WaPo/ABC poll: Check the sample

posted at 9:45 am on September 24, 2008 by Ed Morrissey

Two questions arise from the Washington Post/ABC poll.  Since they find an unusually high gap — an outlier from all other polling — what kind of sample produced it?  And does this really show a shift in likely voters to Barack Obama, when state-by-state polling shows the race tightening in the other direction?  The answer is that a poor sample still could show some movement:

More voters trust Obama to deal with the economy, and he currently has a big edge as the candidate who is more in tune with the economic problems Americans now face. He also has a double-digit advantage on handling the current problems on Wall Street, and as a result, there has been a rise in his overall support. The poll found that, among likely voters, Obama now leads McCain by 52 percent to 43 percent. Two weeks ago, in the days immediately following the Republican National Convention, the race was essentially even, with McCain at 49 percent and Obama at 47 percent.

As a point of comparison, neither of the last two Democratic nominees — John F. Kerry in 2004 or Al Gore in 2000 — recorded support above 50 percent in a pre-election poll by the Post and ABC News.

Last week’s near-meltdown in the financial markets and the subsequent debate in Washington over a proposed government bailout of troubled financial institutions have made the economy even more important in the minds of voters. Fully 50 percent called the economy and jobs the single most important issue that will determine their vote, up from 37 percent two weeks ago. In contrast, just 9 percent cited the Iraq war as their most important issue, its lowest of the campaign.

As some have already noted, the sampling in this poll could explain why.  In the raw numbers, the poll sample consists of 38% Democrats and 28% Republicans.  This doubles the actual spread on party affiliation, last surveyed by the more-reliable Rasmussen at 38.7%-33.6% D-to-R. When the WaPo/ABC poll takes leaners into account, the spread gets even more pronounced: 54-38.  That suggests a rather strong bias towards Obama, and an almost insurmountable hurdle for McCain.

In contrast, the spread was smaller on 9/7, when McCain had a two-point edge.  The base number gave an eight-point advantage to the Dems, but only a nine-point advantage with leaners.  That’s a seven-point shift in two weeks within the sample, which would certainly account for a large shift towards Obama.

But does it account for all of the shift?  That’s a tougher question.  The seven-point shift in leaners should only result in a seven-point shift, maximum, in the end result — but McCain lost eleven points.  The difference could be attributable to the margins of error that are built into the polling, or it could just be that Obama did really gain a few points, although not eleven as the WaPo/ABC poll suggests.

We’ll have to watch other polling (and their samples) to determine whether this reflects reality or a pollster bias.  Rasmussen’s daily tracking poll has Obama up two points for the first time in a while, so there may be movement — but it still looks like anyone’s race.


Related Posts:

Breaking on Hot Air

Blowback

Note from Hot Air management: This section is for comments from Hot Air's community of registered readers. Please don't assume that Hot Air management agrees with or otherwise endorses any particular comment just because we let it stand. A reminder: Anyone who fails to comply with our terms of use may lose their posting privilege.

Trackbacks/Pings

Trackback URL

Comments

How can anyone believe the liberal candidate be the best person to fix the economy since it was his like liberal buddies who screwed it up in the first place. Flipping main stream media is failing…they are hiding the truth and it is simply sickening.

mindhacker on September 24, 2008 at 9:49 AM

Another example of the tail trying to wag the dog. Polls tyring to influence rather than accurately report.

TooTall on September 24, 2008 at 9:49 AM

They have already tried losing in Iraq, they have tried tanking the economy and even now, half the country stubbornly refuses to elect him. I shudder, what are they going to try next?

DEMOCRATS SLOGAN: DEFEAT ABROAD, DEPRESSION AT HOME

promachus on September 24, 2008 at 9:51 AM

I generally just watch Rasmussen and a few others. I say Obama is 2-3 points ahead atm. Can’t see that changing til after the debate.

lodge on September 24, 2008 at 9:52 AM

The distrubing thing is that ANYONE gives Obama credibility on handling the economic mess. Especially after the cr*pola the Democrats are pulling with the bailout bill in Congress.

A nation gone mad.

johnsteele on September 24, 2008 at 9:53 AM

That poll is some serious statistical noise. I know Republicans have lost ground in party identification, but a 16-point gap? That’s just irresponsible research.

