Unfuckingbelievable.

Three out of five of you are willing to sacrifice your beliefs to elect a politician.

Another one of you is willing to accept a politician for what he is.

What else are you willing to sacrifice? Your career? Your family? Your Constitutional rights?

What else are you willing to accept?

Remember the LA Time's poll that had Barack Obama up 12 points but Gallup was tied?  As it turns out, The Great Poblano pinpointed the reason for the huge spread--the respondents in the LA Times poll were primed to lean more towards Obama because of a leading question posed to them prior to their choice in President.  

Obama seemed to win by a landslide when those polled were first reminded about how they feel about the direction of this country.  Their negative emotions towards Bush and the last 8 years served as a catalyst that made them more inclined to vote for the democrat.

This diary aims to place the idea of a nation gone backwards in proper context and presents a chain email linking Bush, McCain and the GOP with high gas prices.  This is the one issue that pisses off even the most passionate Republican voters.

Short diary, but I wanted to pass on the latest GA poll.  It's an Insiders Advantage poll and shows:
McCain   46
Obama   44
Barr        4

http://www.southernpoliticalreport.c...

The poll is of likely voters and was taken by telephone yesterday, July 2.  Georgia is the 9th largest prize for EV's, with 15.    

Along with today's Montana poll, it seems Obama is doing quite well in very different red states -- both west and south.      

Yesterday, wiscmass posted an excellent tutorial on how to read polls and the true meaning of Margins of Error.

I'm no polling expert, but I read the diary with great interest, and then out of curiosity decided to check out just how they figure out what the Margin of Error figure actually is. Some people commented that you just need to increase the sample size in order to reduce the MoE, which is true, but it's NOT a simple matter of "double the sample = half the MoE" by any means.

It turns out to be a more complex formula than that. Here's Wikipedia's entry on it:

http://en.wikipedia.org/...

I was in Montana in May and there were Obama offices everywhere and I was impressed.

Even some of the smallest little towns sometimes had Obama offices (even if it was someone's house serving as a makeshift headquarters).

Is Montana really in play? Yes, I think so.

And the latest Rasmussen Poll is an eye-opener.

Obama 48% McCain 43%.

http://www.rasmussenreports.com/...

Today's poll round-up from 411mania.com shows Florida a mixed bag of polls with:

-- In Florida, Strategic Vision has McCain ahead of Obama 49% to 41%. We now have a situation where since the primaries ended, Strategic Vision and Rasmussen have shown McCain leading in the state while Public Policy, Quinnipac, and ARG have shown Obama leading in the state.

Obama is up big in CT and MA and Barr's polling 3% in Georgia, with McCain up 8. If as per CNN analysis, third party candidates lose half their support by November, that cuts Barr down to 1.5% in November.

I never thought I would see the day.

Barack Obama is leading John McCain by five percentage points in Montana. The latest Rasmussen Reports telephone survey in the state shows Obama attracting 48% of the vote while McCain earns 43%.

Welcome Red State Converts!!  Welcome to the Democratic party!!!

Did you know that McCain has a two month losing streak in head-to-head polls against Barack Obama?

Maybe you've been distracted by beltway politics. That's understandable.  Myself, I can't get excited over all this Wes Clark/FISA/campaign finance/faith-based b.s. This is the end of June, and none of this stuff's gonna matter come August, much less election day. Wes Clark committed a minor gaffe, the media pounced, and it's already starting to backfire on McCain. FISA is bad, but it hasn't passed yet so let's wait and see.  Obama's decision to decline public financing and raise money from small donors was a great decision.  Al Gore supported faith-based charities in 2000, so I don't see how it is controversial that Obama supports it now.

It would do us all some good to step back and admire the landscape, folks.  I've got some good news for you, courtesy of Real Clear Politics: McCain has not led in a national non-tracking poll since May 3rd.

In 2000, the voters preferred Al Gore on the issues.  Just about every issue:  his stance on taxes.  His stance on social security.  His stance on the budget surplus.  However, when they went into the ballot booth, they decided that those issues were unimportant.  They wanted a regular guy as President, not somebody fancy with good ideas, just crackpots ones.  They wanted the man they'd rather have a beer with.  And George W. Bush was swept into office.

Four years later, John Kerry had the same advantages.  Americans wanted him to end the war.  Americans wanted him to help install a national health care system.  But dammit, if he showed up at their bar on Saturday afternoon, they'd just feel too awkward when he ordered a Pinot Grigio.  Gotta go with that regular guy, George W. Bush.

