CEPR

Widely Cited Economic Models Have a Poor Track Record in Projecting the Winners and Losers of Trade Agreements

April 21, 2016

Contact: Dan Beeton, (202) 239-1460

Washington, D.C. - A new research paper from the Center for Economic and Policy Research shows that the model used by the International Trade Commission (ITC) to project the impact of the Korea-U.S. Free Trade Agreement (KORUS) poorly predicted the winners and losers among export and import industries following the agreement. The report, “Trade and Jobs: Can We Trust the Models?” by economists David Rosnick and Dean Baker, notes this finding is consistent with prior research. “These models have not only failed to predict major changes in the trade balance, they also have been largely unsuccessful in identifying the industries that win and lose following the implementation of recent trade pacts.” Therefore, the authors argue that these models’ predictions of the expected economic impact from the TPP should be viewed with skepticism.

“The most widely cited models failed to predict large increases in trade deficits that followed the signing of NAFTA and the Korea-U.S. FTA. The economic impact of the rise in these deficits dwarfed the predicted gains from reduced trade barriers,” paper co-author and CEPR Co-Director Dean Baker said.

The paper notes that following three major trade openings by the United States in the last quarter century — the North American Free Trade Agreement (NAFTA), Permanent Normal Trade Relations with China, and the KORUS — there was a large increase in the U.S. trade deficit with the partner country. While this may have been due to factors other than the agreement itself, it nonetheless overwhelmed the impact of reducing trade barriers projected in computable general equilibrium models. “In other words,” the paper adds, “the unpredicted impact of the rise in the trade deficit was far larger than any of the predicted effects from these models.”

Examining modeling by the International Trade Commission (ITC) of the expected impact of the KORUS, the authors found that under KORUS there has been a large increase in the size of the U.S. trade deficit with South Korea that was similarly not predicted by the model. This trade deficit more than doubled from $13.2 billion in 2011, the last year before the agreement took effect, to $28.3 billion in 2015.

The paper constructs a baseline that assumed the trend changes prior to the implementation of the KORUS continued. There was no correlation between the differences between the actual changes from this baseline and projections from the ITC model. Noting that the ITC model both failed to predict the rise in the trade deficit and which industries would be winners and which losers, the authors conclude the predictions were of little value.

“These models are often used to promote trade deals, as is happening now with the proposed TPP,” Baker said. “But it is important to look at how accurate modeling has been in the past at predicting the economic benefits of these agreements. Too often, the models have failed.”

###