Tag Archives: party ID

2012 National and State “Party ID”

November 13, 2012 – 9:26 am

The Winston Group is updating the early results from this past election for party ID nationally and at the state level:

We’ve updated our charts of national and state Party ID and Ideology breakdowns to include 2012 numbers, and is presidential-level data. There are two items to note: 1) states with a (P) label means the data from that still is still preliminary and is subject to change and 2) not every state has exit polls, so some states’ charts only go up to 2008.

Access the updated .pdf here: National Exit Polls: Party Identification and Ideology Breakdowns

By keithbacker | Posted in General | Comments (70)

Combatting the Polling Problem

November 13, 2012 – 9:18 am

In addition to fixing a clearly broken brand and attracting more voters, Republicans need to address getting blind-sided by their own internal polling:

In the weeks before Election Day, both Republicans and Democrats were nervous about their poll numbers. Both sides of the aisle have smart pollsters, they reasoned, so how could the numbers that Democrats were seeing diverge so sharply from the numbers the Republicans were seeing? Deep down, I wrote at the time, both parties secretly worried that their side was missing the boat.

What went wrong:

“Everyone thought the election was going to be close. How did [Republicans] not know we were going to get our ass kicked?” lamented Rob Jesmer, head of the National Republican Senatorial Committee. “I don’t understand how we didn’t know. That’s the part that’s most puzzling and frustrating and embarrassing.”

The underlying causes of the errant numbers are the assumptions that the pollsters made about the nature of the electorate. Most pollsters believed the electorate would look something like the voters who turned out in 2008, just with slightly lower numbers of African-Americans, younger people, and Hispanics heading to the polls.

But exit polls actually showed a much more diverse electorate than the one forecast. Black turnout stayed consistent with 2008, Hispanic turnout was up, and younger voters made up a higher percentage of the electorate than they had four years ago. White voters made up 72 percent of the electorate, according to the exits, down 2 points from 2008 and a continuation of the two-decade long decline in their share of the electorate.

That meant that even though Mitt Romney scored 59 percent of the white vote — a higher percentage than George W. Bush won in 2000 and 2004, higher than Ronald Reagan in 1980 and matching George H.W. Bush’s 1988 score, when he won 426 electoral votes in 40 states — it wasn’t enough to overcome the 80 percent support that Obama scored among nonwhite voters.

Solutions:

Pollsters should fix voter screens, used to weed out of their samples irregular voters who aren’t likely to vote. Including only likely voters often leads to a more Republican-heavy sample. But in an era of fine-tuned turnout machines and get-out-the-vote drives, even those irregular voters are likely to show up. Polling all registered voters, rather than those most likely to make it to the polls, would at least give Republicans an idea of the worst-case scenario.

Pollsters should also control more for age, gender, and race than for party identification. One prominent party pollster pointed to a late survey conducted for Indiana Republican Richard Mourdock that showed him leading Democrat Joe Donnelly by 2 points. That survey, conducted by McLaughlin & Associates, showed that 56 percent of Indiana’s electorate would be over age 55. Exit polls revealed that number to be vastly overstated; only 43 percent of the electorate was over 50.

The party-identification question gets to the heart of another misperception that pollsters make. Tell almost anyone that Romney would have won self-identified independent voters by 5 points and logic would dictate that Romney would win a clear victory. But Democratic pollsters say that metric is flawed, and that many Republicans remain so disaffected by their own party that they refuse to identify with it. Instead, some say that pollsters should look at self-described ideology, rather than party identification. Indeed, Obama beat Romney among the 41 percent of voters who call themselves moderate by 15 points.

Pollsters also recognize that Americans’ daily routines are changing, something that has an impact on their surveys. About one-third of all households do not have a landline, according to the National Health Interview Survey, meaning that a significant swath of the electorate is available to pollsters only by cell phone. The percentage of younger Americans who don’t have a landline is almost double that. Pollsters who don’t include a sufficient number of cell-phone respondents in their surveys risk missing out on younger voters — voters most likely to back Democrats, thus skewing polls to the right.

By keithbacker | Posted in General | Also tagged , , , , | Comments (52)

The Difference

November 7, 2012 – 10:52 am

Jim Geraghty’s Morning Jolt said it best:

Those Pollsters Were Right; We’re a Much More Democratic County, at Least in Presidential Years

The exit polls indicate a 39 percent Democrat to 33 percent Republican split, only a percentage point behind 2008. I was incorrect in my skepticism that the electorate would be closer to D+3 or D+4. Nate Silver, take a bow. Public Policy Polling, your samples weren’t as wacky as I believed they were.

The Obama campaign has put together a fantastic get-out-the-vote machine. We saw in Virginia and New Jersey in 2009 and Massachusetts in early 2010 and all over the country in the midterms that Obama’s personal charm did not transfer to other candidates like Jon Corzine and Creigh Deeds and Martha Coakley.

Republicans need to confront the fact that because of demographics and a party infrastructure that has gotten very, very good at bringing out the vote in presidential years, Democrats are going to be very, very tough every four years. One of the strange aspects of this year is that I would have argued that Obama wasn’t all that charming. His favorable numbers dipped. He was dismissive in that first debate, snarky and combative in the second, constantly saying things that his campaign had to explain — “you didn’t build that,” “the private sector is doing fine,” “Romnesia,” “voting is the best revenge” . . . and he still won.

Ari Fleischer points out the silver lining is that so far, Romney is winning independents. That’s not a silver lining, that’s worse news: Democrats don’t really need independents anymore.

By keithbacker | Posted in General | Comments (314)

If Rasmussen is Correct (and he has been in the last two Presidential elections)

November 5, 2012 – 1:24 pm

Then we’re looking at the Scott Walker recall night all over again.

Rasmussen Reports’s Party Affiliation for October is R +5.8.  Below is Rasmussen’s month-end party ID for each October in Presidential years. We compared that with the exit polling party ID provided by the Winston Group:

Year Rasmussen Actual
2004 D +1.5 (Dem 38.7, Rep 37.2) D +0 (Dem 38, Rep 38)
2008 D +7.1 (Dem 40.3, Rep 33.3) D +7 (Dem 40, Rep 33)
2012 R +5.8 (Dem 33.3, Rep 39.1) ?????

In the two prior Presidential election years Rasmussen essentially called the party identification and accurately captured the ground swell in favor of Democrats in 2008. Not coincidentally Rasmussen called the 2004 election within 1% and hit the bullseye in the 2008 election.  I don’t know exactly the why behind Rasmussen’s methodology but his affiliation has been consistently R +whatever but he’s been running his polls at D +whatever. I’m assuming D +2 so if we get R +2 (I refuse to even consider R +6) then it is blowout city.

And if that happens we could see a rerun of two of our favorite clips from the Scott Walker recall, delusional Democrats in disbelief and MSNBC anchors live re-enactment of the movie Scanners.

By keithbacker | Posted in General | Also tagged , | Comments (118)

Senator Sharon Angle Agrees With Nate Silver: Barack Obama has an 84% Chance of Winning

November 3, 2012 – 2:37 pm

Nate Silver has his usual spin on outrageously absurd election outcome odds:

President Obama is now better than a 4-in-5 favorite to win the Electoral College, according to the FiveThirtyEight forecast. His chances of winning it increased to 83.7 percent on Friday, his highest figure since the Denver debate and improved from 80.8 percent on Thursday.

He shows a bunch of polls from a murder’s row of bad polling where Obama is leading and maps out three arguments where they could be wrong.  After arguing and dismissing the first two he concludes:

That leaves only the final source of polling error, which is the potential that the polls might simply have been wrong all along because of statistical bias.

You don’t say!

The FiveThirtyEight forecast accounts for this possibility…I do not mean to imply that the polls are biased in Mr. Obama’s favor. But there is the chance that they could be biased in either direction…My argument, rather, is this: we’ve about reached the point where if Mr. Romney wins, it can only be because the polls have been biased against him. Almost all of the chance that Mr. Romney has in the FiveThirtyEight forecast, about 16 percent to win the Electoral College, reflects this possibility.

Silver makes such pronouncements with outlandish statistical weights as if it is nearly unbelievable that the poll results could be wrong.  One of the main purposes of this blog was to look at the exact same polls, analyze the internal data and test whether the poll data match up with the poll results.  We found that time after time after time the results unequivocally do not match up with the internal data.  Thanks to Sean Davis, we are reminded this was the identical situation only 2 years ago is probably the highest profile race where a deeply unpopular Senate Majority leader was behind in nearly every poll yet still won.

Out of 14 polls between October 1 and election day, Sharon Angle led in 12 of those polls.  Her average lead on election day according to Real Clear Politics was +2.6.  She lost by -5.6 points — an 8.2 point swing.  The polls were not just wrong, but WAY wrong.  Could anyone analyzing the internals of these polls see this?  Why yes they could. But even in the highest profile contest of the cycle, almost no one did such an analysis. The few who did, Democrat pollster Mark Mellman, Republican pollster Glen Bolger and liberal reporter/columnist Jon Ralston, all consistently said the polls were wrong — and each was largely ignored until proven correct on election day.  Why did they know this?  Because they looked at the data in the polls and said the internal information does not reflect the top-line results and the Nevada electorate on election day will not reflect what these polls are indicating. They were right and the polls were wrong … by A LOT.

Today we have an identical dichotomy where the stat gurus like Nate Silver say Obama has an 84% chance of winning because that is what the top-line poll numbers tell him.  Nate Silver called the Nevada Senate race incorrectly because the poll data was wrong.  His accuracy is predicated on accurate polls.  Mountains of evidence says today’s Presidential polls are equally as wrong as the Nevada Senate polls.

Critics of the polls on the Right, like myself, of whom even Silver concedes offer “intellectually coherent” critiques say the results on November 6 will be very different. Maybe Nate Silver is correct and Barack Obama will be re-elected President on November 6.  But any analysis of the data in those same state polls he relies on says the voting preference of Independents, the increased turnout of Republicans, the decreased turnout of Democrats, the change in favor of Republicans in early voting, Romney’s favorability on the election’s top issue (economy) and numerous other factors will result in President Romney on November 6.  United States Senator Sharon Angle from Nevada may disagree.

By keithbacker | Posted in General, Nevada | Also tagged , , , , , | Comments (52)

Thoughts on Yesterday’s CNN/ORC International Poll With Obama Leading by 4 in Ohio

October 27, 2012 – 7:29 am

A spate of polls came out yesterday that I left alone for various reasons but the CNN/ORC International poll with Obama leading by 4-points created special consternation among Romney supporters and readers of this blog.  The reasons behind the concern were party ID and the Independent vote.

The party ID was not outrageous. The split was D +3 (Dem 35, Rep 32, Ind 33) when we are usually dealing with Ohio polls at D +8, D +7 and even D +10. As a matter of fact at the same time as the CNN release, American Research Group released an Ohio poll showing Obama leading by 2-points with a party ID split of D +9. In 2008 Ohio turnout was reported as D +8 although this has been disproven in favor of the real split of D +5. The 2004 party ID was R +5 so the fact that this poll ends up somewhere in the middle says we’re at least in an acceptable range. The other factor causing concern among the pro-Romney factions was that Obama was leading in this poll among Independents by 5-points, 49 to 44.  On those two factors alone this should have been a great poll for team Obama since anomalies in those two areas are the most cited reasons to dismiss the regularly erroneous polls we have seen this cycle.

I tried to contact CNN (OK, I tweeted their political director) for the racial make-up of the poll since they conveniently failed to include that in the cross-tabs and this is another prime area for abuse by the polling outfits.  But alas, my tweet went unanswered. Thankfully incredible readers in my comment sections went to great lengths identifying many other credible sources debunking multiple oddities in this ostensibly credible poll.

I’ll state up-front, this poll does not concern me in the slightest over Romney’s prospects in the state. I try to be as objective as possible when running the numbers and my opinions on who is winning are whatever the #s tell me they should be and Ohio is actually looking pretty decent for Romney.  Yes, I said Ohio is looking pretty decent for Romney.

