Party Discipline, Representation, and the Condorcet Jury Theorem

How does district size affect the quality of political representation? We examine a mechanism derived from the Condorcet Jury Theorem (CJT), which predicts that larger legislative delegations more accurately reflect constituents' preferences when individual legislators vote independently. However, nominal seat counts often overstate the number of independent decision-makers, because party discipline causes legislators’ votes to be correlated. We introduce effective district size as a measure that more accurately captures the number of independent politicians within a constituency. Using 263 Swiss referenda (1992-2024) matched to parliamentary decisions, we provide evidence in line with the CJT: the probability that a majority of a delegation reflects its constituents' preferences increases with (effective) district size. Crucially, the CJT aggregation mechanism accurately predicts delegation-constituency congruence only for effective, not nominal, district size. These results provide empirical evidence that the CJT applies to representative democracies and highlight how party discipline constrains the representative benefits of larger district sizes. JEL Classification: D72 , K16 , H11