We describe the main economic arguments posed for and against the North American Free Trade Agreement (NAFTA) during the U.S. policy debate. To evaluate these arguments, we analyze recent trade data and survey post-NAFTA studies. We find that both the U.S. and Mexico benefit from NAFTA, with much larger relative benefits for Mexico. NAFTA also has had little effect on the U.S. labor market. These results confirm the consensus opinion of economists at the time of the debate. Finally, studies find that trade creation greatly exceeds trade diversion in the region under NAFTA, especially in intermediate goods.

We surveyed the empirical literature using multi-country computable general equilibrium (CGE) models to analyse potential and actual regional trade agreements (RTAs). The studies indicate that these RTAs improve welfare, that trade creation greatly exceeds trade diversion, and that they are consistent with further global liberalisation. The welfare gains are bigger when models incorporate aspects of ''new trade theory'' such as increasing returns, imperfect competition, and links between trade liberalisation, total factor productivity growth, and capital accumulation. We also conjectured that an RTA expands market size and stability, allowing firms to pursue economies of fine specialisation, generating additional ''Smithian'' efficiency gains.

C A U.S.-Mexico agreement to form a free trade area (FTA) is analyzed using an 11-sector, three-country, computable general equilibrium (CGE) model that explicitly models farm programs and labor migration. The model also uses a flexible functional form for specifying sectoral import demand functions, which is an empirical improvement over earlier specifications using a constant elasticity of substitution (CES) function. The model identifies the trade-offs among bilateral trade growth, labor migration, and agricultural program expenditures under alternative FTA scenarios. Trade liberalization in agriculture greatly increases rural-urban migration within Mexico and migration from Mexico to the U.S. Migration is reduced if Mexico grows relative to the U.S. and also if Mexico retains farm support programs. However, the more support that is provided to the Mexican agricultural sector, the smaller is bilateral trade growth.

Noting that developing countries may not have the administrative capacity to levy a “pure” carbon tax, we compare the impact of alternative energy taxes with that of a carbon tax in an economy with multiple distortions. We use a disaggregated computable general equilibrium (CGE) model of the South African economy and simulate a range of tax policies that reduce CO2 emissions by 15 percent. Consistent with a “first-best” economy, a carbon tax will have the lowest marginal cost of abatement. But the relationship between a tax on energy commodities and one on pollution-intensive commodities depends critically on other distortions in the system and on structural rigidities in the economy. We demonstrate that if South Africa were able to remove distortions in the labor market, the cost of carbon taxation would be negligible. We conclude that the welfare costs of taxing carbon emissions in developing countries depend more on other distortions than on the country’s own carbon emissions.

No abstract