I’m not super convinced. Their non-empirical part seems tautological, and their empirical part seems like they kind of tried to separate between the subgroups of democracy and “junta leader of the month” by calling one “strong executive constraints” and the other “weak executive constraints,” but it’s not obvious that you’d expect this method of separation to work (though not that when they did split the data this way, the “democracy-esque” subgroup had the best results). Not only are natural experiments thin on the ground here, but it’s very hard to control for confounders, and I see no reason to expect they succeeded.
I’m not super convinced. Their non-empirical part seems tautological, and their empirical part seems like they kind of tried to separate between the subgroups of democracy and “junta leader of the month” by calling one “strong executive constraints” and the other “weak executive constraints,” but it’s not obvious that you’d expect this method of separation to work (though not that when they did split the data this way, the “democracy-esque” subgroup had the best results). Not only are natural experiments thin on the ground here, but it’s very hard to control for confounders, and I see no reason to expect they succeeded.
Also, they cited Thomas Paine as from 1976.