Independents

Thanks to Josh Jordan (@NumbersMuncher), we know that in the seventeen most recent polls in Ohio, only three have Obama leading with Independents and two of those are these same CNN/ORC polls. In the fourteen other polls, Romney’s lead with independents is +11.4. CNN/ORC looks to be the outlier, not the norm by any means.  In that same ARG poll cited above Romney actually had a 21-point lead among Independents. This essentially means Obama is not leading with Independents despite whatever the outlier CNN/ORC says.  When Ohio’s ballots are counted, each side will lock down their base and the differential in partisan turnout will be minimal. Whoever wins Independents will win Ohio. Obama carried Independents in Ohio by 8 percentage points in 2008 and today he is losing Independents by 10+ points.  This is an 18-point swing among the voting group who will decide the state.

Early voting

Early voting is over-represented in the CNN poll. According to @Adrian_Gray in the poll 2/5 of Ohio likely voters have already cast their ballot. County election offices say only 1/5 have voted. Both cannot be right. Extrapolating the CNN poll statewide, Obama leads 59-38 among the 1.4 million that voted early. Romney leads 51-44 among 4.4 million have yet to vote. The math is pretty straightforward where the overwhelming remaining voters support Romney and should overwhelm any early vote advantage CNN found.

2012 is not 2008

Obama’s 2008 early voting secret weapon is gone. He won Ohio in 2008 due to an overwhelming early vote advantage.  On election day more votes were cast for John McCain than Barack Obama but so many votes banked away for Obama that it wasn’t enough. Today, according to the same Adrian Gray: “220,000 fewer Democrats have voted early in Ohio compared with 2008. And 30,000 more Republicans have cast their ballots compared with four years ago. That is a 250,000-vote net increase for a state Obama won by 260,000 votes in 2008″

Undecideds break for the challenger

On average Obama’s support level is always stuck around 47% in the Battleground states.  His current Real Clear Politics average in Ohio thanks in no small party to this CNN poll and the ARG poll stands at 48%.  We have shown across multiple elections against an incumbent President Undecideds break between 66-80% for the challenger. This means there is not a lot of votes left for Obama beyond his base and already accounted for Independents.

Between the reversal of the Independent voting block, the neutralizing of Obama’s early vote advantage and Undecideds always breaking 66-80% for the challenger Obama is looking really bad in Ohio.  And everyone today admits whomever wins Ohio wins the election.  To those who are unnecessarily concerned. rest easy.  Things are looking far better for Romney than is being represented.  The networks need ratings and horse-races increase ratings.  Four years ago at this time the race was over yet you could still see stories about McCain’s momentum.  It was just to keep people tuned in because that’s what pays the bills.  They did the same thing for Dukakis and Dole down the stretch when those races were blow-outs.

This race is far from over and 11 days is an eternity in politics, but every sign says Romney has the initiative and Obama is only responding to whatever Romney does. Obama’s fire-wall keeps getting moved back while Romney keeps pressing forward.  You can’t call a race until the last ballot is cast but I’m fairly optimistic on Romney’s prospects in Ohio at this point.  And I’m a pessimist by nature.

By keithbacker | Posted in Ohio | Also tagged , , , , , , | Comments (35)

CNN/Time Poll Has Turnout at Democrats +9 in Ohio and Obama Leading

October 24, 2012 – 5:36 pm

The new CNN/Time poll.  What are we going to do with it? The top-line of the poll says Obama leads by 5, 49 to 44 but everything underneath says that poll has little basis in reality.

Independents

In 2008 candidate Obama won Ohio by 5 percentage point while winning Independents by 8 points.  In today’s CNN/Time poll, Mitt Romney leads with Independents by an incredible 13-points but is down overall by 5-points.  That is simply not plausible.

Party ID

The party ID in this poll was D +9 (Dem 37, Rep 28, Ind 29). This compares to D +8 in 2008 (Dem 39, Rep 31, Ind 30) and R +5 in 2004 (Dem 35, Rep 40, Ind 25).  There is no chance the Democrat turnout advantage will exceed Obama’s 2008 best in a generation turnout which we write as D +8 based on the CNN party ID generally used.  This is even though the actual 2008 party ID was really only D +5 making this D +9 that much more unrealistic.

Early voting and the likely voter screen

Early voting is creating a unique problem for polling organizations this year. Since Democrats tend to vote early, you see the Democrat candidate typically leading by wide margins in early voting. According to the CNN survey “respondents who say they have already voted, Obama holds a two-to-one lead over Romney, 60% to 30%.”  So we know upfront in early voting there is an overwhelming number of Democrats.

When it comes to polls, all voters who said they already voted make it through the likely voter screen and end up in the final results. Approximately 17 percent of the respondents were early voters while 83% had yet to vote. This means a 17% segment of those polled are guaranteed to make it through the likely voter screen and we know that group supported President Obama by a 2-to-1 margin. This inherently over-samples Democrats which practically guarantees a favorable result for Democrats. The problem with early voting and polling ahead of election day is the results will skew in favor of the party with the higher early turnout, in this case the Democrats.  That’s one of the ways you end up with a party ID of D +9 when there is no chance of that turnout occurring on election day.

It was nice of CNN and Time to spend the money to run a poll of all-important Ohio, but I think this poll did little more give false hope to Democrats who are likely in for a rude awakening on election night thanks to unrealistic polls like this one.

By keithbacker | Posted in Ohio | Also tagged , , , , | Comments (65)

Obama +5 in Ohio with a Party ID of D +9 — CBS/New York Times/Quinnipiac

October 22, 2012 – 7:52 am

You can’t have everything and I guess this poll result is better than the laughable Quinnipiac poll last month showing President Obama with a 10 point lead with the same party breakdown. At least Quinnipiac is consistent in their auditioning for a job in Obama’s second term. They threw me off last week when they put out some fairly sampled polls, but here we are right back to last month’s monstrosity of an Ohio poll over-sampling Democrats by 9%, higher than the 8% margin they enjoyed in 2008. Party ID is D +9 (Dem 35, Rep 26, Ind 34). This compares to D +8 in 2008 (Dem 39, Rep 31, Ind 30) and R +5 in 2004 (Dem 35, Rep 40, Ind 25). As I wrote last month, “This is for a state that less than 12-months ago went to the polls in a very pro-Union turnout and also voted to REPEAL Obamacare by a margin of 66 to 34.” Considering we have the same party ID as last month, this poll is really really bad for the President.  He’s down 7 with Independents, he’s down 7 among men, and his month-over-month comparisons (below) are horrible. Democrats won’t enjoy near the advantage they had in 2008 and this poll claims Ohio is even more Democrat this time around which simply isn’t true.

The early vote issue

Obama leads among early voters 54 to 39 and 20% of those surveyed said they were early voters. The rise of early voting seems to be skewing results in favor of Democrats which was deftly explained by Adrian Gray previously. If Democrats have a propensity to vote early and early voters are polled, more Democrats will get through the likely voter screen and shade results towards Democrats. This is a problem polling forms are going to have to work towards re-calibrating polls reflecting this unintended Democrat over-sampling. It is not a bias in the polling it is just a flaw in polling methods today failing to appropriately take into account this new phenomenon.

Highlights compared to last month:

  • Independents support Romney by 7-points, up from a 1-point lead last month
  • Both sides lock down their base (93%) with nominal/insignificant crossover (5% Romney, 6% Obama)
  • Republicans are decidedly more enthusiastic about voting 52 to 40 — similar to the 10-point gap last month
  • Men support Romney by 7-points, 51 to 44 — similar to the 8-point gap last month
  • Women support Obama by 15-points, 55 to 40 — down from a 25-point gap
  • Youth vote supports Obama by 25-points, 59 to 34 — down from 35-points
  • Seniors support Romney by 7-points, 51 to 44 — reversing an Obama 1-point lead
  • Obama Job Approval good at 50/47
  • The candidates split on who would do better on the economy 48 a piece
By keithbacker | Posted in Ohio | Also tagged , , , | Comments (40)

Ohio Tied if 407,000 White Voters Don’t Show Up — Gravis Marketing

October 20, 2012 – 1:31 pm

I need to get on with my day of college football but this poll was too funny to just ignore. We already blogged today that Ohio isn’t the deficit for Romney polling and media would lead you to believe.  Now we have polls aggressively adjusting the racial composition of the polls to favor Obama.

Polls are tied despite heavy over sampling of non-Whites

Earlier this morning the comical SurveyUSA poll of Florida found a 16 percentage point drop in actual White votes (a Romney demo) yet Obama was only up 1%. Now Gravis Marketing surveys Ohio and finds a 7 percentage point drop in the expected White turnout and President Obama is tied at 47 with 6% Undecided.  Still well below 50% and with Undecideds likely breaking at least 2/3 for the challenger, he’d lose 49 to 51.

Ohio demographics

The racial composition of the Ohio vote in 2008 was: White (83%), Black (11%), Hispanic (4%), Asian (1%), Other (1%)
The racial composition of the Gravis poll is: White: (76%), Black (12%), Hispanic (6%), Asian (1%), Other (5%).
Did 407,000 Whites voters leave Ohio while 232,000 “Other” voters suddenly move to Ohio?  That is what a 7pp drop in White voters and 4pp increase in the “Other” category would mean based on Ohio’s 2008 vote. Something tells me that didn’t happen.

Most polls don’t re-weight by party ID but they do by race

Remember how much the media and polling firms mocked the critics over party ID complaints?  The said they don’t weight by party ID so the criticism is wholly without merit.  They admit they do weight by race though so these laughable demographic compositions are things the polling firm consciously CHOSE.  That is two straight firms, creating voter samples completely unrealistic anywhere beyond David Axelrod’s wildest dreams and yet President Obama can no better than a statistical tie … and that’s before we factor in Undecideds who break heavily for challengers.

Party ID

The party ID in this poll is D +9 (Dem 41, Rep 32, Ind 27).  This exceeds what we have been using for the 2008 party ID of D +8 (Dem 39, Rep 31, Ind 27).  But as @NumbersMuncher proved out, the real 2008 disparity was D +5 (Dem 37.5, Rep 32.5, Ind 30) while in 2004 it was R +5 (Dem 35, Rep 40, Ind 25). Far too many Democrats, but in this instance we see it is too many non-Whites and too few Whites.

The missing White vote

When the racial composition is correct, the Democrat over-sampling in polls like today’s D +9 means they are over-sampling White Democrats which hides the decline in support for Obama among White voters.  Now the polling outfits are fabricating racial demographics favorable to Obama with no reasonable justification and still only find Obama tied.  This is bad, bad, bad for the President. Importantly, the 2008 election racial demographics fail to account for 1.7 million White voters who stayed home in that election but appear to be more than enthusiastic this time around. In 2008 that was over 95,000 Ohio voters with a proclivity towards voting Republican.

If Gravis also gave us the preferences of these racial groups (I asked for them) we could easily rerun the numbers, but alas no such details are given like too many polling groups who don’t want pesky bloggers to blow their biased polls out of the water.

By keithbacker | Posted in Ohio | Also tagged , , | Comments (22)

Is Obama’s Ohio Firewall Really Just a Maginot Line?

October 20, 2012 – 12:17 pm

@NumbersMuncher (real name Josh Jordan) has the best read of the morning over at National Review’s “The Corner” where he dismantles the argument that Obama has any advantage in Ohio. The entire piece is a must read:

First, he breaks down the party ID in Ohio from 2008 of D +8 is wrong; the real spread was D +5.

Last cycle was a wave election and Barack Obama took Ohio by 4.6 percent, 51.5 to 46.9. The exit polls showed a split of 39 percent Democrats, 31 percent Republicans, and 30 percent independents. If that had been the actual turnout, according to exit polls’ measurement of how members of each party said they voted, Obama would have won 52.8 to 45.6, for a 7.2 percent margin victory, substantially bigger than the margin by which he actually won. This means that the exit polls were off a little, which is unsurprising since they are, after all, just polls. But we have actual vote totals to compare these polls to. If you use the exit-poll numbers for reported voting by party and then look at what kind of a turnout by party you’d need to get to the actual state vote tally, you come out with this breakdown: 37.5 percent Democrats, 32.5 percent Republicans, and 30 percent Independents (that gives you a vote of 51.6 percent for Obama and 46.9 percent for McCain — pretty close to actual results). So while the 2008 exit polls show an eight-point Democrat advantage, in reality it was likely closer to five percent.

