Working Paper: CEPR ID: DP1629
Authors: Robert C. Feenstra; Andrew K. Rose
Abstract: This paper develops a procedure to rank-order countries and commodities using disaggregated US imports data. It finds strong evidence that both countries and commodities can be ranked, consistent with the ?product cycle? hypothesis. Countries habitually begin to export goods to the United States according to an ordering; goods are also exported in order. We estimate these orderings using a semi-parametric methodology that takes account of the fact that most goods are not exported by most countries in the sample. The orderings seem sensible, robust and intuitive. For instance, the country rankings derived from disaggregated trade data turn out to be highly correlated with macroeconomic phenomena such as national productivity levels and growth rates.
Keywords: empirical data; SITC; disaggregate; bilateral; semiparametric; rank; missing; country; commodity; product cycle
JEL Codes: F10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| export timing (F17) | productivity levels (O49) |
| export timing (F17) | growth rates (O40) |
| sophistication of exports (F10) | productivity levels (O49) |
| sophistication of exports (F10) | growth rates (O40) |