On the Latest Gallup Poll - Independents Prefer McCain
The latest Gallup Poll has Barack Obama up by 9%. It is for the sixth day in a row that he has such a sizable lead.
Instead of taking the numbers for granted I decided to have a look at the breakdown by party affiliation and ideology. The results were highly interesting, leading me to wonder just how it is possible that Gallup gave Obama such a tremendous lead.
Here is the breakdown:
Note the difference, especially, with “Pure Independents.” Unlike what the media have been telling us for weeks, Obama is not leading among this group of voters.
I cannot help but wonder what sample Gallup uses. It seems to me that it has to use many more Democrats than Independents and Republicans in order to show an Obama lead of 9%, no? Or am I missing something here?
Not trying to play the conspiracy theorist here, but something seems off here. No?
This seems to confirm the suspicion I voiced yesterday (and yeah, I too am a bit concerned about seeming like I’ve donned the tin foil hat but…)
It does seem that the big jump in Obama’s lead coincided with the voter registration dump by ACORN and other groups right at the deadline for registration. We know that this must have increased Democratic voter registration greatly especially in the swing states (which is where the increase in Obama’s poll standings suddenly became very striking.)
Since the polls are weighted by party affiliation, having a lot more Democrats suddenly registered (you know, like Democrats Mickey Mouse, Donald Duck, etc.) would lead the polls to skew heavily to the Democratic candidate.
I haven’t had a chance to parse the numbers to see if the timing and the effect really coincide- and I’m not claiming that all of Obama’s lead is due to this of course (the shift coincided initially with the economic crisis of course.) But I wonder if part of the ACORN goal was to game the polling and create artificial momentum for Obama.
From the Gallup site FAQ on Daily tracking polls:
There is no mention that there is any weighing by political party affiliation. This in any event would be exceedingly complicated, as the proportion of one to another would vary depending on where a person lived. There seems to be every indication that people were weighted by the above criteria and then asked to self-identify affiliation. Given the categories I would bet that the wording wasn’t "how are you registered?" so much as "with which of the following to you feel most identified?"
In such a case then I would suppose that the number of people who are self-identifying as Democrats of one sort or another has gone up, or maybe even that people calling themselves Republican previously have decided to self-identify as Independents. Either outcome would remove weight from McCain’s camp.
The idea that more people are identifying as Democrats is also supported by how the list is constructed. Notice how Democrats come in a total of three categories; liberal, moderate and conservative, while in the Republican side liberal and moderate are put together in one single category. I would expect that would be because there aren’t enough people to make it two categories. Of course "conservative Republican" could simply be huge to compensate.
That’s interesting Claudia and I guess that answers my question- but I do recall that there was controversy over whether or not polls should weight for party ID during the ‘04 and ‘06 cycles. I bookmarked this:
http://www.depts.ttu.edu/hs/hdfs3390/weighting.htm
but haven’t had a chance to read it all yet, but I think they’re discussing the pros and cons of doing it that way, so I do think it’s sometimes done.
I wonder if Claudia might be right on the "who do you most identify with" question. I wonder if most people, particularly at this point in the election season, could call themselves ‘pure independents.’ There may be some, to be sure, but most have probably started to pick sides.
This poll might have been better to add "Democrat-leaning Independent" "Republican-leaning Independent"
This actually coincides with a post I made today. I think there are huge reservations about both tickets, to tell you the truth, and I also think the independents are still very much up for grabs.
It looks like McCain’s key loss has been in "Reagan" Democrats and moderate Republicans
Instead of conspiratorial conjectures, why don’t you just go to the gallup web site and ask? http://www.gallup.com/ContactUs/Default.aspx For my two cents, no plausible rate of voter registration could account for a 5 point shift in 3 weeks. Get real. The answer is sitting on your nose. We’ve had a +20% slide in home values in a year and a +30% drop in our 401(k)s in a month, with no end in sight. Only those in deep denial could not see it’s due to Republican deregulation of Wall St. chicanery. All the hyper-leveraged banks are imploding. So no loans until the global credit bubble deflates. Deep recession looms. We have no equity left in our McMansions, monster trucks and giant SUVs, and they are unsellable. If you expect us to vote Republican again, you are beyond clueless.
Civics lesson Jillian
Thinking individuals would also see that the Democrat lead Congress has been closer to zero approval ratings than any other Congress in history.
Perhaps a little lesson on the powers of Congress.
And lets not forget.
The ball effectively starts rolling down the mountain beginning with Congress. And this Congress has it’s collective heads where the sun doesn’t shine. If you really intend to be serious about it, vote current Congress out.
In my opinion, whenever anyone claims that blame lies only with Republicans or Democrats, it signals either gross ignorance or gross partisan dishonesty. When people like Jillian above spew their partisan talking points as truth, it seems to me outright dangerous. Ignorance fueled by partisanship is a barrier to effective diagnosis and reform that is essential to contain the problem.
