Market Entry and Trade Weighted Import Costs

Trade costs have fallen surprisingly little given the large increase in international trade in the last 50 years. This paper examines whether trade costs are properly measured. I show theoretically that trade weighted measures will underestimate the changes in trade costs when there are fixed market entry costs and quality differences. Newly traded goods enter at higher trade costs than previously traded ones. Lower import costs shift trade to low quality goods with higher measured trade costs. U.S. import costs fall twice as fast as trade weighted measures from 1974 to 2004 when the impacts of shifting and new goods are removed. Once the biases are removed, typical estimates of trade elasticities can explain increasing trade.