Archive for the 'Polling' Category

What does a 41/30/29 sample in a CNN poll add up to?

Monday, November 5th, 2012

A tie. And CNN points to other polls with the same result: Politico/George Washington University, NBC News/Wall Street Journal, and ABC News/Washington Post. We’ll have more to say on this Wednesday.

For re-reading on Tuesday Evening

Sunday, November 4th, 2012

Paul Krugman:

many of the “analysis” articles being published in these final days leave readers worse informed than they were before reading. As Nate Silver (who has lately attracted a remarkable amount of hate — welcome to my world, Nate!) clearly explains, state polling currently points overwhelmingly to an Obama victory. It’s possible that the polls are systematically biased — and this bias has to encompass almost all the polls, since even Rasmussen is now showing Ohio tied. So Romney might yet win. But a knife-edge this really isn’t, and any reporting suggesting that it is makes you stupider. Worse yet, some reporting tells readers things the reporters have to know aren’t true. How many stories have you seen declaring that “both sides think they’re winning”? No, they don’t: the Romney campaign is visibly flailing, trying desperately to find new fronts on which to attack Obama. They clearly know that it will take a miracle — sorry, a last-minute surge — to prevail on Tuesday.

We haven’t been commenting on polling for a while, since the only one that matters is just around the corner. There will be plenty of time for reflection after that.

The opinions of non-voters

Sunday, November 4th, 2012


Related: only a minority of Americans trust the MSM.

Interesting analysis of polling in the current election cycle

Saturday, October 27th, 2012

Bob Krumm:

Most good time-series mathematical modeling is validated against past events in order to predict future unknowns. Furthermore, greater weight is given to more recent events when verifying the model. That is usually the smart way to model a problem–except, that is, when the recent event with the greatest weight happens to be an historical outlier. The 2008 presidential election was an outlier. It was the first election since 1952 when there was neither an incumbent president nor a sitting vice president on the ballot. Since it was a contest unencumbered by incumbency, late-breaking undecideds were not predisposed by external factors to break one way or the other. Going into election day, the RCP average showed about a 7-point lead for Obama over McCain and that’s the way it ended up on election day. In other words, late-breakers broke to each side in about the same proportion as the decided portion of the electorate. However, when there is an incumbent on the ballot, it is uncommon for him to get the late-breaking vote.

Normally, undecideds break against the incumbent. Is there any reason to think that won’t be the case in 2012?

From D+10 to R+1?

Saturday, October 27th, 2012

Gallup says that likely voters — determined by a list of seven questions — have gone from D+10 in 2008 to R+1 now. That would be interesting if true. Such changes are not permanent of course. There was a swing to R+2 in 2004 that completely reversed to D+10 four years later. But the candidate of D+10, the uniter and so forth, is long gone. What’s left is a mess on the economy, foreign affairs, and so much else.

Bush won by 2 points when Gallup’s likely voter sample was R+2 in 2004. So what about this year? The story is the Independents. Independents were +3R in 2004 and +8D in 2008. In 2010 the I’s flipped 33 points to R+15 according to one estimate, as you recall. Nothing much has changed for the better since 2010, so we would expect the I’s to break hard for Romney, which is what most polls report. We’ll see if any of this is true shortly. Polling is a much tougher business today, since 91% of the people don’t respond anymore.

Unusual email from a campaign

Thursday, October 25th, 2012

The Romney campaign sent this email about Ohio:

Some public polling shows the true nature of the race. Rasmussen this week showed it tied at 48%, Suffolk tied at 47%, and Angus-Reid tied at 48%. Our view is that the race is a dead heat with Romney on an unmistakable upward track.

Other public polling continues to vastly overstate Democrat partisan advantages in Ohio. For example, the Time Magazine poll this week shows a nine-point advantage for Democrats in party identification, which would be a stronger Democratic turnout than in either of the last two presidential campaigns in the state. A reasonable look at the political climate tells us the partisan boundaries of this race will be fought somewhere between R+5 (2004) and D+8 (2008). Anything more than D+8 shows a survey to be vastly out of touch with today’s political reality – Obama isn’t as popular, his base isn’t as energized, and Mitt Romney’s supporters are poised to shatter voter contact records in Ohio.

One thing all the public surveys have in common is that Governor Romney is winning among Independent voters. In the 19 polls released in the public realm since the first debate, Governor Romney leads among Independents in 15 of them (two did not include data among INDs). And, if you take an average of his lead among Independents in those 15 polls, the margin is 12 points (49-37).

The Time Magazine poll shows Governor Romney trailing by 5 points statewide, but winning Independent voters by a 53%-38% margin. That’s just not possible. Write it down – if Mitt Romney wins independent voters by 15 points in Ohio, he’ll be the next President of the United States.

ABC: “The Obama campaign, however, believes it is again running up a commanding lead in the early voting, especially in Ohio.” We thought the Romney email was unusual. It seems to be an attempt to counter the Obama team’s early voting claims and to keep the RR team fired up. Meanwhile, Nate Silver gives Obama a 71% probability of re-election. Somebody’s going to have a stroke in a couple of weeks.

