ALL BLOG POSTS AND COMMENTS COPYRIGHT (C) 2003-2014 VOX DAY. ALL RIGHTS RESERVED. REPRODUCTION WITHOUT WRITTEN PERMISSION IS EXPRESSLY PROHIBITED.

Tuesday, September 11, 2012

Ignore the polls

When they're being run by pro-Obama Democrats, it should hardly be surprising that they are wildly inaccurate:
The Pew Research Center and Rasmussen Reports were the most accurate in predicting the results of the 2008 election, according to a new analysis by Fordham University political scientist Costas Panagopoulos. The Fordham analysis ranks 23 survey research organizations on their final, national pre-election polls, as reported on pollster.com.

On average, the polls slightly overestimated Obama’s strength. The final polls showed the Democratic ahead by an average of 7.52 percentage points — 1.37 percentage points above his current 6.15-point popular vote lead. Seventeen of the 23 surveys overstated Obama’s final victory level, while four underestimated it. Only two — Rasmussen and Pew — were spot on.
It's not an accident that 17 of the 23 polls overstated Obama's vote margin. And keep in mind that this is a comparison of their final polls, which is when the more egregiously pro-Democratic organizations suddenly start cleaning up their act a little so they don't get it too noticeably wrong. And where are the Pew and Rasmussen polls today?

Rasmussen has Obama 50, Romney 45. Pew has Clinton 29, Obama 16, and Michelle Obama 15, but their poll concerns who gave the best speech at the DNC so I don't think there is much to be learned from that. What might be informative would be to look at where the 23 polls were in the middle of September 2008 and compare them with the current polls. I may do that at some point, but not today.

Labels:

79 Comments:

Anonymous Tico September 11, 2012 7:43 AM  

Intrade, in which Obama is KILLING Romney?

Anonymous Rantor September 11, 2012 7:53 AM  

Rasmussen showed Romney getting a four point convention bump and that going lower through the Dem convention, Sunday at Rasmussen, Obama was ahead by 4 (I think it was 48 to 44). Monday, Obama was ahead by 5 (50 to 45). In several recent Republican victories, the Republicans were behind at this point in the contest. So things can and do change.

Blogger James Dixon September 11, 2012 8:01 AM  

My rule of thumb is that the polls overstate the democrats position by 5%. That seems to have worked fairly well over the years.

I suspect it's even worse this time, since the press is so overtly in the tank for Obama.

Anonymous physics geek September 11, 2012 8:43 AM  

Re: Intrade: if memory serves, Intrade had McCain ahead until just a couple of weeks before the election, or right around the time he suspended his campaign, at which point Intrade started trading McCain down.

Anonymous Anonymous September 11, 2012 8:48 AM  

Vox, why not analyze the electoral college outlook? If you do it looks bleak for Romney. He essentially has to sweep swing states like Michigan, Wisconsin, PA, Florida, Colorado, and maybe New York to have a chance.

Anonymous Rally September 11, 2012 8:51 AM  

My understanding is that Nate Silver at 538 adjusts these polls for their historical bias and then aggregates them. Doing that he was able to call the 2008 election accurately.

Anonymous VD September 11, 2012 9:05 AM  

My understanding is that Nate Silver at 538 adjusts these polls for their historical bias and then aggregates them. Doing that he was able to call the 2008 election accurately.

Nate Silver makes a new call every single day. He predicted both a McCain win and an Obama win in 2008. He is now calling for an 80 percent chance of an Obama victory. Do you think it will still be anything close to that in November?

Anonymous JP (real one) September 11, 2012 9:13 AM  

You also have to wonder how voter fraud will factor into this. Peter Schiff recently interviewed Hans von Spakovsky of the Heritage Foundation on his show. They talked about how 30,000 dead people were discoverd on North Carolina's rolls, how ObamaCare never would've happened without voter fraud, and how the (In)Justice Dept. bocks voter protection in different states.

Anonymous whatever September 11, 2012 9:17 AM  

While it's great that even polls are nonsense these days, isn't it still funny that the Republicans nominated candidates are so hated by the American people that they can't beat a loser like Obama?

The economy is devastated. Romney shouldn't have to do anything to win. The election commercials for the Democrats don't even have Obama talking, they use Bill Clinton. Yet, yet, it's still sorta close. That's pathetic.

Anonymous Mrs. Pilgrim September 11, 2012 9:22 AM  

"Dewey Defeats Truman".

Polls can be deceiving. As with anything, keep perspective.

Anonymous dh September 11, 2012 9:28 AM  

> they are wildly inaccurate:

The deviation of the average of the polls was 1.37%, and you call this wildly inaccurate? What other standards do you hold to under 2% precision? Where any of the polls outside of the margin of error?

