<![CDATA[This article examines how religious affiliation shapes split-ticket voting in Indonesia by comparing the polarized 2019 election with the less polarized 2024 contest. Using individual-level survey data, we estimate logistic regression models. The main specification pools data across election years and includes interaction terms to assess cross-election differences. We complement this with separate models for each election year to facilitate contextual comparison. The analysis focuses on the effect of religious affiliation (Muslim versus non-Muslim) on two forms of cross-ballot behavior: split-party and split-coalition voting. The results indicate that religion was a strong predictor of voting behavior under polarization but weakened significantly in a less polarized context. In 2019, non-Muslim voters were less likely to split their ballots, reflecting consistent alignment across electoral arenas. By contrast, in 2024—when cross-cutting coalitions emerged—the effect of religion declined, no longer predicting split-coalition voting and only modestly influencing split-party behavior. At the same time, evaluative and candidate-centered considerations became more prominent, consistent with patterns of personalistic politics in weakly institutionalized party systems. These findings suggest that religion operates as a context-dependent cue: its constraining effect diminishes as electoral competition shifts from identity-based polarization toward more candidate-centered and personalized forms of voting. The Indonesian case thus demonstrates how variation in polarization conditions the role of identity in shaping electoral behavior.]]>
Copyrights © 2026
