Can the author justify the accuracy of estimating revolt using a Bernoulli process? At minimum he should provide sufficient reason to believe that the random va

Can the author justify the accuracy of estimating revolt using a Bernoulli process? At minimum he should provide sufficient reason to believe that the random variables he is using as components of his Bernoulli process (which are 1 if a revolt or revolution occurs in a given year and 0 otherwise) are independent and identically distributed Bernoulli trials. These assumptions in the situation he is describing seem unlikely, or at best hard to establish. The author is being deceptive by not providing some sort of warning stating that there are good reasons to believe this is a bad model and the probabilities computed are inaccurate (no confidence intervals provided for instance).