> And keep in mind that this is a comparison of their final polls, which is when the more egregiously
> pro-Democratic organizations

Pollsters run two types of polls - polls meant to drive the narrative - at the behest of their customers - and polls meant to predict what will actually happen. This is the entire business model of Rasmussen and many other local/regional pollsters. When you see a joint operation - one part of the pair is the sponsor, the other is the narrative driving pollster. For example 3 weeks prior to the election in 2010 Rasmussen polled the Hawaii senate race. Prediction was for a 13% margin of victory. Actual result was 53% margin of victory. The challenger was the sponsor of the poll. Guess what the challenger (Cam Cavasso) collected almost enough contributions in the last 3 weeks to retire his entire campaign debt and repay the money he loaned his campaign.

This is how polling exists - it's a for profit enterprise designed to help separate donors from their money. Publicly released polls are press-releases and advertising for polling firms.

Virtually every pollster on this list is doing so to generate profit. Interestingly the university/college based programs like Quinnipac and Marist do very well, and in 2010 Qunnipac and Rassmussen where a very close #1 and #2 (Qunnipac blew the Florida governors race call, which was close but not that close).

Anonymous dh September 11, 2012 9:36 AM  

> They talked about how 30,000 dead people were discovered on North Carolina's rolls

This is a huge red-herring. Dead people don't usually get purged from voter rolls for a few election cycles. Dead people happen everyday.

Even in a notoriously bad ballot security area like Chicago, you need hundreds or thousands of participants to make a dent using "dead people voting" method. And it's not even worth it in most cases, since these worst areas are already machine-party based anyways.

It's much more likely that voter-access issues are going to play a bigger role. In Pennsylvania, for example, the State estimated that 9% of previously eligible voters don't have the proper paperwork or ID to vote under new laws. Ohio, which had a heavy voting percentage in 2008, mostly among minorities, is working around the clock to prevent as much early voting as possible.

Anonymous dh September 11, 2012 9:39 AM  

> American people that they can't beat a loser like Obama?

You got it. Romney is the worst GOP candidate in a few generations. And that's saying something considering his competition includes Bob Dole, John McCain, and Gerald Ford.

Anonymous Google September 11, 2012 9:40 AM  

Rasmussen today, 48 Obama, 45 Romney

Anonymous JP (real one) September 11, 2012 9:46 AM  

"It's much more likely that voter-access issues are going to play a bigger role. In Pennsylvania, for example, the State estimated that 9% of previously eligible voters don't have the proper paperwork or ID to vote under new laws"

Does the proper paperwork include (a) tax returns for the last 3 years and/or (b) proof of real property ownership?

Yes, I jest, but it should. If this stops thousands of people--with no skin in the game--from being bussed in to vote in inner cities, then I'm happy with it.

From what I've seen in the Deep South, a lot more of them vote than blue collar/independent contractor types, who don't want to take time out of their schedules. The idea of elections being "swung" by the SSI/food stamp crowd (thanks to Jessie Jackson & Co.) is repulsive...and certainly not what the Founders intended.

Anonymous Josh September 11, 2012 9:49 AM  

Nate Silver makes a new call every single day. He predicted both a McCain win and an Obama win in 2008. He is now calling for an 80 percent chance of an Obama victory. Do you think it will still be anything close to that in November?

That's because it's a dynamic model that changes as new polls are released.

And, as others have said, the national tracking polls are worthless, what matters are the state polls, because of the electoral college. And the electoral map is tilted in Obama's favor.

Anonymous dh September 11, 2012 9:56 AM  

I think the best part of this election season is seeing the stages of Romney denial.

In the primaries, it was Romney is a great candidate, he's doing great. He's going to unite the base and destroy Obama. He's got millions!

Now, it's "every single pollster including OUR own is biased and out to get us".

It's almost worth watching the country fall apart to see the denial and stages of anger. Because even the best of the best right wing pundits are desperately trying to avoid Romney being tagged with a stench.

Romney's best bet is the debates. TV advertising is looking like diminishing returns has set in.

Anonymous VD September 11, 2012 9:57 AM  

That's because it's a dynamic model that changes as new polls are released.

I'm aware of that. The point is that I wasn't predicting Hillary Clinton was going to win the election on November 1, 2008. And, in fact, I predicted the Democratic candidate was going to win in 2008 regardless of who it was before either nominee was settled.

The deviation of the average of the polls was 1.37%, and you call this wildly inaccurate? What other standards do you hold to under 2% precision?

You are forgetting that the biggest historical difference since 1984 was 8.2%. So, on average, they are off by 16% of the maximum possible difference, which is not 2% precision. It's not reasonable to imply 99-1 is just as in play as 51-49. But "wildly inaccurate" is probably too strong, especially in comparison with the deviations seen in economics statistics.

Anonymous George September 11, 2012 9:58 AM  

In order for Romney to win, he needs to take Florida, VA and OH.

Not gonna happen. And it only gets worse for Romney after he's forced to defend his policies in front of a national audience at the debates.

The only issue left is the disposition of the House of Representatives.

Anonymous VD September 11, 2012 10:01 AM  

I think the best part of this election season is seeing the stages of Romney denial.

It is really remarkable to see how badly he is doing. Any pedestrian Republican candidate would likely be well ahead of Obama at this point in time. I suspect the Mormon thing is hurting him much worse than the professionals thought it would, as is the Republican failure on the debt ceiling. And the Roberts betrayal on Obamacare badly hurt the usual appeal to the Supreme Court argument.