Independents have swung 16-points in favor of Romney versus Obama’s 2008 performance.

In 2008 Obama beat John McCain by 8 percent among independents in Ohio. Of the seven current RCP polls that give independent numbers, Romney is up by an average of 8.7 percent.

Turnout in 2012 will not meet 2008 levels

Of the seven current RCP polls in Ohio, the average Democratic advantage in party ID is 5.5 percent. That is, if we assume 2008 advantage was D+5, as explained above, then the average poll in Ohio right now assumes a 2008-level turnout. While anything is possible on November 6, there are not many people on either side thinking Obama can match his 2008 turnout advantage.

Early voting favors Republicans versus 2008

[F]rom CNN today: “Four years ago, Democrats made up about 42 percent of the early and absentee vote while Republicans made up 22 percent. Through Wednesday, however, the margin has narrowed: Democrats account for 36 percent of the early and absentee vote while Republicans make up for 29 percent.” The current polls have been seriously inflated for Democrats because they’re reporting Obama with 30+ percent leads in early voting (which is then automatically counted in “likely voter” samples), which seems to be vastly overestimating the Democratic advantage among these voters. As CNN explains, Romney is making huge gains from 2008.

Conclusion

Obama won in 2008 largely because of a healthy lead among independents and a highly enthusiastic base’s turning out votes. Right now Romney is leading big with independents, has a more enthusiastic base, and is drawing crowds in Ohio that rival Obama’s. While he is down 2.5 points in the polls, the average poll is assuming 2008 turnout which is unlikely to repeat itself this year. Adding the fact that early voting is trending more Republican than in 2008, there is a lot of reason for optimism that this race is much closer than the current polls suggest. Not bad for a candidate who was declared dead in the state just a few weeks ago.

Just an outstanding deconstruction of the media/Obama spin on his Ohio firewall

By keithbacker | Posted in Ohio | Also tagged , , , | Comments (25)

Radio Interview, Party Identification and Polling Bias

October 19, 2012 – 5:12 pm

Since it’s Friday I decided to post a radio interview I did last month based on the blog post discussing the idea that Obama’s lead in polls was entirely due to over-sampling Democrats (September 16).

Let me say up front I make a great error of omission at the very beginning of this interview.  I knew we were going to discuss the party ID issue and the weighting of polls by party ID which pollsters like Scott Rasmussen do, but I failed miserably to explain that many polling firms do not weight by party ID and explain why their sampling is often equally erroneous.  I have explained the problem many times on this blog but in the linked interview I screwed up and made it seem like every polling firm weights by party ID when they do not.  I effectively restricted the discussion to firms that do weight by party ID when a wider explanation of the polling methodologies would have better served the listeners. It was my own fault because I was so locked in to talk about party ID weighting. At one point I can hear myself try to catch the error but just panicked and plowed ahead with my original point. To give a more complete picture, I excerpted the relevant portions on polling and party IDs from the above linked piece which must be read in companion with what I say in this interview.

September 18 Radio Interview: KID Newsradio 590 am/92.1 fm, Idaho Falls, ID. The Neal Larson Show

Polling bias and Party identification

The majority of polling critiques this election cycle focus almost exclusively on the amount of Democrats versus Republicans surveyed with the observation invariably there are far too many Democrats in the sample. There is much in dispute around this complaint because most polling organizations do not weight polls by the party identification of respondents. Polling organizations argue the disproportionately high amount of Democrats sampled draws a sharp inference there are more Democrats in the overall electorate, not just in the sample size. While it is possible and even probable there are a few more self-identified Democrats in the American electorate (the average in elections since 1984 is 3% more Democrats), the great dispute is the unusually large disparity of Democrats showing up in today’s polls, often as much as 7 to 12% higher than Republicans among the respondents. There are many reasons to challenge this conclusion which I will discuss later, but if we assume these polls have too many Democrats, an interesting phenomenon appears among which Democrats are oversampled.

Most polling methodologies, including how polls are weighted once responses are collected, mirror the Gallup Organization who has been the standard bearer in the US for over 75 years. According to the organization, “After Gallup collects and processes survey data, each respondent is assigned a weight so that the demographic characteristics of the total weighted sample of respondents match the latest estimates of the demographic characteristics of the adult population available from the U.S. Census Bureau. Gallup weights data to census estimates for gender, race, age, educational attainment, and region.” Based on explanations like this there is little reason to suspect intentional political bias in the disparate party weighting, especially when they do not re-weight polls by party identification.

…

When we reflect on accusations of bias in polling based on party identification, it seems hard to justify when most organizations do not adjust their polls based on this metric. These organizations do, however, run the risk of confirmation bias where the media and polling firms have a predilection towards one candidate and upon achieving results they agree with fail to challenge outlier data like unrealistic Democrat turnout levels in 2012. Inconvenient poll compositions like the fantastical party identification of respondents shake the credibility of desired outcomes but no explanation is given for such oddities. This leaves more fair-minded poll watchers uneasy with the factual reporting on data with obvious internal issues while partisans react more strongly with bias accusations not substantiated based on the available data.

This only raises the question of where the polling firms are getting their samples from — possibly heavy Democrat districts — because the end results are party identification breakdowns unrealistic in today’s electorate…polling outfits thus far have consistently sampled an election turnout often greater than candidate Obama’s 2008 best-in-a-generation advantage. [If this is not based on intentional bias] that means something else is going on. But the polling organizations shrug their shoulders and have been found to say the losers in the results are just crying sour grapes. This is even though their sample outcomes have party identification splits unrealistic beyond any stretch of reason. Sadly no credible defense is given for the unusual party split in these results which gives rise to charges of bias whether intentional or accidental. If the polling firms believe today’s electorate will exceed the incredible 2008 advantage Obama achieved they should make the argument to justify results that contain suspect internal data.

[L]ikely due to “confirmation bias” the media and polling organizations report favorable results for President Obama without challenge. There are many explanations for odd internal data in polls as well as the built in accuracy issues that come with the very nature of polling. As Michael Barone writes, “it’s getting much harder for pollsters to get people to respond to interviews. The Pew Research Center reports that it’s getting only 9 percent of the people it contacts to respond to its questions. That’s compared with 36 percent in 1997.” But consistently unrealistic sample outputs give rise to greater scrutiny from the polling outfits and media organizations who report the results uncritically for whatever their reasons may be.

By keithbacker | Posted in General | Also tagged , , | Comments (5)

Election Night Surprise: Why Minnesota Will Turn Red on November 6

October 19, 2012 – 12:55 am

Minnesota has crept into the news cycle recently with senior campaign surrogates stumping in the state and campaign dollars flowing to a state once thought out of reach for Republicans this cycle. I received a lot of push-back over my conclusion regarding Minnesota’s competitiveness based on Rochester, Minnesota being a top 10 ad market this week. Upon closer inspection, however, the evidence keeps piling up that the Land of 10,000 Lakes should be on everyone’s radar for an election night surprise.

The latest is a poll released yesterday from SurveyUSA gives President Obama a 10-point lead over Mitt Romney, 50 to 40.  Romney leads by 3-points among Independents 45 to 42 with 4% are voting 3rd party and 6% are Undecided.  Shouldn’t a 10-point lead definitely mean it is not a Battleground? If you believe that, you must be new to this blog.  A 10-point lead would largely be safe at this juncture if the poll were an honest representation of Minnesota today (and remember other polls have it as close as 4). But this SurveyUSA poll is far from a fair representation of the Minnesota electorate.  Before I get too deep into this flawed poll I give SurveyUSA credit for making all of the data available unlike too many other polling firms.   This allows critics to make their own judgements on what is lying beneath the top-line numbers and justify their criticism with facts and figures and not just flippant calls of bias.

Party ID

The biggest issue with the poll is the Party ID.  I understand polling firm do not weight their polls by party ID, but when they re-weight their polls based on age, race, etc and the outcome of respondents is a party ID disparity that defies all logic and reason, that means something is deeply wrong with the sample group they gleaned their answers from.   This poll specifically has a disgraceful disparity between Democrats and Republicans surveyed. The Party ID is D +10 (Dem 37, Rep 27, Ind 30).  This compares to 2008 of D +4 (Dem 40, Rep 36, Ind 25) and 2004 of D + 3 (Dem 38, Rep 35, Ind 27).  First and foremost we see a Democrat advantage in the sampling 2.5x greater than that the 2008 peak of hopey-changey.  This is ludicrous for many of the enthusiasm reasons often cited:

Basically, for the above reasons there is a dramatically increased probability Republicans will vote with greater propensity in 2012 than they did in 2008 while the inverse is true for Democrats.  But that is only one of the reasons this poll and party ID are ludicrous.

The Changing Minnesota

Minnesota is no ordinary state politically. The most important thing to know about Minnesota is the state is rapidly changing in favor of Republicans and has been for years.  Sean Trende of Real Clear Politics did the seminal work in this regard. His state-by-state analysis of voting trends shows that over the last 8 elections the performance of Republican candidates in Minnesota has steadily gotten closer to the candidate’s national performance. That is, when a Republican Presidential candidate got 50.1% of the vote in 1980, you could expect them to get 43% in Minnesota.  In 2008, the under-performance versus the national shrunk to 1%, meaning had John McCain received 51% of the national vote, he could have expected to get 50% of the vote in Minnesota.

If we apply only the most superficial analysis of Minnesota today and we see Mitt Romney leading  in the national average by only a few points, let alone as much as 7-points, it would seem more than likely he would carry Minnesota based solely that lead and the 30-year trend in Republicans favor shown above.

Party ID Recent History

This brings me back to the Party ID issue. In 2008 Minnesota’s political affiliation change was unlike most of America. Between 2004 and 2008 we saw Battleground States experience wide swings in their party ID as citizens bought into the magnetic story of Barack Obama.  States like Ohio saw its Party ID swing 13-points in favor of Democrats; Nevada swung 12-points in favor of Democrats; North Carolina swung 12-points in favor of Democrats; Virginia swung 10-points; New Hampshire 9-points; and on and on all towards the Democrats.  Minnesota’s party affiliation, however, only swung 1-point towards the Democrats.  This was smaller than every party affiliation move among even the most remote of contested state.  This lack of change during the Democrat tidal wave of 2008 is a major component of hidden Republican strength demonstrated in the chart above.  Minnesota is, and has been, a state trending steadily Republican even in the face of the incredible Democrat wave seen in 2008.

Republicans have surged in local politics

But more than just a macro statistical argument, Republicans have made dramatic strides at the ballot box. Over the last four years Minnesota caught the 2010 midterm wave for the GOP and flipped both houses of its state legislature, and in dramatic fashion:

  • Republicans gained a State Senate majority of 55/45 which was a dramatic shift from the 31/69 disparity previously (based on percentages not actual seats)
  • Republicans also gained a State House e majority of 54/46 which was also a dramatic shift from the 35/65 split previously (based on percentages not actual seats)
  • At the Federal level, Republicans picked up one seat balancing out the Congressional delegation at 4 for each party

Despite all of these substantial moves in favor of the Republican party, polls like SurveyUSA’s D +10 turnout still show up and is the basis for people to argue Minnesota is out of reach this election.

But what if Minnesota is not D +10 or anywhere near that?

Below I breakdown the exact same SurveyUSA poll.  First, as they have it with D +10.  Second with the 2008 party ID of D +4 and then prospectively with an even party breakdown for all of the reasons outlined above.  I use with the same number of Independents for the 2012 estimate that SurveyUSA found in their poll although I suspect Independents will be even higher on election day. Pollsters, campaigns, or individuals can make their own assumptions, these are just mine.

This following needs to be clear up front:

  • Party ID does not equal 100% in the SurveyUSA poll so I make Other 6%. Also, the “Other” category in the survey was unusually high at 6% but instead of eliminating that I shaved 3% from the two major parties for 2008 and 2012 est. Also the Other voters overwhelmingly supported 3rd party candidates in the poll so their impact on this analysis is small.
  • The vote total also does not equal 100% because of Undecideds which are also 6%. Due to space constraints I put Undecideds and Other on the same line which can look, when reading from left-to-right, like there are 106% of voters.  But this is not Cook County, it is just labeled that way so you know where the numbers come from.  “Other” turns out to be +1% for each candidate and Undecideds are expected to break at least 2/3 for the challenger which is +2 for Obama and +4 for Romney. Note: adding in Undecideds to the SurveyUSA poll gives the candidates final totals of Obama 52 and Romney 44.

