Since it's not from subject matter experts, expect things like sampling bias and administrative reporting issues to be completely neglected. That is to say, the uncertainty estimate is simply based on observed variance.