The gaps in living standards are widening among developing countries. Developing countries in East Asia have grown quickly, although they were temporarily set back by the Asian crisis of 1997. For many poor countries in Africa, average incomes have been declining for several decades. And countries in transition from central planning to market economies experienced large declines in output and income during the early years of transition, with most countries of the former Soviet Union experiencing especially large declines.

Developing countries must decide what trade policies to adopt toward primaryproduct exports, industrial imports, and industrial exports. A traditional fear about relying on exports of primary products is that the world market price trends are unfavorable to producers, especially those in developing countries. The evidence shows a downward trend in the relative prices of most primary products, as commonly feared. Two factors lowering the relative price of primary products are Engel's law and the development of modern synthetic substitutes for primary materials. Two opposing forces, which would tend to raise primary-product prices, are natural resource limits and the fact that productivity growth is often slower in the primary sectors than in the rest of the economy.

Joining an international cartel could bring gains to a developing country that exports the cartelized product. The greatest cartel success by far is OPEC's pair of price victories in 1973-1974 and 1979-1980. With all international cartels, even OPEC, success breeds decline. Four forces dictate the speed at which a cartel erodes: the rise in product demand elasticity, the rise in the elasticity of competing supplies, the decline in the share of the cartel in the world market, and the rise in cheating by members of the cartel. Because of supply conditions, it is unlikely that cartels in other primary products could achieve anything close to OPEC's success.

One strategy open to developing countries is that of import-substituting industrialization (ISI). It could raise national skill levels, bring terms-of-trade gains, and allow planners to economize on market information (since they can just take industrial imports themselves as a measure of demand that could be captured with the help of protection). Studies of ISI and related policies, however, show that income growth is negatively correlated with antitrade policies like ISI, and positively correlated with outward-oriented policies that are closer to free trade. The available evidence supports the fears about ISI raised by the analysis of Chapters 8 through 10.

Another strategy is to concentrate on developing exports of manufactured goods, especially those that are intensive in less-skilled labor. This has been a slowly prevailing trend since the 1960s, though ISI also remains practiced in developing countries. Relying on exports of manufactures has its risks, however. Developing nations have rightly complained about import barriers against their new manufactures erected by the industrialized countries. Such barriers have indeed been higher than the barriers on manufactures traded between industrialized countries. Still, evidence shows that an outward-oriented trade policy encouraging exports of manufactures is part of the most promising strategy for most developing countries.