Blogger swiftfoxmark2 September 11, 2012 10:05 AM  

Pollsters run two types of polls - polls meant to drive the narrative - at the behest of their customers - and polls meant to predict what will actually happen. This is the entire business model of Rasmussen and many other local/regional pollsters. When you see a joint operation - one part of the pair is the sponsor, the other is the narrative driving pollster. For example 3 weeks prior to the election in 2010 Rasmussen polled the Hawaii senate race. Prediction was for a 13% margin of victory. Actual result was 53% margin of victory. The challenger was the sponsor of the poll. Guess what the challenger (Cam Cavasso) collected almost enough contributions in the last 3 weeks to retire his entire campaign debt and repay the money he loaned his campaign.

This is how polling exists - it's a for profit enterprise designed to help separate donors from their money. Publicly released polls are press-releases and advertising for polling firms.

Virtually every pollster on this list is doing so to generate profit. Interestingly the university/college based programs like Quinnipac and Marist do very well, and in 2010 Qunnipac and Rassmussen where a very close #1 and #2 (Qunnipac blew the Florida governors race call, which was close but not that close).


Pollsters will take a sample size of a little over 1000 people. If you knew anything about statistics, you would know that the smaller the sample size, the more inaccurate the actual results are. Of course, even if you didn't know anything about statistics, this is common sense.

So we have a nation of 310 million people. We are supposed to assume that a sample size that is less than a thousandth of a percent of the population will accurately reflect the real attitudes of the American people?

Anonymous 691 September 11, 2012 10:06 AM  

Vox, I never got a response from you last week.

Anonymous CJ September 11, 2012 10:09 AM  

The biggest reason to believe Obama will win is the EC map. The only big prize Romney has locked up is TX, while Obama has CA, IL, NY, and probably PA. Romeny has to sweep FL, OH, NC and VA and win two of IA, WI, CO, and NV. If he loses one of NC or VA, he has to win three of the last four.

Obama has more paths to victory and a greater margin for error.

Anonymous Rally September 11, 2012 10:14 AM  

Yeah, good point. It will change back and forth. It might be 80% in November.

That can happen even if the projected popular vote is in the neighborhood of 52-48.

Anonymous Roundtine September 11, 2012 10:15 AM  

The worthless polls are the registered voter polls. The likely voter polls are always more accurate.

The news media constantly release polls skewed Democrat, probably because the Democrats are at home watching TV during the day and answer the phone (that's not a joke). The pollsters are too lazy to adjust the total or throw it out and do another poll. Rasmussen and any other accurate pollster take the time to obtain a good sample.

Anonymous Other Josh September 11, 2012 10:25 AM  

All this shows me is that people are reluctant to embrace Mittens Romney in spite of Obama's gross ineptitude. The Republican party truly deserves to be spit upon.

Anonymous dh September 11, 2012 10:26 AM  

> Obama has more paths to victory and a greater margin for error.

Exactly why all the statistics people are calling this 70-80% probability for Pres. Obama.

This is where "predictions" from VD et all are worthless. Because they are by balls, exclusively. They are not based on evidence, or reality, or anything. People projecting a large Gov. Romney win haven't sat down to calculate what would have to happen. What will it take for Gov. Romney to win in Nevada? Ohio? Florida? Pennsylvania? How much movement has to happen in how many days?

It's 50-ish days to the election. Erick Erickson from Red State says it takes "10-days" for the polling to reflect reality. There are a limited number of 10-day periods left.

Anonymous dh September 11, 2012 10:27 AM  

> The worthless polls are the registered voter polls. The likely voter polls are always more accurate.

Most pollsters switch to likely voters in October. It is much more time consuming and costly to do a LV poll, and to develop a model that actually filters likely-voters from registered voters.


There are lot of a troubling signs for Romney besides just the electoral projects. Favorability is a major problem for Romney as well. He's under 50%, and his last best chance to raise it - the convention - is over.

Anonymous dh September 11, 2012 10:31 AM  

> So we have a nation of 310 million people. We are supposed to assume that a sample size that is less
> than a thousandth of a percent of the population will accurately reflect the real attitudes of the
> American people?

Yes, history and statistics tells us it will. Adding more people does not necessarily improve your result. Adding statistically significant people will.

The confidence interval on most polls is 95% (including Rassmusen, Gallup, Ipos, Reuters, CNN, etc).

If you are a stats geek, highly recommend Prof. Sam Wang from Princeton. His methods are 100% transparent, with source code (http://election.princeton.edu/for-fellow-geeks/) and statistically speaking he is not a lightweight like Nate Silver. Very solid. I would defy you to find a flaw in his methodology, model, or implementation.

Blogger Nate September 11, 2012 10:35 AM  

I'm sorry... Vox... you have to see this Rand Paul quote...

"When it comes to federal employment numbers, zombies must've ate Paul Krugman's brain." - Rand Paul

I am amused.