Slublog on September 24, 2008 at 9:53 AM

How come YOU posted this poll Ed? now people don’t get to abuse alphapundit

neuquenguy on September 24, 2008 at 9:54 AM

The more I see of these polls, the more I believe that either a) McCain in a squeaker or b) McCain in a landslide (or at least enough of a margin to declare a mandate).

The media is rigged and for the most part the media does the polling. Believe nothing and just VOTE!!!!!!!

J.J. Sefton on September 24, 2008 at 9:55 AM

It’s a weekend poll as well. 9/19-9/22. So its further biased against republicans.

lorien1973 on September 24, 2008 at 9:56 AM

did I say “alphapundit”? I need my coffee

neuquenguy on September 24, 2008 at 9:56 AM

Run, run for the hills and hide! Game over, it’s game flajocking OVER, man!

Someone call Huck, we need to reorganize.

Bishop on September 24, 2008 at 9:57 AM

Rasmussen has it today at 2% for Obama, up 1% from yesterday.

And the other polls showed whatever edge Obama had as shrinking a bit. It seems to me that if we had a 16 point swing as indicated by this poll that that would have shown up somewhere in the other polls.

This is clearly an outlier although my guess is that the financial crisis is benefitting Obama more than McCain.

SteveMG on September 24, 2008 at 9:58 AM

I think the Charlie Gibson sample was high.

Abby Adams on September 24, 2008 at 10:01 AM

Polls merely reflect the MSM’s effectiveness in advancing the Dem agenda. When their candidate falls behind, they adjust the spin/attack machines accordingly.

Conservatives should never, ever talk to pollsters. Period. If we held to that principle uniformly, polls would become irrelevant.

Akzed on September 24, 2008 at 10:01 AM

It’s a scam. Pollsters are trying to influence the election.

SaintOlaf on September 24, 2008 at 10:02 AM

The answer is that a poor sample still could show some movement:

The “movement” is the crap running down the legs of these biased pollsters who overweight their polls like a sagging pair of diapers.

But who’s really surprised at these tactics?

Rovin on September 24, 2008 at 10:06 AM

It’s looking like the only state McCain will win is Alaska….and that’s looking like a squeaker too.

John Doe on September 24, 2008 at 10:06 AM

Speaking of wagging the dog… I think the yahoo Buzz thing is a bust for us conservatives. Everytime I go on there, the top buzz is always some anti-McCain garbage from the Huffington Post or the like. We know we are outnumbered on the net by “liberal loons” (to steal a phrase from the tshirt dad from yesterday). We need to fight back people!

RobertCSampson on September 24, 2008 at 10:08 AM

Who exactly are these polls asking anyway?

I sure haven’t heard from any of them, and I asked my friends, and no one has gotten asked a question by these polls either.

So where exactly are these polls getting their information from anyway?

pilamaye on September 24, 2008 at 10:09 AM

The McCain campaign needs to hammer on Obama’s ties to Fannie Mae through Raines etc. Screw the racism attacks. And then there is this just begging to be an ad….

Top Recipients of Fannie Mae and Freddie Mac
Campaign Contributions, 1989-2008

1. Dodd, Christopher J D-CT $133,900

2. Kerry, John D-MA $111,000

3. Obama, Barack D-IL $105,849

4. Clinton, Hillary D-NY $75,550

5. Kanjorski, Paul E D-PA $65,500

Hat Tip to RS at http://www.wildturkeys.net/phpBB3/viewtopic.php?f=12&t=17068&p=76739#p76739

Deanna on September 24, 2008 at 10:10 AM

Keep AP away from the knife drawer, Captain Ed.

andycanuck on September 24, 2008 at 10:10 AM

Wow! Even Drudge is in the tank for Obama.

Calm Before the Storm on September 24, 2008 at 10:13 AM

As i pointed out over at Ace’s, the Washington Post sample is more slanted than the sample used for the Daily Kos poll.

WaPo: 38% Dems and 28% Republicans (10 points for the Dems)

Daily Kos: 35% Dems and 26% Republicans (9 points for the Dems)

Check it out: http://www.dailykos.com/dailypoll/2008/09/24

DelD on September 24, 2008 at 10:14 AM

Calm Before the Storm on September 24, 2008 at 10:13 AM

How so? For 3 days, drudge has been linking to “democrats caused this” stories.

Today, Drudge links the NYT story, McCain’s denial, then points out that Obama got $100,000 from Fannie Mae.

lorien1973 on September 24, 2008 at 10:14 AM

Another example of the tail trying to wag the dog. Polls tyring to influence rather than accurately report.

