Rasmussen released a poll of the Kentucky Senate race that showed Mitch McConnell leading Alison Lundergan Grimes 52%-44%. Here is why the poll is another piece of propaganda that is designed to discourage Democrats from voting for Grimes.
The Rasmussen poll that has Mitch McConnell leading Alison Lundergan Grimes by eight points comes with more than a few statistical red flags that should give anyone who bothered to look some serious cause for concern. Rasmussen poll gave Mitch McConnell a positive job approval rating of a net (+2) 49%/47%. The same poll gave Grimes an unfavorable rating of a net (-4) 46%/50% For the last few years, Mitch McConnell has been one of the least popular senators in the country, so any poll that shows McConnell with a positive approval rating is clearly oversampling Republicans.
The Rasmussen poll also has Mitch McConnell getting 24% support from Democrats. For the sake of comparison, the insanely biased Fox News poll that had McConnell up by four points has the incumbent getting 12% Democratic support. It is clear that Rasmussen built McConnell’s eight-point advantage in their poll by doubling the Republican’s level of Democratic support.
To get more stories like this, subscribe to our newsletter The Daily.
In both 2010 and 2012, Rasmussen was at the bottom of 538’s pollster rankings because of their high margin of error, and Republican bias. Rasmussen has had a margin of error as high as six points, and they have maintained a four-point bias towards Republicans.
The problem with Rasmussen is that they don’t poll cell phone users. Instead of polling cell users, they use an online panel and landline phones to compile their data. The problem with not polling cell phone only users is that the poll leaves out younger voters who are more likely to support Democrats.
It is safe to conclude that the Rasmussen Reports poll of the Kentucky Senate race is another in a string of recent polls from Republican biased pollsters who are trying to discourage Democratic turnout while influencing the polling models. A close examination of the recent data leads to the suspicion that Republican pollsters have increased the level of bias in their polling in order to negate the built in bias subtraction of the polling models. For example, if 538 calculates a four point Republican bias, the pollster responds with a poll that contains a 6-8 point level of bias, so that their favored candidate still leads and the polling result influences the average of all polls.
The Rasmussen poll is a bad poll from a Republican pollster with a history of being inaccurate and wrong. The Senate election in Kentucky is going to be close. The deciding factor will be turnout and the impact of the Grimes ground game. Media members who write a story giving these Republican polls legitimacy are committing journalist malpractice. Members of the media don’t need to be polling experts, but it would be nice if the could tell the difference between a good poll and a bad poll. The Rasmussen poll is a very bad poll.
The numbers can be shaped to lie. Mitch McConnell is a lot closer to losing his seat than he is to leading by eight points.
Jason is the managing editor. He is also a White House Press Pool and a Congressional correspondent for PoliticusUSA. Jason has a Bachelor’s Degree in Political Science. His graduate work focused on public policy, with a specialization in social reform movements.
Awards and Professional Memberships
Member of the Society of Professional Journalists and The American Political Science Association