Anonymous Angel September 11, 2012 10:35 AM  

I think Vegas odds would be a bit more honest and telling at this point, no?

Anonymous JP (real one) September 11, 2012 10:42 AM  

"All this shows me is that people are reluctant to embrace Mittens Romney in spite of Obama's gross ineptitude. The Republican party truly deserves to be spit upon."

Yeah, other than being younger and better looking than Dole, Romney fits the same utterly bland, moderate profile.

It's really sad that bizarre "Biden with bikers" photos can emerge and it doesn't make a difference. It's like the Dem ticket is dropping its hands and sticking its chin out in a boxing/MMA fight--taunting a foe that can't win.

Anonymous dh September 11, 2012 10:44 AM  

> It is really remarkable to see how badly he is doing. Any pedestrian Republican candidate would
> likely be well ahead of Obama at this point in time. I suspect the Mormon thing is hurting him much
> worse than the professionals thought it would, as is the Republican failure on the debt ceiling. And
> the Roberts betrayal on Obamacare badly hurt the usual appeal to the Supreme Court argument.

It's just demographics. The GOP coalition is so narrow that there are only a few percentage points of possible wiggle that will get any GOP candidate over the top. The GOP will never be able to compete with race and group based politics.

There just aren't enough white people to keep putting GOP candidates over the top. 2012 is probably the last election that it will be possible on paper for a white-anglo to win a general election. And in 8-12 years when Texas flips from a generation of hispanic voters being able to vote 80-20 for liberals, it won't even be a contest.

The GOP's best hope is that Romney wins in 2012, and immediately sets up proportional representation, in order to fracture the Democrats into 2 or 3 liberal parties. At least in the future a right/center coalition could squeak out 51%.

If nothing changes, no matter what the laws are, watch for billionaires to abandon their new favorite pastime of meddling in politics. These people are businessmen, they want return on investment not excuses. If there is no return, they'll go back to being wallflowers.

Anonymous Stickwick September 11, 2012 10:51 AM  

So, whaddya do if a polling company doesn't toe the line? You let your true Chicago thug colors shine. The DOJ is suing Gallup after unfavorable poll results:

In a scene right out of a typical authoritarian regime, Fox News reports that “employees at the venerable Gallup polling firm suggested they felt threatened by Obama campaign adviser David Axelrod when he questioned the methodology of a mid-April poll showing Mitt Romney leading the president – according to internal emails published Thursday.”

According to the email chain titled “Axelrod vs. Gallup,” the White House in addition asked that a Gallup staffer “come over and explain our methodology,” which was apparently perceived as a subtle threat.

Fox News reported that “a Gallup official said in an email he thought Axelrod’s pressure ‘sounds a little like a Godfather situation.’”

Gallup refused to change its methodology to suit the White House.

And the Justice Department intervention in the whistleblower suit came three months later.

Anonymous DonReynolds September 11, 2012 11:00 AM  

Big money donors to political campaigns already know that the polling organizations are partisan and can reliably be expected to fudge the numbers in favor of their candidate. The public trusts the pollsters much more. Nixon once remarked how a slight rise in the polls really makes the donations pour in. A nice little paperback on this subject is "Lies, Damned Lies, and Statistics."

Anonymous VD September 11, 2012 11:06 AM  

Vox, I never got a response from you last week.

Refresh my memory, preferably with a link, please.

Anonymous JP (real one) September 11, 2012 11:11 AM  

"And the Justice Department intervention in the whistleblower suit came three months later."

What do you expect the DoJ to do? Go after a real criminal (with DoJ connections) like Corzine instead? Under Holder, they are the most 1984-like agency I've ever seen. Their name and actual agenda are at 180 odds.

Anonymous DonReynolds September 11, 2012 11:25 AM  

dh..."There just aren't enough white people to keep putting GOP candidates over the top. 2012 is probably the last election that it will be possible on paper for a white-anglo to win a general election."

This is the argument used to get Republicans to pander to minorities. Of course, I disagree.

There is no shortage of white voters in this country....79% of the votes are cast by white voters. Obviously, they can elect anyone they please. The problem is simple....white voters are divided ideologically and do not like each other very much. This division is as old as the nation itself, even leading to civil war at one point. (Very likely, that war would still have occurred, even if there had been no slavery.) In truth, the whites in the USA are fairly evenly divided into two camps and they oppose each other every day. There has been a long history of either side recruiting black and hispanic minority voters to tip the scales in general elections.

More importantly, and many pundits refuse to see this, is the dynamic in those states where whites are not a majority. When whites are not an overwhelming majority, they tend to gravitate to a single party in order to retain power. The party that becomes the white party does this by moving to the center of the political spectrum, while the other party becomes even more Leftist to monopolize the minority vote.

The Democrat Party in Texas is beginning to show this as well.....aside from a few white LBJ liberals, the Democrat Party in Texas is overwhelmingly hispanic and black. The Republican Party is white with a big tent to include more of a diversity of viewpoints (out of necessity). You can see the same thing in Louisiana, Mississippi, New Mexico and South Carolina. Demographics will not force the GOP into oblivion but it may force them to the middle of the political spectrum in order to sheep-steal more white voters from the Democrat Party. Do we see that today? Of course, this is very bad news for the Religious Right wing of the Republican Party, and the conservatives and the Libertarians.