Now, obviously, one might protest that Bush wouldn't have the beer you were offering him, since he doesn't drink.  But that didn't stop anybody.

I'm a social scientist. I study all kinds of things and use lots of different methods in my research, but mostly I make my living analyzing large amounts of data collected through surveys. By necessity, that means I'm pretty good at using advanced statistical techniques to figure out what large populations think and how they behave.

A political poll is essentially nothing more than a public opinion survey. Lots of people think conducting a survey is easy. Lots of people are mistaken.

Below the jump, I'll explain what I mean, and I'll also explain (without resorting to the technical jargon characteristic of my field) why margin of error is important, how it works, and why so many people misinterpret it.

It’s hot here in Gomorrah today, so my friends and I languish out on the balcony of my 700-square-foot loft apartment while the sounds of All Things Considered waft through the energy-efficient windows. My friends and I play "Rock, Paper, Scissors" to pass the time. As my fist tightens to form the symbolic fist of a "rock", I said to my friends,

"Hey look guys I’m playing BARACK, Paper, Scissors!"

Let’s get one thing straight: Barack Obama is not the "most liberal member of the US Senate" despite what the National Journal says.  And John McCain doesn’t think so either.
 
John McCain is known to point out that Obama supports the President at least 50 percent of the time.  The National Journal disagrees.  In its analysis and rating of senators’ votes in 2007, the Journal concluded Obama is the "most liberal member of the Senate."

Although nearly all legitimate polls show Barack ahead of McCain (some by double digits) this (unscientific) bizarro-world AOL straw poll, has McCain over Obama by 62% to 38%!

http://news.aol.com/...

My Congressperson is a...

1. Democrat who has a safe seat

2. Democrat in a possibly Republican challengable seat; help keep this one!

3. Republican in an, alas, safe seat.

4. Republican in a possibly Democratic challengable seat; help change this one!

This connects to a couple different things...

Why you come to Daily Kos?  Some come here just for the politics, anti-Bush and anti-Iraq war.  Some are here because they are from an otherwise deeply Repuglican area and need the community. Others like me come here from an already all Blue area, hoping to help reach out and be able to do something useful.

Also I want to refocus some attention from the Presidential race and towards the need for more & better Democrats in the House. We may wish to consider giving money to strategically targeted House races via ActBlue, to those Democrats where our limited dollars can have the most bang for the buck in increasing our margin in Congress.

If you think that this an imortant and timely discussion, please Rec this diary.

The polls are looking good, the economy is in the tank, Iraq is a millstone around the neck; the Republican brand is in trouble. But. . .

UPDATED:

It's Presidential election season, and that means lots of political horse race polls, with head to head numbers between Barack Obama and John McCain. Some national polls are different than the others – CNN/Opinion Research, USAT/Gallup, ABC/Washington Post, NBC/Wall Street Journal and CBS/NY Times – because they drive narrative. It’s not that they’re more venerable, more accurate or more prestigious (taking nothing away from them, that might also be said with justification about Pew.) It’s because the networks use their own (maybe a little too heavily) to drive their story line, while selectively ignoring the others.

If you were getting your political information last week from the nightly network news, or the cable networks, for example, you'd know that Obama has an insignificant lead on McCain but Obama's in trouble despite Bush's unpopularity and the state of the economy because he (a) did not get much of a bounce after Hillary Clinton withdrew, and (b) is having trouble consolidating Latinos, women, white Catholics and independents. This from ABC:

Bottom Line: Obama Leads, McCain Winning Key Groups

On the other hand, if you got your news from the internet, you'd know that Obama has a small but sustained lead in the horse race, and is in strong position because he has wide leads with women, Latinos, the electoral college, and enthusiasm of supporters. In addition, his strength elsewhere is masked by relative McCain strength in the South, McCain women supporters are vulnerable to rejecting him on the choice issue, and there are more Democrats and less Republicans this year. Because of all of that, most people think he's going to win (especially women).