TooTall on September 24, 2008 at 9:49 AM

Ding ding ding, we have a winner!!

thirteen28 on September 24, 2008 at 10:17 AM

Apparently, it is standard practice for polling firms to adjust their weighting of their results. Sampling may be another issue. But the likely more modest gains by Obama have more to do with a full court press by the MSM, a thuggish ground game by team Obama and a smothering of markets with Obama adds in a concerted and fairly co-ordinated disinformation offensive. The conventions proved that given an even shot, the Republican Message reverberates with the American people but those kind of openings will be few and far between. Now comes more financial scandal. One can only hope the debates will prove the difference but talk radio and conservative press may not be enough alone.

Nyog_of_the_Bog on September 24, 2008 at 10:18 AM

I know I shouldn’t be surprised, but the level of attempted press manipulation of this election is pretty amazing.

Slublog on September 24, 2008 at 10:18 AM

Polling results are incapable of predicting the impact of a higher than expected turnout among given groups of voters. When you consider this fact together with the obvious bias in this poll, why is anyone paying any attention to such polls? I believe that the Palin selection may result in a significantly higher turnout among conservatives and perhaps other groups that will be voting for real change during this election cycle. She is the first politician in my lifetime that actually represents the middle class. I think that her candidacy stands a good chance of motivating many in the middle class that have been used to staying home on election day. Because of this, the polls may all be out-to-lunch this year. Of course, if this happens and results in a McCain landslide, the Left will cry foul, say they were cheated and proclaim that McCain is an illegitimate President. They believe more in the polls than they do in the actual tabulation of ballots.

NuclearPhysicist on September 24, 2008 at 10:22 AM

Mr. Clemens, living in a more innocent time, missed it by one. Allow me to amend his list:

There are lies, damned lies and statistics…and then there are polls.

Satan is called the “father of lies”. He works for Zogby.

Puritan1648 on September 24, 2008 at 10:24 AM

This may seem like a dumb question, or an apocalyptic answer, but, if as another post today points out, more people are considering themselves democrats, wouldn’t the sampling show the same mix as well?

If we always had a 50/50 poll, wouldn’t we always get 50/50 results?

cntrlfrk on September 24, 2008 at 10:25 AM

I had my brother-in-law, a young business owner living in Las Vegas, on the phone and convinced him that McCain is a far better pick for him than Obama. He was a true undecided, so I did my political evangelism for the week.

Listen all, if we want to win this election, we need to understand that our vote DOES NOT COUNT unless we bring over to our side at least one undecided, and make sure that he/she votes. This is especially true for those of us in swing states.

Get out there and change some undecided minds.

Pope Linus on September 24, 2008 at 10:26 AM

The media is rigged and for the most part the media does the polling. Believe nothing and just VOTE!!!!!!!

J.J. Sefton

…sage advice…couldn’t’ve said it better myself….

…well, I could possibly have said it as well, but it would’ve taken four times as long….

Puritan1648 on September 24, 2008 at 10:29 AM

Wow! Even Drudge is in the tank for Obama.

Calm Before the Storm on September 24, 2008 at 10:13 AM

Drudge is about what ever he thinks brings the most traffic. Post convention for example he was all about Palinmania.

Nyog_of_the_Bog on September 24, 2008 at 10:29 AM

It may actually be advantageous to McCain that polls show Obama with a comfortable lead, because it could make democrats more lackadaisical on Election Day.

Logic on September 24, 2008 at 10:30 AM

This may seem like a dumb question, or an apocalyptic answer, but, if as another post today points out, more people are considering themselves democrats, wouldn’t the sampling show the same mix as well?

Most reputable polling firms weigh their sample based on the moving target of party identification. Both Rasmussen and Gallup sample more Democrats than Republicans, but by a much smaller gap. Most polls show a 3-4 point Democrat advantage on party ID.

Slublog on September 24, 2008 at 10:31 AM

Since the MSM won’t report the active participation of the Democrats in creating this crisis, it takes longer for the truth to spread… It will come back.

CC

CapedConservative on September 24, 2008 at 10:32 AM

I’m more worried about Rassmussen. They has him up +2 in a 3 day sample which was tied yesterday which means today’s sample was something around +5.

Aristotle on September 24, 2008 at 10:32 AM

Since the MSM won’t report the active participation of the Democrats in creating this crisis, it takes longer for the truth to spread… It will come back.

