Spatial and temporal interdependence in binary outcomes is common in social sciences, but often overlooked in empirical models. This study introduces new methods to account for these interdependencies, using simulation-based approaches like recursive-importance-sampling and Bayesian Markov-chain Monte-Carlo. By comparing different estimation strategies, the researchers show that these new methods are more effective in capturing spatial effects and temporal dependence in binary outcome models. The study applies these methods to analyze U.S. states' adoption of legislative term-limits and great-power decisions in World War I.