I do know that polls can be skewed in any desired direction depending on how questions are phrased.
Would you prefer to:
a. pet a cute and cuddly cat or
b. pet a smelly and dirty homeless person?
Answer - a. Therefore more people care more about animals than they do about people.
Or something like that...
In the RealClearPolitics website, I see four poll takers. They can’t all be making the same mistake in their questions.