Abstract: This paper empirically evaluates four conjectures derived from Hermida-Rivera (2025), who argues that party discipline may make constitutional voting rules self-stable even when individual legislators would support changing them. We construct an original country-legislature dataset that records the voting thresholds used for ordinary legislation and constitutional amendments, the composition of legislative chambers, and party seat distributions. Using the conditions established in Hermida-Rivera’s Theorem 2, we classify constitutions at both the individual-legislator and political-party levels and examine whether these classifications are related to subsequent changes in constitutional voting rules. Specifically, we test whether (1) most constitutions are not self-stable when legislators are treated as independent decision-making units; (2) most are self-stable when political parties are treated as coordinated decision-making units; (3) party-level self-stability is associated with the preservation of constitutional voting rules; and (4) party-level non-self-stability is associated with their amendment. By translating formal stability conditions into comparable institutional measures, the paper provides a systematic empirical bridge between theories of constitutional self-stability, party discipline, and observed constitutional change.
Author: Myrto KASIOUMI, Héctor HERMIDA-RIVERA
Keywords: Constitutional self-stability, Political parties, Party discipline, Voting rules, Constitutional amendments, Voting power, Constitutional design