Anonymous JartStar September 11, 2012 11:36 AM  

If Obama is reelected I have to wonder under what scenario a Republican which isn’t just like Obama policy wise could possibly win going forward. Bush Senior was ousted over a minor recession and Obama survives the Great Recession? Hard to believe, but if he does then having 1/3+ of the country on the government dole is the apparent way to stay in power. If Obama does win, the GOP won't be more conservative, but rather much more liberal in hopes to win next time.

Blogger James Dixon September 11, 2012 11:39 AM  

> ...isn't it still funny that the Republicans nominated candidates are so hated by the American people that they can't beat a loser like Obama?

Well, it's deliberate, but I wouldn't call it funny.

Blogger JartStar September 11, 2012 11:42 AM  

It’s possible that the GOP is throwing the election, but Romney and Ryan didn’t get the memo apparently. McCain and Palin was an obvious punt, but R/R at least appear serious, and the bankers are backing them.

Anonymous DonReynolds September 11, 2012 11:45 AM  

Bush Sr. lost the election in 1992 in large party by DENYING there was any recession, or an economic problem, and by vetoing Congressional extension of the unemployment benefits THREE TIMES. (He was the first president to veto an extension.) It was only after Congress overrode his veto and made the unemployment benefits extension retroactive that the empass was broken.

Obama does not deny there is an economic problem. He may be silly and say it would have been worse, had HE not been president the past four years. But his re-election bid seems to ride on whether enough voters are willing to give him more time to deal with the economy. FDR managed to get elected president four times by saying he was doing everything he could and the voters believed him. Will they believe Obama?

Blogger James Dixon September 11, 2012 11:48 AM  

> Yes, history and statistics tells us it will.

Not in the history I'm familiar with. From memory, Reagan was supposed to lose, Bush the elder was supposed to lose, Clinton was supposed to win by over 10%, Bush the younger was supposed to lose (both times, I believe), and Obama was supposed to win by something close to 10%. None of those happened.

Anonymous JP (real one) September 11, 2012 11:58 AM  

"Demographics will not force the GOP into oblivion but it may force them to the middle of the political spectrum in order to sheep-steal more white voters from the Democrat Party. Do we see that today? Of course, this is very bad news for the Religious Right wing of the Republican Party, and the conservatives and the Libertarians."

More "middle of the spectrum" than the Bushes, Dole, McCain and Romney? Is that even possible?

Heaven help us.

Anonymous dh September 11, 2012 12:17 PM  

> Not in the history I'm familiar with. From memory, Reagan was supposed to lose, Bush the elder was
> supposed to lose, Clinton was supposed to win by over 10%, Bush the younger was supposed to lose
> (both times, I believe), and Obama was supposed to win by something close to 10%. None of those
> happened.

None of this is true, except the Clinton stuff which is close to true.

1. Carter was winning through the conventions (that were 6 weeks earlier than now); after the last debate (10 days before election) Reagan had thoroughly destroyed Carter, and Carter's base shifted to Anderson. Gallup correctly predicted the win, the final poll was 3.8% off (it was a three way race) on 4.5% margin of error.

2. "Bush the Elder" was correctly predicted to win by Gallup; Gallup over stated Bush's margin by 3%.

3. As for Clinton - another 3 way race - the final Gallup poll had Perot at 14%, he ended up with 19%, which all came from Clinton. So his predicted margin of 12% was reduced to around 6%.

4. Both elections for GW Bush were called correctly. Of course the Gore debacle ended up that they predicted Bush to have more popular votes, when it didn't come out that way (but the prediction was right and the popular vote was right).

5. Gallup predicted a 9% Obama win, he won by 7%.

All of the results have been the margin of error, and especially the last 4 cycles, where technique and methods have really been excellent in regards to correlating.

What exactly is your complaint? That your memory is bad? Most likely you remember all of the interminable months of horse-race coverage from the media, which is almost 100% useless. October is it. The only month that matters other than the day. The conventions are over, the debates happen - it's when the few undecideds make up their mind.

Anonymous dh September 11, 2012 12:23 PM  

> There is no shortage of white voters in this country....79%

Wishful thinking:

Year - White - Black - Hispanic - Asian
1988 - 84.9%/9.8%/3.6%/0%
1992 - 84.6%/9.9%/3.8%/1.2%
1996 - 82.5%/10.6%/4.7%/1.6%
2000 - 80.7%/11.5%/5.4%/1.8%
2004 - 79.2%/11.0%/6.0%/2.3%
2008 - 76.3%/12.1%/7.4%/2.5%

In 1988, HW Bush won 60% of the white vote. This year, for Romney to collect the same actual number of votes, he needs closer to 67% of the white vote

You can't only win 60% of the white vote, while the demographic trend goes the way it goes.

