Some commenters are pushing back against my last post on manufacturing. U.S. manufacturing output is still climbing, they note.

That is certainly true, but it doesn’t change the nature of my point – there has been sharp erosion in manufacturing employment since 1999. This means that, at a minimum, these folks will have to find something else to do, and that puts a structural burden on the U.S. employment situation.

It’s also important because out-of-work manufacturing workers are, generally speaking, what a recession looks like. Thus, especially in manufacturing-heavy areas of the country, it has looked like a recession for more than a decade now.

I also fingered international trade as the culprit. The stock economics answer is that manufacturing employment is decreasing because productivity is increasing. That’s true. Yet, it can give poor intuition about what is happening.

However, mechanically outsourcing increases productivity. Suppose that I have a factory with 1,000 workers who produce $1 million worth of goods. Now, suppose I restructure my factory so that I outsource half the work to China or India. I keep my 500 most skilled employees and have them focus on the most value-intensive work.

My new reorganized factory produces $1.5 million in output, of which $500,000 comes from outsourced workers. The value added by my factory is still $1.5 million – $500,000 or $1 million. However, since I have 500 workers, average worker productivity has doubled.

If you were to simply read the statistics, you might say, well, trade contributed somewhat to those 500 job losses, but it looks like the real driver was an increase in worker productivity. Yet, it is trade that facilitated that increase in worker productivity. It was trade that allowed me to fire a bunch of my regular line staff and keep my engineers and highly skilled machine operators.

There is reason to believe this is what is happening. Check out “Offshoring Bias in U.S. Manufacturing.” It’s an economic journal piece but readable for anyone who’s comfortable with numbers.

Now, in the long run, there are reasons to think this is all for the best. However, that is no reason to pretend it doesn’t create enormous hardship in the present.

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