The short answer is, no, I can't think of one.

As to suitability for your Thesis, the information you provide seems to me, possibly correlated but not necessarily causal. If the data are not or cannot be demonstrated to be causal then you are in for a long walk in the sun.

It is difficult for me to fit those governments into one box called Democracy, as many of them are at odds with what you and I might think of as a democracy... so you got that definitional thing.(Maybe the "spectrum" of democratic characteristics might have some impact on the term of the insurgency?)

I would be leery of making the commitment to this topic until I had found some relationship that looks causal. The tough part about a Thesis (IMHO) is picking a subject that you can run with and use that yet allows a narrow enough focus to allow a clear and succinct examination of your hypothesis. This is what you need for a clean product.

I think that to make this work, you'll need to tunnel deeper and discover something that might (or might not) be demonstrable as a causal factor. If it was my assignment and I could not identify that prospective relationship, I'd shift targets.