Did we watch the same debate as the pundits?

Tuesday, October 23rd, 2012

It was interesting to watch the debate analysis on MSNBC and Fox, but we came away wondering whether we watched the same debate. Fox was ecstatic about this and that in Romney’s performance. We weren’t. He let go of Benghazi. It was a club and Romney didn’t use it to beat his opponent. MSNBC went on about Romney’s flip-flops but seemed very upset about Obama’s performance. They have become unhinged. Not what we saw. Obama was very effective in the first half-hour with five points about Libya and three points about Iran and that Israel is our best friend (repeat 10x).

Interestingly, both candidates circled back to the economy, and like Jim Lehrer, Bob Schieffer left them alone for the most part. Excellent moderating, unlike the other two debates.

So Romney held back, presumably to act presidential, and Obama attacked and was forceful. Obama did a good job in that regard. But the pundits on Fox were ecstatic and those on MSNBC were funereal. Rasmussen explains: “In the 11 swing states, Mitt Romney earns 50% of the vote to Obama’s 45%…This is now the third time Romney has hit the 50% mark in the combined swing states in the past four days and is the biggest lead either candidate has held in nearly three weeks.” No doubt the internal polling by each candidate confirms this.

Everybody knew about Benghazi immediately, including the CIA

Friday, October 19th, 2012

AP reports what the CIA knew about the Benghazi attack on 9-11:

The CIA station chief in Libya reported to Washington within 24 hours of last month’s deadly attack on the U.S. Consulate that there was evidence it was carried out by militants, not a spontaneous mob upset about an American-made video ridiculing Islam’s Prophet Muhammad, U.S. officials have told The Associated Press. It is unclear who, if anyone, saw the cable

Just like the State Department, the CIA knew all about Benghazi day one. Everybody knew. And yet the next day the president prominently alluded to the video (words 177-185) and demoted terror to a more generalized reference (word 620).

The cover-up began immediately. Reuters reported on 9-13 that it was “clear” that an angry mob gathered at 7pm to protest the movie, when not one of those facts was true. We think the cover-up was mostly about preserving a chosen electoral narrative. If so, it has been the most spectacularly unsuccessful cover-up of all time.

Interesting results in PPP poll

Friday, October 19th, 2012

A labor-funded group sponsored a PPP poll. 14% of Democrat likely voters disapprove of their incumbent’s performance. 50% of those polled overall disapprove of the president’s performance. Of course they manage to torture the data to produce a tie, but the internals don’t tend to support their conclusion.

Bonus fum: Nate Silver takes on Gallup.

Say “Get the transcript” and the transcript appears!

Thursday, October 18th, 2012

From what we discussed the other day:

Romney: I think it’s interesting the president just said something, which is that on the day after the attack he went in the Rose Garden and said that this was an attack of terror.

Obama: That’s what I said.

Romney: You said in the Rose Garden the day after the attack it was an act of terror? It was not a spontaneous demonstration?

Obama: Please proceed.

Romney: Is that what you’re saying?

Obama: Please proceed, Governor.

Romney: I want to make sure we get that for the record, because it took the President 14 days before he called the attack in Benghazi an act of terror.

Obama: Get the transcript.

Crowley: He did, in fact, sir. So let me call it an act of terror in the Rose Garden. He used the word–

Obama: Can you say that a little louder, Candy? (Applause.)

How convenient that Crowley happened to have the transcript handy, and knew that word 620 of the 801 words in it was “terror.” It produced a gotcha moment at the very end of the debate and rule-breaking applause started by an interested party. It’s probably all just a happy coincidence however that the president said “get the transcript” and Crowley just happened to have it in front of her and knew all about word 620 of 801. Taranto provides a plausible, even likely, solution.

Jeffrey Lord thinks the Crowley episode is a tipping point of sorts. We’re not sure about that. 70% of Independents and Republicans have already tuned out the Democrats with press passes and are living, as best they can, in a universe with an alternative narrative. The funny part is we well recognize the other side’s universe, but they are mostly ignorant of ours.

Preference cascade at 2:45

Wednesday, October 17th, 2012

Watch this video. It’s a Frank Luntz focus group of 23 or so voters in Las Vegas, the majority of whom voted for Obama in 2008. They really liked Romney last night, and the majority of the former Obama voters are now going to vote Romney. We’re not surprised by that. What did surprise us is the surprisingly salty and unbleeped remark by one gentleman at 2:45 of the video. He was really ticked off at falling for a con man. This kind of talk was unimaginable not so long ago. Maybe this other Las Vegas resident has a point.

Who lost the debate?

Wednesday, October 17th, 2012

The media lost of course. In our view, at the beginning Romney hit it out of the park again on the first question, whereas Obama seemed energized but robotic. Romney engaged the questioner — asking him to contact him when he graduates in 2014 — and Obama said “you’re in college, good, but I want everyone to get a good education” or some such, and droned on about clean energy. After that, we paid some, but not very much, attention.

