If, as has been reported, the Democrat / Republican preference balance heading into the 2018 midterms now shows only a 3% margin for the Democrats, what the pollsters would discover if they were telepathic is how greatly distrusted they are by their polling targets.
It’s become generally known that political surveys, which are usually paid for by a partisan organization, will tend to over-sample those who share that partisanry. While this can be corrected for, it doesn’t happen often. Still, some pollsters make the effort to assemble a sample that reflects the overall balance of political allegiances in the U.S., and they deserve credit for that.
What pollsters can’t do is detoxify the information-collecting process itself.
A pollster who knocks on your door or call you on the phone already has enough information about you to make your life hell, if he so desires. Inasmuch as it is also generally known that the Left, in its political instantiation as the Democrat Party, is quite willing to harass, harangue, slander, and even assault persons known to be in the Right, the latter group is justified in distrusting anyone who asks for his political opinions, whether about party alignment, policy partialities, or candidate preferences.
There is no way to know how many Americans of conservative or libertarian inclinations keep those inclinations to themselves out of pure self-defense. Thus any poll that shows a margin for the Democrats should be assumed to be inaccurate in the Democrats’ favor.
However, there is a way to learn something akin to that. It consists in a rather different kind of “poll.” Let the “pollster” knock on random doors and, if the knock is answered, offer to buy any campaign buttons, signs, banners, or other memorabilia from previous campaigns. There’d be a lot more trustworthy information available that way, at least as regards prior political postures...and while it’s not a perfect predictor, political preferences do persist more often than not, even through the generations.
Food for thought.
