The recent rise in the US trade deficit has led to widespread questioning of the benefits of free trade and increased non- tariff barriers to trade, of which the most important have been voluntary export restraints (VERs) on imports of steel and of Japanese autos, and the quotas on textiles and clothing imports negotiated under the multi-fibre arrangement. These have been the subject of several recent studies using partial equilibrium models, which have provided useful estimates of the costs of individual non-tariff barriers. Aggregating the costs of such restrictions requires a general equilibrium approach, however, for three reasons. First, it removes an upward bias by including a balance-of-trade constraint; second, it accounts for the effects of income transfers to and from the rest of the world, particularly the effects of quota rents on resource allocation; and third, it takes into account economy-wide resource constraints and inter-industry linkages.
In Discussion Paper No. 401, Research Fellow Jaime de Melo and David Tarr use a static ten-sector computable general equilibrium model of the US economy to estimate the costs of these protectionist measures using previous estimates of the elasticities of capital/labour substitution, import demand and export supply, and of the `premium rates' paid
to the exporters of the three types of good facing trade restrictions.

The authors break up the costs of restrictions into the direct costs of income or rent transfer to foreigners and the distortionary costs arising from the divergence of the prices faced by US consumers (and producers) from the world prices they would otherwise face. They find that capturing the rents accruing to foreigners for the three sectors considered, taken together, would result in an annual welfare gain of about $15 billion. Eliminating the distortionary costs of quantitative restrictions would result in a further $6 billion welfare gain, while eliminating the remaining tariff protection would yield a further $0.5 billion.
They also estimate that the simultaneous removal of the restrictions on all three sectors would lead to the loss of some 170,000 jobs in textiles and steel, but that the autos sector would gain 1,700 jobs, first because of the reduction in the price of steel, and second because of the increased demand for autos in response to the effects on consumer income of the removal of restrictions. They estimate the net benefits of removing restrictions to amount to $104 billion over six years.
Finally, they estimate that the distortionary (and total) costs of the quantitative restrictions estimated for the three sectors are equivalent to those implied by average tariff levels of 13% (and 20%), which have not been seen since the early days of multilateral tariff negotiations.

Welfare Costs of US Quotas in Textiles, Steel and Autos
Jaime de Melo and David Tarr

Discussion Paper No. 401 (IT)