Ideological positions in the chilean chamber of deputies (2002–2026): A legislative roll-call dataset

This dataset provides ideological position estimates for members of the Chilean Chamber of Deputies across six consecutive legislative periods, spanning from 2002 to 2026 in four-year intervals. The estimates are based exclusively on roll-call votes related to bill proposals debated and voted on in the plenary sessions. It excludes other types of votes, such as those on non-binding agreements or investigative commissions, which do not directly contribute to lawmaking. To estimate ideological positions, the dataset includes results from three distinct methods. The first uses the classic DW-NOMINATE model, primarily for comparative purposes. The second applies a Bayesian model to estimate latent ideological positions within each legislative period. The third introduces a dynamic model suitable for large scale data that links estimates across time, allowing direct comparisons of ideological positions even when legislators served in different periods. In addition to ideological estimates, the dataset includes information on each deputy's party affiliations throughout their time in office, as well as their age and date of birth. A major electoral reform enacted during the middle of this time frame increased the number of deputies from 120 to 155. As a result, earlier periods contain fewer legislators than later ones-a structural change that enhances the dataset's value for studying institutional evolution and its effects on legislative dynamics. Each methodological approach offers a different lens to examine ideological variation and change within the Chilean legislature. The dynamic model, in particular, supports longitudinal research by enabling consistent ideological comparisons across more than two decades. Researchers can use this dataset to explore topics such as political polarization, coalition dynamics, party-system development, and the relationship between institutional reforms and representation. The structure and transparency of the dataset make it suitable for integration with other legislative or electoral databases, offering additional opportunities for analysis in political science, public policy, and institutional studies.