uncertainty.
Parties to transnational litigation can be uncertain about the dictates of the
underlying rule, or about whether the rule will apply extraterritorially—or both.
uncertainty is confined to whether or not the rule applies extraterritorially, Priest and
Klein’s model operates straightforwardly. Where uncertainty surrounds both the scope
and the content of the rule, however, the plaintiff’s expected rate of success is a mere 25
percent. Except where the stakes for the plaintiff are very high, the likely result would be
no suit. Where litigating parties are uncertain about the content of a rule, but not its
extraterritorial applicability, the dispute is unlikely to appear in my sample, since the
defendant would not raise a jurisdictional challenge on this basis. For the vast majority
of cases in my sample, therefore, the key dimension of uncertainty is likely to involve the
scope of the U.S. jurisdiction, and a baseline expectation of 50 percent plaintiff success
(meaning a U.S. court agrees to regulate) is appropriate.
Summary of results
To examine the effects of particular explanatory variables and their combinations
on jurisdictional outcomes, I tested several models using multivariate logistic regression.
The results of these tests are summarized in the Table below.
Model A contains all of the variables described above. In this model, both
favored explanatory variables perform quite well. The ‘institutional threat’ coefficient
appears to account for some variation in the outcomes with a 95% level of confidence.
The coefficient for threats to basic rights, however, is almost four times as large, and
significant with 99% confidence. The signs for most coefficients are as expected.
Somewhat surprisingly, having an American citizen on the plaintiff side appears to
account for even more variation than domestic threat, and does so with a 99% degree of
confidence. Conduct partially in U.S. territory appears also to matter, although this result
just misses the threshold for statistical significance. The variable tracking conduct in an
OECD member state proved to have some explanatory weight at the 90% confidence
level—albeit in the opposite direction anticipated by Slaughter. And finally, having the
47
Jurisdictional questions are also a factor in domestic litigation, both between federal courts and state
courts, and the courts of different states. Priest and Klein (1984), however, do not incorporate this source of
uncertainty into their model.
48
Uncertainty regarding jurisdictional outcomes is more consequential for transnational regulatory disputes
than for domestic disputes, because a federal court’s failure to find or exercise extraterritorial jurisdiction in
most cases terminates the claim in the U.S. court system.
49
‘Year suit commenced’ is weakly negative, and the presence of an American citizen among the
defendants actually reduces the likelihood a U.S. court will regulate extraterritorially. However, neither
coefficient is large, nor does either meet the threshold of statistical significance.
19
Convention