Sunday, September 21, 2008
Major Disconnect on Daily Kos / Research 2000 Palin Favorability Figures
To take September 19th as an example, the R2000 poll shows 41% favorable, 46% unfavorable and 13% no opinion for Palin. But on the same date, the Diageo/Hotline poll shows figures of 50% favorable and 35% unfavorable (with presumably 15% undecided). This is a huge difference that cannot be accounted for by normal variations or differences in party ID splits. And while the Diageo/Hotline numbers are down from where they were a couple of weeks ago, they are up from earlier in the week.
The recent CBS / NY Times poll gave Palin 40% favorable and 30% unfavorable, with 15% undecided and 15% "haven't heard enough". The September 18th Pew poll likewise continues to show strong favorables for Palin.
The only reasonable conclusion I can come to is that the favorables numbers in the DK/R2K poll are messed up. This, combined with the obvious problems in their voter splits (too many Democrats, too many women and too many Hispanics, all of which help Obama) lends credence to the view that this poll, despite not being conducted by Daily Kos, is biased towards Obama.
Saturday, September 20, 2008
Poll Details and Observations: CBS / NY Times Poll of September 17, 2008
As we head into the stretch of the 2008 presidential race, we are bombarded with poll results on a daily basis. But usually we only see the “bottom line” numbers: what percentage of voters as a whole favor each candidate. These are obviously the most important figures, but while they tell us the general state of the race, they don’t provide much information about each candidates’ strengths and weaknesses, and where their messages are penetrating or being rejected.
To get the more detailed and useful information in polls, it is necessary to go beyond the bottom line and number and look into the poll internals. The two most important internals are the answers to specific questions about the candidate, and crosstabs, which show how answers to questions break down by political affiliation, gender, race and other factors.
Since I am always curious about what’s going on behind the scenes, I’ve decided I will regularly dig into poll internals and report my observations. Polling firms vary greatly in terms of whether or not they release their internals. Some do not, and others do only for subscribers, but some provide full information for their polls. One of these is the CBS / NY Times poll, which has a large number of questions and answers.
I’ve decided to use the CBS/NYT September 17th poll (based on polling done from Sept 12-16) for this analysis. This poll shows Obama leading McCain 48%-43% among registered voters, and 49%-44% among likely voters.
Note that while I am obviously partisan in this election, I do make an effort to be objective in my observations in this sort of analysis.
Here are some of the more noteworthy items I found in this poll’s internals.
1. The biggest swing in overall support between this poll and the one taken on September 8 was in independent voters: they had previously been 55-29 in McCain’s favor, but are now 46-41 for Obama. Frankly, this strikes me as an excessively large shift and makes me suspect that something is off a bit here. Also a bit strange is Democrat support for McCain increasing from 6% to 10%, though this could just be noise.
2. Both McCain and Obama are seen as bringing change to
3. On “maverickness”, Obama actually does better than McCain. He is seen as a “different type of Democrat” by 47%, whereas 48% see him as “typical”; McCain’s figures are 40% and 57% respectively. Note that Kerry’s scores were 30% and 55% in July 2004, likely one of many important reasons why Kerry lost.
4. Obama continues to consistently lead in the “soft” appeal type questions, such as who can you relate to and who understands voters’ needs and problems. On the latter question he trounces McCain 60%-33%, which is a bit surprising. McCain has actually lost ground in this area.
5. Conversely, McCain continues to beat Obama soundly in his strong suit areas: experience and military leadership. A plurality of voters now believe Obama is prepared to be president, but only barely: 48% to 46%. He has a lot more work to do in this area. Only 26% feel that Obama would be very likely to be an effective commander-in-chief, while 52% feel McCain would. Obama is not making any headway in this area, and it’s a serious weakness for him.
6. Much has been made about how Sarah Palin increased enthusiasm among McCain voters, and it is true that he has improved in this area, but he still trails Obama by a large margin. 61% of Obama supporters are enthusiastic, while 23% have reservations (difference of 38%); McCain’s figures are 47% and 37% (difference of 10%).
7. The “McCain is out of touch” meme seems to be succeeding with independent voters, but the “McCain is not a plain talker” meme is not nearly as effective; in fact, Obama scores worse on this than McCain. Independents say Obama understands their needs and problems by a 57%-35% margin, but McCain scores just 48%-45%. Asked if he says what he believes or not, independents rate McCain 58%-38% but Obama gets an anemic 47%-49%.
8. Many people have tried to claim that Obama’s effort to tie McCain to Bush is not effective: the numbers suggest otherwise. A full 46% believe McCain would continue Bush policies (up from 42% in a prior poll), including 38% of independents.
9. On VP picks, voters see McCain’s pick of Sarah Palin clearly: a whopping 75% of them said he chose her to try to win the election, and only 17% because she’s well qualified. Biden’s scores are 57% for qualifications and 31% to help Obama win.
