MyDD :: Poll Numbers

Poll Numbers

by Jerome Armstrong, Thu May 01, 2008 at 10:35:35 AM EST

I usually stay away from the tracking polls, but Rasmussen's numbers are just shocking:

Day Polled      Clinton    Obama

4/27            41         49       
4/28            41         49       (Wright on Press Club Q & A)
4/29            43         47       (Obama throws Wright under Bus)
4/30            46         44       (Yesterday)

That's a 10 point swing which includes 2 days of polling after Obama having held a press conference and spoke about Wright without defending him again, but instead, joined the average American in being outraged.

In NH, McCain is up by 3% over Clinton in NH, 47-44, and McCain leads Obama outside the MOE, by a 51-41 margin in NH. We picked up two seats in NH in 2006, and have a possible pickup in the US Senate there as well. If the GOP finds that they can win by linking candidates to Obama, in the special elections going on right now in LA and MS, they'll extend the template across the nation.

In NC, Mason-Dixon has polled the primary. They find Obama leading Clinton by a 49-42 margin.

87 percent of African Americans plan to vote for Obama, while 62 percent of whites said they will vote for Clinton. There has been very little evidence suggesting either candidate can cut into those numbers before Tuesday.

That's very polarized. Just to give you an indication of how far Obama has fallen with white Democratic voters. Compare this to the Virginia results, which matches up fairly well with NC. In the VA primary, Obama pulled in 90% of the black vote, and it looks like he'll reach a similar amount in NC. Among white voters in VA, Obama won by a 52-47 margin.

The 62 percent of white voters that Clinton has now reached in NC is higher than SUSA's poll (61%) that put the race at 5 percent, and IA's poll (54%), that showed Clinton ahead by 2 percent.

Clinton's problem in NC is similar to a problem that Obama showed in Ohio and Pennsylvania-- her highest poll numbers have been at 44 percent. She'll have to win over the undecideds in order to make a surprise showing in NC.

In IN, Rasmussen has their first poll out, showing Clinton ahead 46-41 over Obama. In the three polls from Indiana taken since 4/25, over the last week. Clinton leads Obama 49.3 - 42, for a 7 percent margin. The previous weeks poll of polls numbers showed Obama slightly ahead of Clinton 45.3 - 45.

In North Carolina, the race is polarizing, but Clinton appears to have a ceiling of support and would need undecideds to break her way again, and in Indiana, Clinton appears to pulling away.

Tags: 2008 election (all tags)

Permalink :: 199 Comments :: Post a Comment

Re: Poll Numbers (2.00 / 6)

People don't get it. Democrats have two choices, nominate Clinton and maybe beat McCain and have the White House or nominate Obama and lose.

Its sad but Jeremiah Wright sunk him and will go down in history as the "Willie Horton" moment of this campaign

by rossinatl on Thu May 01, 2008 at 11:17:30 AM EST

Re: Poll Numbers (2.00 / 1)

I completely disagree. That's a very biased angle on this situation. Both Democrats will have a huge advantage on McCain when the nominee is chosen.

I'll say this every time a poll comes out showing a pro-Obama result or a pro-Clinton result. It means next to nothing. The polling is inherently flawed. It doesn't account for cell phone only users, which is a growing demographic. I don't want to use this space to spell out the dozen or so other problems that exist, but the accuracy of daily polling is the most questionable.

In fact, getting all worked up over a weekly poll is silly. If multiple polls show a trend over multiple weeks it may be that there's a trend. In this case, there may be a story or there may not be a story. Treating those numbers as any kind of accurate predictor of reality is a bit dangerous.

by mikeplugh on Thu May 01, 2008 at 11:23:36 AM EST
[ Parent ]

Changes are informative (none / 0)

I agree with your points in general about the inherent biases in polls. However, the changes are unlikely to be biased becauase the biases should stay constant and cancel out when you compute the day-to-day change.

The fact Obama's poll numbers are dropping is telling us that the Wright controversy and his mis-handling of it are hurting.

Is this a temporary bump or a permanent effect?

That only time can tell.

by BigB on Thu May 01, 2008 at 11:43:51 AM EST
[ Parent ]

Or the changes (none / 0)

Are due to margin of error. Rasmussen doesn't call the same 1600 people everyday, as far as I'm aware.

by Covin on Thu May 01, 2008 at 01:08:13 PM EST
[ Parent ]

Actually Gallup specifies that they call cells too (none / 0)

"Results are based on telephone interviews with 6,117 registered voters, aged 18 and older, conducted April 21-27, 2008. For results based on the total sample of registered voters, one can say with 95% confidence that the maximum margin of sampling error is ±1 percentage point.

Interviews are conducted with respondents on land-line telephones (for respondents with a land-line telephone) and cellular phones (for respondents who are cell-phone only)."

....and they have her up over Obama against McCain by as much as 17 points, 14 point lead among Democrats generally ... and even a little lead now among the young too.

http://www.gallup.com/poll/106960/Clinto ns-vs-Obamas-Strengths-General-Election. aspx

by dotcommodity on Thu May 01, 2008 at 04:34:42 PM EST
[ Parent ]