Every two years more and more people ask me whether the polls aren’t inaccurate because they don’t contact people who only use cell phones. And each year I’ve said, in theory, yes, but in practice it just hasn’t been the case. Until this year. This was the year when the cell phone issue finally came of evidence. And Pew has a new study out showing that yes, polls that only surveyed landlines were significantly and consistently more friendly to the GOP.