Last night, I reported that Marquette University had overturned an entire year’s worth of its polling and changed the partisan sample for its latest poll from D+2-4 to D+11. Its a massive change in a partisan sample from one month. In the poll before the switch in the sample, Obama was leading Romney by just 3 points. After the sample switch, Obama’s lead grew to 14. I called BS on the poll. But, I was wrong. I thought the poll was designed to juice support for Obama. Now, I think, there was another motive at work.
The Marquette University poll is, historically a good poll with a great pulse on voters in Wisconsin. It was very accurate on the Scott Walker recall election just a couple months ago. Right before that election, Marquette produced a poll with a D+5 sample that just about nailed the final results of the recall. Throughout the entire year they have used a similar sample, never exceeding the D+6 sample that reflected the WI electorate in 2008. All very reasonable.
So, I was genuinely surprised when their latest poll used a sample that was D+11. That strains the limits of credibility. There is simply no way that, in 2012, the electorate is going to become more Democrat than it was in 2008. If such a Democrat wave were building, we would see it in donations, rally attendance, yard signs, etc. It simply isn’t happening.
But, there may be something weirder going on. An alert reader–and polling expert whose name I won’t divulge–drew my attention to several problems with the poll. This isn’t the place to go into all his concerns, save one: This poll had a lot of questions about global warming.
No other poll conducted this year by Marquette University touched on the issue of global warming. It isn’t even something that is highly relevant to the politics or policies of the day. Yet, this poll asked half-a-dozen questions about global warming. That is, shall we say, odd.
Unsurprisingly, given the massive Democrat skew in this poll, the respondents sided with the alarmists on global warming. But why, with seven weeks to go in the presidential campaign, did the poll decide to divert to assess attitudes on global warming?
Here are a few of the questions I have for Marquette University. Did someone who supports government action on global warming sponsor this poll? Why the sudden need, when nothing on the issue is relevant, did you feel the need to ask these questions? Worse, did you adjust the sample to ensure that the results would match the viewpoints of the donors to the poll?
Marquette polling has been very consistent–and accurate–throughout the year. Out of the blue, they ask questions about global warming and make a huge shift in their partisan sample. A shift that. in itself, would be a major political story. A shift, also, by the way, that ensures the answers to their questions on global warming would fall in line with the views of someone who would sponsor such a poll.
He who pays the piper calls the tune.