It does not add up

I was looking at the poll results from the last 24 hours and find them puzzling to say the least. Let me explain:

1. CBS is out with polls of Wisconsin, Virginia and Colorado. Each shows a small lead for Obama over Romney. So far it all seems fine. Then I looked at the internals of the polls. In Virginia, Obama has a 4% lead which is the same as the last CBS poll in that state taken in early August. The problem is that since early August, the poll also reports that Romney has moved up his performance among men and among women. Among men his lead has moved from being ahead by 5% to being ahead by 6%. Among women in Virginia, Obama's lead among women has fallen from 14% a month ago to just 11% now. That's right, Romney is doing better among men and among women, but the overall results have not changed. What this means is that the poll this month obviously included many more women than a month ago. Indeed, if you do the math, there have to be a major swing in the number of women included and a big drop in the number of men. I think we can all agree that the percentage of women in Virginia's population did not change between August and September. Something is wrong with the poll.

2. Two pollsters have tracking polls of the big swing states. These include Colorado, Florida, Iowa, Michigan, Pennsylvania, North Carolina, Virginia, Wisconsin, Ohio, New Hampshire and Nevada for Rasmussen and the same states plus New Mexico for Gallup. Rasmussen tracks likely voters while Gallup tracks registered voters, so one would expect Obama to do about 2% better in the Gallup poll. The results today show Romney leading by 1% in the Rasmussen poll and Obama up by 2% in the Gallup poll. Those results are completely consistent. But here's the problem. According to the polls of the main stream media, Obama is leading in every one of the twelve states covered by these polls except for North Carolina where Romney is supposed to have a tiny lead. Again, according to the main stream media polls, Obama is ahead in Pennsylvania, New Mexico, Michigan and Wisconsin by about 8% or more. Obama is also supposed to be way ahead in Ohio according to the main stream media. That is impossible. You cannot average eleven or twelve states, most of which supposedly show big leads for Obama with one state with a small Romney lead and find the race tied or with Romney ahead. As they said at the Democrat Convention, you need to do the arithmetic.

And it is not just the polling numbers that don't make sense. The reporting is also slanted. Just this morning, I heard the CBS director of polling tell listeners about the results in Wisconsin. Wisconsin is a state that Obama carried by 14% in 2008. It was never expected to be a swing state in 2012; six months ago all of the pundits would have put it in the likely Obama column. Romney was given no chance of victory there. But this morning, CBS was reporting that Wisconsin is a state that Romney has always needed in order to win the election. So, it is not enough to come out with slanted polls, CBS is also reporting the results in a way that make it seem like Romney cannot win the election. It clearly looks like CBS is trying to demoralize Republicans so that they will not vote in November. After all, if they think it is a lost cause, the is no need to vote.