Effect of Electoral Seat Bias on Political Polarization: A Computational Perspective

Research on the causes of political polarization points towards multiple drivers of the problem, from social and psychological to economic and technological. However, political institutions stand out, because – while capable of exacerbating or alleviating polarization – they can be re-engineered more readily than others. Accordingly, we analyze one class of such institutions – electoral systems – investigating whether the large-party seat bias found in many common systems (particularly plurality and Jefferson-D’Hondt) exacerbates polarization. Cross-national empirical data being relatively sparse and heavily confounded, we use computational methods: an agent-based Monte Carlo simulation. We model voter behavior over multiple electoral cycles, building upon the classic spatial model, but incorporating other known voter behavior patterns, such as the bandwagon effect, strategic voting, preference updating, retrospective voting, and the thermostatic effect. We confirm our hypothesis that electoral systems with a stronger large-party bias exhibit significantly higher polarization, as measured by the Mehlhaff index.

💡 Research Summary

The paper investigates whether electoral seat bias—the systematic advantage given to larger parties in translating votes into seats—exorts a causal influence on political polarization. The authors argue that large‑party bias creates two reinforcing mechanisms: a “threshold effect” that discourages support for small or niche parties, and a “concentration effect” that makes votes for large parties more “valuable” in terms of policy influence. Both mechanisms should, ceteris paribus, push the electorate toward a more polarized equilibrium. Because cross‑national empirical work is hampered by endogeneity and confounding variables, the study adopts a computational, agent‑based Monte Carlo approach.

The model builds on the classic spatial voting framework, placing voters and parties in a two‑dimensional policy space. It augments the basic distance‑based utility with five empirically documented voter behaviors: bandwagon effect, strategic voting, retrospective voting, preference updating, and the thermostatic effect. Electoral rules are parameterized by a seat‑bias coefficient, allowing the simulation of systems ranging from highly proportional (low bias) to strongly majoritarian (high bias) such as plurality or Jefferson‑D’Hondt. Over many simulated election cycles, the authors compute the Mehlhaff index—a measure of ideological dispersion—to capture polarization levels.

Results consistently show that higher seat bias yields higher Mehlhaff scores. In low‑bias settings polarization remains modest, while as bias increases the index rises almost linearly, with a noticeable acceleration once bias passes a certain threshold. The authors interpret this pattern as evidence that strategic voter coordination on large parties, combined with the systematic under‑representation of small parties, amplifies both ideological distance and affective hostility among the electorate.

Methodologically, the study’s strength lies in its clean experimental design: all non‑institutional parameters are held constant while only the seat‑bias variable is manipulated, thereby isolating its causal impact. This circumvents the typical identification problems of observational comparative research. However, the model assumes homogeneous voter decision rules and a spherically symmetric policy space, which may oversimplify real‑world heterogeneity. Moreover, reliance on a single polarization metric (Mehlhaff index) limits the robustness of the findings, as it conflates ideological and affective dimensions.

The paper concludes that electoral engineering—specifically reducing large‑party seat bias through more proportional formulas or compensatory mechanisms for small parties—offers a viable pathway to mitigate political polarization. Future work should incorporate network‑based opinion dynamics, richer multidimensional issue spaces, and alternative polarization measures to test the external validity of the simulation outcomes.