Legalization and dispute settlement benefits: The case of the GATT/WTO

Participation itself is also important because there is “no disadvantage for developing countries in terms of winning verdicts or compliance from defendants” under the WTO (Busch and Pelc 2015: 400).

See Appendix Table 8.

Under the late GATT there are 17, 15, 15 and 4 user countries and 95, 15, 15 and 4 member countries, under the early WTO there are 28, 12, 19 and 9 user countries and 110, 12, 19 and 9 member countries, and under the late WTO there are 40, 10, 18 and 5 user countries and 128, 10, 18 and 5 member countries in the first, second, third and fourth categories respectively.

The GNI refers to the total incomes of all of the residents of a country, including the incomes obtained from outside of the country, but excluding the domestic incomes generated by foreigners.

Data are from World Bank’s ‘World Development Indicators’: http://​data.​worldbank.​org/​data-catalog/​world-development-indicators.

The data regarding the IO memberships are from The Correlates of War Project’s ‘Intergovernmental Organizations’: http://​www.​correlatesofwar.​org.

For the regressions for the first dependent variable, the indices of independent variables of each user or member country are their average values for each period, and for the regressions for the second dependent variable, the indices of independent variables of each user country are those of the beginning years of the respective cases.

Seemingly unrelated estimation is similar to seemingly unrelated regression that tests OLS models. Yet, it is useful to test the regressions such as ordered probit regression and binary logit regression that this study implements. It uses SUEST command in STATA to test this estimation (Mandemakers and Dykstra 2008: 499; Van Solinge and Henkens 2008: 425–426).

In this study, developed countries are the members of the Organization for Economic Cooperation and Development (OECD) with per capita GNI over $20,000 in 2005, and developing countries including the least developed countries are all the other countries. This condition of per capita GNI is necessary since not all of the members of the OECD are considered as developed countries. In general, per capita GNI shows the level of economic development better than GNI. If we apply this definition of developed and developing countries to the users of the dispute settlement mechanism under the GATT/WTO, then the developed countries include the USA, EC and EU countries, Canada, Japan, New Zealand, Australia, Norway, Switzerland, and Denmark, whereas developing countries are Brazil, Mexico, India, Argentine, South Korea, Thailand, China, Chile, Guatemala, Indonesia, Honduras, Panama, Columbia, Costa Rica, Hungary, Philippines, Taiwan, Pakistan, Peru, Poland, Ukraine, Ecuador, Turkey, Vietnam, Russia, Antigua & Barbuda, Bangladesh, Cuba, Czech Republic, Dominican Republic, El Salvador, Hong Kong, Malaysia, Moldova, Nicaragua, Singapore, Sri Lanka, Uruguay, and Venezuela. These developing countries were developing countries during the period of the GATT, but one may ask whether South Korea, Hong Kong, and Singapore still remain as developing countries. First, South Korea has been a member of the OECD since 1996, but its per capita GNI was $18,511, half of Japan’s $36,615, in 2005, and $24,758, approximately two thirds of Japan’s $37,583 in 2014. Second, Hong Kong and Singapore are not members of the OECD, but their respective per capita GNIs were $26,790 and $27,894 in 2005, and $35,134 and $37,120 in 2014, which are as high as Japan’s per capita GNI. How should we classify these three countries? South Korea’s per capita GNI does not belong to that of the developed countries. The per capita GNIs of Hong Kong and Singapore belong to those of developed countries, and one may argue that they should be classified as developed countries. However, even if we accept this argument, the classification of these two countries as developing countries minimally distorts our statistical data about the use of the dispute settlement mechanism. Hong Kong and Singapore has filed a complaint only once throughout the entire period of the WTO. In addition, South Korea, Hong Kong, and Singapore have all declared themselves as belonging to developing countries within the WTO, and the WTO acknowledges their declaration. All the per capita GNI data are from the World Bank’s databank.

The EEC is counted as a developed country.

The data from 1947 to 1989 can be found at Hudec (1991: 587-608), while the data from 1990 to 1994 can be found at the GATT Digital Library: 1947–1994 (http://​gatt.​stanford.​edu/​page/​home). There are a total of 249 cases including 12 cases with multiple complainants and two cases with multiple respondents.

The total number of the complainants and respondents exceeds the number of the total dispute settlement cases, since there are multiple complainants and respondents for the individual dispute settlement cases of the GATT.

The data are from the WTO homepage (http://​www.​wto.​org). There are a total of 488 dispute settlement cases from 1995 to 2014. The total number of the complainants and respondents exceeds the number of the total dispute settlement cases, which include seven cases with multiple complainants and five cases with multiple respondents.

The EC/EU is counted as one developed country, and this is the reason why the number of developed countries is relatively small.

The number of complainants and respondents exceeds the number of total cases since there are multiple complainants and respondents for the individual dispute settlement cases of the WTO.

The results of these Model 1 and Model 2 are similar to those of the negative binomial regressions using the continuous dependent variable shown in Appendix Table 9.

The coefficients of world trade share differ significantly (p < 0.05) and those of previous experiences do so (p < 0.1) for the three different periods.

The coefficients of previous experiences differ significantly (p < 0.05) for the three different periods.

The results of Model 5 are almost the same as those of the independent regressions of the respective periods. See Appendix Table 10.