This paper challenges categorical typologies by analyzing historical data on dozens of autocratic features. Using latent dimension analysis, we identify three time-varying dimensions orthogonal to standard democracy measures.
How We Did It
We developed a novel method leveraging extensive historical datasets covering autocratic characteristics across multiple contexts and over time.
The Findings
Our research reveals distinct ideal types: party dominance, military rule, and personalism. This quantitative approach demonstrates that these dimensions are statistically orthogonal to conventional democracy scales.
What's New?
uThis paper uniquely captures the dynamic nature of autocratic features, particularly personalism, which previous typologies missed entirely.
Real-World Relevancen
Personalism emerges as a crucial predictor in conflict initiation and regime collapse scenarios.
