Between them, North America and Western Europe account for two-thirds of world trade and one-half of world income. What benefits and costs would accrue – and to whom – if they became one giant North Atlantic trading bloc? This was the question posed by Joseph Francois (Erasmus Universiteit, Rotterdam, and CEPR) at a CEPR/European Institute meeting in Washington on 7 April 1997. As Francois noted, the issue was not simply whether pursuit of a trade and investment agreement between the two regions would be justified by the potential benefits. Equally important was whether they would benefit more from a preferential bilateral initiative or from joint support for multilateral initiatives.

Francois drew on his recent CEPR Discussion Paper (co-authored with Richard Baldwin), which had provided a quantitative assessment of the implications of preferential trade liberalization by the North Atlantic economies. This included an examination of the pattern of production, trade and import protection in both North America and Western Europe, and the likely trade and income effects of trade liberalization.

Three broad conclusions had emerged from this work. First, ‘narrow’ bilateral preferential agreements would have little discernible impact on national incomes in the two regions. Second, a ‘deeper’, but still only bilateral, trade agreement would generate modest increases in income and wages for both regions, but at the cost of generating welfare losses for North Africa and the Middle East. Third, the economic benefits arising from such bilateral preferential trade liberalizations would be easily swamped by the potential gains from comparable multilateral liberalizations. For other regions – particularly developing countries – the implications of such multilateral liberalization depend critically on their own participation in the liberalization process.

Francois noted that trade between North America and Western Europe consisted largely of exchanges of similar industrial products. In 1994, almost 40% of this trade was in machinery, cars, car parts and other transport equipment. Out of $276 billion in total transatlantic merchandise trade, very little, in relative terms, was in politically sensitive industries, such as textiles, clothing, steel and agricultural goods, where trade frictions tend to be concentrated.

While the North Atlantic economies accounted for a large share of world trade, the analysis suggested that they offered relatively limited opportunities for further trade liberalization. This was because of the combination of Uruguay Round commitments to reduce tariffs, and the Information Technology Agreement reached at the Singapore Ministerial of the World Trade Organization. Taken together, these developments meant that, even without a preferential trade agreement, most transatlantic trade would face either relatively low tariffs (generally less than 2.5%) or zero tariffs by 2005.

Thus reduction in industrial tariffs alone, without a deeper liberalization initiative, was likely to have little – if any – impact on trade and incomes. Peaks of protection (tariff equivalents of over 30%) did affect EU and US exports of fishery and mining products, textiles, clothing, fabricated metal products, transport equipment and machinery, but these rates were generally found in Asia, Latin America and Africa, not in Western Europe or North America. (There was a notable exception, of course, in continued EU and US protection of sensitive products in their agricultural, textiles and fisheries sectors.)

With residual industrial protection relatively low, a ‘narrow’ preferential agreement would be likely to have only a very small positive effect on national incomes in North America, with near-zero gains for EU incomes. The research simulations suggest, in contrast, that a ‘deeper’ agreement, including bilateral elimination of industrial tariffs, agricultural protection and anti-dumping remedies, would generate modest increases in income and wages for the two regions. This would entail a broadening of the coverage of the plurilateral government procurement agreement, and a reduction in trading costs through mutual recognition and harmonization of standards.

The implications of a preferential agreement for third countries, however, vary according to the degree of liberalization. One qualitative result which was common to most of the simulations was that welfare losses would be implied for North Africa and the Middle East. The explanation was that a preferential agreement between the EU and North America would erode these other countries’ existing tariff preferences in EU markets. The losses were generally of a magnitude roughly comparable – as a percentage of GDP – to the likely gains for the EU. Indeed, this result was quite general, in that preferential liberalizations solely among developed countries, whether involving only the North Atlantic economies or the entire OECD, were likely to imply adverse effects for various developing country regions (especially in Africa, but also in Latin America).

The research further suggested, however, that any potential economic benefits to Western Europe and North America from preferential bilateral liberalizations would be easily swamped by the potential gains from comparable multilateral liberalizations. For example, in the simulations, further trade liberalization also involving developing country tariff reductions implied significant gains for developed and developing regions alike. Moreover, these gains for non-OECD regions flowed largely from import liberalization by the developing regions themselves. Thus, within a multilateral framework, the gains that an individual region will enjoy as a result of a liberalization depend on the extent to which the region liberalizes itself.