CC

CapedConservative on September 24, 2008 at 10:32 AM

It would also help if McCain told the truth instead of falsely blaming Bush and Wall Street. It’s his fault. Entirely.

Aristotle on September 24, 2008 at 10:34 AM

NuclearPhysicist on September 24, 2008 at 10:22 AM

Only a landslide by Mac and Palin will prevent the cries of diebold and lawsuits of another “stolen election”. The liberals just cant face the fact that this nation is not ready for their socialist/marxist policies that will destroy us all.

Rovin on September 24, 2008 at 10:35 AM

These polls would be more fair if they stopped calling people on the DNC list.

moonsbreath on September 24, 2008 at 10:39 AM

NuclearPhysicist on September 24, 2008 at 10:22 AM

Spot On!

Except I believe we can see the beginings of a race based arguement being made for Obama losing.

Stuck on stupid.

Marine_Bio on September 24, 2008 at 10:40 AM

It would also help if McCain told the truth instead of falsely blaming Bush and Wall Street. It’s his fault. Entirely.

Aristotle on September 24, 2008 at 10:34 AM

John McCain on the Senate Floor:

“I join as a cosponsor of the Federal Housing Enterprise Regulatory Reform Act of 2005, S. 190, to underscore my support for quick passage of GSE regulatory reform legislation. If Congress does not act, American taxpayers will continue to be exposed to the enormous risk that Fannie Mae and Freddie Mac pose to the housing market, the overall financial system, and the economy as a whole.

I urge my colleagues to support swift action on this GSE reform legislation.”

Any more “insights” Aristotal-Not

Rovin on September 24, 2008 at 10:44 AM

As some have already noted, the sampling in this poll could explain why. In the raw numbers, the poll sample consists of 38% Democrats and 28% Republicans. This doubles the actual spread on party affiliation, last surveyed by the more-reliable Rasmussen at 38.7%-33.6% D-to-R. When the WaPo/ABC poll takes leaners into account, the spread gets even more pronounced: 54-38. That suggests a rather strong bias towards Obama, and an almost insurmountable hurdle for McCain.

There are two ways you can take a poll.

1. You can take an honest sample (random picks without knowing the background of the people polled) and after they complete the poll, collect the democgraphics from the people polled. Then use the actual demographics of the U.S. population from a full population census census (as if these are valid) and weight the sample according to a specific demographic strata. (for instance choose party affiliation with three choices DEM, GOP, IND, or choose age groups (16-21, 22-32, 32-42 etc) , or choose income groups).

If the characteristic by which you determine weighting strata has few categories (DEM, GOP, IND) or are too complex (DEM + high income, DEM = low income, etc) the analysis gets foggier and foggier

Honestly document the entire weighting procedure including the test used to determine which strata the voter belongs to, and the numbers used to weight the sample. Report the calculated probabilities and document the formulas used to calculate the probabilities of the results

or

2. You can use leading questions to determine the demographics of the people polled and then decide if you want to let them take the poll on the pretense they do or do not fall into the strata. Then evaluate their answers and decide if you still want to keep their answers. Then if the answers in your culled group still don’t work for manipulating public opinion, throw out the hanging chads until you get the right count invent a criteria for margin of error

This is such a stinking fake poll. They had to used the special phone book to get a sample rigged to more DEMs than GOPs not counting the fact that a goodly percent of DEMs will not admit to being DEMs but call themselves INDs

The self described DEM is the left wing of the libs and the self described IND stealth DEMs are the soft wing

entagor on September 24, 2008 at 10:45 AM

Several points. First, McMaverick had a bad week flipping and flopping all over the place on the financial meltdown, sounding more like Huey Long than a thoughtful crisis manager. Second, the campaign has sent its best D.C. lawyers to Alaska to screw up the Tazergate investigation that could have easily been discarded as tainted by the leading Democrat’s earlier promise of an October surprise. Finally, all the polls are going to reflect a magnification of the Bradley effect because the Obama campaign and the MSM are playing the race card with a vengeance–”If the One loses, you are a racist nation!” How many folks on the other end of the pollster’s call suddenly took the easy way out? Relax, there is a long way to go here.

FalseProfit on September 24, 2008 at 10:45 AM

Rovin on September 24, 2008 at 10:35 AM

A McCain landslide would be great. It would create a maelstrom of weeping and wailing and gnashing of teeth among liberals that would be way too fun to watch.

