摘要

This paper develops a mathematical programming model to reconcile trade statistics subject to a set of global consistency conditions in the presence of an entrepot. Initial data reliability serves a key function for governing the magnitude of adjustment. Through a two-stage optimization procedure, the adjusted trade statistics are achieved as solutions to a system of simultaneous equations that minimize a quadratic penalty function. As an empirical illustration. the model is applied to reconcile the 2004 trade statistics reported by China, Hong Kong and their major trading partners, initialized with detailed estimates of bilateral trade flows, re-export markups, cif/fob ratios and data reliability indexes. Published by Elsevier Inc.