The (appallingly biased) media lost again, Crowley being arguably even worse than the moderator from ABC. She interrupted Romney 28 times versus 9 for Obama. She gave Romney less time than Obama. She strategically chose to bury the Benghazi debacle at the very end of the debate and then covered for Obama when she suddenly became a fact-checker who didn’t have her facts quite straight.

Of course what we think about the debate matters not. But some undecided voters at MSNBC of all places (why is anyone undecided by the way?) broke for Romney. And Romney won on the economy by 18-21 points according to CBS and CNN insta-polling, though these same people judged the overall debate for Obama. Yawn.

The only thing that got our goat was the discussion of pay for women. And it’s not the bit about women employed by the White House making 18% less than the men. It was more Romney’s going on and on to pander and to treat the matter as some serious question. Try this for a response instead: if companies in America can pay 20% less for female employees than males, you’d have to be nuts to hire a man when a woman will do the same job for less money. Companies aren’t irrational and want to maximize profit, so why does any man have a job in this country?

Another poll

Monday, October 15th, 2012

WaPo / ABC. D+9 sample, versus D+8 actual in 2008. I’s only R+6, versus R+15 in 2010. Go figure.


Friday, October 12th, 2012

In a sense the debate was a waste of 90 minutes of life. Its structure and the appalling moderator meant that each side would perform brief (and sometimes bizarre) set pieces without any meaningful dialogue between the participants. All this showed up in the CNN insta-poll. But TV is transient and the internet is forever. There’s a fun ad showing the goofiness on display last night for example. And Benghazi isn’t going away. We can’t imagine that any opinions formed after the first debate were changed by anything that went on last night.

Final point: here’s the entire debate in less than 8 minutes. Much better.

From a referendum to a choice election?

Tuesday, October 9th, 2012

There have been lots of “this race is over” type articles, and we won’t do that. Anything can happen, as should be clear from the past week. But we have thought that Romney was comfortably ahead in the race since May, based on a close reading of the polls. An important element of this was that the president couldn’t seem to clear the mid-forties and that does appear to have changed in his favor since the conventions, up until the debate.

Frankly we’re not sure how much to pay attention to the polls, since 91% of the people now refuse to participate. But current polling results seem consistent with what we observed last spring. Much has ben made of the dramatic shifts in the polls following the debate and whether such a massive shift is credible. But suppose we look at it another way. Back in the spring you had a similar number of people who said they were not for Obama. What the debate did was to give them a reason not to just be against Obama, but a convincing reason to be pro-Romney and to go public with that sentiment.

Anything can still happen, but the current situation seems rather similar to what we’ve seen for much of the year. A referendum has apparently become a choice.

Interesting strategy: keep reminding people about the debate

Tuesday, October 9th, 2012

At least they’re consistent. Romney won the debate 49/39 among Democrats and 70/20 among independents. How to deal with that? Create ads that remind everybody about the debate. Okay, that’ll work.

The quality of the writing has deteriorated

Monday, October 8th, 2012

Like most of you, we oppose many of the editorial positions of the NYT. But this editorial on Mitt Romney’s foreign policy speech seems poorly written as well as, in our opinion, wrongheaded. It has typos too. We’re not going to even bother to quote from it. Maybe they’re still discombobulated from last week. (As for the content of the editorial, ask Lara Logan what she thinks.)

BTW, the speech is viewable here. It seemed fine to us if unexceptional. But then again, we thought the last four years have been bizarre — Honduras, the Saudi bow, the movement for freedom in Iran, etc. How is it possible to be on the wrong side of so many issues?

70 million Americans saw this

Monday, October 8th, 2012

70-19 among independents. Pretty impressive. Worth watching again.

Watching the debate, again

Saturday, October 6th, 2012

It’s very instructive to watch the debate again, a debate viewed by 70 million people or more. Governor Romney was clearly on top of his game, but we didn’t think the president did a bad job. He just doesn’t know anything about business, and hasn’t bothered to learn anything in the last four years. A typical instance was the president’s saying that government health care was superior because studies show the government has lower administrative costs and the government doesn’t have to make a profit. That’s the sort of thing he actually believes. He really has no clue about how competition works in the private sector to improve quality, multiply choices and lower costs. Final point: it is surreal that while those polled thought that Romney won the debate 67-25 on the most important issue of our day, electoral polls continue to show a deadlocked race. How can that be?

A number of trends in exactly the same direction

Sunday, September 30th, 2012

As we’ve discussed, 91% of the people don’t respond anymore to pollsters, though 4x that number did fifteen years ago in 1997. 40% of people have confidence in the media now (only 28% among R’s and I’s); it was 53% back than. A quarter of LV’s get their main political news from the internet. It was about zero back then. Newspaper advertising in real dollars is about where it was in 1950. The old media and the pollsters have been in decline for the last 15 years, just as access to new information technologies has taken off. (Michael Barone has a good piece on polling.) We won’t be surprised if the 9% who still deign to participate in polls turn out not to be representative of the other 91%. In any event, we’ll know soon.