The truth is that nearly everyone owns a piece of this financial meltdown:
1) Democrats (led by Barney Frank) mandated in the Community Redevelopment Act that banks give subprime mortgages to address the problem of "redlining" (i.e. refusal to lend to subprime markets that disproportionately affected minority groups). By definition, these groups were worse credit risks.
2) Republicans refused time and again to extend the same kinds of regulations that governed money markets and stock trades to derivatives trading (e.g. so-called "mortgage-backed securities" and "default credit swaps", both of which are the overleveraged components central to the current meltdown).
3) LOOK IN A MIRROR. The American public ran up credit cards, cashed out equity during the housing bubble to cover it, and ran up more. Rinse and repeat. As a result, tens of millions of regular Americans (including me, I’m not blind to my own personal bouts of overleveraging) who are now casting around for someone to blame should start at home.
Regarding the weighting of polls, it is done on a number of factors, but mostly age and race where voter turnout by demographic is more easily tracked and has historical data to support. As for party identification, Gallup I believe allows for self-identification, whereas Rasmussen using actual registrations, which leads to occasional disparities, but both have the largest tracking samples and good records for the national samples, so they should be relatively accurate.
As a point of interest, the areas where they may be weighting samples too lightly are in areas which favour Obama over McCain, such as Obama’s ability to turn out black and youth voters at levels much higher than historical norms, in addition to a Get Out The Vote ground game that is by far the superior of anything the Democrats have ever had before.
As to CStanley’s point regarding voter registrations, two points:
First, you might also want to note at some point the possibly illegal purges of tens of thousands of voters that seem to be a quadrennial event that seems to be getting a lot less play here than the ACORN story.
Second, do a little math. Assume for a moment that ACORN managed to register an additional 100,000 voters illegally. For a national tracking poll, the number of eligible voters is around 200,000,000, meaning the effect of an additional 100,000 is all of 0.05%, certainly not enough to affect the tracking polls by several percentage points assuming they even changed their weightings in response. (This is not to say such numbers couldn’t be significant in close races at the state level, since Florida in 2000 particularly was decided by a very small margin, but such numbers remain insignificant at the national level). Basically, for the margin to affected by even a percentage point or two requires the addition of several million voters. To tie things up would require 15 to 20 million additional Democrats illegally added to the voter’s lists. Not exactly likely given they’ve noted disparities of a few hundred in places.
I think Jason’s post in on target. Every party has their own part in causing this crisis, politicians and regular folks.
I don’t see any foul play involved here. There are simply more people who identify themselves as democrats than republicans these days.
I actually would not be surprised if the polls are under-representing Obama’s lead. First-time voters, I think, are much more likely to go for Obama than McCain, and my understanding is they are often included in polling.
Don’t interpret my comment as an endorsement for Obama. I just think republicans trying to spin the polls is wishful thinking. It is notable, however, as you say, that independents do seem to be supporting McCain more than Obama. But even if it turns out that way, it doesn’t look like it will be enough.
And for the person who noted that "liberal/moderate republicans" are grouped, I agree that it’s probably because there weren’t enough to split them out. I can probably count the number of self-proclaimed "liberal republicans" that were polled on one finger.
@C Stanley:
"having a lot more Democrats suddenly registered (you know, like Democrats Mickey Mouse, Donald Duck, etc.)"
You do realize that groups assisting in voter registration are required to turn in all registration cards submitted to them, even if they’re obviously invalid, don’t you? ACORN may have done some questionable things, such as the way they were paying ‘volunteers’ assisting in registration, but they did the appropriate thing by turning in the fraudulent cards. ACORN, just like any other such group, would have turned in 2 sets of registration cards–cards they believed to be valid and cards they believed to be invalid. Both sets would have been clearly marked.
actually you probably mean the Community Reinvestment Act Jason.
And I see myself smiling back at me. I have no credit card debt and live very well within my means. Where’s my bailout?
Rasmussen reports has new partisan weighting targets: 39.3% Democratic, 33.0% Republican, and 27.7% unaffiliated.
http://www.rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/new_rasmussen_reports_partisan_weighting_targets_39_3_democrat_33_0_republican
Honestly though, over the time I’ve been watching elections, I’ve seen some polls, where if you get a hold of the raw data you see how skewed they can be. One poll I saw asked the participants who they voted for in the 1992 election, and as I remember, about 60% said Clinton, 35% Bush, and 5% Perot. They didn’t include that poll question in any of the news reports, and, not surprisingly, the poll results were favorable towards Democrats. The margin of error was reported at something like +=3%
+-3%
I agree that there appears to be a big weight problem with these polls. I’ve found that on average they are weighting repub registered voters at about 8-10 % lower then dems. Yet, historically, the day after election have shown the disparity to be a consistent 2-3 percent on the dem plus side. Even if we ignore the Acorn Fraud Registrations and Assume that Dem registrations has doubled, the end of the day number should be at 4-6 percent. What this means is that the Obama leads in the polls are being inflated by 4-6 percent. Therefore, if i see that obama is up at 6, i’m considering it a statistical tie. If obama is showned up at 10 then we are looking at a very small lead that can be insignificant.