Calculating the vote

  • The vote totals are calculated through the following formula: (Rep Party ID x Rep Party support %) + (Dem Party ID x Dem Party support %) + (Ind Party ID x Ind Party support %) + (Other Party ID x Other Party support %) + (Undecided x Expected %) = vote total
  • For example, with  President Obama in the SurveyUSA Party ID section this breaks down as (27 x .06) + (37 x .93) + (30 x .42) + (6 x .17) + (6 x .33) = 51.7 or 52

Scenario 1: Survey USA Party ID 27 37 30 6 [6] Vote Total
Republicans % Democrats % Independents % Other % Undecided %
Obama 6 93 42 17 33  51.7
Romney 89 4 45 17 66  44.0
Scenario 2: 2008 Party ID 33 37 25 6 [6]
Republicans % Democrats % Independents %  Other % Undecided %
Obama 6 93 42 17 33  49.9
Romney 89 4 45 17 66  47.1
Scenario 3: 2012 Party ID est. 32 32 30 6  [6]
Republicans % Democrats % Independents % Other %  Undecided %
Obama 6 93 42 17 33 47.3
Romney 89 4 45 17 66 48.2

Our three scenarios produce the following results:

  • Party ID  D +10: Obama wins by 8
  • Party ID D +4: Obama wins by 3
  • Party ID even: Romney wins by 1

The point of running these scenarios is the initial read of an Obama 10-point lead based on a D +10 party affiliation is folly. With Undecideds factored in that lead drop to 8 even in this unrealistic scenario. If there is no party affiliation shift from 2008 despite the overwhelming evidence provided, Romney is only down 2.8 points with an unconsolidated base (think a visit might help?) as well as conservative estimates on Undecideds.  If, however, Republicans have burnished their brand and the enthusiasm issue is as meaningful as polling would indicate, the decades-long steady rise in Republican performance in Minnesota should deliver a victory for Romney on November 6.  Enhancing every one of these scenarios is the prospect of a decided national popular vote victory for Romney evidenced by the national tracking polls from Gallup and Rasmussen Reports. If that happens, deep purple Minnesota will turn red on election night.

By keithbacker | Posted in Columns, General | Also tagged , , , , | Comments (53)

Romney Actually Leading Based on Today’s ABC/Washington Post Poll

October 15, 2012 – 8:59 am

I’m hesitant to do this with every poll but after the below monstrosity from ABC/Washington Post, it was worth the time to rework the poll with a more reasonable election turnout.  The largest bone of contention is that the party identification had a Democrat advantage of 9 percentage points more Democrats surveyed than Republicans which in shorthand is D +9. This exceeds the best-in-a generation turnout advantage Obama had in 2008 which was 7 percentage points more Democrats or D +7.  But what if that absurd disparity in turnout was a more reasonable turnout of D +3 which is also the historic average over the last 7 Presidential elections?

First I will map out the poll as reported by ABC and the Washington Post

The party ID and vote totals according to ABC/Washington Post:

Democrat Republican Independent Other Total
Washington Post/ABC Party ID D +9 35 26 33 6 100
% of: Democrat Vote Republican Vote Independent Vote Other Vote
Barack Obama 91 7 42 24.5
Mitt Romney 8 93 48 53.0
Party ID * % vote Party ID * % vote Party ID * % vote Party ID * % vote
Obama Vote 31.85 1.82 13.86 1.47 49
Romney Vote 2.8 24.18 15.84 3.18 46

We see in the final column on the right, we end up with the exact vote split of 49 to 46 favoring Obama according to ABC/Washington Post. The only issue with the above calculation is the “Other” category for support which breaks for Romney 53 to 24.5. I had to back into those #s since they are not provided. If someone sees the actual breakdown I’d be happy to adjust the above calculations. The differences can only be a couple tenths of a percent from what I have  and it won’t affect the outcome but I wanted to point that out up front.

Now what happens if we leave fantasy land where Democrats outnumber Republicans by 9 percentage points and we simply go with the historical average of D +3?

Below I map out the exact same poll but with a more reasonable party ID based on today’s electorate.

Democrat Republican Independent Other Total
Washington Post/ABC Party ID D +3 32 29 33 6 100
% of: Democrat Vote Republican Vote Independent Vote Other Vote
Barack Obama 91 7 42 24.5
Mitt Romney 8 93 48 53.0
Party ID * % vote Party ID * % vote Party ID * % vote Party ID * % vote
Obama Vote 29.12 2.03 13.86 1.47 46.48
Romney Vote 2.56 26.97 15.84 3.18 48.55

Now we see the lead has flip-flopped and coincidentally, if we use rounded numbers, the vote totals are the exact opposite of the ABC/Washington Post results and Mitt Romney is leading 49 to 46. But I included the decimal places because you can see the real lead is only 2.07-points and the rounding misleads on the overall margin.

This is why voter registration, enthusiasm and turnout are all essential to any successful campaign. Mitt Romney is almost certainly leading in this election but if Democrats meet or exceed their advantage at the ballot box like they achieved in 2008, they will almost certainly win the election. Thankfully as we have shown countless times, as in the take down of the ABC/Washington Post poll below, all the evidence points to a dramatically different turnout on election day in favor of Republicans relative to 2008.

By keithbacker | Posted in General | Also tagged , | Comments (21)

ABC/Washington Post Poll Makes In-Kind Contribution to Obama Re-Election Efforts

October 15, 2012 – 1:49 am

Not to be outdone by the ludicrous NBC/WSJ/Marist Battleground State polls from last week, ABC/Washington Post reveal their national poll today showing President Obama with a 3-point lead 49 to 46. Mitt Romney leads among Independents by 6 points (48 to 42) and locks down his base more so than Obama — Reps support Romney 93 to 7 while Dems support Obama 91 to 8.  Yet Romney trails by 3.  How?  Incredibly, they polled 9% more Democrats than Republicans.  This is not a new phenomenon as I outlined in mid-September in the post “Obama’s National Lead Based Entirely on Over-Sampling Democrats.”  Today’s ABC/Washington Post poll is the crowning achievement this cycle in unrealistic national polls only 3 weeks out from the election.  But these types of advocacy “polling games” are nothing new.

The party identification in the survey is D +9 (Dem 35, Rep 26, Ind 33). This compares to 2008 when party ID was D +7 (Dem 39, Rep 32, Ind 29) and 2004 when party ID split evenly (Dem 37, Rep 37, Ind 26).  Making matters even worse, in their poll just over two weeks ago that survey had a party ID of D +3 (Dem 33, Rep 30, Ind 33).  Did the public tune in to Barack Obama’s debate performance and just have a groundswell of love for Democrat passivity and listlessness and embrace the Donkey Party? According to the Washington Post, pre-Debate the race was 48 to 46 in favor of Obama.  Post-debate the race is 49 to 46 in favor of Obama.  Must have been an uneventful debate right?  Here is over-the-top liberal Democrat Andrew Sullivan’s blog yesterday on post debate polls:

If anyone thought that the feisty Biden debate undid the massive damage the president did to himself in the first debate, the news isn’t great. Biden does seem to have reversed the speed of Obama’s free-fall but not the decline itself. Romney’s debate obliteration of Obama – something that, in my view, irreparably damages a sitting president – does not seem to be a bounce, but a resilient jump. It’s not going away by itself. That is: not a bounce.

Sullivan also provides a devastating chart showing the post-debate Romney surge in polls (red line) and Obama free fall (blue line):

But today the Washington Post and ABC see fit to publish a poll with Democrat affiliation 9 percentage points greater than Republicans. This blog has hammered the issue of party ID time and again. Basically there is a zero percent change the Democrat’s advantage at the polls in 2012 will be superior to their advantage in 2008. Here is what I wrote on October 1st when critiquing the large disparities in party identification:

In 2008 seven percent more Democrats than Republicans identified themselves as such on election day, well above the historic average of 3%. This was a big change from 2004 when party identification was evenly split between the Democrats and Republicans. But there were many reasons for the strong Democrat turnout that do not exist today. The top of the ticket was a historic candidate (first Black President), America had war and Bush fatigue, the financial meltdown created an anti-Republican wave, and his opponent wasn’t the strongest (good biography, bad and underfunded candidate). These factors led to a strong Democrat self-identification advantage at the voting booth in 2008. But in the 2012 election, none of the advantages outlined above are there for Obama and many of those factors are now largely working against the President: 8%+ unemployment for three years, sub-2% GDP, 23 million unemployed, Arab Spring blowing up and casting the historic vote in 2008 is yesterday’s news. Additionally the Romney campaign ground game has exceeded the McCain campaign across many metrics as much as 10- to 15-fold. Despite the stark changes in each of these factors, polling outfits thus far have consistently sampled an election turnout often greater than candidate Obama’s 2008 best-in-a-generation advantage.

Over the last month we have seen:

Interestingly many of these above trends actually show up in the ABC/Washington Post poll.  President Obama’s support among Non-Whites is a surprisingly low 73%.  His support is typically closer to 80% so this drop of is a major red flag in the President’s re-election efforts.  But this is where the Democrat over-sampling comes in to save the President.  I went to great lengths to demonstrate that these polls that over-sample Democrats are not simply over-sampling generic Democrats, these polls very specifically over-sample White Democrats.  And in this survey we see Barack Obama’s support among White voters at 43%, the same level he achieved in 2008. If that percentage was accurate Obama would almost certainly be re-elected.  Unfortunately for him that support level is not accurate based on the unrealistic disparity in party identification and the over-sampling of Democrats masks what is far more likely support for Obama among Whites closer to 36 or 37% as I explained in the previous post here.

Despite the mountain of evidence above completely undermining the unrealistic voter turnout models presented by ABC, the Washington Post and others, major news organizations pass off these unserious polls as credible when neither sense nor reason supports such claims.  Today’s disaster is only the latest example of major news organizations weakening the public’s trust by publishing fantastical polls whose sole purpose is to advocate for one candidate over the other.

By keithbacker | Posted in General | Also tagged , , , , , , , | Comments (24)

Will the Party ID really look like R + 2 or R +3?

October 8, 2012 – 4:52 pm

The just released Pew Research poll showing Mitt Romney up 4-points has the political blogosphere and twitterverse (can we get one name for the two?) up in arms over the party identification of respondents which was R +2.75 (Dem 32.5, Rep 35.25, Ind 29.5). But is that any more correct than the D +7 or D +10 samples we on the Right have indignantly complained about all election cycle? We have asked this question before and have seen detailed analysis of the trends and historical data which on average has been D +3 since 1980.

But what struck me in the Pew Poll was the party identification is  oddly similar to the current Rasmussen party identification that he releases on a monthly basis. Why do we care how close it is to Rasmussen?  Because he nailed the last two elections despite wild swings in the electorate’s preference.

Rasmussen Reports released his party ID for September last week and commenter “blcjr” took the raw data Rasmussen makes available to show you how trends have changed over time from 2004 through today.  Below is Rasmussen’s month-end party ID for each October in Presidential years and therefore the result immediately before the actual election.  We compared that with the exit polling party ID provided by the Winston Group:

Year Rasmussen Actual
2004 D +1.5 (Dem 38.7, Rep 37.2) D +0 (Dem 38, Rep 38)
2008 D +7.1 (Dem 40.3, Rep 33.3) D +7 (Dem 40, Rep 33)
2012 (Sep) R +2.6 (Dem 34.2, Rep 36.8) ?????

In the two prior Presidential election years Rasmussen essentially nailed the party identification and accurately captured the ground swell in favor of Democrats in 2008.  Not coincidentally Rasmussen called the 2004 election within 1% and nailed the 2008 election on the nose.  If party identification is to Republicans advantaged by near +2.6%, Romney should win overwhelmingly, much like the results we see in today’s Pew Poll with a party identification of R +2.75.

