80 Hathaway combines multiple sources, resulting in a total of 3,119 agreements over the 1980-2000 period. See Hathaway, supra note 1, at 1258-60. In contrast, the data set used here contains 6,148 agreements over the same period. The U.S. Supreme Court Pink (1942) found that international agreements, which were concluded in law, have the same legal status as treaties and do not require Senate approval. To Reid v. Concealed (1957), the Tribunal, while reaffirming the President`s ability to enter into executive agreements, found that such agreements could not be contrary to existing federal law or the Constitution. The table shows the coefficient on the contract indicator for different model specifications. “Ex Post” compares contracts to ex-post congressional executive agreements. “Other” compares contracts with other executive agreements. The view that treaties and agreements between Congress and the executive branch can be considered legal substitutes, of course, raises the question of why the United States needs two legal instruments to regulate the same types of international relations. Indeed, some commentators have asked why the United States should not abandon the treaty in favour of the agreement between Congress and the executive branch. Footnote 38 Contracts and executive agreements are instruments of national law.
These procedures allow the United States to be part of an international agreement. Nevertheless, it cannot be concluded that there is certainly no difference between treaties and ex post-Congress executive agreements. Failure to reject the zero hypothesis differs from the evidence of the zero hypothesis. The number of ex post executive agreements in the sample is small. Therefore, the failure to reject the zero hypothesis may be due simply to large standard errors due to data scarcity. This is especially true for models (2) (5), which include a large number of covariates, resulting in data savings in many subgroups. The fact that almost all model specifications provide negative coefficients certainly allows a larger number of data to obtain a statistically significant difference, although a small one. 35 Koh, above in Note 5, 91-93 (perfect legal substitutability describes it as a “long-term dominant vision” and indicates that legal science has rejected conflicting conclusions); Koh, supra note 26, to 339 (describes the debate as “settled a long time ago”). In Table 5, the same model specifications are executed with the competing complementary protocol protocol. Again, the results consistently show that treaty agreements are those that are concluded in the form of executive agreements. The results therefore do not depend on the specific characteristics of the Cox model, but are also robust for other model specifications. 123 John D.
Veau – Ross L. Prentice, The Statistical Analysis of Failure Time Data 47 (2d ed. 2002). The explanation refers to the permanent proportional risk model in which observations were grouped over time.