Gallup: McCain Breaks A Long Losing Streak, Barely

What a twist! Last week Obama dominated the news and jumped to his biggest lead ever in Gallup’s tracking poll over the weekend. He also leads in a number of national polls released Monday. So who would expect that yesterday McCain would break a long losing streak and grab his first lead in a national poll since early May? (How long a losing streak McCain just ended is evident with a quick look at RCP’s list of all national polls.) Although yesterday's poll means relatively little, the Republican's campaign will no doubt seize upon it as an island of hope in a sea of troubles. It helps that the poll was conducted by Gallup, as reputable a name among American pollsters as any. (Note: The poll was conducted for USA Today and is not today’s edition of Gallup’s tracking poll.) Of course, we're talking about a 1% lead -- and that lead tallied only among likely voters:

• From a 6% deficit last month, McCain is now leading 49% to 45% among likely voters. Among registered voters, Obama leads 47% to 44% -- down from a 6% lead last month.
• The poll was taken over the week-end.

That was the only good polling news for McCain. Other national polls released this week find Obama clearly ahead, suggesting that the USA Today poll is the exception to the consistent but narrow Obama lead we have been seeing for weeks:

• First, Gallup’s tracking poll finds Obama leading 48% to 40%, a one point decline since yesterday but the biggest lead Obama has had in the tracking poll since March except for yesterday.

• Rasmussen’s tracking poll, on the other hand, finds Obama leading only 48% to 45%, down from 6% two days ago and 5% sunday.

• The fourth poll taken over the week-end is Research 2000’s first national poll: It finds Obama leading 51% to 39% with 3% to Bob Barr and 2% to Ralph Nader. The poll tested likely voters. An interesting finding is that the biggest regional swing in the past 4 years is the Midwest, where Obama now leads 53% to 37%, a 17% swing. Given the high concentration of battlegrounds in this region, Obama would surely win the election if he were to gain more in the Midwest than in the rest of the country.

• Two other national polls were released Monday but were taken early last week, making them a bit less valuable as the timeliness of surveys seems more important than usual. First, a Democracy Corps poll taken from the 21st to the 24th finds Obama leading 50% to 45%, 48% to 42% if Barr and Nader are included, a significant factor.

• A YouGov poll for The Economist finds Obama leading 41% to 38%. This one was taken from the 22nd to the 24th.

So what should we make of the USA Today/Gallup poll if five other polls today found Obama with a lead ranging from 3% to 12%? Gallup’s USA Today results are not what is interesting here: McCain’s lead is within the margin of error, and the lead it tracks is limited to the likely voter model. (Gallup’s editor says “registered voters are much more important at the moment,” though the poll does go against the conventional wisdom that Democrats are much more enthusiastic this year).

It is the psychological effect of a poll like this that matters: for McCain to come out with however slim a lead in a major national poll at the height of Obama’s media coverage must be a huge relief for Republicans, and it blunts the building narrative of Obama wildly successful international trip -- whatever other polls find in the coming days. The GOP desperately needs more energy if it is going to win the White House and the first step is for voters and strategists to begin believing the party can win.

Furthermore, that McCain is performing much stronger than he seems to have earned the right to be performing and that he is holding on in battleground states such as Ohio and Michigan despite slipping among base states such as Arizona has long been documented, including on this blog. This USA Today/Gallup’s poll feeds the storyline that Obama can not break through, despite incredibly strong campaigning weeks, like the one that just ended. That the numbers appear to be tightening once more in the tracking polls strengthen that narrative further.

Read more at Daniel Nichanian's blog Campaign Diaries.

Comments

[JXJASON]

Who cares about the polls...I don't. I am voting for Senator Obama.

McCain is only going to continue to make a fool of himself. If McCain BELIEVES he is winning then McCain will probable do more stupid things.

McCain is not going to win because the country knows that McCain is part of the problem.

Go to Moveon.org for your free OBAMA button.

OBAMA 08

[Thron]

The sound of a Gallup in a dishonest gallop. This is the Pappy sphere of lies and deceit disguised as a straight talker.
If you can make an election number says anything you dream of , why not a Gallup poll.

[research]

For all we know, Obama has 90% of the vote.

Land line phone polls paid for by the MSM are

BS.

The conservative MSM owners, want the polls to show it's close, so they can get away with stealing the election.

[SirReal1]

I've posted this comment so many times I'm simply going to save it in Word format so I can "cut & paste" it from now on.

