Perceived to slack: secondary securitization and multilateral treaty ratification in Israel

ABSTRACT

This study emphasizes the place that cognitive processes rather than objective concerns have in ratification of multilateral treaties. We argue that secondary securitization by non-security experts hinders treaty ratification. When security is at stake, the potential costs of undesired action by the treaty’s IO are deemed higher, risk-aversion increases, and asymmetry among the member states’ policy perceptions is greater. Thus, our secondary securitization model improves over existing explanations of multilateral treaty ratification by assuming that national selfishness drives treaty (non)ratification, but not necessarily in a rational way. We support our argument with survival analysis regarding the ratification process in Israel of 243 treaties, based on documents retrieved from official archives, and controlling for a variety of competing explanations. We break securitization into objective and subjective components and correct for the possibility of undocumented acts of securitization. Our results are robust to all this. We follow with discourse and content analysis of official discussions of three human rights treaties (ICCPR, ICESCR, and CEDAW). We innovate theoretically by distinguishing secondary from primary securitization, and by combining Securitization and Principal-Agent theories. We believe our results travel well for other countries in which security concerns overshadow aspect of civilian life, and IOs are regarded with suspicion.

KEYWORDS:

• Securitization
• principal-agent
• multilateralism
• Israel

Acknowledgments

The authors are especially grateful to Mark Blyth and Galia Press-Bar-Nathan for their invaluable support and suggestions. They are also indebted to Michael Barnett, Mark Copelovitch, Yoram Haftel, Mark Hallerberg, Lior Herman, Mark Kayser, Andrew Lugg, Magnus Lundgren, Thomas Risse, Michal Shamir, Yael Shomer and participants at the 2017 PEIO meeting, the 2016 American Political Science Association and the 2017 meeting of the Israeli Association for International Studies for their helpful comments. Nizan Feldman shared data on UN voting patterns. Tal Jana, Benjamin Paley, Naama Rivlin and Eran Rubinstein provided diligent research assistance. Eyal Rubinson is also grateful for the support of the School of Political Science, Government and International Affairs at Tel Aviv University, the Gideon Doron fellowship, and DAAD.

Supplementary Material

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

Notes

¹ We study treaty ratification rather than signature because only ratified treaties become binding.

² Mandat International is a non-governmental organization. It operates a legal search engine (What Convention) that enables access, search and analysis throughout many international legal instruments.

³ A notable exception includes Fischhendler and Nathan (2014), who applied this platform to analyze Israel’s decision-making on natural gas exports.

⁴ The vast majority of multilateral treaties concentrates on technical and procedural matters, that are of little interest to the general public. Moreover, with a few exceptions, notably some EU treaties and some particular countries, most treaty ratifications do not entail referendums or other forms of comprehensive public consent, making securitization efforts aimed at the wider public less likely. Hence, speech acts directed to the domestic arena by political parties are uncommon, and the use of mass media for securitization is less likely.

⁵ Note that we are not saying that any and all undesired action represents a threat to a member state’s security, only that policy-makers are likelier to perceive actions as undesired when the issue-area is securitized.

⁶ Independent treaties are such that membership in one does not formally oblige or forbid membership in another.

⁷ The archives did contain material regarding 96 of these 178 treaties, but no record of ratification-related discussions or exchanges.

⁸ Indeed, in nine of the 243 treaties (three of which are included in the 65) we are aware that some documents remained confidential even after the limitation period.

⁹ The following discussion of the variables is brief. The bracketed signs following their names indicate the expected relationship with the likelihood of ratification by 1988. For more detailed explanations of the data coding see online Appendix 2. For descriptive statistics see online Appendix 3.

¹⁰ Koremenos (2016) codes IO autonomy, but covers only part of the treaties in this study.

¹¹ UNITY_GOVERNMENT is dropped because it is a perfect predictor of SEC.SECURITIZATION (no such acts were documented under a unity government).

¹² The residual values are negative under non-zero predicted values and SEC.SECURITIZATION = 0, reflecting a potential for resistance to objective concerns against ratification.

¹³ While coefficient of SEC.SECURITIZATION in Regression (2) suggests a 46% (1–0.54) decrease in likelihood, Figure 2 compounds the effects of other variables.

¹⁴ The documents are all in Hebrew. The translation provided below is ours.