A NOTE ON THE AGGREGATION BIAS IN CURRENT PROCEDURES FOR THE MEASUREMENT OF TRADE BARRIERS

For purposes ranging from practical commercial policy matters to the testing of theoretical trade models, economists have utilized aggregate data relating to tariffs and non-tariff barriers (NTBs) across countries. However, the statistics employed in these analyses are generally derived using aggregation procedures which are based on weights relating to trade that occurs under restrictions. While it is recognized that this approach incorporates a bias in the resulting averages, due to the fact that the products facing the highest tariffs or NTBs enter the calculation with relatively low (actual) trade weights, little information is available for quantifying the magnitude of this bias. This study provides such empirical evidence and also proposes indices that overcome some shortcomings of the traditional measures.