Most surveys are propaganda: a sample size of just a thousand people allows to get any result in any area. I could get 85% respondents to support ktulhu as the official deity by handpicking the right respondents. Even if I really wanted to know how much support ktulhu really has, I'd be very troubled because accountants in Idaho are going to view ktulhu differently than mechanics in Florida. For this reason, when I see "N% support X", I read it as "the media agency wants me to think that X has N% support" and I wonder why. Most of the time the answer isn't terribly complex: it's the party affiliation of the agency.
Assuming it's a properly conducted and accurately reported survey, I would say no, except in the very postmodern sense that everything is "propaganda" of a sort, since what is chosen to report and how to report it is determined by the media, and is at least significantly influenced by those who own and control the media.