Chamber of Commerce critiques own method in trigger-happy response to Public Citizen
We fully expected a Chamber of Commerce response to our new report "Lies, Damn Lies and Export Statistics" ... I just didn't expect it to be in the form of an argument we already shredded in the report itself.
Does anyone read anything in this town?
For those just tuning in, our new report shows that the export growth rate to non-FTA countries is twice that to FTA countries. This should put a real damper on any calls for Obama to adopt Bush's FTAs as a way to boost exports. As David Dayen from Firedoglake put it,
I like this report because it challenges neoliberalism on its own terms. It puts aside for a moment the 5 million manufacturing jobs lost since NAFTA and its clones in 1994 (but only for a moment – this painful adjustment is reason enough to be skeptical of FTAs), and it puts the lie to the prevailing wisdom that this job loss is balanced by increased trade and “open markets” and economic growth, which lifts up the US economy in other sectors.
In the New York Times story on our report this morning, two Peterson Institute scholars (Robert Lawrence and Jeff Schott) respond to our report with a variety of non-sequiturs.
Lawrence cites a statistic on the export growth rate to Canada following the U.S.-Canada FTA, but fails to compare that with anything. As it turns out, sure, U.S. export to Canada grew in the first six years of the deal... but exports to non-FTA countries grew faster. Gotta have a control group, which is what our study is all about.
Schott says: "groups like Public Citizen were correct that there was not enough of a safety net for workers." Thanks for the kudos, but we're talking about exports, not safety nets.
Now, as the Chamber's John Murphy puts in their blog, and the Washington Post reports, they don't like that we exclude re-exports from our study. For the uninitiated, "re-exports" are goods that are not made by U.S. workers that are passing through U.S. ports. If you include these in your calculation (as unfortunately many superficial calculations by trade agencies do), you artificially inflate the export picture. For multinationals dedicated to overseas production and more and more FTAs, you can see how using this stat would have some appeal.
There's just one catch: our finding of an "FTA export penalty" doesn't go away if re-exports are included.
And there's just one other catch: the export data we use for
calculating the trade deficit that the Chamber doesn't like is also the export data
used in the Chamber's
own study (see page 7 - where it
describes its GTAP trade flow database) - not to mention that the non-partisan,
International Trade Commission uses it for calculating the trade balance.*
I was actually giving kudos to the Chamber all day for this
one correct point for being better than
A little trigger happy on the blog post, guys?***
There are some more problems with the Chamber's post, revealed after the jump...
The Chamber also states:
However, a July 2009 GAO report entitled “Four Free Trade Agreements GAO Reviewed Have Resulted in Commercial Benefits, but Challenges on Labor and Environment Remain” found that U.S. exports to Jordan, Morocco, Chile, and Singapore rose by 167%, 190%, 365%, and 72%, respectively, from the year prior to FTA implementation through 2008. (They were implemented in 2001, 2005, 2003, and 2003, respectively.)
A U.S. Department of Commerce Report entitled “Top U.S. Export Markets: Free Trade Agreement and Country Fact Sheets 2009” finds that U.S. exports have grown rapidly after implementation of an FTA. For instance, U.S. exports to Australia grew by 59% from 2004-2008, and U.S. exports to Central American and the Dominican Republic grew by 50% between 2005 and 2008. It also confirms the GAO numbers.
Whoa, fuzzy math alert!
Again, one major problem with the Chamber's research on trade is that they compare apples and oranges. We address these problems in pages 18-35 of our report, again, for those with the patience.As they state, the growth rates from 2001-08, 2005-08, 2003-08, and 2003-08. In case digits ain't your thang, note that these are different start dates. Not only are these numbers not comparable, but the Chamber makes no attempt to compare them.
A key plank of social science is having an adequate control group: one group that they "event" happens to, and another that the "event" doesn't happen to. Then, you have to make sure that the two groups are alike in other ways. In this case, FTA and non-FTA countries would be the groups, the event is the "FTA", and one example of the likeness of the two is that they be examined over the same time period.
The Chamber refuses to recognize that the salient measure of the record of FTAs isn’t the statistics representing a few cherry-picked FTA partners, but rather the statistics representing all of the FTAs. If Public Citizen wished to show that the growth of exports to non-FTA partners has been stellar by cherry picking some non-FTA partners, there is no shortage of examples to choose from.
For the growth statistics cited by the Chamber above, we can compare the United Arab Emirates to Jordan, Brazil to Morocco, Belgium-Luxembourg to Chile, and Turkey to Singapore, in terms of U.S. export growth over the same periods. The U.S. exports to the United Arab Emirates grew by 381 percent over 2001-2008, more than twice as fast as exports to Jordan over the same period. The U.S. exports to Brazil grew by 94 percent over 2005-2008. U.S. exports to Belgium-Luxembourg grew by 68 percent over 2003-2008. Finally, U.S. exports to Turkey grew by 211 percent over 2003-2008, more than twice as fast as exports to Singapore.
This exercise is demonstrated above of course does not in itself show that the growth of exports to non-FTA countries has outpaced the growth of exports to FTA countries. Rather, the exercise is done to show that this latest Chamber cherry picking of stats is methodologically unsound. The best way to measure the performance of exports to FTA vs. non-FTA partners is to consider FTA partners as a whole and non-FTA partners as a whole, which we do in the report.
Written by Todd Tucker and Travis McArthur, coauthors of "Lies, Damn Lies and Export Statistics"
* See the page 1 footnote which describes what's used in the
GTAP program, which the Chamber study says it uses. The U.S.
International Trade Commission makes clear that the GTAP database only
contains domestic exports. Furthermore, note that the trade balance tables in
the USITC’s annual “Shifts in U.S. Merchandise Trade” report use domestic exports and imports for consumption. Also notice that the
historical export, import, and trade balance numbers cited in USITC’s oft-quoted
study on the Korea FTA are
the domestic exports and imports for consumption data, not the data including
**Okay, we’re accuracy nuts. A previous version of this post
failed to exhaustively explain what the Chamber got right and what the Chamber omitted.
What the Chamber got right (in their report, see Table 2) is that their study
contained the accurate measure of domestic exports in both their overall model
and in their computations of historical export growth rates in Table 2 –
although note that they used a weighted method for non-FTA countries and a
unweighted method for FTA countries, which we critique in our report. We know
they used domestic exports in their Table 2 of their May 2010 report because we
were able to almost exactly reproduce the export growth figures in Table 2 when
we used “domestic exports.” See footnote 34 of our report.
But what the Chamber omitted (including in Tables 4 and 5 where they
tout the supposed export gains under the FTAs but no import estimates) is that
they don’t even attempt to measure the deficit with FTA countries since the FTA
story gets worse when one considers the deficit.
The fact remains, our method of calculating the trade
deficit is what the USITC uses, and our finding of an FTA “export penalty” holds
whether re-exports are included or not, as explained in the first page of our
*** As for Vargo’s worry about transshipped goods on the import side of the ledger, we account for that by using “imports for consumption,” rather than “general imports,” as does the USITC.