Daniel, and any other "reporter" please take note:

POLLS are conducted with people who have "land-line" telephone numbers.

Several "reports" over the last several years indicate that a significant percentage of the 18 to 35 year old demographic have forgone the "land-line" and use their cell phone as their exclusive communications choice.

I can not say with certainty, but as of the last time I looked into it, VoiP numbers also did not show up on the pollsters "contact lists" due to the same "block numbering" system that is prevalent in cell phone systems. This may have changed, but having worked in Telecommunications for 20 years in the Army, I tend to stay up on these types of changes, and I've not seen any indication that any changes occurred.

All this information must INFORM the analysis of "Polling Data". The largest bloc of Obama supporters are likely NOT BEING CONTACTED. A significant bloc of likely Obama supporters (those who embrace "new technologies" are likely not being contacted.

Of the remaining "potential voters" (e.g. those who STILL have a land-line telephone for the pollsters to call) there is a near 50/50 split on who they will likely vote for.

[Drumbeato]

I find it hard to believe that when poll results are discussed, no one ever mentions the fact that there is a MAJOR group of voters that will have a big effect in November, that are not a part of these polls. This group is made up of voters under the age of 35, who are overwhelmingly for Obama. The reason they are not included in these polls is simple, most of them do not have land line phones, they use cell phones. The poll takers don't call people on their cell phones. This means a big chunk of Obama voters are not even a part of the polling results. Factor them in and I bet you Obama would have a commanding 10 to 15% lead.

[shades3]

When Obama is shown to be ahead in a poll by eight points, or more, the MSM brushes it off as too soon to be any indicator.
When McCain "surges" ahead, by ONE POINT, it's seen by those same people as a game changer.
What the.......????

[DoTheMath]

All semblance of REPORTING the news is gone. Now the MSM is just making an all-out attempt to CREATE/CONTROL the news.

[DRaymond]

The thing that is talked about even less than margin of error in polls is their confidence level. The 'industry standard' in calculating confidence level is 95%.

The technical definition of margin of error is that if the poll says 45-55 with a 3 percent margin of error and a 95% confidence factor then additional polls of the same population would fall from 42-58 to 48-52 ninety five percent of the time. The other five percent? Somewhere outside, and definitly bogus. But it also means that there is a 5% chance that YOUR poll was the one that wound up being the bogus results. Now think, how many presidential polls have there been lately? Enough to presume that at least one of them has fallen outside the confidence level? And how would we recognize that one?

[baznyc]

Whatever the poll, everything at present, points to a close race.
And that's not good news for Obama (or the country/world), as we know what the GOP can do in close races.
The most troubling aspect of these polls, to me, is that it illustrates the effect the Media is having, by refusing to scrutinize McCain.
No matter what he says, or does, he remains the Maverick, Hero, GOP-defying, Straight-shooting, Independent, Foreign Policy whiz.
And sadly, the Obama Campaign, looks more & more like the Kerry Campaign - - doing little to refute all the attacks & lies. All they seem to be doing, is sending out Press Releases... and we know how the Media handles anything negative about McCain - - ignored.

[MaybeMilo]

Good post!

[riley85]

Go John---I know it won't last but for one day it made me happy!

[DoTheMath]

This is great. Just keep doing these "likely voter" polls to convince Republicans they don't need a plan B. In the meantime, Obama's movement continues to register enthusiastic unlikely voters.

[SpoxLogic]

More reasons why the politicians and people should pay no attention to polls. They can be easily manipulated into what ever the pollsters want them to say. This isn't the first time this has been done this election season - recall a recent polls citing Obama's "white voters" problem. Turns out they left out key data that wiould've gone against the conclusion the pollsters were trying to portray.

And since these polls are taken every day, haven't they lost their relevance?

[shades3]

YES. Anyone who has been telephone-polled knows that the questions asked have no nuance and are designed to elicit a particular response.
The only real indicator of who supports which candidate is the November Presidential Election.
Having said that, the voting process should be of much more concern that the DAILY fluctuations of any poll.
Voters should be checking, NOW, whether they are on the voters' list and also where they are to go to vote.
The voting machines have NOT been changed, despite a lawsuit that DIEBOLD lost, in response to which they simply changed the company name to PREMIER ELECTIONS SYSTEMS.
They are the SAME machines, with the same hackability.
There are still three months to go. Time enough to make sure THIS TIME that the Rove chicanery is not going to decide the election again.