This crisis began decades ago when a new wave of technology — things like satellite communications, container ships, computers and eventually the Internet — made it cheaper for American employers to use low-wage labor abroad or labor-replacing software here at home than to continue paying the typical worker a middle-class wage. Even though the American economy kept growing, hourly wages flattened. The median male worker earns less today, adjusted for inflation, than he did 30 years ago.

But for years American families kept spending as if their incomes were keeping pace with overall economic growth. And their spending fueled continued growth. How did families manage this trick? First, women streamed into the paid work force. By the late 1990s, more than 60 percent of mothers with young children worked outside the home (in 1966, only 24 percent did).

Second, everyone put in more hours. What families didn’t receive in wage increases they made up for in work increases. By the mid-2000s, the typical male worker was putting in roughly 100 hours more each year than two decades before, and the typical female worker about 200 hours more.

When American families couldn’t squeeze any more income out of these two coping mechanisms, they embarked on a third: going ever deeper into debt. This seemed painless — as long as home prices were soaring. From 2002 to 2007, American households extracted $2.3 trillion from their homes.

Eventually, of course, the debt bubble burst — and with it, the last coping mechanism. Now we’re left to deal with the underlying problem that we’ve avoided for decades. Even if nearly everyone was employed, the vast middle class still wouldn’t have enough money to buy what the economy is capable of producing.

I don't know if Reich is right about the cause, but I do agree with his three-phase analysis of why the problem has been submerged. You don't see this discussed so often--the enormous increase in GDP from two-income families.

Another step: workers who lose their jobs and have to settle for positions that pay less could qualify for “earnings insurance” that would pay half the salary difference for two years; such a program would probably prove less expensive than extended unemployment benefits.

This is also a very thoughtful idea. I think there is a pretty well-established conclusion that a strong side-effect of unemployment benefits is that people resist--right up until their benefits run out--making career changes that involve lower pay. Even if such a change is all-but-inevitable. This alternative approach might improve that "structural" problem.