I have always found the Caribbean fascinating because it is a microcosm of the world, with tiny countries trying to get a workable government without the economies of scale the rest of the world enjoys. But as a readers of Economics research, the presence of this myriad of too-small countries lead to many frustrations, as they bias results in cross-country regressions. But sometimes, these micro-countries can be useful for research.
Roland Craigwell and Alain Maurin use them to study whether there is convergence in the Caribbean. It is well established that there is no convergence on world-wide country data, but it is very visible on subsets, such as US states. In the later case, all US states are under the same currency and roughly the same laws and government systems, there is some cross-state redistribution and no trade barriers. As you drop these features, which one gets you lack of convergence. In the case of the Caribbean, there is a partial monetary and trade union, laws and governments are more dissimilar than in the US and there is no redistribution. And as Craigwell and Maurin show, that is sufficient to make convergence disappear. Once more, it looks like once more institutions and to a lesser extent globalization are the keys to development for the poorest economies.