counterfactual economy would look like with the reversions, we then compare the results to the current
economy to size up the impact of actual policy reforms. Essentially, the logic of our CGE analysis follows
the logic set out in our partial equilibrium analysis. CGE models are powerful tools for this type of
calculation because they can evaluate the impact of shocks on the entire economic system. CGE models
encompass both the direct impact of a policy or cost change and myriad indirect impacts through the vast
network of economic linkages. For example, the repeal of NAFTA would certainly affect trade between
the United States, Canada, and Mexico, but it would also have indirect impact on commerce within the
United States and trade with non-NAFTA countries.
Our CGE analysis begins with the GTAP7 database, released in November 2008. The base year
for the GTAP7 model is 2004. From the base year we run the same six scenarios that are outlined in the
partial equilibrium analysis section. The scenarios are as follows:
n 1: Reversion to Uruguay Round bound tariff rates
n 2: Reversion to Tokyo Round bound tariff rates
n 3: Return to circa 1980 transportation costs
n 4: Removal of preferential tariff rates
n 5: Reversion to circa 1990 NTB levels
n 6: Reversion to circa 1990 MFN applied tariff rates
The starting or “present” data for each scenario, with the exception of Scenario 5, is wholly
contained within the GTAP7 database. For Scenario 5 we must first augment the GTAP7 model with
“present” NTB rates. For each of the six scenarios we shock the GTAP7 model by essentially replacing
the “present” data with “past” data. The data are disaggregated versions of information used in the partial
equilibrium analysis. Rather than the aggregate numbers displayed in table 4, however, we use specific
data points for the United States and the 17 major partners across the 42 merchandise trade sectors
included in GTAP7.15 Due to the structure of the GTAP model, our results actually include the indirect
impact on trade in services, even though we only include tariff, NTB, and transportation cost changes
for merchandise goods. The general equilibrium model has the added benefit of analyzing the impact
on more than one country at once. We stress the results for the United States in this paper, but we also
display, and discuss in a few instances, the results from the other 17 countries.
Tables 8 and 9 show the estimated impact on exports for each of the six scenarios. A reversion in the
NTB levels of protection that prevailed in the early 1990s in the 18 countries would reduce US exports
by roughly $140 billion a year, or 13 percent of total goods and services exports. A reversion to Tokyo
Round bound rates would have a similar impact, $130 billion or 12 percent of exports. The reversion to
MFN applied tariff rates in the early 1990s would have a slightly smaller impact, a decrease of roughly
15. See table 12 for a list of sectors.
10