Click picture for larger version:

So, what gives? Why are the two narratives so different? It's not the polling directors. ABC's Gary Langer  – Close Race: McCain Stays in Range Amid Challenges for Obama – and CBS's Kathy Frankovic know their stuff, and each writes frequently about polling issues, while NBC and WSJ use a combination of well-respected pollsters (Democrat Peter D Hart and Republican Neil Newhouse) for their polls. It's not always the political people (MSNBC's staff wrote up the Latino lead for Obama, at least on line — "This was a case of conventional wisdom that was never based on fact, just semi-informed speculation based on primary exit polling and bad stereotypes of Latinos.") But something about the way network polling is commissioned and presented to the general public suggests several things:

1. networks tout their own polls to the exclusion of others

2. the networks are invested in having a close race

3. media narrative changes slowly

networks tout their own polls to the exclusion of others

One of the great things for the public that media does is commission and release periodic polling so that a snapshot of 'things as they are today' can be shared. Whether it's CNN or NBC, the poll of the day is likely the lead story on a well-viewed hour the day the poll is released. Unfortunately, if there are other, similar polls, they tend not to be discussed. When the 6/15/08 ABC/Washington Post poll was released, one of ABC's headline was 'Danger Signs' as Clinton Supporters Resist Obama, while the Sunday prior, we (and others) noted that

Obama leads women by 12 points in the latest Rasmussen polling (including leaners) and McCain has a 3 point lead among men (subscription required for crosstabs). In 2004, it was the opposite (Bush 11 point lead amongst men, Kerry 3 points with women.) In 2004, 54% of voters were female.

Now, anyone who followed the WSJ/NBC poll here (and not on network news) would be aware that Obama beats McCain 52-33 with women, and 61-19 with those who say they voted for Clinton, making Clinton supporters a stronger demographic for Obama than women in general.

By the way, the networks tend to exclude polls done by automated calls (like Rasmussen and Survey USA), even with their excellent track record. That unwarranted bias eliminates a large amount of perfectly detailed, publicly available and timely polling data that would enhance network coverage.

the networks are invested in having a close race

Now, this is tough to prove, but it isn't exactly rocket science. A blowout election is not ratings sustainable the way a close race is. The narrative that "the race is close" is actually true, and a need for caution this early out is laudable, but it doesn't mean that non-network sponsored individual state races need be ignored just because network polling is horse race polling. In fact there are multiple internet sites that look at state polling (like pollster.com and Real Clear Politics, as well as FiveThirtyEight.com), and the trend this week is one-sided in Obama's favor (states like PA and WI aren't looking like swing states this week, while Obama gained in OH and FL). However, stories like that and McCain's support among GOP women shows cracks don't make network news, even when the information filters into the written media (McCain Appears to Trail In Trio of Battleground States.)

There are times that polls like Newsweek and LA Times/Bloomberg will get play (their large Obama leads contradicted network polls), but they were frequently introduced in the context of being "outliers" rather than getting into the fascinating question of Party ID and what it means this year. This is from a National Journal column by Mark Blumenthal:

At about the same time as my talk, Los Angeles Times/Bloomberg released a new poll [PDF] showing Barack Obama leading John McCain by a 12-point margin (49 percent to 37 percent). Immediately, reporters and political junkies were asking, "Is this result real?" Once again, we started focusing intently on just one poll, overlooking other recent surveys showing a closer race.

Given enough time (a key factor for television), the topic does get introduced on occasion, so it can be done (this by Chuck Todd):

For instance, in this poll, John McCain is down 12 points but winning independents.  Why?  Because more Republican voters don‘t want to say they‘re Republican, so they‘re saying they‘re independent.  And more independent—Democratic-leaning independents are saying, Hey, I don‘t mind being called a Democrat right now.

So you‘re looking at that and you say, This is just an enthusiasm gap.  Maybe it‘s not 12.  Maybe it‘s 8.  You know, maybe it‘s not 15.  But it‘s a lead.  And we know that the low end is 6, the high end is 15.  Even if you decide it‘s somewhere in the middle, this "LA Times"...

In fact, both Newsweek and the LA Times needed to publish somewhat defensive pieces because of the public criticism of the polls.

June polls of a horserace that ends in November aren't "reliably predictive," as the survey experts say. So why are Republicans so ballistic about new surveys, including Newsweek's, that show Sen. Barack Obama with a big lead?

Too large a Democratic advantage in new L.A. Times/Bloomberg poll?

To be fair, there's reasonable caution in not suggesting any one poll or series of polls now means anything predictive for November. Still, anyone bringing up Michael Dukakis' 17 point lead in 1988 ought to remember the Dukakis lead was after the Democratic National Convention in July and not pre-convention in mid-June (networks are supposed to provide 'context' with the numbers, and that's one contextual piece of information that is important to include.)