Chart compliments of commenter “blcjr”:

Party identification over time since 2004 through September 2012

By keithbacker | Posted in General | Also tagged , , , , | Comments (13)

Can We Believe the Presidential Polls?

October 4, 2012 – 2:17 pm

Karl Rove knows a lot more about polls than just about anyone on the planet and he has a lucid column on the state of Presidential polling with some great references to 1980 and 2004. The truth of the matter is today’s contest is a close race with momentum waxing and waning between the two camps.  Right now Team Romney is riding high but expect the President to come back hard in the next debate.  No one gives up the crown without a fight and to expect anything less would be a great disservice to what will likely be a struggle to the last day. Also, based on Rove’s examples below he must read this blog ;)

First, the open embrace of pro-Democrat polls to help Jimmy Carter in 2980:

On Oct. 8, 1980, the New York Times released its poll on the presidential race in Texas, one of 10 battlegrounds. (Yes, the Lone Star State was then a battleground.) According to the Times, the contest was “a virtual dead heat,” with President Jimmy Carter ahead despite earlier surveys showing Ronald Reagan winning… Then came more hard punches. On Oct. 13, Gallup put the race nationally at Carter 44%, Reagan 40%. The bottom appeared to fall out two weeks later when a new national Gallup poll had Carter 47%, Reagan 39%. That produced more than a few empty chairs in phone banks across Texas. But most volunteers, grim and stoic, hung on, determined to stay until the bitter end. Only Election Day was not so bitter. Reagan carried all 10 of the Times’ battleground states and defeated Mr. Carter by nearly 10 points.

2012 polling

In the past 30 days, there were 91 national polls (including each Gallup and Rasmussen daily tracking survey). Mr. Obama was at or above the magic number of 50% in just 20. His average was 47.9%. Mr. Romney’s was 45.5%.

2004 polling (the last time an incumbent was running for re-election)

There were 40 national polls over the same period in 2004. President George W. Bush was 50% or higher in 18. His average was 49%; Massachusetts Sen. John Kerry was at 43.8%.

Media misrepresentations versus reality

An Oct. 4, 2004, story in the New York Times declared the Bush/Kerry race “a dead heat” and asked “whether Mr. Bush can regain the advantage.” Mr. Bush was hitting the vital 50% mark in almost half the polls (unlike Mr. Obama) and had a lead over Mr. Kerry twice as large as the one Mr. Obama now holds over Mr. Romney. So why was the 2004 race “a dead heat” while many commentators today say Mr. Obama is the clear favorite? The reality is that 2012 is a horse race and will remain so. An incumbent below 50% is in grave danger. On Election Day he’ll usually receive less than his final poll number. That’s because his detractors are more likely to turn out, and undecideds are more resistant to voting for him.

Unrealistic state polls

Then there is the tsunami of state-level polls. Last week, there were 46 polls in 22 states; the week before, 52 polls in 18 states; and the week before that, 41 polls in 20 states. They’re endowed by the media with a scientific precision they simply don’t have.

Take last week’s CBS/New York Times Florida survey, which had Mr. Obama leading Mr. Romney by nine points. The poll sampled more Democrats than Republicans—nine percentage points more. Yet the Democratic advantage in the 2008 presidential exit polls was three percentage points. Does it seem probable that Florida Democrats will turn out in higher numbers in 2012, especially when their registration edge over Republicans dropped by 22% in the past four years?

On Aug. 2, radio talk-show host Hugh Hewitt asked Peter Brown, assistant director of the Quinnipiac University polling organization—which runs the CBS/NYT battleground state polls, including last week’s Florida poll—if he expected a Democratic advantage in the Sunshine State three times what it was last time. Mr. Brown responded that “I think it is probably unlikely,” but defended his polling organization’s record.

By keithbacker | Posted in Florida, General | Also tagged , , | Comments (10)

Obama Enthusiasm Down 10pts, Romney Enthusiasm Up 10pts

October 2, 2012 – 6:00 pm

I have hammered this point countless times when enumerating reason after reason why the turnout results in polling is unrealistic and not reflective of today’s electorate. Obama had a record advantage in 2008 and the enthusiasm among his troops is meaningfully lower today while the enthusiasm among his opponent’s troops is equally higher.  According to the latest Washington Post/ABC News poll, the two camps are approximately even in enthusiasm:

President Obama may win re-election, but if he does it appears it will be without the legions of fervent supporters and big enthusiasm that propelled him to victory four years ago. In the latest Washington Post-ABC News poll, Mitt Romney’s supporters are about as enthusiastic about supporting him in November as are those who back Obama. This is a stark contrast from the Post-ABC poll earlier this month, as well as throughout 2008, when Obama had huge leads over Sen. John McCain in this area.

In the new poll, 51 percent of  voters who support Obama are “very enthusiastic” about supporting him, as are 48 percent of Romney backers who are very enthused. Romney came close to Obama on enthusiasm once before, closing the gap to six percentage points in August.  Compared with four years ago, Obama’s supporters are less enthusiastic, while Romney’s are clearly more so than were McCain’s. Obama’s level of intense enthusiasm is down 10 percentage points from late September 2008, when 61 percent were “very enthusiastic.” Strong enthusiasm for Romney is 10 points higher than it was for McCain at the same point in the campaign.

The four-year drop-off in Obama enthusiasm has been most stark among ideological moderates. More than six in 10 moderate Obama voters were “very enthusiastic” about supporting him In a Post-ABC poll at this point in the 2008 campaign. Today, just 42 percent of Obama’s moderate voters are highly enthusiastic. Liberals, by contrast, are just as enthusiastic about Obama now (65 percent very enthusiastic) as four years ago (61 percent). For Romney, the improvements over McCain are among conservatives, older voters and among evangelicals. For instance, a 56 percent majority of white evangelical Protestants are “very enthusiastic” about voting for Romney, up from 36 percent who were jazzed about McCain. The spike among this group may come as a surprise, given that evangelical Christians were among Romney’s weaker groups in the Republican primaries this year.

By keithbacker | Posted in General | Also tagged , , , | Comments (5)

We Ask America’s Fabulously Awful Poll of Nevada (and Missouri)

October 2, 2012 – 12:06 pm

It is still comical to see these gawd-awful poll internals presented as credible polls. When We Ask America first burst onto the scene Leftys cried they were a partisan pro-Republican polling form because they were owned by the Illinois Manufacturers’ Association. By that they cried foul because this was sponsored by a business association of manufacturers and anyone pro-Business had to be a biased Republican outfit.

Now We Ask America time and again comes out with poll results impossible to believe that get passed off as credible. In Missouri, where Mitt Romney allegedly only leads by 3-points, he outpaces Obama among Independents by 16-points. This must be the huge Democrat wave we are seeing in the Show Me State that led President Obama to not even campaign in a state he lost by only 4,000 votes in 2008. A state with such a Democrat fervor the most unelectable Republican this cycle has pulled to even with the incumbent Democrat in the Senate race.

But we don’t blog Missouri so we’ll move on to Nevada with a comical lead for Obama of 10.5-points with Mitt Romney favored among Independents by 15.4-points. The Nevada Democrat party is doing laps around the state GOP party right now on voter registration erasing all of the gains the GOP made ahead of the Ron Paul takeover, but this is ludicrous. Nevada’s Democrat advantage at the voting booth was 8-points in 2008 (Dem 38, Rep 30, Ind 12) flipping the GOP advantage in 2004 of 4-points (Dem 35, Rep 39, Ind 26). Based on the absurd Obama lead and Romney’s margin with Independents, the survey has to have a 20+ point advantage of Democrats surveyed. Truly astonishingly awful. Definite winner of most in-the-bag pollster of the day.

By keithbacker | Posted in Nevada | Also tagged , , , | Comments (13)

Hiding the Decline: What Polls Over-Sampling Democrats Mask

October 1, 2012 – 12:49 pm

The over-sampling of Democrats in today’s polls most likely hides a sharp decline in support for President Obama among White voters. If President Obama’s support level among White voters dips a single percent or two below 40, his road to re-election would be in jeopardy. The national polling results today showing President Obama with support levels among Whites between 40-44% likely over-sample support for President Obama by 4% to 8% among this demographic. Poll re-weighting by race achieves an accurate demographic make-up for the United States in 2012 but almost certainly a wholly unrealistic split between self-identified Democrats and Republicans. Because the accurate re-weighting of polls by race often achieves political splits that are not credible, polling organizations give rise to accusations of bias when in reality better selected sample inputs would most likely achieve more credible end results but also meaningfully worse results for the President.

Problems with polls

The majority of polling critiques this election cycle focus almost exclusively on the amount of Democrats versus Republicans surveyed with the observation invariably there are far too many Democrats in the sample. There is much in dispute around this complaint because most polling organizations do not weight polls by the party identification of respondents. Polling organizations argue the disproportionately high amount of Democrats sampled draws a sharp inference there are more Democrats in the overall electorate, not just in the sample size. While it is possible and even probable there are a few more self-identified Democrats in the American electorate (the average in elections since 1984 is 3% more Democrats), the great dispute is the unusually large disparity of Democrats showing up in today’s polls, often as much as 7 to 12% higher than Republicans among the respondents. There are many reasons to challenge this conclusion which I will discuss later, but if we assume these polls have too many Democrats, an interesting phenomenon appears among which Democrats are oversampled.

Most polling methodologies, including how polls are weighted once responses are collected, mirror the Gallup Organization who has been the standard bearer in the US for over 75 years. According to the organization, “After Gallup collects and processes survey data, each respondent is assigned a weight so that the demographic characteristics of the total weighted sample of respondents match the latest estimates of the demographic characteristics of the adult population available from the U.S. Census Bureau. Gallup weights data to census estimates for gender, race, age, educational attainment, and region.” Based on explanations like this there is little reason to suspect intentional political bias in the disparate party weighting, especially when they do not re-weight polls by party identification. Importantly, though, they do re-weight polls by race. This gives rise to some curious issues regarding support levels for the President today.

Racial demographics and voting preference in the US

In the 2008 election, the racial breakdown of the national voting public was 74% White, 12% Black, 9% Hispanic, 2% Asian, 3% Other. When a survey is conducted polling organization re-weight the respondent answers to ensure the each of these groups has accurate representation in the final results. Most national polls reflect this reality usually within a 1% variation for any group.

If you look how each one of these groups vote, you find outsized rates of support for Obama among the non-White groups: Blacks ~95%, Hispanics ~70, Asians ~65%. These levels of support for Democrats are consistent with most modern elections although President Obama has been able to boost these levels slightly above historic averages. In aggregate, non-White support for Obama is roughly 80% in nearly every survey. At the same time Obama, like Democrat Presidential candidates before him, struggles with the White vote. In these same polls, Obama typically averages 40-44% support among White voters. His 2008 support level was 43% and it is widely believed by the Obama campaign among others that he needs support of at least 40% Whites to win the election.

Low potential for over-sampling non-White support for Obama

If we consider the idea that polling today has large over-samples of Democrats, the consistently high percentage of support for Obama among non-Whites makes it almost impossible to over-sample minority groups. First there is not a lot of room for support increases and second, data on the voting trends in non-White groups is often achieved through demographic specific polling of solely Blacks or Hispanics for example. Hence, any over-samples in the non-White demographic would meaningfully alter the already high levels of support for Obama and reveal itself as inconsistent with independent polling. Additionally, any over-samples in the non-White demographic would almost certainly change the racial make-up of the survey and set off red-flags to anyone scrutinizing polls. Therefore it is highly unlikely over-sampled Democrat polls contain an excess amount of non-White voters.

White Democrats

This leaves only White Democrats as the over-represented respondent in these polls that arguably over-sample Democrats. If the average in election turnout since 1984 is 3% more Democrats and these polls have 7 to 11% more Democrats, that means the polls specifically have 4 to 8% more White Democrats surveyed in their likely voter results. The problem for the Obama campaign is if his support level among White voters (74% of the voting public) is between 40%-44% and that support is based on a sampling that over-states his support 4 to 8%, his real level of support is probably closer to 36% or 37%. This is meaningfully below the campaign’s own magic level of 40% and is a huge danger zone for any Presidential candidate no matter how much anyone may spin the demographic changes in today’s America.

