We intuitively understand that by asking bent questions, a pollster can get the result they want. A classic of the genre is asking people if lawbreakers should be punished to make things Americans overwhelmingly support (e.g. weed and abortions) poll lower by contextualizing that support as support for criminals. But Gallup set about demonstrating that even subtle differences in wording can make all the difference.
It concludes from a review of recent polls that about 6 in 10 Americans want "assault weapons" banned outright (an amazing 92% are in favor of more gun control), which simply gets us to the obvious situation we're in: a political establishment completely unresponsive to public opinion and driven instead by other, more perverse incentives.
But look how easy it is to flip that 6.
The overall results of these experiments confirm that asking Americans if they are "for or against a law which would make it illegal to sell or possess assault rifles" produces a lower level of support than asking them about such actions using a specific reference to a "ban" or by simplifying wording to ask directly about making ownership or manufacture of assault weapons illegal. It may be that the first wording is confusing to some respondents, particularly for those with less formal education. The use of more direct wording points to the conclusion evident from the preponderance of other research conducted in 2018 and 2019 — that a majority of Americans support a ban on assault weapons at this time.
If Gallup was on duty at the McDonalds window: "Do you think you should or should not order, procure or consume the shoestring-sliced oil-immersed tuber cultivars known as fries with that?"
