that the pattern of economic success in these cases is influenced by incentives that the framers of the
dispute settlement provisions may or may not have envisioned, we can perhaps identify reforms that
might lead to a more economically successful dispute resolution framework.7
Our approach also complements and contributes to the existing empirical literature on trade
disputes, which we characterize as falling into two categories. The first area also looks at the outcomes
of trade disputes, but this research has focused almost exclusively on United States use of Section 301
of its trade law, where American exporters can petition the government to initiate a dispute against
a foreign country. For example, Bayard and Elliott (1992,1994) and Elliott and Richardson (1997)
examine when the U.S. use of Section 301 resulted in market opening versus market closing. Kherallah
and Bhegin (1998) also focus on U.S. trade disputes and identify economic and political factors that
increase the likelihood of the petition ending in a trade war as opposed to an agreement.8 In each of
these papers, the outcome of the dispute was characterized as a categorical variable, interpreted by
the researchers and from the perspective of the plaintiff country. Our approach differs in that we look
at measures of resulting trade liberalization as our indicator of the dispute’s economic resolution.9
Relative to papers that consider only cases in which the U.S. is a plaintiff, our approach is also much
wider in scope in that we consider a set of trade disputes involving many developed and developing
countries in the GATT/WTO system.
The second area of the empirical literature relates to the initiation of GATT/WTO trade disputes
and includes investigations by Horn et al. (1999) and Bown (forthcoming).10 Bown (forthcoming) uses
data on disputes and safeguard measures under the GATT regime to determine what factors influence
a country’s decision of whether to provide import protection through the agreement’s safeguard