Preventing a War of All Against All: Geographic Sorting in Open-List Proportional Representation Electoral Systems

In open-list proportional representation (OLPR) systems, candidates must obtain personal votes to succeed. A general expectation about these systems is that these rules induce intraparty competition and, ultimately, lead to weak political parties. Recent work has challenged this view, suggesting that intraparty competition in OLPR, even under very permissive conditions, is in fact constrained.

In this paper, Cheibub and colleagues use data from elections for the Brazilian Chamber of Deputies at the precinct level to map the distribution of votes among competitive candidates. They create a measure of spatial overlap, and show that being from the same party reduces the degree of competition between any two pairs of candidates. Our models control for attributes of individual and pairs of candidates, parties, and polling stations.

José Antonio Cheibub (Ph.D. University of Chicago) is the Andrew W. Mellon Professor of Political Science at the University of Pittsburgh. He joined Pitt in 2023 and has previously taught at the University of Pennsylvania, Yale University, the University of Illinois at Champaign-Urbana, where he was the Boeschenstein Professor in Public Policy and Political Economy, and Texas A&M University, where he was the Mary Thomas Marshall Professor in Liberal Arts. His research focuses on democratisation, the emergence and effects of specific democratic institutions, and political economy. Currently, he is engaged in projects pertaining to electoral strategies under open-list proportional representation systems, electoral personalism, political instability in Latin America, caretaker administrations in parliamentary democracies, and the effect of anti-majoritarian institutions on economic outcomes.