Introduction Panel data contains measurements from multiple subjects measured over multiple time points. Such data can be encountered in many social science applications such as when analysing register data or cohort studies (for example). Often the aim is to perform causal inference based on such observational data (instead of randomized control trials). A new rOpensci-reviewed R package dynamite available on CRAN implements a new class of panel models called the Bayesian dynamic multivariate panel model (DMPM) which …