And thanks to the electoral college, in 2020, when the tipping point in Texas is reached, the entire election will be over before it starts - California, New York, Texas, and Florida will so dominate the electoral college that it will be very close to impossible for the GOP *of today* to even get close to a winning pattern.

Anonymous JI September 11, 2012 12:57 PM  

Why have the Republicans, for the past two elections, nominated the one person who is incapable of beating Obama? First McCain and now Romney. It's almost like the Republican primaries are rigged to give a candidate who will lose.

Anonymous Stilicho September 11, 2012 1:25 PM  

Of course, this is very bad news for the Religious Right wing of the Republican Party, and the conservatives and the Libertarians.

Interesting perspective, Don. Of course, at some point white Democrats may have to choose between joining a fairly conservative white voting bloc or being metaphorically (and physically?) eaten by their fellow-traveler minorities. In other words, it will not be a matter of the "white party" becoming more liberal to attract white liberals although it may ultimately trend that way due to the infusion of liberals, but rather a matter of many white liberals joining the conservative voting bloc as a matter of self preservation in the face of increasingly antagonistic minority voting blocs that will not tolerate them.

Anonymous RedJack September 11, 2012 1:30 PM  

Been telling my parents about the demographic shift, and how it doesn't matter anymore. Made my Mom so mad she hung up on me.

We will fall, and to be honest have fell. The US is about to go hard socialist. The only question is will it be international or national socialist.

Anonymous RedJack September 11, 2012 1:32 PM  

Stilicho,
Their world view doesn't allow for that. Our foriegn policy the last few decades shows that they will not see the problem. They keep thinking that Iraqi's will think the same as people from New York.

Rome didn't see it either.

Blogger JartStar September 11, 2012 1:44 PM  

dh - That's the demise of the GOP right there. I gave up voting back in 2004 because I realized the pointless nature of it. Just like the US financial collapse is inevitable without a seismic historical event, the US will continue to drift towards the left at the same time.

Anonymous Stilicho September 11, 2012 1:47 PM  

RedJack:

Under normal circumstances, that's true. However, facing the cannibal's pot can radically alter one's outlook in favor of whatever will help one avoid being eaten. That does not mean that they won't revert to type once they feel safe, but that may take a while in the face of societal collapse.

Anonymous dh September 11, 2012 1:50 PM  

> , but rather a matter of many white liberals joining the conservative voting bloc as a matter of self
> preservation in the face of increasingly antagonistic minority voting blocs that will not tolerate
> them.

The productive anglo's will tolerated by the minor voting blocks so long as they are productive and don't step off the plantation.

Anonymous Stilicho September 11, 2012 1:50 PM  

They keep thinking that Iraqi's will think the same as people from New York.

Both groups are tribalistic rent-seekers who believe that the only way to succeed is at someone else's expense, so, in a sense, they are close to the truth. One group is considerably more honest about their nature though.

Anonymous Stilicho September 11, 2012 1:53 PM  

The productive anglo's will tolerated by the minor voting blocks so long as they are productive and don't step off the plantation.

The productive Anglos are not the group which is generally voting with the barbarian hordes.

Blogger Thomas Schauer September 11, 2012 1:53 PM  

"What might be informative would be to look at where the 23 polls were in the middle of September 2008 and compare them with the current polls."

Although admittedly not capturing all polls, the RealClearPolitics Poll Average on 9/11/2008 had McCain up 2.5

Anonymous George September 11, 2012 2:09 PM  

VD Said:

"It is really remarkable to see how badly he is doing. Any pedestrian Republican candidate would likely be well ahead of Obama at this point in time."

I'm not so sure. What is a "pedestrian" republican? The problem is with the base of the party and the necessary positions a candidate must take to get through the primary. Look how far right Romney had to go. Now it's coming back to bite him.

American's would vote for a Republican with right leaning economic views and centrist/left social views. But the small base that determines primaries don't allow that.

Plus, look at the shitty group of candidates. Bachmann? Really? Santorum? Paul? Perry? The smartest one among them wasn't smart enough not to declare he had wrapped up the nomination before Florida even voted.

Anonymous Todd September 11, 2012 2:10 PM  

Overstating the poll numbers helps when you're planning on stuffing the boxes with votes
from the deceased or illegal.

What mob runs Chicago now? Is it the russian mafia? If it is, a lot things make sense ...
like how commie the politicians there are.

Anonymous George September 11, 2012 2:12 PM  

DH Wrote:

"The GOP's best hope is that Romney wins in 2012, and immediately sets up proportional representation"

Really? Romney's going to do this?

I am amused.

Anonymous whatever September 11, 2012 2:15 PM  


Been telling my parents about the demographic shift, and how it doesn't matter anymore. Made my Mom so mad she hung up on me.

We will fall, and to be honest have fell. The US is about to go hard socialist. The only question is will it be international or national socialist.


So mentally ill. Obama=Hard Right Austerity Wackjob in Europe

Anonymous George September 11, 2012 2:15 PM  

JL wrote:

"Why have the Republicans, for the past two elections, nominated the one person who is incapable of beating Obama? First McCain and now Romney. It's almost like the Republican primaries are rigged to give a candidate who will lose."

