Empty Promises and Nonincorporation in Mercosur

ABSTRACT

Why do the member states of the world’s fourth-largest trading block incorporate only about two-thirds of all policies they adopt? This article argues that empty promises are an important reason for Mercosur’s incorporation problems and that Mercosur’s institutional design furthers such a defective behavior. Member governments easily sign agreements whenever they are rewarded for the mere act of doing so. However, if they expect high costs from implementing these policies, they try to avoid incorporation. Since only the last state to incorporate a policy triggers its overall legal validity, Mercosur’s members can easily veto any agreement ex post. In addition to empty promises, mismanaged drafting and incorporation and the abuse of negotiation power also pose important obstacles to incorporation. Free riding, however, does not play a role. Due to the incorporation rules, there can be no externalities that incentivize unilateral defection. The article substantiates the arguments empirically with the multivariate analysis of the complete incorporation record of 1,033 policies adopted between 1994 and 2008.

KEYWORDS:

• Comparitive regionalism
• compliance
• empty promises
• Mercosur
• ratification

Acknowledgment

This research benefited tremendously from debates with friends and colleagues during my time at the Graduate School for Economic and Social Sciences at the University of Mannheim. For detailed comments, criticisms, and support at different stages of the project, I am indebted to Michael Becher, Alvaro Coronel, Ina Drepper, Thomas Gschwend, Andres Malamud, Alexander Pinz, Berthold Rittberger, Duncan Snidal, Daniel Stegmüller, and Deisy Ventura. I also thank the editor and the anonymous reviewers for very useful comments and suggestions. Kim Bauer and Renke Deckarm provided invaluable research assistance.

Funding

This project was financially supported by the Deutsche Forschungsgemeinschaft, the Friedrich-Ebert Stiftung, the Landesstiftung Baden-Württemberg, and the Emil-Kömmerling-Stiftung.

Supplemental data for this article can be accessed on the publisher’s website.

Notes

¹ See Bouzas and Soltz (2001a); Bouzas, Gratius, Soltz, and Sberro (2008); Geneyro (2003); Peña (2003); Pena and Rozemberg (2005); Perotti, Stark, Vaillant, and Ventura (2004); Ventura and Perotti (2004); Zalduendo (1998).

² See Bergamaschine Mata Diz (2005); Bouzas et al. (2008); Caetano and Perina (2003); Geneyro (2003); Justo Nascimento (2004); Pena and Rozemberg (2005); Perotti, Stark, Vaillant, and Ventura (2005).

³ See Bouzas and Soltz (2001b); Bouzas et al. (2008); Geneyro (2003); Pena and Rozemberg (2005); Ventura and Perotti (2004); Zalduendo (1998).

⁴ While this had been good practice from early on in Mercosur, decision CMC 20/02 made it formally compulsory.

⁵ The Secretariat’s website allows for querying the status quo of single regulations only.

⁶ Governments may even strategically use ratification—rather than signing an agreement—to bolster their legitimacy (Hafner-Burton and Tsutsui 2005; Schimmelfennig 2000).

⁷ See Leuffen, Rittberger, and Schimmelfenning (2013) for the dimensions of differentiated integration.

⁸ For the effect of an increase in intraregional conflict on the adoption of policies regarding “deeper” cooperation, please see the online appendix.

⁹ For an overview over relevant mechanisms to incorporation and ratification processes in other contexts than Mercosur, see Simmons (2010); Treib (2008); and von Stein (2013).

¹⁰ In general, international agreements may foresee different rules as to when exactly policies become legally binding. In some cases—like the European Union—a member government who incorporates a directive suffices to trigger immediate validity in this particular country. In other cases, policies take effect only after a previously agreed threshold has been reached. For example, the Kyoto Protocol foresees in its Article 25 that at least 55 states that also account for at least 55% of CO₂ emissions need to deposit their instruments of ratification. In the extreme case, like in Mercosur, all countries need to legally commit before a policy enters into force in all signatory states. These incorporation rules have important consequences for the strategic character of international cooperation. The respective thresholds determine when domestic courts can enforce international law and when exactly signatory states jointly produce collective goods (Frohlich and Oppenheimer 1970; Schelling 1973).

¹¹ Not all of Mercosur’s policies have to be incorporated by all four members, which is why the overall number of incorporation cases is not 4 x 1,033 = 4,132.

¹² In contrast to other regional organizations (see Hartlapp and Falkner 2009), successful incorporation and its timing can be exactly identified for 92.95% of all cases. Usually, Mercosur’s members use one legal instrument for incorporation. Since 2002, decision CMC 20/02 obliges the members to use only one single domestic legal instrument. Governments occasionally made use of several legal measures only in early years. I rely on the first legal measure notified in those cases.

¹³ For an overview, see Table 1 and also Table A1 in the online appendix.

¹⁴ Please see online appendix for the recoding of the respective questions.

¹⁵ Both values are log transformed to account for their skewed distributions.

¹⁶ The principal component score was calculated using all Mercosur regulations—meaning those that require incorporation but also those that do not explicitly demand it.

¹⁷ The distribution assumes proportional effects from the covariates on the baseline rate. I examine this assumption for a Cox-model with the Therneau-Grambsch test (see Grambsch and Therneau 1994). The global test reports its violation for all variables. The standard remedy would be interactions with some transformation of time (Keele 2010; Licht 2011). It does not mitigate the problem in any of the cases. Further robustness checks with specifications that do not require either assumptions lead to similar results in magnitude and error margin. In addition, from a theoretical perspective, a monotonic duration dependence is a plausible way to think about the incorporation process. The Weibull distribution therefore seems to be an appropriate distribution to model nonincorporation.

Additional information

Funding

This project was financially supported by the Deutsche Forschungsgemeinschaft, the Friedrich-Ebert Stiftung, the Landesstiftung Baden-Württemberg, and the Emil-Kömmerling-Stiftung.