Production-weighted Estimates of Aggregate Protection in Rich Countries toward Developing Countries

A challenge in the development of aggregate indexes of trade protection is finding weights to put on various tariffs that a) reflect their importance to exporters and b) are not endogenous to the protection being measured. One common basis for weights is actual imports; but these, as is well-known, are endogenous. Various authors have worked to correct this endogeneity, but doing so is difficult in product areas where protection is both high and widespread. For this reason, I develop a new set of estimates of overall protection in rich countries with respect to developing ones that eschews import weights as much as possible in favor of weights based on the value of exporter’s total production in each product area. The results are generally much higher than those from the Bouët et al. (2004) “MAcMap” data set; there, weights are based on imports of large reference groups of countries. I conclude that product areas in which protection is high and widespread are systematically de-emphasized when using pure MAcMap weights to aggregate across major product groups. In particular, when gauging rich-country protection with respect to developing countries, agriculture is de-emphasized. I also develop estimates of trade-distorting subsidies by country and commodity and translate these into tariffequivalents with the methodology of Cline (2004) in order to estimate overall protection levels. Agricultural tariffs dominate subsidies in trade-distorting effect, and agricultural protection in turn dominates goods protection generally. Japan is most protective, largely because of rice tariffs near 900%, followed by Norway and Switzerland. Because of their greater reliance on agriculture, the poorest countries face higher trade barriers than wealthier developing countries, despite tariff preferences.