A Bayesian spatial voting model to characterize the legislative behavior of the Colombian Senate 2010-2014

Area of Science:

• Political Science
• Computational Social Science
• Econometrics

Background:

• Understanding legislative behavior is crucial for analyzing political dynamics.
• Colombian Senate operates within a multi-party system, presenting unique challenges for analysis.
• Previous studies often lack robust methods for individual preference retrieval from roll-call data.

Purpose of the Study:

• To characterize the voting behavior of legislators in the Colombian Senate between 2010 and 2014.
• To retrieve individual political preferences from roll-call voting records.
• To analyze the latent trait underlying voting decisions and map legislators' positions in political space.

Main Methods:

• Implementation of a one-dimensional standard Bayesian ideal point estimator.
• Utilization of Markov chain Monte Carlo (MCMC) algorithms for estimation.
• Application to nominal voting data from the Colombian Senate.

Main Results:

• Successful retrieval of individual legislators' ideal points, reflecting their political preferences.
• Identification of the primary dimension structuring legislative votes.
• Characterization of legislators' positions within the Colombian political spectrum.

Conclusions:

• The study provides a robust method for analyzing legislative behavior in unbalanced parliaments.
• The findings offer insights into the political dynamics and individual preferences within the Colombian Senate.
• Methodological contributions enhance the analysis of nominal voting data in multi-party systems.