For the economy as a whole, labor’s share of national income has stayed roughly constant at just above 70 percent. What has changed is that highly skilled laborers earn more labor income than low-skilled workers.
Thomas McCraw's new biography of Schumpeter, which Tyler rated as a must-read and I rated as just shy of that, points out that Schumpeter thought that it was wrong to view labor as a homogeneous class. Instead, the differences in returns to skilled and unskilled labor are a major source of inequality.
Some of Tyler's main points (read the whole column):
Income distribution thus depends on the balance between technological progress and access to college and postgraduate study. The problem isn’t so much capitalism as it is that American lower education does not prepare enough people to receive gains from American higher education.
...Technology is advancing faster than our ability to educate. So even if inequality declines today, it may well intensify in the future. Even if American education improves at every level, the largely not-for-profit educational sector may simply be less dynamic than the progress of new technologies.
The lesson is this: Economists are homing in on the key to the inequality problem, but don’t think any solution will necessarily last for long.
I am going to quibble with the "homing in" statement.
Until we can figure out the cause of the differential earnings of college graduates, we are not homing in. And the story is far from clear.
I would say that the least promising source of the differential is valued added from college education. Of course, if I am wrong about that, then policies to boost enrollment in and completion of college would be attractive.
Another view, much debated on this blog, is that college works as a signal. In my opinion, the story of college as a signal of IQ is weak, because cheaper alternative signals are available. My co-blogger has a better case when he views college as a signal of conformity. As a former middle manager, I know what it's like to deal with a bright employee who cannot internalize deadlines or organizational priorities. If bright people who do not go to college are so independent-minded that they are high-maintenance employees from a management perspective, then, sure, I'd rather hire the person with the sheepskin.
Finally, there is my view that college is a tribal ritual for the affluent. People who strongly desire to belong to the tribe of the affluent (or, perhaps more accurately, children whose parents want them to belong to that tribe) will (a) tend to go to college and (b) tend to be affluent. When we observe (a) and (b) together, we are seeing correlation, not causality.
What the tribal-ritual story suggests is that people who do not go to college are just not as motivated to become affluent as those who do not. In that case, policies to push more people through college are a strange form of paternalism--changing people's behavior because you think they are not putting a sufficiently high value on income and status.