Top of the News: The Institute for Supply Management's manufacturing index seems consistent with recovery, coming in at 53.6 for November; any number above 50 signifies expansion in the sector. Unfortunately, the reading sagged from 55.7 the month before, tripping up what economists had hoped would be a steady climb out of recession.

A deeper look shows that the index provides no relief for the biggest immediate problem facing Americans, unemployment. Only six of 18 manufacturing industries reported growth in employment. Only 11.7 million Americans worked manufacturing as of October. That compares with 17.3 million in October 1999.

Not only do manufacturing jobs pay better than their counterparts in service industries, they tend to add real value to economic activity (as opposed to selling mortgage swindles). They are also twined with our trade issues. Even fewer manufacturing jobs are now in industries that export, a key part of our huge manufacturing trade deficit.

Unfortunately, this phenomenon was happening even before the Great Recession. A report from the Economic Policy Institute shows that manufacturing employment between 1965 and 2000 never dipped below 16.5 million. This even as manufacturing shrank as a share of the economy (take out Boeing and it would be much smaller). This changed as imports surged after China joined the WTO and other Asian factory centers upped their game. By 2004, the number was lower than any time since 1950.

"It is often claimed that declines in manufacturing employment stem entirely from productivity growth," according to EPI economist L. Josh Bivens. "However, rapid productivity growth is the norm, not the exception, in manufacturing. What is new about the manufacturing job crisis of the last four years is the sharp downturn in the ratio of domestic production to demand."

Indeed, American steelmakers are shrinking yet again.

The Back Story: The official unemployment rate including discouraged workers and part-timers seeking full-time work is 17.5 percent. But Shadow Government Statistics, a provocative and reliable site, argues even this underestimates the problem. Try...22 percent.

Today's Econ Haiku:

Before the big crash
Stocks suffered a lost decade
So who's a bull now?