10. In terms of readiness to serve as president, 65% say Biden would be and 24% say he would not. But for Palin, the numbers are 33% “qualified” and 62% “concerned”. This may be part of why Obama has regained some of his lead. Even 32% of McCain supporters are concerned about Palin. The Alaskan governor is now considered by 42% to be prepared to be VP while 52% say she is not; an earlier September poll had 47% saying prepared and 42% unprepared – a bad trend for McCain.
11. Biden’s net favorability is +21% with a high 45% undecided; this is largely unchanged since he was picked. Palin has a net +10% with 30% undecided, down from +22% in a September 8 poll. Independents still have a +14% favorability rating of Palin, which is better news for McCain.
12. If McCain intended the Palin pick to appeal to Hillary Clinton voters, he failed miserably (no surprise to me or most other observers, of course). Palin’s net favorability among former
13. Another interesting but unsurprising observation: 2% of McCain supporters said they were “angry” about Obama’s selection of Joe Biden; 20% of Obama supporters said this about McCain’s choice of Sarah Palin.
14. So far, at least, Palin is not being penalized very much by voters for allegations of misconduct as governor or claims about her honesty – very few voters state these as reasons for disliking her.
15. Palin is seen as someone voters can relate to by a 55% to 42% margin; this is down from 60%-34% in a prior poll.
16. A near majority (48%) of Americans rate the economy and jobs as the most important election issue, including 58% of Democrats, 41% of Republicans and 40% of independents. Voters have confidence in Obama to deal with the economy by a 60%-39% margin; for McCain the figures are 53%-46%.
17. Neither candidate is having much luck convincing Americans that their taxes will go down if he’s elected, but Obama is failing miserably on this count. 34% of Americans think McCain will raise their taxes, but 49% think Obama will. Conversely, 15% think Obama will lower their taxes, and just 8% feel this about McCain. Obama does much worse here among independents than McCain: 53% think Obama will raise taxes and 12% lower them, compared to 31%-13% scores for McCain. This may well be Obama’s single biggest failure to communicate of the election (or McCain’s greatest success).
18. On
19. 62% of Republicans approve of the job George W. Bush is doing? Really? Um. Wow.
Wednesday, September 17, 2008
Favorability Update: McCain and Palin Continue to Slide
I just updated my graph based on two more days of polling figures, and they show that the divergence between the Democrats' and the Republicans' favorables is continuing:

Again, these numbers are "net favorability" for each candidate, meaning, what percentage of those polled have a positive view of them, minus how many have a negative view.
As I surmised on the 15th, these numbers tend to be leading indicators for poll movements, and we are seeing those occur now. I would expect the McCain figures to bounce back; I think most of his hit here is due to his negative campaigning, and if he stops that, he should recover.
But frankly, despite Palin's net favorability now being negative, I think her figures have far more room to drop, at least in the short run. Simply put, more and more people are figuring out every day that she's a dishonest, power-hungry fraud -- and her continued cowardice in avoiding the media almost three weeks after selection makes people nervous about her competence.
Monday, September 15, 2008
There's Hope for Us After All: Americans *Do* Seem to Care About Honesty and Issues
Anyway, what I like about the DK/R2000 polls is that they also do favorability each day, and show the results. In watching these, I noticed a rather dramatic drop-off in favorability for the McCain/Palin side of the ticket, with little change for Obama and Biden:

These numbers are "net favorability" for each candidate, meaning, what percentage of those polled have a positive view of them, minus how many have a negative view. The numbers for McCain and Palin have fallen sharply -- far more sharply than the poll itself. But this may be a "leading indicator" and is certainly good to see (from someone who thinks that McCain/Palin is arguably the worst major party ticket he can ever recall seeing).
Friday, September 12, 2008
More Random Thoughts from the Internals on the Newsweek Poll
1. McCain has 91% of Republicans, Obama has only 85% of Democrats. This is not great news for McCain, but I'm not sure Obama can capitalize on the extra potential for him to draw Dems, since a big chunk of those are single-issue voters or racists. Still, that's what his surrogates will be after.
2. Obama still leads 44-43% among independents. I think that will shift Obama's way once Palinmania wears off.
3. McCain is killing Obama among evangelicals, which I expected, but also Catholics (probably due to the abortion issue).
4. McCain's biggest gains are in strength of support, which means people are voting because of Palin, not him.
5. Was surprised to see 77% of Obama supporters and 88% of McCain supporters say they considered VP important. This is a risk area to McCain, again, because Palin's support is based on myths and "celebrity".
6. Favorability ratings of Obama/McCain nearly the same.
7. Hillary Clinton, even out of the limelight, *still* has a 43% unfavorable rating, which really puts the dampers on some of the "Hillary to the rescue" scenarios.
8. People think McCain/Palin can bring change but that's not why they are voting for them.
9. This is a biggie: everyone talks about the economy being the big issue. But while 55% of Obama supporters list it as such, only 23% of McCain supporters do. McCain leads 22-7% among those who list taxes and spending as their primary concern, suggesting Obama is failing on educating voters about his tax plan.
10. They are about evenly split on ability to handle energy issues, which is why I think the "drill drill drill" nonsense is on the back burner.