Another contextual piece is that Republicans have attacked the LA Times poll before. None of that should preclude discussion of what the polls (all the polls) actually say, and both contextual pieces (the Dukakis lead being after the convention, and a previous track record of fierce Republican attacks on polls that show Democratic leads) are important to add to tell the full political story.

media narrative changes slowly

How slowly does media narrative change? This (McCain Gains from Clinton-Obama Feud), from March, is still the narrative from June. That and "Democrats in Disarray", a favorite since 2000, is seen in the coverage here (Danger Signs' as Clinton Supporters Resist Obama) even though all the other polls say otherwise. In fact, in this remarkable example, ABC lists all of Obama's issues and none of McCain's, even though Obama is leading in their own poll. A balanced view of both candidates' issues is nowhere to be found.

CHARLIE GIBSON: There are some key constituency groups, though, with which John McCain is actually doing pretty well with.

GEORGE STEPHANOPOULOS: And these are the groups that go with the winners. Uh, independents, right now, it's basically a dead heat for Obama. You think he would be doing better than that. Again, given the conditions of the country. Also, McCain (is) doing very well with the group that's been right the last eight elections - white Catholics - he's leading by 14 points among those groups.

You might get the false impression from this that the trailer in the poll has more strengths than the leader, and is aligned with 'winner' groups. In fact, that seems to be the purpose of the dialogue, and is especially interesting in light of more recent polls on religion and politics Update [2008-6-29 12:11:11 by DemFromCT]:: BruinKid points out that white Catholics and Catholics are not precisely the same demographic:  

The report disputed the belief -- based only on exit polls -- that Catholics have "a near perfect winning streak" in backing the winner in recent presidential elections. By also using data from the American National Election Studies and Gallup polls, CARA found that Catholics voted for the winner in the overall popular vote 71 percent of the time between 1952 and 2004 and the Electoral College winner 64 percent of the time.

   "In the past five elections, a majority of Catholics voted with the winning candidate only when Bill Clinton was on the ballot," the report said.

Two weeks ago, Frank Rich wrote:

NBC Nightly News" was so focused on these supposedly devastating Obama shortfalls that there was no mention that the Democrat beat Mr. McCain (and outperformed Mr. Kerry) in every other group that had been in doubt: independents, Catholics, blue-collar workers and Hispanics. Indeed, the evidence that pro-Clinton Hispanics are flocking to Mr. McCain is as nonexistent as the evidence of a female stampede.

One would have to ignore all the evidence to come to the conclusion that Obama was in trouble with Clinton supporters or women, but it does make the race seem closer to do so, and Obama 'troubles' with both women voters and Clinton voters, as with Latino voters, was a very popular theme in the media, and won't disappear easily despite abundant polling evidence to the contrary. In fact, the Unity rally with Clinton and Obama Friday only sparked more stories about internal conflict, Bill's 'sulking', and familiar story lines that nonetheless have little to do with what's happening as measured by the polls. The Chris Matthews Show today devoted most of the time to this topic. Some media narrative changes glacially.

there are alternatives

It's not surprising that the networks do what they do. They invest money in polls, they're sensitive to criticism of "bias", and they have an eye on ratings. But there's a reason that those who are interested in politics are increasingly turning to the internet.

A record-breaking 46% of Americans have used the internet, email or cell phone text messaging to get news about the campaign, share their views and mobilize others.

The use of the internet to get political news is growing. At times, it's the only place you can get the news. And as long as networks insist on relying only on their own polls, or pushing an old narrative that's not supported by the newest data, the American public is not going to get a clear and fair picture of what's going on by relying on broadcast news alone. Not unless the networks do a better job in reading polls.

AUTHORS NOTE: I had written this to post it yesterday afternoon. For some reason, it would not let me publish, saying I had already written a diary (it had been deleted). I chose not to edit it, so some of the references ("today", "yesterday") are a bit off. Apologies...

There was another bevy of polls this week, both on the presidential race and downstream, as well. Unfortunately, I do not have all of them at my disposal, as I put them in a notebook which I cannot, at present, find.

Nevertheless, I do have numbers on 22 separate contests, so let's head past the jump for the numerical goodness, which includes two polls just released in the past few hours!

Big news...  New polls show Obama has big leads in four battleground states: Minnesota, Michigan, Colorada and Wisconsin.  Ummmm, Thanks.  But didn't we already sort of know that, for three of these four states anyway?

I'd like to respectfully suggest that Minnesota, Michigan and Wisconsin are no longer swing states (if they ever were).  Thanks for the data on Colorado though.

Danger, rant against silly polling below the fold.