Hiding the decline

The issue with the suspect polling internals and media embrace of the figures is the consistent lead for Obama would be immediately challenged if his support levels dropped dramatically among the outlined racial groups. Support levels of 60% among Hispanics (9% of the voting public) or 80% among Blacks (12% of the public) would jump off the page to poll watchers. The same holds true for support levels of 36/37% among Whites (74% of the voting public). It would be near impossible for Obama to win the Presidency with support levels like the ones I just outlined. Unfortunately support for President Obama among White voters has declined from 43% in 2008 to apparently as low as 36%-37% in today’s polls absent unrealistically high levels of self-identified Democrats. With White voters making up 73-74% of the electorate and support levels in the upper 30s, it is inconceivable President Obama has the advantage these polls lead readers to believe. But the results largely go unchallenged in the media despite the impractical internal party identification make-up.

Polling bias and Party identification

When we reflect on accusations of bias in polling based on party identification, it seems hard to justify when most organizations do not adjust their polls based on this metric. These organizations do, however, run the risk of confirmation bias where the media and polling firms have a predilection towards one candidate and upon achieving results they agree with fail to challenge outlier data like unrealistic Democrat turnout levels in 2012. Inconvenient poll compositions like the fantastical party identification of respondents shake the credibility of desired outcomes but no explanation is given for such oddities. This leaves more fair-minded poll watchers uneasy with the factual reporting on data with obvious internal issues while partisans react more strongly with bias accusations not substantiated based on the available data. The over-sampling of Democrats may not be showing the bias of polling organizations but it is likely hiding the decline of dwindling White support for Obama.

This only raises the question of where the polling firms are getting their samples from — possibly heavy Democrat districts — because the end results are party identification breakdowns unrealistic in today’s electorate. In 2008 seven percent more Democrats than Republicans identified themselves as such on election day, well above the historic average of 3%. This was a big change from 2004 when party identification was evenly split between the Democrats and Republicans. But there were many reasons for the strong Democrat turnout that do not exist today. The top of the ticket was a historic candidate (first Black President), America had war and Bush fatigue, the financial meltdown created an anti-Republican wave, and his opponent wasn’t the strongest (good biography, bad and underfunded candidate). These factors led to a strong Democrat self-identification advantage at the voting booth in 2008. But in the 2012 election, none of the advantages outlined above are there for Obama and many of those factors are now largely working against the President: 8%+ unemployment for three years, sub-2% GDP, 23 million unemployed, Arab Spring blowing up and casting the historic vote in 2008 is yesterday’s news. Additionally the Romney campaign ground game has exceeded the McCain campaign across many metrics as much as 10- to 15-fold.

Despite the stark changes in each of these factors, polling outfits thus far have consistently sampled an election turnout often greater than candidate Obama’s 2008 best-in-a-generation advantage.

That means something else is going on. But the polling organizations shrug their shoulders and have been found to say the losers in the results are just crying sour grapes. This is even though their sample outcomes have party identification splits unrealistic beyond any stretch of reason. Sadly no credible defense is given for the unusual party split in these results which gives rise to charges of bias whether intentional or accidental. If the polling firms believe today’s electorate will exceed the incredible 2008 advantage Obama achieved they should make the argument to justify results that contain suspect internal data. But they would also have to explain why the 2008 election gave Democrats massive majorities in the House of Representatives yet today’s electorate will likely return massive majorities in the House to Republicans. It defies all logic. But very likely due to “confirmation bias” the media and polling organizations report favorable results for President Obama without challenge.

There are many explanations for odd internal data in polls as well as the built in accuracy issues that come with the very nature of polling. As Michael Barone writes, “it’s getting much harder for pollsters to get people to respond to interviews. The Pew Research Center reports that it’s getting only 9 percent of the people it contacts to respond to its questions. That’s compared with 36 percent in 1997.” But consistently unrealistic sample outputs give rise to greater scrutiny from the polling outfits and media organizations who report the results uncritically for whatever their reasons may be.

By keithbacker | Posted in Columns, General | Also tagged , , , , , , , | Comments (33)

When it comes to polls: readers beware — Michael Barone

September 30, 2012 – 1:56 pm

The dean on conservative election watching and an authority on all American politics, Michael Barone has probably forgotten more about polling and elections than every TV talking head, blogger and reader combined. In his latest column he lays out the trends and controversy in today’s polling with his usual insights and lucid reasoning.  I excerpted the closing portion of his column, but it’s worth reading the whole thing:

I don’t believe that any of the media pollsters have been tilting their results in order to demoralize Republicans, though I do look with suspicion on the work of some partisan pollsters.

But I do have my doubts about whether samples with more Democratic Party identification than in 2008 are accurate representations of the electorate. Many states with party registration have shown big drops in registered Democrats since then.

Pollster Scott Rasmussen, who weights his robocall results by party identification, adjusted monthly, has shown a much closer race than most pollsters who leave party identification numbers unweighted. So has the Susquehanna poll in Pennsylvania.

It may be that we’re seeing the phenomenon we’ve seen for years in exit polls, which have consistently skewed Democratic (and toward Barack Obama in the 2008 primaries). Part of that is interviewer error: Exit poll pioneer Warren Mitofsky found the biggest discrepancies between exit polls and actual results were in precincts where the interviewers were female graduate students.

But he also found that Democrats were simply more willing to fill out the exit poll. Which raises the question: Are we seeing the same thing in this month’s polls?

By keithbacker | Posted in General | Also tagged | Comments (18)

Hugh Hewitt Takes on the Pollsters

September 29, 2012 – 1:23 pm

Radio host and conservative columnist Hugh Hewitt has done fantastic work this season interviewing pollsters and asking the tough questions looking for answers how allegedly reliable polls have such unrealistic internal make-ups.   In his Townhall column he lays out numerous arguments to challenge the data including two objective data points that make the Ohio sampling more obviously incorrect:

There are plenty of data points to encourage Republicans, and these are genuine data points as opposed to the junk food offered up by Quinnipiac and Marist, which derived their predictions from samples that included enormous Democratic voter margins in key states, pro-Democratic turnout margins that were even greater than those achieved in Obama’s blowout year of 2008..

Two data points that warm GOP hearts and undermine the junk polls: (1) Absentee requests in Ohio by Democrats are trailing their 2008 totals –often by a lot in key Democratic counties like Cuyahoga County; and (2) overall voter registration for Democrats in the Buckeye State is down dramatically from 2008.

These two bits of info undermine the credibility of the Obama booster polls, as did the interviews I conducted with key leadership from both polls and with other informed observers.

In addition to doing the media’s job actually finding the data to challenge the assumptions, Hewitt has used his radio shows to go right to the sources on polling and how we should interpret the data.  After numerous interviews Hewitt provides five major takeaways:

  • The pro-Obama pollsters don’t have answers as to why their skewed samples are trustworthy beyond the fact that they think their approach to randomness is a guarantee of fairness, and they seem to resent greatly that the questions are even asked. Like [Convicted fraud Bernie] Madoff would have resented questions about his stunning rate of return.
  • Barone notes that percentage turnout by party in a presidential year hasn’t been much greater for the president’s party than it was in the preceding off-year, which makes samples outstripping even the 2008 model of Democratic participation “inherently suspicious.”.
  • Cost notes that Romney is winning the independent vote in every poll, which also makes big Obama leads suspect.
  • And my conversation with Mr. Shepard, whose employer National Journal has a reputation for the best non-partisan work inside the Beltway, didn’t find any academic, disinterested support for the proposition that party identification cannot be weighted because of the inherent instability of the marker.
  • The biggest unanswered question of all: If party ID is so subject to change that it should not be weighted according to an estimate of turnout, why ask about it at all? And if it is for the purpose of detecting big moves, as Mr. Shepard argued, why not report that “big move” in the stories that depend upon the polling?

There are a number of reasons polling organizations could offer for their curious sampling but they offer no defense of these results other than it is consistent with the prior election which fails to take into account that admitted notion that party identification changes every election and the current samplings do not reflect the reality of today’s electorate.

By keithbacker | Posted in General, Ohio | Also tagged , , , , , | Comments (45)

Why the Q-poll is a qrock — New York Post

September 27, 2012 – 4:00 am

If you’re feeling motivated, please visit the actual column linked here. Leave a comment on the Post’s website regarding the column (if available), email the NY Post link to the universe and tweet it out from the New York Post website as many times as you like if you have a twitter account.  It’s all I’ll ever ask of you. — “Keith”

Why the Q-poll is a qrock

By KEVIN PATRICK

Yesterday’s Quinnipiac Poll results were great news for President Obama: The Q-poll has him ahead by substantial margins in three battleground states — up 9 points in Florida, 10 in Ohio and 11 in Pennsylvania. Take a closer look: These numbers seem less like a scientific effort to measure how the campaign is going, and more like a drive to push it in that direction.

Simply put, the Quinnipiac surveys oversample Democrats; they only make sense if we believe that President Obama’s supporters are going to turn out in even greater numbers (relative to Republicans) than they did in 2008. Questions about the partisan makeup of poll samples have been rampant this election season, and Quinnipiac (which recently partnered up with CBS and The New York Times for polling) is at one extreme of that debate.

The 2008 election was a banner year for Obama and Democrats in general. The top of the ticket was a historic candidate (now our nation’s first black president); America had war and Bush fatigue, and the financial meltdown created an anti-Republican wave. Plus, Obama’s opponent, despite a great biography, was a poor candidate — a foreign-affairs and military expert running when the economy was the issue. John McCain’s campaign also had far less cash to spend.

All these factors helped yield to a strong Democratic advantage at the voting booth — where more voters identified themselves as Democrats than Republicans by a remarkable 7 points, 39 percent to 32 percent. This was the best advantage for Democrats in over a generation; in polling shorthand, we refer to it as D +7. For comparison, the 2004 split in party ID was perfectly even at 38 percent apiece (the GOP’s best showing in any recent presidential contest), and the average split in modern elections is D +3.6.

But Obama’s advantages are clearly less strong this year. He’s given us 8-plus percent unemployment for three years, economic growth below 2 percent and 23 million unemployed — and now the American flag is being burned across the Muslim world. Plus, Mitt Romney’s ground game far exceeds McCain’s 2008 effort.

But the Q-poll is having none of it. Like some other polling outfits, it is consistently and systematically insisting that Election Day turnout will favor Democrats as much or more than Obama’s 2008 best-in-a-generation advantage. In the Q-polls released yesterday, the spread between Democrats and Republicans each exceeded Obama’s 2008 advantage.

In Florida, the 2008 actual result was D +3; yesterday’s Q-poll had it at D +9. In Ohio, it was D +8 in 2008, D +9 in the Q-poll. And Quinnipiac gave us D +11 in Pennsylvania, versus a 2008 result of D +7. Mind you, each of these states has seen dramatic changes in party preferences since 2008 — electing Republican governors, flipping the state legislatures to GOP control, etc. Quinnipiac and others have given us polling that reflects a Democratic edge exceeding 2008 all year — including in last week’s Q-polls on Colorado, Virginia and Wisconsin.

Doug Schwartz, the director of Quinnipiac Polling, recently addressed these criticisms, citing the change from 2004 to 2008 to justify his sampling: “A good example for why pollsters shouldn’t weight by party ID is if you look at the 2008 presidential election and compare it to the 2004 presidential election, there was a 7-point change in the party-ID gap.” Um, so why is Schwarz assuming that the trend from 2004 to 2008 will continue in 2012?

Again, we have a president mired in a weak economy — with the economy remaining voters’ top issue (no survey even shows a close second). And Obama’s 2008 voting coalition is less enthusiastic in 2012, especially Hispanics and the youth vote.

But Quinnipiac uses 2008 as the norm — and then adds in even more Democrats, because, says Schwartz, there are “more people who want to identify with the Democratic Party right now than the Republican Party.” Yes, more people identify with Democrats — that’s why national polls should reflect a sample that’s D +2 or D +3, and state polls should reflect a sample between the 2004 and 2008 electorates — not the unusually and likely unrealistically large Democratic advantage that Quinnipiac is awarding President Obama.

