Paul Craig Roberts Misses the Mark

In an opinion piece published this week, Paul Craig Roberts takes exception to a conclusion in my recent Cato paper about the state of U.S. manufacturing.  I usually welcome disagreement as an opportunity to elaborate or persuade.  But it’s quite evident that Roberts is not interested in elaboration and is beyond persuasion.  The purpose of his dissent was to construct a straw man against which he could present his skeptical, and empirically refutable, views about trade.  

In my paper, Roberts identifies what he believes is an “extraordinary mistake [which] results in an incorrect conclusion.”  He argues that my failure to distinguish imports produced by U.S. companies abroad (offshored production) from imports produced by foreign firms abroad (import competition) leads me to the erroneous conclusion that “the health of U.S. manufacturing [is attributable] to import competition.”  

First of all, nowhere in my paper do I attribute the health of U.S. manufacturing to import competition.  The only passage from which such an interpretation might be drawn (by a careless reader, I would add) is this one: “Revenues, profits, output, value added, and even compensation rose the most for industries most exposed to import competition, and they rose the least for those industries experiencing the smallest increases in imports.”  That is just a statement of fact, as gleaned from the data.  It assigns no causation to import competition.   

I also write: “Exposure to trade, as evidenced by the relationship between imports and exports and operating performance, has been an important component of the success of U.S. manufacturing industries.”  This statement at least implies some degree of causation, which is supported by the fact that profit growth (operating performance) is a function of revenue growth (expanding exports) and cost reduction (increasing imports of production inputs). 

Second, my failure to distinguish between sources of imports in no way undermines the central points of my paper.  The purpose of my paper (“Thriving in a Global Economy: The Truth about U.S. Manufacturing and Trade”) was simply to evaluate the health of the U.S. manufacturing sector.  The conventional wisdom holds that U.S. manufacturing is eroding, the country is de-industrializing, and that import competition is the driving force behind this trend.  We hear this all the time.  Politicians tell us.  Op-ed page writers remind us.  Lou Dobbs warns us.  And members of Congress have proposed all sorts of punitive trade legislation under the banner of arresting and reversing manufacturing decline. 

I set out simply to assess the credibility of the premise.  My approach was straightforward, honest, and devoid of ideology.  There was no shell game or sleight of hand.  I found the most relevant, comprehensible, comprehensive, objective statistics that speak to the health of the sector, presented those data, and offered conclusions that are easily verifiable (i.e., not confused by economic modeling or econometrics or the debatable assumptions upon which such approaches often rely). 

What the data show very clearly is that U.S. manufacturing is far from declining; it is, in fact, thriving.  Output, revenue, profits, profit rates, return on investment, and exports have all been trending upward since the nadir of the manufacturing recession in 2002 and all reached record highs in 2006.  If that doesn’t constitute “thriving,” I’d be interested to learn what does. 

Since data and trends pertaining to the sector as a whole might mask different conditions in particular industries, I drilled down to explore the health of individual U.S. manufacturing industries (broadly defined at the 3-digit NAICS level).  Out of 18 broadly-defined industries, I found that 12 are doing very well and that 6 are struggling, according to the same metrics used to assess the sector as a whole.  With the exception of the auto industry, those industries that are faring poorly are generally low-technology, low-wage, and labor-intensive. 

With the manufacturing sector and the majority of its component industries found to be doing very well, the purpose of the paper was fulfilled.  The conventional wisdom was refuted.  If manufacturing is thriving – and not declining – then it is moot to demonstrate that trade has not been an important cause of manufacturing decline.  But since trade is so demonized, and the data so exonerating, it was pertinent to describe the relationship observed between trade and the various performance metrics. 

Here are some of my observations:  

  • “the rising level of U.S. imports and exports has been associated with positive developments in key manufacturing performance indicia”;
  • “As manufactured imports declined in 2001 and 2002, manufacturing output, exports, and revenues declined as well.  When imports began to pick up again as the manufacturing recession was ending, all of those real variables tracked upwards, adding more data points to the line that confirms a strong positive correlation”;
  • “As manufacturing imports have achieved new heights, manufacturing output, revenues, exports, and profits have all set records, too.”
  • “The premise that U.S. manufacturing is under duress from imports is not supported by the data”;

It is noted in the paper that industries that experienced higher levels of import growth fared better than industries that experienced lower levels of import growth.  I suggest that access to imported raw materials, components, and capital equipment helps keep the lid on costs of production for U.S. producers.  I mention that 55 percent of all 2006 import value was of intermediate products – precisely those products consumed by industry in its own production processes.  I mention that manufacturing export growth has been strong in recent years and that foreign markets are likely to be even more important to U.S. manufacturers in the year’s ahead since that’s where the dynamic growth is.  All of those conclusions (implications, if you prefer) counsel in favor of treading lightly on the trade protection front. 

The fact that some proportion of U.S. imports might have been offshored U.S. production making its way back to the United States in no way undermines any of the key points in my paper.  Even if all imports were of U.S. offshored production, the fact remains that trade has been an important part of that success story.  To the extent that the import figures reflect offshored U.S. production, rising profitability affirms the wisdom of that decision.  But the proportion of imports attributable to offshored production is likely quite small, and not “substantial,” as Roberts suggests. 

Third, my failure to distinguish between offshored U.S. production and foreign production in the import data is insignificant because other data presented affirm the limited importance of offshored production.  Roberts asserts that offshoring simply substitutes imports for domestic production.  If that were the case and if offshoring constitutes more than an insignificant portion of U.S. imports, then we should see that in the data.  But we don’t.   

Instead, we see that U.S. factory output and U.S. value added increased the most for industries that also experienced the largest increases in imports.  In other words, if those imports are offshored U.S. production, they aren’t having a discernible substitution effect, as U.S. output, value added, and profits are rising too.  We also see that U.S. factories accounted for 21.1 percent of the world’s manufacturing output in 2005 (2.5 times greater than Chinese factories), which is virtually unchanged from the 1993 figure of 21.4 percent.  In absolute terms U.S. manufacturing output and value added have been rising virtually year-after-year, as has world manufacturing output.  Yet, the United States has somehow managed to preserve its share of world manufacturing output for at least 13 years.  If Roberts’ concerns about my data presentation had any merit, we would not observe the correlation between imports and output, nor would we see U.S. share of world output that has been remarkably stable. 

I usually don’t mind disagreement with my point of view.  It happens frequently.  But I find it offensive when someone disparages and dismisses my work without a coherent basis for doing so.