Fact-Checking the Fact-Checker: Washington Post Gets It Wrong on Bogus Trade-Pact Jobs Claims
As the Obama administration seeks to Fast Track the controversial Trans-Pacific Partnership (TPP) through Congress over public and congressional opposition, it has resorted to a familiar tactic – promising job gains from the deal on the basis of unfounded assumptions. We have repeatedly warned against dubious TPP jobs claims by spotlighting the inaccuracy of the administration’s job claims for the last major trade pact – the 2012 Korea “free trade” agreement (FTA). The Korea FTA was used as the U.S. template for the TPP.
The White House promised that the Korea FTA would create 70,000 jobs based on a report issued by the U.S. International Trade Commission that projected an increase in U.S. goods exports and a decrease in the U.S. goods trade deficit with Korea. In contrast, in the first three years of the Korea pact, U.S. goods exports to Korea have fallen while our goods trade deficit with Korea has surged.
We have shown that, plugging the actual U.S. government trade data into the ratio that the administration used to project job gains from the pact, the first three years of the Korea FTA’s outcomes defy the administration’s “more exports, more jobs” slogan for the deal, providing a cautionary tale for the job claims the administration is currently using to sell the TPP.
The Washington Post’s Fact Checker columnist Glenn Kessler has similarly taken on the administration’s TPP job claims, describing them as a baseless fabrication. But in today’s Post, Kessler takes issue with our fact-checking of the administration’s Korea FTA jobs promise.
Kessler seems to think that we are producing our own jobs projection for the Korea FTA, when we are actually fact-checking the administration’s jobs projection – a projection that Kessler acknowledges “would have earned Pinocchios if it had come to our attention at the time.” It is unclear why Kessler is now assigning Pinocchios to us for calling out the administration’s bogus Korea FTA jobs claim, rather than to those who actually made the claim.
In our recent press release on the third anniversary of the Korea FTA, we stated:
The U.S. goods trade deficit with Korea has ballooned an estimated 84 percent, or $12.7 billion, in the first three years of the Korea FTA (comparing the year before the FTA took effect to the projected third full year of implementation)…The surge in the U.S. trade deficit with Korea under the FTA equates to the loss of nearly 85,000 American jobs, according to the trade-jobs ratio that the administration used to promise job gains from the deal.
Kessler’s primary critique of this statement appears to be that we did not replicate the administration’s flawed methodology of only counting exports and disregarding imports when debunking the administration’s jobs projection. Such a misleading approach – ignoring half of the trade equation – contradicts standard economics and empirical data showing the job-displacing impacts of imports, which Kessler even acknowledges in his column today.
As an example of the administration’s one-sided trade accounting, the Office of the U.S. Trade Representative’s (USTR) factsheet on the third anniversary of the Korea FTA touted that 21,255 additional vehicles were exported to Korea under the FTA. It made no mention of the 461,408 additional vehicles imported from Korea during the same time period (using the administration’s same cut of the automotive trade data). That is, for every additional vehicle the United States exported to Korea under the FTA, it imported more than 21 additional vehicles from Korea. The net effect has clearly been a loss for U.S. auto workers. It is not only contrary to mainstream economics, but common sense, to only look at exports and ignore imports when assessing trade’s impact on jobs.
Indeed, when Kessler did a Fact Checker column in January debunking the administration’s “concocted” claim that the TPP would create 650,000 U.S. jobs, he stated that the study used as the basis for that claim actually showed that “the net number of new jobs [projected for the TPP] is zero” because the study “found that imports would increase by virtually the same amount as exports.” That is, Kessler presumed that $1 in imports had a job-displacing effect equivalent to the job-supporting effect of $1 in exports, which matches the calculation used in our press release. (For what it’s worth, respected economists estimate the dollar-for-dollar job-displacing effect of imports may be even greater than the job-supporting effect of exports.)
But even if we were to abandon the common-sense approach that Kessler presumed in January, defy standard trade accounting, and only count exports, U.S. goods exports to Korea fell by more than $2 billion in the first three years of the FTA. Were we to replicate the administration’s exports-only approach, then our fact-check of the administration’s Korea FTA promises would need to say something to the effect of:
The estimated $2.6 billion decline in U.S. goods exports to Korea in the FTA’s first three years equates to the loss of 17,400 American jobs, according to the trade-jobs ratio and exports-only methodology that the administration used to promise job gains from the deal. That methodology ignores the impact of imports and the significant increase in the U.S. trade deficit with Korea since the FTA was implemented. Including imports, the surge in the U.S. trade deficit with Korea under the FTA equates to the loss of 85,000 American jobs, according to the trade-jobs ratio that the administration used to promise job gains from the deal.