Like to The New York Post column: http://www.nypost.com/p/news/opinion/opedcolumnists/why_the_poll_is_qrock_56LUGAzegy7yn9zE8PsymK#ixzz27dv2FZLG

By keithbacker | Posted in Columns, General | Also tagged , , , , | Comments (7)

Why Party ID Matters

September 26, 2012 – 5:56 pm

I meant to blog this a couple days ago but I’ve been “busy” …

Jim Geraghty at National Review’s Campaign Spot nicely laid out the argument why party ID in matters in polling. This is a hotly contested issue especially since some pollsters weight their polls based on party identification and some do not.  Geraghty also does a great job laying out the mischaracterization many in the media make to defend this practice. The don’t have a good answer for our critique so they change the question to answer something they can defend.

First, the basics: Most Democrats are going to vote for the Democratic candidate, most Republicans are going to vote for the Republican candidate, and the independents are usually going to split somewhat evenly. So the proportion of the three groups more or less determines which candidate the polls are going to show ahead. No one claims to be able to predict, with absolute certainty, what the partisan makeup is going to be on Election Day. But we do have a range from recent history — from even in 2004 and a seven percentage point advantage for Democrats in 2008.

Now Geraghty lays out the biggest gap in defenders of current sampling

If a pollster believes that the electorate will be even more heavily Democratic in 2012 than it was in 2008, I’m willing to hear those arguments, but I think it’s a tough case to make.

Now the argument for why turnout will not meet or exceed 2008.  Something I have argued many times:

The last presidential cycle was a perfect storm for Democrats — an unpopular GOP incumbent, frustration over wars overseas, a terrifying economic meltdown, a Republican nominee who had spent much of his recent career fighting his own party and who openly admitted he wasn’t focused on economics, the first African-American major party nominee in U.S. history… This year you have Obama running as an incumbent in an extremely tough economy, a more aggressive GOP nominee, the grassroots energy of the Tea Parties, a huge change in each party’s financial resources, the novelty of making history wearing off, and now our embassies under siege in the Middle East… And then there is the shift in the number of voters who are registered voters in each party…

Defenders of the polling mischaracterizing our complaints about sample sizes (I think I got so fired up I commented in Cilizza’s comment section here)

Yet our objection keeps getting misstated and twisted by poll defenders time and again. Here’s Chris Cillizza, claiming the complaints are “a series of false assumptions none bigger than that because the country has been virtually evenly divided on partisan lines for the past decade or so that the party identification question should result in something close to a 50-50 split between Democrats and Republicans.” No. I do not demand an even split, nor do any of the other folks paying attention to this factor. I do think that a split of D+7 or more is excessive, and that a D+3 or D+4 divide — halfway between the GOP peaks of 2004 and 2010, and the Democrat peak of 2008 — is more likely.

The reality

Because for much of this year, in quite a few national polls, we’ve seen Romney winning almost all the Republicans and hold a lead among independents, and still trail Obama, because Obama is winning almost all the Democrats, and Democrats make up such a large share of the sample. Possible? Sure, anything’s possible. But if a pollster is going to offer a hypothetical electorate that looks different from everything we’ve seen before, I’d like to know why they think this is the case…What we see in the most disproportionate poll samples is confirmation bias. To many people covering this race, Romney should be trailing badly, Republicans are flailing desperately, and Obama is running an exponentially better campaign. Thus, it makes perfect sense for the electorate to be even more heavily Democrat than it was in 2008.

By keithbacker | Posted in General | Also tagged , | Comments (28)

Jay Cost Breaks Down the Party ID Conundrum

September 26, 2012 – 9:53 am

This blog has hammered the party ID problem in polling from the beginning.  The incomparable Jay Cost puts the pieces together in his megaphone at The Weekly Standard lays out how the polls split on voter turnout and how that drives whether they find the race a blowout in favor of Obama or a dead heat:

Republicans, by and large, are frustrated with recent polls of the presidential election because they think Democrats are being oversampled. Many pollsters respond by saying that “weighing” the polls for partisan identification creates its own problems and might end up skewing the polls in the wrong direction.

A lack of weighing creates its own problems, which many pollsters often fail to acknowledge. Specifically, many polls have, in my judgment, overestimated the Democrats’ standing right now. I base this conclusion not on a secret, black box statistical methodology or some crystal ball, but rather on a read of American electoral history going back to 1972. If I am right, then some of the polls are giving a false sense of the true state of the race, and will likely correct themselves at some point or another.

One important “tell” in my opinion, is this president’s continued weak position with independent voters, who remain the true swing vote. Obama’s average overall margin over Romney in these same polls is roughly 4 percent.

Bottom line: You do not get a four-point lead overall with a tie among independents, unless you are squeezing substantially more votes out of your base than your opponent is. And more generally, you are not “winning” an election in any meaningful sense of the word when 3/5ths of unaffiliated voters are either undecided or against you.

 

 

 

Cost sees “two ways the polls are tilted in favor of the president.”

First, many of the polls are guessing that Democrats are set to turn out at levels that match or sometimes exceed 2008. Take two examples – recent polls in Ohio and Florida. I’ve included the 2008 and 2004 exit polls as a baseline for consideration.

The midpoint between 2004 and 2008 is D+1.5. You’ll notice that Gravis, Washington Post, and Fox basically see a replay of 2008 while Rasmussen and the Purple Poll see roughly something in between 2004 and 2008. Relatedly, the polls on the high end for Democrats see a 5-point lead or better for the president (with Gravis being a strange exception), and Obama at or near 50 percent. The polls that see a tighter partisan split basically see a toss-up.

 

 

Cost then proves out the same point with Florida and points out the trend holds for Virginia, Colorado and likely Nevada (a lack of polling impacts the analysis for NV). He concludes his party ID discussion with the following:

All told, we see a statistically significant relationship between Obama’s margin and the Democratic advantage in partisan identification. In other words, there appears to be a bimodal distribution of the polls. They are not converging around a single point. Instead, some (notably Rasmussen, Purple Strategies, Survey USA, and Mason-Dixon) see Obama ahead by just 1 to 3 points in the key swing states, while others (notably the Washington Post, Fox News, PPP, and NBC News/Marist) see an Obama lead that ranges between 4 and 8 points. And the difference looks to be built around how many Democrats are included in the polling samples.

If it comes down to whether or not this will be a repeat of 2008 — which is basically what the latter camp of pollsters is suggesting — then my money is on no. Of course, it is possible that I am wrong. I have no crystal ball looking forward. All I can do is look back through history, where I see on average a nationwide Democratic identification edge of about 3 points, which is also roughly the midpoint between 2004 and 2008. That is my guess about 2012. It is an informed guess, but it is still a guess. If I’m right, then Rasmussen, Purple Poll, Mason-Dixon, and Survey USA are closer to the mark. But I could be wrong, in which case Fox, PPP andWashington Post are closer to the mark.

Importantly, the pollsters are guessing, too. They are guessing via the myriad of choices they make about when to poll, whom to poll, and how to poll. By Election Day, polling will be much more “scientific” than it is today; but now there is quite a bit of “art.” That’s how we wind up with two points of convergence, instead of just one.

…

As I mentioned earlier, a big “tell” here is that Obama cannot build any kind of a lead among independent voters. That suggests to me that his advantage is built entirely on Democratic enthusiasm, which right now is above its historical trends and clearly on a post-DNC bump. Nobody in the postwar era has won the presidency by carrying less than 49 percent of independents, and Obama is quite a ways below that mark, even if some polls show him at or above 50 percent nationwide and in the key swing states.

By keithbacker | Posted in General | Also tagged , | Comments (5)

It’s Official: Quinnipiac Says We’re All Democrats, Obama Still Losing Independents

September 26, 2012 – 8:57 am

There is no question in my mind that Barack Obama has promised to hire Quinnipiac to be his official pollster in his next election.  That’s the only explanation for polls consistently sampling an electorate that only exists in David Axelrod’s fantasy world. It’s nothing new that the Obama campaign aggressively lobbies polling outfits for turnout models they favor but Quinnipiac gets the gold star for going the extra mile in the audition to be Obama’s pollster for his 3rd term.  Today’s polls would have you believe we as a nation are so full of Justin Bieber level love for the Democrat party, we’re about to send 300+ into the House of Representatives not 190. But we should have seen this coming considering the way Quinnipiac indicted itself in the earlier discussion of party ID.

It’s latest swing state poll of Florida, Ohio and Pennsylvania is the most consistently absurd sampling of any I’ve seen this cycle. With results like these CBS and New York Times must find their joint effort with Quinnipiac to be the most smashing marriage since Kim Kardashian and Kris Humphries.

Florida

Party ID is D +9, (Dem 36, Rep 27, Ind 33).  This compares to D +3 in 2008 (Dem 37, Rep 34, Ind 29) and R +4 in 2004 (Dem 37, Rep 41, Ind 23). With a sampling like this Obama should have a 15-point lead. But Romney leads by 3 with Independents so Obama’s lead is only 9. Also check out the favorable/unfavorable of the two political parties.  Democrats are above water +6 at 49 Favorable/43 Unfavorable whereas Republicans are deeply under water -12 at 38 Favorable/50 Unfavorable. This must mean its Congressional delegation to the House of Representatives is overwhelmingly Democrat right?  But the Florida House delegation is made up of 19 Republicans and and 6 Democrats. And a Democrat Tsunami isn’t hitting the Florida delegation any time soon.  This electorate used to sample for Quinnipiac poll won’t exist on election day in Florida any time soon despite the hopiest of hopey wishes from Team Obama.

Ohio

Party ID is D +9 (Dem 35, Rep 26, Ind 35). This compares to D +8 in 2008 (Dem 39, Rep 31, Ind 30) and R +5 in 2004 (Dem 35, Rep 40, Ind 25).  Romney leads with Independents by 1 but is down overall by 10 because Quinnipiac called all the shop stewards at GM plants across Ohio since the Obama re-election team was busy answering the Washington Post’s survey. This is for a state that less than 12-months ago went to the polls in a very pro-Union turnout and also voted to REPEAL Obamacare by a margin of 66 to 34.  Suddenly they’re all rushing into the arms of Obama because his war on coal keeps Ohioans unemployed by banning fracking of the Utica shale?  Ohio may be tough terrain for Romney but this sampling reflects as much reality as Kate Upton coming over to wake me up every the morning … It ain’t happening.

Pennsylvania

Party ID  is Dem +11 (Dem 39, Rep  28, Ind 27). This compares to D +7 in 2008 (Dem 44, Rep 37, Ind 18) and D +2 in 2004 (Dem 41, Rep 39, Ind 20). Well someone needed to challenge that Susquehanna poll showing a two point race in Pennsylvania because after the vigorous defense of their polling methodology someone may get the notion that Barack Obama isn’t most popular of popular Presidents in the history of all Presidents throughout the entire Universe. Thankfully for Team Obama Quinnipiac was up to the challenge finding an electorate right out of Ed Rendell’s party machine making it seem like Pennsylvania didn’t reject nearly every Democrat state-wide and turn over the House and Senate to Republicans. Romney leads by 1-point with Independents but is down 11 in this poll not worthy of blogging but I needed the exercise.

By keithbacker | Posted in Florida, Ohio, Pennsylvania | Also tagged , , , , | Comments (4)

Washington Post Surveys Obama Re-election Team in Ohio, Finds Obama Leading by 8

September 25, 2012 – 9:33 am

You have to laugh at the absurdity of polls like these as I broke down in the post below:

  • 2008 was a best-in-a-generation advantage for Democrats
  • Obama’s job approval is locked below 50%
  • The unemployment rate has been above 8% for 3 years and it’s actually higher if you count the people so despondent they simply quit looking for a job
  • There isn’t one economic indicator that is positive for President Obama and the economy always surveys as by far the #1 issue for voters (there is never even a close 2nd)
  • Every single survey shows Obama’s 2008 coalition is less enthusiastic in 2012 than in 2008 (especially Hispanics and the youth vote)

At the same time the GOP ground game has improved over its 2008 performance by at least 10-fold (that’s no exaggeration).

