This study examined U.S. demand for salmon imports differentiated by source (Canada, Chile, and the rest of the world [ROW]), product cut (fillets and other salmon products), and form (fresh and frozen). The Rotterdam model was used in estimation, and source-aggregation tests were performed to determine the significance of source differentiation in analysis. We also performed separability tests to determine if import preferences were source-wise dependent or source independent. Test results strongly reject source aggregation; however, source-wise dependence could not be rejected. Furthermore, source-aggregated demand was significantly more price-elastic when compared to source-wise dependent demand. Results show that import preferences are not homogeneous across exporting countries, and there is significant information loss when source differentiation is not considered.

JEL Classification Code: F14, Q11, Q17, Q22