This is rather odd... more than once I've had a random event where the good choice has a smaller penalty for chosing it than the neutral option.
(And I don't just mean a different type of penalty that isn't as important (like money vs research), it's the same type, just a smaller negative number.)