I think it is a bit unfair to ask for a specific implementation detail on policy which the general population overwhelmingly supports. You can always arbitrarily make a policy look less popular by taking only the subset of the general population that supports a specific implementation.
An example: Say 90% of the people of Madeupistan supports some sort of increased autonomy from the Roman empire (which just became a democracy after 2000 years of imperial rule). But a disingenuous pollster wants to make it look like the support isn’t that great. So instead they will point to the fact that only about 40% of the population supports independence. This is ignoring the fact that 30% of the population want something similar to home rule, and 20% want to be in a federation with neighboring states.
An example: Say 90% of the people of Madeupistan supports some sort of increased autonomy from the Roman empire (which just became a democracy after 2000 years of imperial rule). But a disingenuous pollster wants to make it look like the support isn’t that great. So instead they will point to the fact that only about 40% of the population supports independence. This is ignoring the fact that 30% of the population want something similar to home rule, and 20% want to be in a federation with neighboring states.