Pew Research had an interesting piece on poll accuracy a while back. One of the main drivers of what I call 'bad info' is the steady decline of people who will participate in the poll. Busy people do not participate...which brings up all kinds of questions...are they working people? Do those who do participate fit in a category of stay at homes, retirees, soap opera day tv watchers (maybe on the govt dole)? One thing researchers noted, those who willingly participate are community spirited volunteers, raising a whole other set of questions. What demographics have a landline anymore? For pollsters to access cell phones, the numbers, by law, must be manually dialed....so cell polling is time consuming and expensive.
If I were to have to condense the whole article, it sounds like polling is inaccurate and getting worse every year.
