| Abstract: | Conclusions The necessary and sufficient condition suggested by Hillman [1980] for the index of RCA, when used in cross-country comparisons, to provide a one-to-one relationship between pre-trade comparative advantage and revealed comparative advantage is fulfilled for the great majority of the “commodities” traded in 1985 by 118 developing economies. Given the low level of data aggregation (the lowest for which statistical information is currently available), the large sample of countries and the high percentage of developing economies’ total exports captured in this research, we can conclude that Balassa’s export-performance index, for cross-country comparisons, is a good indicator of comparative advantage as reflected by pre-trade prices. In other words, Hillman’s condition is a useful indicator of the presence of monotonicity in indices of RCA: we have observed that at a 5-digit level of commodity aggregation, increases in Balassa’s export performance index of RCA are likely to correspond to increases in export levels. Aggregation of commodities at a 3-digit and at a 1-digit level suggests that Hillman’s condition is unlikely to be violated if the cause for values of the Hillman’s Index less than one is due to export specialization; on the contrary, the number of cases of a Hillman Index smaller than one due to a large share of world markets is expected to be negatively related to the level of aggregation. Our results suggest that Hillman’s index should be calculated in any empirical investigation trying to assess the long-term implications of trade liberalization negotiations using an export-performance index of RCA. There is evidence that, if used at a disaggregated level, the HI is a tool that may help flag cases in which the RCA index can be a misleading indicator of countries’ comparative advantage, even in cross-country comparisons. It may also help reduce disagreements regarding the most appropriate coefficient of RCA. Further research should be directed toward enlarging the sample of countries to include the entire world and to incorporate at least three years of trade so as to be able to eliminate the possible influence of cycles. |
