How large was your prior for "insurance helps some and harms others, and we should try to figure out which is which" before that was one possible way of rescuing insurance from this study? That sort of argument is, I respectfully suggest, a warning signal which should make you consider whether your bottom line is already written.