In this paper I identify new goods in U.S. imports data at the product levels of 7-digit TSUSA or 10-digit HS codes. I show that consistent with product cycles, the North’s new goods exports relative to old goods exports grow faster than the South’s, and then the South catches up with the North. The catch-up takes 13 ~ 18 years and happens in the 1990s, an age of globalization and outsourcing. Since the double difference includes old goods exports within the same 4-digit mSIC industry as controls, the evidence is not tainted by the myriad forces that also affect the North’s and the South’s new goods exports. Importantly, I find these results only when new goods are properly identified. When I assign new goods and old goods randomly in the data the evidence for product cycles goes away.
new goods, product cycles, double difference
Date of this Version