private schools and high-priced colleges. It consists of generous contributions by their employers to health insurance. Maybe it consists of some expensive goods and services, such as gourmet meals with fine wine at fancy restaurants.
If you do not crave such things, maybe you can get by working on and off, rather than in a steady job.
One point about the Great Substitution of leisure for work is that leisure escapes taxation. That raises all sorts of issues.