Developing countries experienced a revolution in trade policy in the 1980s and 1990s, but it is unclear how much real openness increased. After all, they had started with multiple, often redundant, trade restrictions. And it is unclear how changes in openness should be measured. The most appropriate measure of openness is based on imports of consumer goods, argue the authors, since these imports commonly face the biggest trade barriers. After developing several such measures, including a measure of the change in tariff equivalent protection, they explore the recent evolution of trade policy, using readily available trade data. Openness has developed incrementally rather than overnight. In the early adjustment stages, barriers to imports tended not to be reduced much. At first, the net reduction of incentives to produce import substitutes was minor, especially when currency depreciation is considered. Recently import barriers have been reduced more substantially, and since there has been little currency depreciation, incentives to produce import substitutes have declined. Shock therapy was uncommon. A few countries moved quickly to eliminate nontariff barriers to imports and to adopt low, fairly uniform tariffs. But most countries tended to peel away redundant layers of trade barriers, one at a time. They usually began with the barriers embodied in rationing and exchange controls, proceeded to nontariff measures, and finally reduced tariffs. Each step may have reduced protection a bit but the big reductions apparently came only in later stages. Still, even gradual reform helped open up those economies. The Asian countries tended to be most open both early and late. They were also above-average in reform efforts, by some measures, so their strong growth performance (based on exports) was unsurprising. The African countries, whose trade policies were probably worst to begin with, made relatively modest progress initially. In recent years their progress has been substantial; whether they have improved as much as other countries depends on which measure is used. Countries tied to the French franc (for whom real devaluation was more difficult) showed less progress than nonfranc countries, illustrating the importance of the connection between devaluation and trade reform. There is no evidence that rapid trade reform resulted in Africa's de-industrialization.
