From December of 2007 to December of 2009, the U.S. economy lost about 2.2 million manufacturing jobs. That’s over a quarter of the total job loss over that same period, about 8.5 million. Manufacturing is always hammered heavily during a recession. What made this time different is that manufacturing came in on its knees. It had never recovered from the 2001 recession.

Indeed, in early 2007, UCLA forecaster Ed Leamer’s computer models were screaming recession. However, Leamer chose to ignore them because recessions are typically about a loss in manufacturing jobs. What’s left to lose, he wondered? It turned out plenty.

Leamer had reason to be skeptical. Manufacturing jobs go up and down with the business cycle. Yet since 1999 they had only gone one way: down.

Here are U.S. manufacturing jobs through the ’80s and ’90s, right up until the eve of the Great Recession.

Here is the picture continued until today.

It’s understandable that Leamer found it hard to imagine how manufacturing jobs would further erode when they have never turned around from the collapse of 2000. And without a collapse in manufacturing, recessions are difficult to imagine.

Still, manufacturing did collapse. Yet, it’s hard not to look at the graph above and think that the real manufacturing recession began in 1999 and simply never stopped. What’s amazing is that we had any recovery at all.

Looking at the graph one might guess that we had been in a recession ever since the dot-com bubble burst.  The graph is just manufacturing. but manufacturing is where the job losses are concentrated during most downturns.

If we take a step back, then we see our current employment situation as in large part an attempt to cope with the loss of U.S. manufacturing employment. Had manufacturing returned to 1999 levels, there would be 6 million more jobs in the U.S. today. That doesn’t assume employment growth. We haven’t expected that in manufacturing in a long time. It only assumes that we, as we had in the years before 1999, returned to roughly the same number of jobs that we had before the recession began.

The lesson is that it seems unlikely this trend was caused solely by the financial crisis or housing collapse. The job loss that began in 1999 has continued at a greater or lesser pace ever since. More likely this is the result of globalization.

It’s anathema in many economics circles to speak ill of international trade. Indeed, I am not even willing to go that far. What I am saying is that the story of this recession is a part of the larger story of globalization and its effects on the U.S. labor market.

Karl Smith is an assistant professor of economics and government at the University of North Carolina School of Government and a blogger at