The fear I have is that the media will run on the Brandley effect when Obama looses without ever doing the due diligent to understand that it is really due to the Poll/Acorn effect.
Poor Mccain, never seem to get a fair break.
Deregulation has given us prosperity for almost 3 decades. Reagan, Bush Sr., Bill Clinton, and Shrub (Bush Jr.) all were deregulators. The problem is that the people with oversite were not enforcing (regulating) the laws in place. I’m sure when we look at this problem we will find holes that need to be filled for regulation, but if no one is enforcing such regulations it does no good. I love the finger pointing game, especially from the Dems, like they weren’t even in the room when this was going on. We are all to blame and if we as american citizens allow them ( the parties) to divide us, then they will have won their cause and we will pay. This has been going on for much longer than 8 years and we are all to blame.
BJ: First, you might also want to note at some point the possibly illegal purges of tens of thousands of voters that seem to be a quadrennial event that seems to be getting a lot less play here than the ACORN story.
Yeah, we never hear about ‘possibly’ illegal purging, do we? {rolls eyes}
You do realize that purging is generally required by law, so that we remove the potential placeholders for voters who don’t have pulses and such? Certainly purging has potential for fraud and abuse, just as voter registration drives do. If a nonprofit group funded by GOP partisans, and with an avowed partisan agenda, also got taxpayer funding to go around and selectively enforce purging laws in districts where doing so might give them an advantage, would you have a problem with that tactic? If so, then why shouldn’t Republicans cry foul to what ACORN is doing which has exactly the same effect- giving an advantage to the party they support even though they claim to be supporting a general GOTV effort rather than a partisan one?
You do realize that groups assisting in voter registration are required to turn in all registration cards submitted to them, even if they’re obviously invalid, don’t you? ACORN may have done some questionable things, such as the way they were paying ‘volunteers’ assisting in registration, but they did the appropriate thing by turning in the fraudulent cards. ACORN, just like any other such group, would have turned in 2 sets of registration cards–cards they believed to be valid and cards they believed to be invalid. Both sets would have been clearly marked.
Which is why we have a systemic problem as well as a specific example of an organization that is either corrupt at its core or incompetent in preventing corruption.
Quite honestly I don’t see how any group could be trusted to be nonpartisan in registering voters or trying to turn them out for voting. There’s nothing wrong with partisan groups doing that, but they should have to declare their partisanship, be transparent in their efforts and in their funding apparatus, and they must never, ever, get taxpayer funding. There’s simply no way to avoid the bias that comes from an internal culture of an organization- even one that starts out with bipartisan goals will shift one way or another over time as people aggregate themselves with people who hold similar ideology.
Ask yourself- do you believe that ACORN workers, no matter how well meaning, would purposely target a neighborhood that had a lot of more conservative leaning voters? If not, do you think they’re purposely targeting neighborhoods that lean to the Democratic party? And are you OK with taxpayers supporting an organization that has such a bias?
Would you be OK with taxpayer funds going to a group that canvassed predominantly white, evangelical churches in the South? Why or why not? And even leaving the taxpayer funds aside, would you think it was bogus for Democrats to scrutinize a group that was doing that and call for investigations if they were turning in thousands of new registrations from those areas?
By the way, there are reports that the ‘two piles’ of voter registrations is not the way this is happening at least in some places. Some districts are reporting that the county officials are the ones who are detecting these obvious duplicate name and duplicate handwriting registrations, and that those were not flagged by ACORN. Until the facts come out, I’ll refrain from judging what really has happened, but it’s a smokescreen to declare that ACORN has only done what it was required to do by law by turning in those forms, when we don’t in fact know if that’s the case.
Wow, some of you people are either blind or naive. There’s no way ACORN would screw itself by doing something so stupid. What you have here is a few goose-stepping right-wing fringe Obama haters who posed as voter registration workers, and then proceeded to wreak havoc on a legitimate organization. Doesn’t take a genius to figure that one out.
They will do anything to win, as the McCain internal polls are showing much worse than the ones we are all seeing. Obama state by state is actually kicking butt, doesn’t matter what the real breakdown is of Dem., Rep. or Ind. It could change but given the sterling McCain campaign and the economy it isn’t likely.
That’s right. It’s a conspiracy!!!
Please, do go ahead and show us your evidence that a longstanding leftist organization was deeply infiltrated by far-right activists in 9 different states simultaneously in a massive campaign to generate 1.3 million fraudulent voter registration cards just to “get” Obama….
Doesn’t take a genius to figure that one out.
Hah, that part, at least, is accurate.
admin: Name-calling and BDS conspiracy theories are not welcome here.
have you ever wondered why the democrats are alwys leading in the polls ,right up to the finish. Well, I don`t know about any other elections but I checked on this one,and It shows, Polls taken In Majority black cities,like Philadelphia,Pa. where there are over 650 thousands blacks,Denver,where in colo .thats the only city with majority blacks,New orleans,Houston,Chicaco, are you getting the picture??????? Obamma will lose this election because of the bradley effect.