USDA's FTA Report Repeats Errors of Previous Flawed Studies
Earlier this week, the USDA released a report attempting to estimate the effects of the Korea, Colombia, and Panama FTAs upon U.S. agricultural trade. It also examined possible effects of the ASEAN-China and ASEAN-Australia-New Zealand FTAs upon the U.S.
Unfortunately, the USDA estimated the effects through a computable general equilibrium (CGE) model, which has a shoddy track record, to say the least. A 1999 U.S. International Trade Commission (USITC) study on the likely effects of China’s tariff offer for WTO accession used a CGE model to estimate that the U.S. trade deficit with China would increase by only $1 billion dollars due to China’s accession. In reality, the trade deficit with China skyrocketed by $167 billion between 2001 and 2008.
Similar studies on NAFTA were also way off the mark. An economist at the Federal Reserve concluded that a CGE-based study of NAFTA underestimated NAFTA’s impact upon U.S. imports by ten times the actual effect of NAFTA. He concluded his study with a recommendation: “If a modeling approach is not capable of reproducing what has happened, we should discard it.”
Not accounting for currency manipulation is a chief problem of CGE models, as Rob Scott at the Economic Policy Institute has demonstrated. The USDA's report even acknowledges the devastating effect currency devaluation can have upon U.S. agricultural exports:
In 1997, U.S. apple exports to Southeast Asia peaked at 150,000 tons, just as the Asian financial crisis struck. The crisis led to sharp devaluations of Southeast Asian currencies that raised the prices of imported apples and income losses that further discouraged apple buying, triggering a dramatic decrease in U.S. apple exports to the region.
As we discuss in a factsheet, Korea is only one of three countries to have ever been placed on the Treasury Department’s list of currency manipulators, having repeatedly manipulated its currency in the past. The Korea FTA contains no prohibition against currency manipulation, so the Korean government could effectively negate the tariff cuts mandated under the FTA through currency manipulation. Despite the long history of Korea manipulating its currency, the USDA’s estimates do not attempt to account for the very real possibility of another devaluation, even though they could have done so through estimating alternative scenarios.
Given that fair trade opponents have touted beef as a major winner in the Korea FTA, a close examination of the USDA’s beef assumptions is warranted. The USDA assumes a very optimistic scenario for beef like the 2007 USITC study, although it is slightly less optimistic. (Who knows? Maybe the next study will have realistic expectations.) The 2007 USITC study assumed that U.S. beef exports would be unimpeded by current Korean regulations prohibiting the import of U.S. beef from cattle older than 30 months and it set the initial beef export level in the model at $1.1 billion, even though actual U.S. beef exports to Korea were tiny at the time the study was conducted. This USDA study assumes that the initial beef export level is $701 million. Actual 2010 U.S. beef exports to Korea amounted to about $350 million, far off the level of these optimistic scenarios. The initial level of exports is one of the primary determinants of the results of the CGE model, and higher initial exports result in greater predicted exports from tariff reductions. With the slightly lower initial beef export assumption, the USDA report projected increased beef exports to Korea of $563 million, compared to the USITC’s projection of $628-1,792 million greater beef exports. Realistic beef export data would result in even lower projected gains.
Finally, even though the report is some fifty pages, the USDA is highly stingy with its presentation of results. It presents the projections on the bilateral changes in trade flows with Korea and Colombia under the FTAs, but not how the FTAs will affect overall U.S. trade with the world. (The projected impact upon overall U.S. trade can be quite different from the bilateral results due to the trade diversion effects of bilateral trade pacts.) The decision to exclude the global results is especially puzzling given that the report focused on the effects of trade diversion in its assessment of the ASEAN-China and ASEAN-Australia-New Zealand FTAs.
It is likely that leaving the global results out of the report conceals the fact that the overall trade balance in several agricultural sectors is expected to worsen under the FTAs. The 2007 USITC report on the Korea FTA, which used the same model as the USDA study, did report both bilateral and global changes and found that several sectors expected to have improved bilateral trade balances would have worsening balances with the world, such as wheat, oilseeds, and miscellaneous crops. The global trade balance is what matters since it indicates the total effect of goods exiting and entering the U.S. upon farmers’ livelihoods. Thus, this USDA report does not contain sufficient information on projected imports and exports to evaluate the effect of the FTAs on U.S. farmers, even in terms of its own flawed CGE model.
In sum, at first glance this report projects significant gains for agriculture from the Korea, Colombia, and Panama FTAs, but a close examination reveals that methodological flaws render the report unreliable. Instead of these predicted gains, we could see a repeat of the NAFTA experience in which U.S. exports of beef and pork to Mexico in the first three years of NAFTA were 13 and 20 percent lower, respectively, than beef and pork exports in the three years before NAFTA was enacted, partly due to currency devaluation.