“Rasmussen Reports uses randomly-selected, computer-generated calls, often referred to as robo-calls,” said McMaster pollster Richard Quinn in a campaign memo. “We’ve seen in past elections that Rasmussen poll numbers are less reliable than ones conducted using the traditional method of live callers speaking one-on-one to registered voters with a history Republican primary participation.”
At his blog, Moulitsas authored a post titled “Rasmussen’s (dishonest) game,” accusing the pollster of caring “only about setting the narrative that Democrats are doomed” by eschewing polls on this week’s races for polls on races that showed Republicans romping, notably by missing out on the Democratic victory in PA-12.
Scott Rasmussen, the firm’s president, laughed off the criticism. “I assume you recall that we were first to show Sestak catching Specter and our final showed Sestak by 5,” he said. “On PA-12, we never poll House districts. We focus on statewide polls. We cover every senate and gov race in the country. The more competitive ones get more coverage.” And when I’ve spoken to Rasmussen in the past, he’s said that the firm decides what issues and races to cover the way a newspaper decides what to cover — what’s going to be newsy, what will set the narrative.
Join the conversation as a VIP Member