There are so many potential biases involved, that I'm not sure I would use "hearing it on the news" as a statistical metic to drive my personal evaluation.

Obviously, if you're using it to set policy, some kind of study would be valuable.