Who among those running would have easily beat Obama?

Blogger James Dixon September 11, 2012 2:26 PM  

....Lots of comments about the final polls, which we're not dealing with here.

> > All of the results have been the margin of error, and especially the last 4 cycles...

Since the margin of error is usually +/- 3-4%, that would be hard not to manage.

> What exactly is your complaint? That your memory is bad?

Everyone's memory is bad. Which is why I noted that it was from memory. And yes, I was talking the overall polls, not the final ones.

> Most likely you remember all of the interminable months of horse-race coverage from the media, which is almost 100% useless.

Which is exactly what we're dealing with here.

> October is it. The only month that matters other than the day. The conventions are over, the debates happen - it's when the few undecideds make up their mind.

If that's the case, why are you saying Obama is an almost certain lock now. The October polls aren't here yet.

Anonymous JP (real one) September 11, 2012 2:33 PM  

"Look how far right Romney had to go"

How far right did he go, George? About 1/46,000 of a millimeter?

The Left's spin on the political continuum is moronic...

Anonymous George September 11, 2012 2:41 PM  

JP,

No it's not. There is a significant difference in both economic policy and social policy when it comes to a number of issues. If you look at Romney's previous policies, particularly those espoused while Gov. of MA and what he was forced to say to gain votes from the base in the primaries, you can see he was dragged to the right.

Anonymous JP (real one) September 11, 2012 2:48 PM  

Compared to what? His policies in MA would be considered leftward by some Southern democrats.

Based on what's historically been considered left/right, Romney's convention rhetoric was moderate. Yes, Obama has spewed out non-stop class warfare drivel and tried to drage the center far to the left, but I'm not buying it.

Anonymous JP (real one) September 11, 2012 2:49 PM  

drag the center line far to the left.

Anonymous DonReynolds September 11, 2012 2:54 PM  

By my experience, Americans seldom vote FOR anybody. More often than not, they are voting AGAINST the other sonofabitch. If that still holds true, the majority vote will go with Romney. As long as this election is a referenda on Obama, he will lose. He has few pluses in his column and a lot of minuses. While Romney may be relatively unknown, that may not matter much. Bill Clinton and Jimmy Carter were obscure former governors when they were elected president. Carter had the gift of Watergate vote and Clinton had the gift of Ross Perot pulling conservatives from Bush Sr.

There is a natural reluctance on the part of American voters to give the President a second term. Don't Re-Elect Anybody is a pretty popular sentiment. While Obama does not have the high profile screwups of Jimmy Carter, there is still plenty to be unhappy about. Four more years of the same thing? I doubt it and many voters hope not.

While Romney himself may not be prone to the charge of racism, the Republican Party has had real difficulty shaking off that label, and that has made them reluctant to campaign hard against Obama. The same was true of McCain. He campaigned harder against his own base than he did Obama. Both men have avoided anything and everything that could remotely be constrused by a hostile press to be tainted by the scent of racism. Of course, too many Democrats insist that any criticism of the Obama agenda and policies is overt racism.

Anonymous JP (real one) September 11, 2012 3:05 PM  

This is from Mark Skousen, who happens to be a Mormon economist/financial newsletter guy...FWIW.
----

Right now Intrade, the political futures market, shows Mr. Obama as the heavy favorite by 58% to 42%.

But much could change between now and November 6, and Intrade has had a checkered past in its ability to predict winners...

A better record has been recorded by several economists who have developed economic models that have been surprisingly accurate.

And all of them agree: President Obama is in serious trouble and is likely to lose in November!

Here are the results of four of the best forecasters in the business:

1. Two political scientists at the University of Colorado, Ken Bickers and Michael Berry, used their economic indicator model to predict who would win in November. Their model has correctly predicted the outcome of the past eight elections, and this year they predict a big Romney victory with nearly 53% of the vote, with Romney winning almost all the swing states.

2. Nigel Gault, the chief U.S. economist at IHS Global Insights, a Boston-based research firm, uses proprietary methods and various economic variables to show the president in even worse shape. The high unemployment rate (8.1%) is a "crucial variable," he states, and based on his model, Obama will garner only 45.4% of the popular vote.

3. Two economists, David Rothschild of Microsoft Research and Patrick Hummel of Google, have created a model that is both political and economic, and they too show President Obama losing.

4. Ray C. Fair, the Yale economics professor, has developed a successful prediction model that covers elections since 1916. His model depends entirely on the strength of the economy. Right now he has Romney leading by only one percentage point, 50.5% to Obama's 49.5%, but he says there is a 2.5% "standard error," so the election is "too close to call."

Anonymous Salt September 11, 2012 3:16 PM  

Rush says Obama is Jimmy Carter's second term.

Anonymous Boris September 11, 2012 3:30 PM  

"The final polls showed the Democratic ahead by an average of 7.52 percentage points — 1.37 percentage points above his current 6.15-point popular vote lead."

