Paul Krugman really doesn’t like the possibility that there is a structural shift in employment, because it weakens the argument for the massive Keynesian spending spree he’d like to see the government initiate.  To that end, he published this piece on his blog February 13th.

Before we go on, some readers may wonder what a structural shift is and why it weakens the argument for Keynesian spending.  A structural shift is when employment permanently shifts (Well, as much as anything is permanent in economics) from one economic sector to another, say from construction to healthcare.

The reason that a structural shift weakens the Keynesian’s argument is that moving workers from one sector to another takes time.  They may need retrained.  They may need to move to another location.  Think of our construction worker moving to health care.  He or she probably doesn’t have the skills to be immediately employable in health care.  Some sort of education or training has to happen first.

This poses a problem for Keynesian expansionists, because their argument is that the only problem is a drop in aggregate demand (consumer spending) brought about by….well, animal spirits.  Since there is no real problem, government can increase spending (it doesn’t matter what you spend the money on.  You could dig holes and fill them back up.), fool the consumer into thinking she is better off, and voilá, aggregate demand goes up with the government spending.

Problem solved.  It’s a beautiful thing.

However, spending can’t solve the problem of unemployment brought about by a structural shift.  It takes time to retrain the affected workers.  There are things government can do to speed the process, but spending willy-nilly is not one of them.

Hope that clears things up.  Let’s get back to Krugman’s piece.

He claims that unemployment in every sector has just about doubled since the recession began, and that this is proof that no structural shift is going on.  He has a nice chart to show the increase in unemployment by sector.

There is a problem though.  The Bureau of Labor Statistics—the same source that Krugman claims originated his data—reports that construction jobs fell by 2 million, or 26.7 percent, from December 2007 through December 2010, while education and healthcare jobs grew by1.2 million, or 6.5 percent.

This appears to contradict Krugman’s data, but it is possible that both sets of data are true.  If they are both true, then Krugman is being no less dishonest than if he created his numbers out of thin air.

If Krugman is telling the truth when he presents a graph showing that unemployment approximately doubled from 2007 to 2010 in both the construction and the education and healthcare sector, then is must be that large numbers of unemployed construction workers migrated to being unemployed education and healthcare workers.

There is no other possible explanation.

This, of course, completely contradicts Krugman’s argument.  If his data are true, he’s using data that confirms a structural shift to argue that there is no structural shift, by neglecting to disclose the jobs data I’ve disclosed above.

Krugman is not a dumb guy.  He has a well-deserved Nobel Prize for his work on international economics.  He has a career of looking at data, in depth and with insight.  His failure to provide the entire story has to be considered something besides an oversight.  We have to conclude that he’s purposely being deceitful.

I don’t know why a guy with all of Krugman’s gifts and accomplishments would use data deceitfully.  It is a shame, though, that an economist at the top of his profession and with the New York Times bullhorn uses that bullhorn to confuse instead of to enlighten.

This previously appeared at