Could A “Polling Bubble” Explain Why This Race Appears To Be So Close?

by John Hawkins | October 31, 2012 4:52 am

This is an extremely difficult election to forecast because, unlike the last few cycles, there has been so much weird and unexplainable polling data.

Going just by the top of the line, final product numbers, this election is neck-and-neck going down the stretch. On the other hand, going by the crosstabs, Romney is going to win this election just as handily as Obama won in 2008.

Why do I say that?

Early voting[1].

My pal Guy Benson found a juicy nugget that helps to bring more clarity to the news from Gallup yesterday that shows Romney leading Obama in the early vote by a full seven points, 52-45%. Almost exactly four years ago (October 28, 2008), according to Gallup, Obama was massacring John McCain among early voters with a fifteen-point lead, 55-40%. That means, at least according to Gallup, that Obama’s early vote advantage has dropped 22 points when compared to ’08.

Independent voters.

In the last three releases of the tracking poll conducted by The Washington Post and ABC News, Obama has trailed former Massachusetts governor Mitt Romney among independent voters by between 16 and 20 percentage points[2].

That’s a striking reversal from 2008, when Obama won independent voters, who made up 29 percent of the electorate, by eight points over Sen. John McCain of Arizona.

Favorability.

“Fifty-one percent (51%) of voters have a favorable opinion of Romney and 48% say the same of Obama.” — Rasmussen Reports[3]

Swing state polling.

“In the 11 swing states, Mitt Romney earns 50% of the vote to Obama’s 46%. Two percent (2%) like another candidate in the race, and another two percent (2%) are undecided.

Romney has now led for 12 straight days with margins of four to six points most of that time.

In 2008, Obama won these states by a combined margin of 53% to 46%, virtually identical to his national margin.” — Rasmussen Reports[4]

Most importantly, there’s the breakdown of the electorate.[5]

In 2008, the Democrats had a 39-29 (D+10) advantage in hard party ID, and a 54-42 (D+12) advantage with leaners. In 2012 though, we’re in the post-TEA party era. Republicans now show a 36-35 (R+1) hard party ID advantage, and a 49-46 (R+3) lead with leaners. This gives us a range of party ID swings from 2008, from R+11 to R+15.

The chances of Mitt Romney winning the early vote, independent voters by 8-10 points or more, having a higher favorability rating than Obama and winning the swing states by four or more points and losing the election has to be very close to zero. In fact, going by those numbers, Mitt Romney should win just as handily as Barack Obama won in 2008 — and make no mistake about it, people were very aware that Barack Obama was going to beat McCain in 2008.

Here’s the breakdown of predictions from the day before the election.

Hot Air (Ed): McCain 273 vs. Obama 265
The Campaign Spot: Obama 286 vs. McCain 252
Right Wing News: Obama 311 vs. McCain 227
Hot Air (Allah): Obama 318 vs. McCain 220
Outside the Beltway: Obama 325 vs. McCain 213
Election Projection: Obama 338 vs. McCain 200
Karl Rove: Obama 338 vs. McCain 200
Matthew Dowd Obama 338 vs. McCain 200
Donna Brazile: Obama 343 vs. McCain 195
FiveThirtyEight: Obama 349 vs. McCain 189
Electoral-Vote: Obama 353 vs. McCain 185
George Stephanopoulos: Obama 353 vs. McCain 185
Larry Sabato: Obama 364 vs. McCain 174
Obama 365 vs. McCain 173 (Hawkins’ note: Actual results inserted)
George Will: Obama 378 vs. McCain 160
Kos: Obama 390 vs. McCain 148

Although most of the predictions were on the low side, they were generally on the money. That’s because there was a general agreement amongst the pollsters of note. This time, what we’re seeing is that most of the pollsters seem to be projecting that the Democrats are going to have just as big of a partisan edge as they did in 2008. On the other hand, Rasmussen and Gallup seem to be assuming that the 2012 breakdown will be much more Republican than it was in 2008. Given that the GOP had its best year in half a century in 2010 while Mitt has the advantage in early voting and has completely reversed Obama’s lead with independents, Gallup and Rasmussen seem more likely to be right.