And yet polls like the latest from the Washington Post get published with a Gomer from the Andy Griffith Show feel of “well Goll-ly look how many Democrats there are, Obama must sure be popular.”  It is professionally incompetent and completely partisan advocacy to survey states or the nation this way but there isn’t a reader out there who doesn’t already know the press is in the bag for the Obama re-election team.  The partisan breakdown of the poll was D +11 for adults, D +8 for registered voters and D +7 for likely voters.  In 2008 Obama enjoyed an advantage of D +8 in Ohio (Dem 39, Rep 31, Ind 30).  In 2004 this was a Republican advantage of R +5 (Dem 35, Rep 40, Ind 25).  Not the number cruncher @numbersmuncher broke down the Ohio returns and found that the party ID in Ohio was actually D +5 on election day, making this sampling even worse. Romney leads by 1% among Independents but the Washington Post finds Team Obama in full hopey-changey frenzy and by sampling 7% more Democrats than Republicans comes out with a result that Obama leads in Ohio by 8-points, 52 to 44:

For President Percent
Barack Obama 52
Mitt Romney 44
Other/Unsure 4
By keithbacker | Posted in Ohio | Also tagged , | Comments (9)

Quinnipiac Indicts Itself in Polling ID Debate

September 25, 2012 – 8:56 am

National Journal interviewed pollsters on the great party ID debate and Republican complaints that far too many Democrats are being sampled in the surveys. They gave fair hearings to both the polling outfits and critics like Rick Wilson in yesterday’s New York Daily news who observed thusly:

“Far too many of the public and media polls have set their likely voter screens and models to something looking more optimistic than the 2008 turnout model,” GOP consultant Rick Wilson wrote in Sunday’s New York Daily News, “which even Obama’s most dedicated partisans think is highly unlikely.”

But the real loser in this debate was Doug Schwartz, director of Quinnipiac Polling. In addition to childishly dismissing the complaints as sour grapes by Republicans, Schwartz indicts himself in explaining the controversy:

Schwartz, whose institute conducts polls in battleground states for CBS News and The New York Times, asserts that pollsters who weight according to party identification could miss the sorts of important shifts in the electorate that could be determinative.

“A good example for why pollsters shouldn’t weight by party ID is if you look at the 2008 presidential election and compared it to the 2004 presidential election, there was a 7-point change in the party ID gap,” Schwartz said. Democrats and Republicans represented equal portions of the 2004 electorate, according to exit polls. But, in 2008, the percentage of the electorate identifying as Democrats increased by 2 percentage points, to 39 percent, while Republicans dropped 5 points, to 32 percent.

Asked specifically about GOP complaints regarding the party-ID composition of public surveys, Schwartz said: “They’re the ones trailing in our swing-state polls.” “There are more people who want to identify with the Democratic Party right now than the Republican Party,” he added.

Our entire point is party ID changes from election to election yet Quinnipiac and the other polling outfits act like 2008 is the baseline and they are adjusting upwards in favor of Democrats from there.  2008 was a best-in-a-generation advantage for Democrats.  Obama’s job approval is locked below 50%.  The unemployment rate has been above 8% for 3 years and it’s actually higher if you count the people so despondent they simply quit looking for a job. There isn’t one economic indicator that is positive for President Obama and the economy always surveys as by far the #1 issue for voters (there is never even a close 2nd). Every single survey shows Obama’s 2008 coalition is less enthusiastic in 2012 than in 2008 (especially Hispanics and the youth vote).  But Schwartz and other lemmings use 2008 as the norm and add in more Democrats for flavor because according to Schwartz “more people who want to identify with the Democratic Party right now than the Republican Party.”  Few dispute more people identify with the Democrats which is why polls should be D +2 or D +3.  However, no sane person looks at the litany of reasons outlined above and says I think Obama’s going to have even a bigger advantage in 2012 than he had in 2008.  It’s professionally incompetent to reach such an absurd conclusion only the most politically partisan person could reach.

By keithbacker | Posted in General | Also tagged , , , , , | Comments (2)

Obama +8 in Pennsylvania in a poll oversampling Democrats by 10%– Mercyhurst

September 24, 2012 – 7:46 pm

Note: thanks to commenter MikeN, in my haste I misread one of the questions.  They were surveying vote intensity between the the two parties not the voter split of how many Democrats are voting for Obama and how many Republicans are voting for Romney.  It still means Independents probably favor Romney but we still don’t know the breakdown.  Sorry for the confusion and thanks to the commenter.

[Begin original post]And we’re supposed to call this poll credible but Susquehanna’s is not?  Come on.

The earlier Susquehanna poll showing Obama only ahead by 2-points generated a bit of a furor leading the Susquehanna Polling people to explain in great detail their methodology in the poll.  Basically by not sampling absurd turnout levels beyond the once-in-a-generation peak Obama achieved in 2008 Susquehanna arrived at a poll showing the race deeply competitive.  Mercyhurst Center for Applied Politics (MCAP)  had no such qualms in their poll and ran will full gusto into the arms of the Obama re-election team sampling the Pennsylvania electorate with a self identification over-sampling of Democrats by 10% (D +10 in Question 36).  The party ID breakdown was Dem 52, Rep 42, Ind 4 versus D +7 in 2008 Dem 44, Rep 37, Ind 18 and D +2 in 2004 Dem 41, Rep 39, Ind 20. So we have dramatically more Democrats and a disappearing Independent electorate for what is probably the fastest growing segment of the three. In a poll with a dramatic Democrat turnout advantage President Obama leads Mitt Romney 48 to 40 with 6 Undecided.

Highlights:

  • Both candidates lock down their base with 91% of the Democrat/Republican vote
  • In a poll with 52% self-identified Democrats Obama still can’t crack 50%
  • This would mean Barack Obama got almost zero votes from Independents (91% of 52 = 47.3; Obama’s vote total was 48)
  • Of the Independents it appears 40% lean Republican and 37% lean Democrat (Q39)
  • No mention of how Independents voted although the math says they voted 82 to 18 for Romney
  • Voters favor Pennsylvania’s Voter ID law 57 to 33
For President Percent
Barack Obama 48
Mitt Romney 40
Other 5
Undecided 6
By keithbacker | Posted in Pennsylvania | Also tagged , | Comments (4)

What if Barack Obama is NOT winning in Minnesota?

September 24, 2012 – 12:10 pm

This is a state where Mitt Romney has only a small presence but is does lurk in the nether regions of Battleground States that has undeniably trended Republican over the last two decades.  You likely can’t win if you don’t compete, but the latest Minnesota poll would not give me comfort if I were Team Obama given the alleged 8-point lead in the survey.  Here are the details and why Obama is more vulnerable in Minnesota than you might think:

  • President Obama leads Mitt Romney 48 to 40; the headline problem is Obama can’t crack 50% — a big problem in an allegedly safe state
  • The party ID of the poll was D +13 (Dem 41, Rep 28, Ind 31).  This compares to D +6 in 2008 (Dem 40, Rep 36, Ind 25) and D +3 in 2004 (Dem 38, Rep 35, Ind 27).
  • Many problems with that voter make-up:
  1. A 7-point advantage over Obama’s 2008 monster year will not occur due to meaningfully reduced enthusiasm plus voters simply switching parties
  2. The entirety of Obama’s lead is based solely on the party make-up of the survey
  3. The poll has Democrat make-up above its 2008 peak but reduces the GOP make-up 8 points despite Republicans using a wave of pickups to take over both houses of the state legislature in 2010
  4. The party ID means Romney is winning with Independents by a sizable margin but exact #s not available
  5. In this poll massively over-sampling Democrats, Obama’s sub-50% vote looks even worse
  • Obama won the state by 10-points in 2008 which is four less than his win in neighboring Wisconsin where he is definitely vulnerable
  • Although Minnesota unemployment is comparatively low at 5.9%, it has ticked up 0.3% over the last two months and union strong-arming is fast becoming a thorny issue in the state
  • This isn’t the first poll that skews Democrat showing Obama with a single-digit lead

The bottom line is Minnesota is not on the immediate radar but these numbers demonstrate and incredible vulnerability for Obama in Minnesota that could be exploited by an aggressive Romney campaign.

By keithbacker | Posted in General | Also tagged , | Comments (3)

About those Susquehanna Polls Saying Pennsylvania is in Play — Must Read

September 24, 2012 – 11:22 am

Plenty of people on the Right were emboldened when they read the latest poll from Susquehanna showing Romney within 2-points in Pennsylvania while nearly the entirety of the Left dismissed the polls as inaccurate.  Thankfully Susquehanna Polling and Research took to the interwebs and explained their methodology.  Bottom line they are using a D +6 model (PA was D +7 in 2008 and D +2 in 2004) but they are meaningfully lowering the amount of youth vote they expect to make up the Pennsylvania electorate this Fall.  Also note, embedded in the explanation is the reality that yes, the Obama re-election team actively lobbies pollsters for voter turnout models far to their advantage but not necessarily accurate portrayals of the voting public:

Our vote model for gauging the number of interviews conducted with voters of different demographic groups (things like party affiliation, racial background and age range, etc.) is a blend of turnout models from both the 2008 and 2004 presidential elections, but leans more towards 2004 VTO and is predicated on the belief that turnout this November will not be anywhere near ’08 levels when 5.9 million votes were cast.

First, our ratio of interviews conducted with Republicans and Democrats in our recent polls (49D – 43R) gives Democrats a 6-point advantage based on the fact that Democrats outnumber Republicans in actual registration.  However, this ratio is slightly more Republican based on both national and state polling showing that Republicans are more likely to vote than Democrats this year given high intensity among Republicans who strongly disapprove of the President’s job performance.  Nonetheless, this +6 Democratic advantage is only one point less Democrat than the 7-point advantage these same exit polls gave Democrats in the 2008 presidential election.  Besides, simply conducting more surveys with Democratic voters (as some have suggested) doesn’t necessarily translate into more votes for President Obama when you consider that Mitt Romney is winning Democratic-leaning counties in Western Pennsylvania by ten or more percentage points.  Nonetheless, it is entirely appropriate to sample Republicans one or two points higher than in 2008 if you believe as we do that voter turnout this November will have little resemblance to the last presidential election.

Second, our ratio of younger to older voters reflects turnout that is likely to be slightly higher with older voters given the lack of enthusiasm from younger voters.  In our surveys, 18-44 yr. olds make up 30% of all interviews and voters 45 years of age and older represent the remaining seventy percent.  For instance, according to 2008 exit polls voter turnout among 18-29 year olds peaked at 18%, but national and state polling proves interest among younger voters down sharply this year due to higher unemployment with younger voters and college graduates in particular.  So conducting approximately ten percent of surveys with 18-29 year olds is a reflection of this lower anticipated turnout among these less-enthusiastic voters.  Besides, the fact that Obama backers have suggested that over sampling older voters skews results in favor of Mitt Romney is a striking revelation in a state like Pennsylvania known for having the 5th largest population of senior citizens in the country.

Third, recent polls showing a double-digit lead for Obama are not believable, and are probably using the 2008 voter turnout as the basis of their survey model.  It is simply unrealistic to think Obama can or will win the Keystone State by the same double-digit margin he won by four years ago when you consider that most state and national polls continue to show most voters unhappy with the direction of the country after two straight years of unemployment at 8% or higher.  This is why our statewide polls conducted every month since the primaries shows the President failing to hit fifty percent in most key measurements like favorable name ID, job approval and his ballot score.  Plus, polling we have conducted in dozens of state senate and house races on behalf of incumbent legislators and other candidates, PACs and other special interest groups shows Obama’s support down an average of seven percent when compared with his vote margins in these same districts four years ago.  We estimate this 7-point drop off could mean up to 434,000 fewer votes cast for Obama this November, leaving a margin of less than 200,000 votes between the candidates.  Based on this, perhaps the Phil’ Inquirer poll showing Obama winning by a bigger margin than he won by four years ago is the real outlier.

For these reasons and others we fully stand by our results, and all indications are that the upcoming election will be closer than many others suggest.

Boom!

By keithbacker | Posted in Pennsylvania | Also tagged , , , | Comments (6)
Follow

Get every new post delivered to your Inbox.

Join 218 other followers