Good points. As I see it, the trouble arises when advocates on either side of the issue use poll results arising from a loaded question to predict actual results. For example, Hunter Bacot of Elon argues here the poll results indicate North Carolina may be more moderate on the issue. In reality, however, pollsters would probably find similar results in states that have already approved marriage amendments if they framed the poll question in a rights-denying way.
The same could be true for other issues. A poll surveying voter attitudes on lowering taxes and cutting government programs is a good illustration. The results would be vastly different were the question phrased “do you support or oppose lowering income taxes?” versus “do you support or oppose cutting government programs for needy families?”
It all depends on how the issue is framed. That’s why it’s risky to put too much stock in polls. The sad part is that many readers catch the headline but don’t follow through to learn the poll’s methodology, sample group, and wording. But that’s a separate problem.
Another issue is how the Elon poll singles out same-sex marriage rights. That issue has the most political relevance right now, obviously, but what about other minority living arrangements? If a poll question must be phrased in the negative, a better option would be to ask respondents whether they would support an amendment that would prohibit all relationships except those between a man and a woman from being recognized as a civil marriage.
Not that Elon is likely to take my advice …