So, what would explain so many of the other pollsters using 2008 turnout numbers to drive their projections?

Certainly, you have to accept the possibility that they’re right while Gallup and Rasmussen are wrong. I don’t think that’s the case, but it could be.

Could it be a conspiracy? We haven’t seen anything like that previously and it would ultimately be impossible to cover-up. So, that seems very unlikely.

Are they all just rooting for Obama to win and trying to help him? Some of them probably are, but polling agencies live and die on their reputations. Most of them would not voluntarily slant their numbers to help a candidate.

Of course, there’s another possibility: market bubble psychology.

The reason this comes to mind is that as of yet, ABSOLUTELY NO ONE has been able to give a plausible explanation for why the 2012 party ID numbers should be roughly similar to those in 2008. True, Obama is on the top of the ticket, but it’s as a poorly performing incumbent with a bad economy — instead of as a challenger without a record, up against a Republican, when the incumbent was also a member of the GOP with a 25% approval rate. Yet, that is EXACTLY the assumption most of the pollsters are making.

Here’s a little breakdown of how it works with a market bubble[6] (Highlighting mine).

First, similar to any decision, even one as mundane as choosing a fry-up instead of cereal, an investor forecasts that buying one stock will lead to a “higher level of satisfaction” than if he chooses to keep his money or make another investment selection.

In stage two, an investor grows over-confident about his investment decision. He also thinks too much about recent positive results, failing to take long-term information into account. “It is through such a dynamic that forecasts of future asset values — whether mortgages or internet stocks or other financial instruments — become increasingly skewed to the positive,” opine Vanguard strategists.

In stage three, things turn pear-shaped as more individuals adopt the same view on a particular asset. How does an entire market become overconfident? A necessary condition of a bubble is a surge in enthusiasm for a given asset among a majority of market participants, according to the Vanguard study. All of these market participants suffer from two conditions. The first is group-think, whereby they begin to feel invulnerable and rationalise their behaviour and also to ignore contradictory information. The second problem is that groups tend to make riskier decisions than individuals make on their own.

Early on, it may have been defensible to start with the idea that Obama would have similar numbers to 2008. Then, as it went on, no one was changing the breakdown. The longer it went on, the more difficult it was to explain, but it also became harder to change. Eventually, the explanation for the numbers became, “Well, all of these other pollsters are doing it, so it must be right.”

If that is what’s happening, there are going to be a lot of pollsters with egg on their face next week when the bubble bursts and Mitt Romney wins easily. On the other hand, if they’re right, Rasmussen and Gallup are going to look foolish, and this election is going to be as random as a coin flip.

Endnotes:
  1. Early voting: http://www.breitbart.com/Big-Government/2012/10/30/Gallup-Shows-22-point-drop-in-obama-early-voting-advantage
  2. former Massachusetts governor Mitt Romney among independent voters by between 16 and 20 percentage points: http://tpnn.com/how-could-barack-obama-lose-independents-by-numbers-like-these-and-win/
  3. Rasmussen Reports: http://www.rasmussenreports.com/public_content/politics/obama_administration/daily_presidential_tracking_poll
  4. Rasmussen Reports: http://www.rasmussenreports.com/public_content/politics/obama_administration/daily_swing_state_tracking_poll
  5. the breakdown of the electorate.: http://www.redstate.com/2012/10/26/swingometer-gallup-party-id-figures-predict-solid-romney-win/
  6. how it works with a market bubble: http://www.ft.com/intl/cms/s/0/409ef5f8-888b-11e0-afe1-00144feabdc0.html#axzz2ArPJilJq

Source URL: https://rightwingnews.com/john-hawkins/could-a-polling-bubble-explain-why-this-race-appears-to-be-so-close/