Referendums and Peace Processes in South Africa and Cyprus

Abstract

While conflict resolution literature has dealt extensively with failures in peace processes, few studies have addressed contentious voting in negotiated settlements more specifically peace referendums. The article builds on Putnam’s two-level game theory model to address this gap. While questioning the broader efficacy of peace referendums, it also demonstrates how the latter might be frequently necessitated by constitutional or normative constraints. Recognizing that popular vote could be often unavoidable or even the preferred option of local actors, the article presents alternative designs and timing options aiming to minimize the negative downsides of contentious peace voting. Drawing evidence across divided societies and particularly the contrasting cases of South Africa and Cyprus, the article argues that early mandate referendums, which focus initially on domestic constituencies, enable leaders to finalize an international peace settlement at a later time. Mandate referendums have three important advantages: they incorporate early feedback into the negotiations and prepare the public for a settlement; they reduce the uncertainties of post-settlement referendums while trading credible commitments with regards to implementation for favorable concessions; and finally, they help safeguard the peace process from subsequent and often unavoidable anti-settlement grassroots mobilizations. The study of mandate referendums has important implications for research on democratization and peace-building since it suggests mechanisms through which leaders could minimize failed ratifications of desirable and necessary treaties, including not only peace settlements but also international agreements on global warming, humanitarian intervention, and debt-crisis recovery.