The takeaway is the same. Any (reasonable) way you slice it, the administration’s job gains promise for the Korea FTA is the opposite of the deal’s outcome thus far. Indeed, over the Korea FTA’s first three years, the actual results have been consistently the opposite of the specific export growth and job gain figures that the Obama administration used to sell the Korea FTA. And that gets back to our main point: rely on similar promises now being made for the TPP to your own peril.
Kessler also takes issue with our time frame, implying that we should have counted the trade data from January and February of 2012 as part of the results of the Korea FTA, despite the fact that the FTA actually took effect on March 15, 2012. Our measurement compares U.S. trade with Korea in the 12 months before the Korea FTA took effect (April 2011 to March 2012) with the third full year of the FTA’s implementation (April 2014 to March 2015).
Kessler dislikes this approach, which he criticizes as “trying to be very precise.” Instead, he states it would be more “appropriate” to compare calendar years. But the selective timeframe for which Kessler advocates would errantly count pre-FTA months as occurring since the FTA, while eliminating the most recent months of actual Korea FTA data.
The Fact Checker column also bizarrely faults us for adjusting for inflation. Typically, trade flow studies are criticized if they fail to perform the standard adjustment for inflation, since this would misleadingly count price increases as export or import increases. Nonetheless, Kessler considers the adjustment for inflation as “an effort to manipulate the data further.” Why? Because “the price of goods could decrease.” It is precisely because the price of goods could decrease (or increase) that it is important to adjust for inflation. We do so by using a standard inflation adjustment from the U.S. Bureau of Labor Statistics, thereby eliminating the effect of shifts in goods prices. (And, for what it’s worth, if we were trying to “manipulate” the data rather than be scrupulous with it, we would not be the ones controlling for inflation. Failing to adjust for inflation would make the increase in the U.S. goods trade deficit with Korea during the FTA’s first three years appear even larger than it is in real terms, not smaller.)
After deciding to replicate the administration’s usual data distortions of not counting imports, counting non-FTA months as occurring since the FTA, and not adjusting for inflation, Kessler then performs his own assessment of the Korea FTA outcome, claiming that U.S. exports increased by $2.3 billion in the FTA’s first three years.
This claim is in need of a fact check, as we have not been able to replicate it with any cut of the data. If you use the selective timeframe preferred by Kessler (comparing calendar years 2011 and 2014) and skip the inflation adjustment, you get a $750,000 increase in U.S. exports, not a $2.3 billion increase. (And that "increase" owes entirely to price increases - after a standard inflation adjustment, it becomes a real export decline of more than $1 billion.) Even if you misleadingly count foreign-produced goods as “U.S. exports,” as the administration often does, you get a $1 billion increase in exports – still less than half of Kessler’s claim (and still a real decline in U.S. exports merely by properly accounting for inflation).
The only way we see to get a $2.3 billion increase in U.S. exports to Korea is by mistakenly axing an entire year of the FTA and comparing 2014 with 2012 (as if the FTA did not take effect until 2013 – one year later than its actual implementation date). Even with this blunt error, you would still need to use the arbitrary calendar year timeframe, fail to adjust for inflation, count foreign-produced goods as “U.S. exports,” and ignore imports altogether.
Using the accurate FTA time period and excluding foreign-made goods, our total goods exports to Korea actually have fallen since the Korea FTA took effect (whether or not one properly controls for inflation). Rather than acknowledge this aggregate outcome, the Fact Checker column highlights a few specific products as having “shown real gains in the past three years” of the Korea FTA. This is another common maneuver of USTR: while overall U.S. agricultural exports to Korea increased an estimated zero percent in the FTA’s first three years, for example, USTR focuses on a $78 million gain in cherry exports. The cherry-picking in today’s Fact Checker column, however, includes products, such as apparel, that have actually not seen export gains. U.S. apparel exports to Korea actually have fallen $43 million, or 37 percent, in the Korea FTA’s first three years.
Unfortunately, this exports decline has been more the rule than the exception under the Korea FTA, as overall U.S. goods exports to Korea have fallen. Meanwhile, imports from Korea have risen, and the U.S. trade deficit with Korea has surged. The point that we have repeatedly made stands: the outcome of the FTA thus far is a far cry from the administration’s promise of “more exports, more jobs.” We would do well to keep that failed promise in mind as we now hear its echoes in the administration’s sales pitch for the TPP.