The TPR data enabled us to estimate frequencies in the application of non-tariff measures at the harmonized system (HS) two-digit level for 97 product categories. Thus the frequency ratios (f) calculated from the TPRs relate to the proportion of HS2 product categories out of the total affected by a particular measure. The weakness of this indicator is that it gives equal weight to the presence of a measure in a country that could affect only one or a few lines in an HS2 category, for example HS72 (iron and steel), while the presence of the same measure in another country might affect a large number of tariff lines, for example all steel products.
Formally, let Nqm be a non-tariff measure imposed by country m on a product or group of products q. The frequency ratio for that measure, fnm = ΣNqm/ΣQm where Qm is the total number of products, measured in total tariff lines or product groups. Thus for the calculation of f, using the HS2 product breakdown employed in most of the analysis, Qm = 97. Where tariff line information was available, for example in the case of antidumping measures, for the calculation of f′ a standard HS six-digit tariff line classification of approximately 5200 lines was used.
It may appear at first glance that f will always be larger than f′. This is not the case, however. The two different frequency ratios show different aspects of a country's trade regime: if a specific non-tariff measure involves a large number of tariff lines concentrated in one or two groups of products, f may be smaller than f′; the reverse will be the case if a particular measure applies to a few products in a large number of groups. A simple example from one of the countries – Thailand, for which tariff line and broader category measures are available for the same year – can be used to illustrate this point. In 1997 Thailand applied non-automatic licensing on a total of 25 product categories, involving 713 tariff lines. In this case f = 26 per cent while f′ = 14 per cent. In the same year, Thailand's prohibitions were concentrated in six product categories involving 613 tariff lines. In this case f = 6 per cent and f′ = 12 per cent.
An effort was made to complement the TPR analysis of non-tariff measures with data obtained from the UNCTAD TRAINS data base, which permits the calculation of frequency ratios (f′) at the tariff line level. The TRAINS data are available for a fewer number of countries (22), and only in six cases was information available for the same country over a period of time. The f′ ratios for similar non-tariff measures as those calculated from the TPRs but based on tariff line data from TRAINS are shown in Table A2.1. A comparison with Table 4.8 suggests that there is a pretty good correlation between the frequency ratios in countries that apply non-tariff measures on just a few products and those that apply them on a very large number of products, but there appears to be little correlation between the two frequency measures for countries in between.
On the other hand, when looking at the evolution of frequency ratios over time, the frequency ratios for Chile, Colombia and Thailand increased between the first period and the second, but using the TPR information they declined for the same countries. On closer investigation it appears that the reason for the increase was the introduction in all three countries of licensing and/or a prohibition on the
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Book title: Developing Countries in the WTO. Contributors: Constantine Michalopoulos - Author. Publisher: Palgrave. Place of publication: New York. Publication year: 2001. Page number: 258.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.