The final popular vote was Obama +7.2%, so the +7.52% poll average was essentially correct. And there were a lot of polls at Obama +7, so instead of 17 of 23 overstating Obama it was more like 50/50.

So, in actuality, polls are pretty good. Or they have been. They could be wrong this year, of course.

Anonymous Stilicho September 11, 2012 3:41 PM  

Rush says Obama is Jimmy Carter's second term.

Fair enough, but a Romney presidency would be Richard Nixon's third term.

Blogger James Higham September 11, 2012 3:42 PM  

Bias is as endemic there as here.

Anonymous George September 11, 2012 4:00 PM  

JP Wrote:

"Compared to what? His policies in MA would be considered leftward by some Southern democrats."

There are still southern democrats?


"Based on what's historically been considered left/right, Romney's convention rhetoric was moderate."

That's the trick, isn't it; pivoting to the middle as the general campaign ramps up. Which leads to the question: "Who is this guy and what does he beleive?" Don't be surprised when that becomes a theme of the debates, because it's a legitimate question.

Anonymous George September 11, 2012 4:05 PM  

Don Reynolds said:

"There is a natural reluctance on the part of American voters to give the President a second term."

If you go back to FDR, we find that He, Truman, Eisenhouer, Kennedy (with Johnson), Nixon, Reagan, Bush II and Clinton all got second terms. Only two (Carter and Bush I) failed to get their second term.

I think it's more factual to say that American's don't seem very willing to change horses in mid-stream.

Anonymous dh September 11, 2012 4:12 PM  

> Here are the results of four of the best forecasters in the business:

I am not familiar with any but the first, but for the Colorado pair, is not a predictive model. I.e., they didn't develop it eight years ago, and test it over eight elections. It is retroactively developed. There is no way to know which variables they are use are actual factors driving behavior, and which are noise.

This is same stuff that scientists try to use to "prove" global warming. Take some data, create a model, then apply forward.

The Colorado model is not credible for a lot of reasons. It has the entire east coast from Pennsylvania to Florida, and inland to Indiana all going to Gov. Romney. Every swing state but Iowa goes to Romney.

Just Pennsylvania alone is enough to cast doubt on the model - according to the model 5.6% unemployment is the point where the incumbent being a Democrat just won't/can't win. Yet, we are 8% official and real unemployment closer to 11% - 14%, and Obama has been up in 20 out of 21 polls, and is comfortably 8-9% ahead of Romney. So is it possible that in the next 55 days there is going to be 12-13% movement in the polls?

Yes, of course. It is possible. Anything could happen. But their model doesn't have a factor based on "anything could happen". And it doesn't have a factor for the quality of opposition.

Anonymous dh September 11, 2012 4:27 PM  

> Although admittedly not capturing all polls, the RealClearPolitics Poll Average on 9/11/2008 had
> McCain up 2.5

Also very difficult to capture anything and compare, since the RNC convention last time around ended September 4th, meaning that the polling in that week from 9/4 to 9/11 captured all of the convention bounce, including the full Palin bounce from her convention speech.

RCP show McCain wining for two time periods:

- January time-frame, when Sen. Obama and Sen. Clinton were locked in a heated primary, and there was lots of anti-Obama media happening. Sen. Obama was at his lowest Democratic support ebb before he locked up the nomination.

- Post convention, with the full Palin "game change" in effect, for a few precious days of goodwill and positive coverage.

Right now the GOP is freaking out and all the polls are supposedly wrong not because they show Obama "suddenly" winning, but because of favorability. Pres. Obama's has been going up - to around 53% - despite $200 million in negative advertising. Despite the GOP convention. And Gov. Romney's has never really broken 50%.

The polls show that the race hasn't moved much in months, which is why everyone is freaking out. In 2008, the race was more fluid, but McCain saw the same thing Romney did, which is that the race was slipping out of his fingers and he was at a small but persistent disadvantage.

Anonymous Anonymous September 11, 2012 5:46 PM  

"The only question is will it be international or national socialist."

The national variety suits me fine. NAMs won't be involved, too much discipline required.

Anonymous Luke September 12, 2012 4:47 AM  

Two predictions about future Presidential elections (those that there are):

1) If Barack Obongo loses, neither party will nominate even a part-black (e.g., quatroon or more) for >20 years;

and,

2) if Mitt Robama loses, the Repukes will never again nominate a Mormon instead of a Christian. The Dims will of course never ever nominate a Mormonist in any event, no matter what happens.

Anonymous JP (real one) September 13, 2012 1:11 PM  

Bennie (Bernanke) and the Fed Jets just gave their unofficial vote for Obama in the form of QE3.

Has the Fed ever done anything like this so close to an election?! OK, maybe 2008, during the depth of the financial crisis, but in a year when the Dow and S&P were already up over 10%?

Ridiculous...especially since Producer Price Inflation came in at 1.7% month- over-month gain this morning--much higher than expected.

Deflation vs. inflation debate should pick back up after this.

Post a Comment

NO ANONYMOUS COMMENTS. Anonymous comments will be deleted.

Links to this post:

Create a Link

<< Home

Newer Posts Older Posts