However, a landslide by McCain with polls showing Obama ahead just before the election would undoubtedly result in unfounded claims of cheating by liberals that might result in angry reactions by the Obamatons.

NuclearPhysicist on September 24, 2008 at 10:49 AM

Rovin on September 24, 2008 at 10:44 AM

You only put emphasis on McCain’s utter populist stupidity to repeat lies instead of telling the truth. He knows exactly who caused the crisis, but yet refuses to tell it like it is order to be seen as bipartisan. But instead he looks like a clueless panderer – and the polls show it.

Aristotle on September 24, 2008 at 10:51 AM

Well if the intent of this poll was to discourage me. It worked.

petunia on September 24, 2008 at 10:54 AM

Someone call Huck, we need to reorganize.

Bishop on September 24, 2008 at 9:57 AM

I’m sorry but that is just uncalled for profanity!

petunia on September 24, 2008 at 11:00 AM

…that might result in angry reactions by the Obamatons.

NuclearPhysicist on September 24, 2008 at 10:49 AM

Would this be considered a pre-detonation or post-detonation? Either way, I can wait to see the implosion followed by a total meltdown of the Democrat party.

Rovin on September 24, 2008 at 11:07 AM

Is it too early for me to order my Obama coin?

GarandFan on September 24, 2008 at 11:29 AM

Jimmy Carter for a new century. Palin 2012.

Angry Dumbo on September 24, 2008 at 11:35 AM

As usual, you don’t need to see the actual poll results to know who is moving up in the polls. You just need to see who is complaining about the pollsters. ;-)

Jazz Shaw on September 24, 2008 at 11:51 AM

This is already’s yesterday’s news.

Just a thought – maybe it’s good to have polls skewed to Obama for now. Then the Dopler Effect (Red Shift) will be more surprising and effective in November.

connertown on September 24, 2008 at 11:53 AM

Splunge!

shick on September 24, 2008 at 11:54 AM

Why oh why dear Lord, year after year, do people think that the Democrats are the ones who can better handle the economy and to bring forth jobs to the American people?

Lack of common sense?

garry on September 24, 2008 at 11:56 AM

I’m curious. It’s my impression that McCain has owned the opinion survey questions on who would handle the War in Iraq better, owned meaning, at least, by an outside the MOE amount, i.e., McCain consistently leading by 4% or more. In fact, this poll alone reflects this, although narrower than the 4 I suggest.

I aske because in this WaPo/ABC poll, the numbers are reversed:

9. (ASKED OF REGISTERED VOTERS) Regardless of who you may support, who do you trust more to handle [ITEM] – (Obama) or (McCain)?
Obama McCain
c. The war in Iraq 49 45

If my impression is right, would this answer be an indicator that the random selection of who was called appears to be very skewed?

Dusty on September 24, 2008 at 12:22 PM

How can anyone believe the liberal candidate be the best person to fix the economy since it was his like liberal buddies who screwed it up in the first place. Flipping main stream media is failing…they are hiding the truth and it is simply sickening.

mindhacker on September 24, 2008 at 9:49 AM

Because your average voter has been told Bush is to blame for the current problems?

18-1 on September 24, 2008 at 12:29 PM

At least FOX has been running with the Ayers stuff, and the astroturfing stuff, and the e-mail hacking stuff, and the Biden stuff, and the NRA stuff, and the Democraps caused the entire stinking bailout stuff, AND thank God that Obama’s IDIOTIC helper-boy, Bill Burton just keeps showing up on FOX so Megyn Kelly can rip him a new one! That’s always a good time-

anniekc on September 24, 2008 at 12:58 PM

Just about every poll is from an organization that is either publicly, substantively, or quietly for Obama.

Work hard in the race, NEVER look back, and run like you’ve never run before, and the numbers will take care of themselves.

Gallup is a joge because they are overemphasizing with a subjective view of the hype factor and overplaying the party affiliation.

Add to that 1-2% Bradley effect possibly in some battleground states, and McCain is actually probably winning this thing slightly.

The economy and uncertainty of the bailout plan is what is helping Obama right now.

Sapwolf on September 24, 2008 at 1:14 PM

Well, if there are that many more Dems, why would they ever lose a national election?

blue13326 on September 24, 2008 at 1:20 PM

I will add that the more reliable Battleground Poll has McCain up by 2.

http://www.realclearpolitics.com/RCP_PDF/BG_2-way-ballot-trender.pdf

Mutnodjmet on September 24, 2008 at 1:28 PM