The Three R’s of CEDAW Commitment: Ratification, Reservation, and Rejection

Abstract

Human rights treaties have mixed success records for improving member states’ human rights behavior. Yet the Convention on the Elimination of all forms of Discrimination Against Women (CEDAW) has been successful for improving women’s rights. However, few CEDAW analyses account for the influence of reservations on state compliance with the treaty. This study develops a theoretical approach drawn from the opposability (vs. permissibility) school of reservations in international law. We argue that states ratifying CEDAW with reservations are less likely to see improvements in women’s rights than states ratifying CEDAW without reservations. We also show that states with worse women’s rights place reservations on core articles most central to the treaty’s success, while states with better women’s rights place reservations on procedural or non-core articles. Ordered logit, regression, and Heckman selection models of state-year data from 1981 to 2019 show that reservations in CEDAW are detrimental to the success of the institution.

Los tratados de derechos humanos tienen un historial de éxito desigual en los que se refiere a la mejora de los comportamientos por parte de los Estados miembros en materia de derechos humanos. Sin embargo, la Convención sobre la Eliminación de Todas las Formas de Discriminación Contra la Mujer (CETFDCM) sí que ha tenido éxito a la hora de mejorar los derechos de las mujeres. Sin embargo, pocos de los análisis que se han llevado a cabo sobre la CETFDCM tienen en cuenta la influencia que han tenido las reservas por parte de los Estados sobre el cumplimiento del tratado. Este estudio desarrolla un enfoque teórico, derivado de la escuela de la oponibilidad (frente a la permisibilidad), de las reservas en el derecho internacional. Argumentamos que los Estados que ratifican la CETFDCM con reservas tienen menos probabilidades de ver mejoras en los derechos de las mujeres que aquellos Estados que ratifican la CETFDCM sin reservas. También demostramos que los Estados en los que las mujeres tienen peores derechos formulan reservas sobre los artículos fundamentales y de mayor importancia para el éxito del tratado, mientras que los Estados con mejores derechos para las mujeres formulan reservas sobre artículos procedimentales o no esenciales. Usamos los modelos logit ordenado, de regresión y de selección de Heckman sobre los datos de los Estados entre los años 1981 y 2019 con el fin de demostrar que las reservas en la CETFDCM son perjudiciales para el éxito de la institución.

Force est de constater que les traités des droits de l’homme ne sont pas toujours couronnés de succès quand il s’agit d’améliorer le comportement des États membres sur ce point. Pourtant, la Convention sur l’élimination de toutes les formes de discrimination à l’égard des femmes (CEDEF) s’est avérée efficace pour améliorer les droits des femmes. Néanmoins, peu d’analyses de la CEDEF prennent en compte l’influence des réserves qui s’appliquent à la conformité des États au traité. Cette étude développe une approche théorique tirée de l’école de l’opposabilité (par opposition à la permissibilité) des réserves en droit international. Nous affirmons que les améliorations vis-à-vis des droits des femmes sont moins probables chez les États qui ratifient la CEDEF en émettant des réserves que chez ceux qui n’en émettent pas. Nous montrons par ailleurs que les États où les droits des femmes sont plus médiocres émettent des réserves quant à des articles centraux et essentiels à la réussite du traité, tandis que les États où les droits des femmes se portent mieux choisiront d’appliquer leurs réserves à des articles procéduraux et secondaires. Un classement de modèles logit, de régression et de sélection d’Heckman de données sur l’année d’entrée des États dans la CEDEF entre 1981 et 2019 montre que les réserves nuisent à la réussite de cette institution.

Keywords:

• CEDAW
• commitment
• ratification
• reservations
• women’s rights

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The general reservation most discussed by the Committee relates to Article 29, which articulates conflict management procedures (e.g., arbitration).

2 “Reservations are the means by which nations alter or exclude treaty terms to individually limit the obligation or effect of the treaty” (Piper 1985296).

3 Unlike most other treaty areas where reservations can be applied by other states via reciprocity, human rights treaties are not reciprocal because even if a State Member objects to another state’s treaty reservations, it does not alter their own obligations under the treaty (Korkelia 2002, 439).

4 This map demonstrates the commitment levels of states to CEDAW. It shows states in the year 2021 that have fully ratified, ratified with reservations and then withdrawn those reservations, ratified with reservations, and not ratified the treaty.

5 We find a similar pattern in Supplemental Table A4 where we estimate a Heckman probit model with CEDAW ratification as stage one and the use of reservations as stage two. We show that more democratic countries ratify CEDAW at higher rates and are less likely to place reservations on the treaty. Yet, wealthier states join less often and use more reservations, a pattern consistent with what Simmons observed (2009, 101) for high rule of law states.

6 The core articles (2, 7, 9, 16, and 29) have an average of 29 states making reservations versus 7 states making reservations on average for non-core articles (1, 5, 10, 11, 13, 15, 28). Articles with no reservations are excluded from these lists.

7 “The Pan-American rule provided that if state A ratified a treaty including a reservation to which state B objected and state C accepted, the treaty would be in force between states A and C subject to the reservation and between states B and C unrestricted by the reservation. The treaty would not be in force between states A and B” (Piper 1985308).

8 See https://www.ohchr.org/en/professionalinterest/pages/opcedaw.aspx. We coded one for each year in our state dataset if the years CEDAW members had ratified the Optional Protocol (OHCHR 2024). Data were coded from the organization’s website: https://indicators.ohchr.org/.

9 CIRIGHTS generates scores based on states’ human rights practices (rather than stated policies) using information on annual human rights reports from Amnesty International and the US State Department. See https://cirights.com/.

10 The V-Dem and CIRIGHTS social rights measures are correlated at 0.53.

11 The social rights indicator in this table is from CIRIRIGHTS and not the social indicator from V-Dem used in our multivariate models. This measure only spans from 2005 to 2021, which is why we employ a V-Dem indicator in our broader analyses. Supplemental Table A3 in the Appendix presents descriptive statistics for the dependent variables.

12 In Supplemental Table A3, we add a dummy variable to our main models indicating if a reservation was made through a declaration and our results are robust. On the importance for RUD distinctions, see McKibben and Western (2020) and Zvogbo, Sandholtz, and Mulesky (2020).

13 Civil and interstate war measures are from the Major Episodes of Political Violence Dataset (Center for Systemic Peace INSCR Data Page 2024); https://www.systemicpeace.org/inscr/MEPVcodebook2018.pdf. Population size and wealth are updated through 2019 using Whitten-Woodring’s (2016) replication dataset. The women’s civil society organizations measure is taken from V-Dem. Legal tradition data comes from Mitchell and Powell (2011), while Polity data comes from https://www.systemicpeace.org/inscrdata.html.

14 These graphs are generated using the margins command in Stata and setting all other variables at their mean.