What Deindustrialization?

Many critics of free trade argue that globalization has led to the US becoming deindustrialized; That is, trade is eroding our manufacturing base.  Economists refute this claim by pointing to the fact that US industrial production is near the record highs set in 2018 or that the manufacturing component, while off the highs set in 2008, is still at a very high level of production.  Indeed, just last quarter, the US manufacturing sector produced $7.3 trillion worth of goods.  This is hardly the picture of a manufacturing sector that has been gutted by international trade.  

“But wait!” the clever protectionist claims. “We need to consider the counterfactual.  Think of how much higher manufacturing would be if it weren’t for globalization!”

This objection is reasonable.  Counterfactuals are always difficult to consider.  By definition, the counterfactual does not exist, so we can never empirically show what the “proper” counterfactual is.  Theory helps guide us, but we can also look at other evidence to suggest what the counterfactual is.  If we were experiencing a declining manufacturing base, it should show up in employment numbers.  After all, factories shouldn’t be hiring; they’d be doing replacement hiring, sure, but that’s about it.  Job openings should be fairly low compared to historical trends, layoffs/discharges fairly high.  

Taking a look at employment numbers, we see data inconsistent with the “deindustrialization” argument.  Job openings in manufacturing in August 2024 (latest data as of this writing) were 505,000.  There are half a million job openings in the US right now for manufacturing jobs.  That is down from the post-pandemic rehiring jump, where openings hit 997,000, but well above the pre-pandemic average of 293,000.  Manufacturers in the US need workers and the demand is generally high.  In a deindustralizing economy, one would not expect to see increasing demand for manufacturing workers.

The trend in job openings is interesting too.  Other than two declines from the 2001 and 2008 recessions, the trend in job openings is generally rising.  The only non-recession exception is in 2018 when the Trump trade war started.  Odd that…if free trade was deindustrializing and tariffs were industrializing, one would not expect to see job openings fall as tariffs go into effect.

Likewise, layoffs are very low.  Indeed, since 2001, firm layoffs have been generally steady at a very low level.  We do not see any mass layoffs (recessions excepted).  Indeed, during the “China Shock,” the number of people laid off fell, not rose.  If the decline in manufacturing during this time was due to China, we should have expected to see layoffs increase.  Indeed, a falling number of layoffs suggests that the decline in manufacturing at the time was likely due heavily to attrition (people quitting/retiring and not being replaced).  

One final note: wages for manufacturing (production and nonsupervisory) workers have generally been rising faster than inflation, suggesting real wages have been rising.  Again, if the demand for manufacturing workers was falling due to deindustrialization, we should see wages fall, not rise.  

Putting these employment figures together, we can start to see a counterfactual emerge.  US manufacturing production has hit a ceiling, yes.  But it is not due to trade.  It appears more due to the fact that firms cannot hire!  They want employees, they need employees, they are willing to pay for employees, but they cannot get them (for whatever reason).  The data do not show a deindustrializing country.  It shows an economy still industrialized but hitting some constraints.  Rather than burdening US manufacturing with more constraints via tariffs, “Buy American,” and other restrictions, policymakers should explore why manufacturing jobs are hard to fill.  

In short, protectionism will not industrialize a deindustrializing base.  It’ll deindustrialize an industrializing base.  And all because they have the wrong counterfactual.