Rich countries engage in international trade more than poor countries do. They also have more diversified export baskets. That much is pretty well known. But how exactly is this increase in the volume of trade and greater diversification achieved? Is it through greater volumes of exports of the traditional kind, or through new products that were not previously exported? There is now a small cottage industry that has been looking at these questions, shedding light on the role that international trade and its transformation plays in the economic growth process.
An interesting paper by Céline Carrère, Vanessa Strauss-Kahn, and Olivier Cadot generates some new results. These authors analyze 6-digit trade statistics to decompose the average pattern of export diversification during economic development into a "within" and "between" component: The first of these ("within") refers to the component of concentration that is due to concentration within traditional export categories, whereas the second component ("between") captures the part due to new exports. The figure below shows the trends for these.
Note first that overall export concentration falls during much of economic development, and then begins to rise again once a country reaches a high-income level. This is similar to what Imbs and Wacziag have found for production and employment in a paper that used much more aggregate data.
Second, even though the "within" component accounts for the major part of overall concentration, it is largely the "between" component that accounts for the U-shaped trend in concentration over the development process. This is important because we know from other work (e.g. that of Lederman and Maloney) that export diversification is good for growth.
This is the strongest evidence to date that I have on the role that new products play in shaping the pattern of export composition during the development process.