Revise and Resubmit at Political Analysis
This dataset addresses significant methodological challenges in measuring regime types and democratic transitions by developing a unified framework that integrates information from approximately 40 regime indicators. The approach integrates the classification of regime types and identification of regime transitions into a single model, directly capturing short and long-term temporal dynamics in a latent variable framework. Rather than imposing theoretical priors on how regime indicators should relate to underlying regime types, the methodology empirically estimates these relationships and reveals non-linearities that purely deductive aggregation frameworks might miss. Instead of relying on predetermined cut-offs to identify transitions, the approach probabilistically identifies transitions and estimates the most likely points in time for their occurrence. The methodology accounts for measurement uncertainty and efficiently handles missingness, making it possible to adjust inferences accordingly when these measures are used in empirical analyses. The resulting dataset provides estimated regime probabilities, transition probabilities, and uncertainty estimates for countries worldwide from 1800 to 2017. The data includes regime classifications and probabilities, transition timing estimates, and multi-year transition probabilities that enable analysis of gradual regime change and democratic erosion processes.
This dataset addresses significant methodological challenges in measuring regime types and democratic transitions by developing a unified framework that integrates information from approximately 40 regime indicators. The approach integrates the classification of regime types and identification of regime transitions into a single model, directly capturing short and long-term temporal dynamics in a latent variable framework. Rather than imposing theoretical priors on how regime indicators should relate to underlying regime types, the methodology empirically estimates these relationships and reveals non-linearities that purely deductive aggregation frameworks might miss. Instead of relying on predetermined cut-offs to identify transitions, the approach probabilistically identifies transitions and estimates the most likely points in time for their occurrence. The methodology accounts for measurement uncertainty and efficiently handles missingness, making it possible to adjust inferences accordingly when these measures are used in empirical analyses. The resulting dataset provides estimated regime probabilities, transition probabilities, and uncertainty estimates for countries worldwide from 1800 to 2017. The data includes regime classifications and probabilities, transition timing estimates, and multi-year transition probabilities that enable analysis of gradual regime change and democratic erosion processes.