Abstract
Theoretical arguments and empirical evidence suggest that non-tariff measures (NTMs) generate a mixture of trade effects. Using the maximum level of disaggregation provided by the UNCTAD NTMs database (four-digit), and focusing on those measures applied by China in its pork trade, a gravity equation is estimated with the poisson pseudo maximum likelihood estimator. Results confirm both restricting and promoting effects that higher levels of NTM aggregation mask. Compared to the average tariff applied by main importers (9 per cent) and China (14 per cent), the most restrictive NTMs are more stringent, with ad-valorem equivalents between 15 and 33 per cent. Simulations illustrate the upper bound bilateral trade gain from NTM removal and the resulting repercussions for China’s pork import shares.
On the Loglinear Poisson and Gamma Model