Pollster.com

Articles and Analysis

 

CA: 49% Boxer, 42% Fiorina (Rasmussen 7/12)

Topics: California , poll

Rasmussen
7/12/10; 500 likely voters, 4.5% margin of error
Mode: Automated phone
(Rasmussen release)

California

2010 Senate
49% Boxer, 42% Fiorina (chart)

Favorable / Unfavorable
Barbara Boxer: 49 / 47 (chart)
Carly Fiorina: 48 / 41

Job Approval / Disapproval
Pres. Obama: 57 / 42 (chart)
Gov. Schwarzenegger: 28 / 71 (chart)

 

Comments
jack:

Democrats can be overheard saying: "Rasmussen is a Very Good Pollster today."

____________________

CHRIS MERKEY:

Yeah that's why i think it is finally raining in PA. No, we just wonder whether his likely voter model is right or not. Another poll had Fiorina ahead but actually they are both pretty close to each other. THis is going to be a close race. Boxer wins because Democrats have an advantage in CA.

____________________

SC Guy:

Exactly, Jack. People beat up Rasmussen because they're releasing the polls that show the Dems being in trouble but they are!!

____________________

Paleo:

No, still a lousy pollster. Just an anomaly.

____________________

Stillow:

Hahahah, that si rgiht, for the libs, today is a good Rass day....3 cheers for Rass they are singing in their parents basements.

____________________

SC Guy:

Paleo, if Rasmussen is a lousy pollster why do they seem to be pretty accurate? If you read towards the end of their release they recap what there last figures were before the 2008 election (in the individual state that's featured) and they generally are quite close, with some exceptions. No pollster is perfect but I think Scott Rasmussen deserves more respect than he's given.

____________________

Paleo:

The head to head numbers were close, but if you look at the state polling, there was a clear Republican tilt. And, as Nate Silver has pointed out, since 2008 Rasmussen has deviated more and more in the Republican direction.

____________________

You right wingers need to understand that you have lost big time the last two elections, yet you act like you speak for the majority of americans. Democrats have WON 4 of the last 5 presidential elections.
Stillow was saying in 2008 that the polls that showed Obama beating McCain were all wrong, and you act like this did not happen. You guys never let facts get in the way of your opinions.

____________________

You right wingers need to understand that you have lost big time the last two elections, yet you act like you speak for the majority of americans. Democrats have WON 4 of the last 5 presidential elections.
Stillow was saying in 2008 that the polls that showed Obama beating McCain were all wrong, and you act like this did not happen. You guys never let facts get in the way of your opinions.

____________________

You right wingers need to understand that you have lost big time the last two elections, yet you act like you speak for the majority of americans. Democrats have WON 4 of the last 5 presidential elections.
Stillow was saying in 2008 that the polls that showed Obama beating McCain were all wrong, and you act like this did not happen. You guys never let facts get in the way of your opinions.

____________________

melvin:

Boxer must be leading by a wide margin if Rasmussen has her up 7%.The GOP has a major problem that is about to explode,the Republican politicians in Utah sent out 1500 forms telling the illegal immigrants to go home,but Cnn is reporting about 400 of those forms was sent out to American born Citizens.This was the worst case of racial profiling you can think of,so you can imagine when the ARZ law takes affect in a few weeks how many racial profiling cases are going to take place.The GOP has just given the Democrats 80 to 90% of the Hispanic vote.If the Republicans was smart they would not let that law in ARZ take place in 2 weeks, because we just seen in Utah what can really happen when you profile a race of people.

____________________

Stillow:

paleo - Silver is ntohing more than a liberal ideologue driven by his desire to twist and spin polling results to favor his point of view. He is very good at spin to his credit, but to quote him or his analysis as a credible source is laughable. he is just another liberal who twists things around to suit his view of reality.

Outside liberal circles, silver does not have a lot of credibility.

____________________

Stillow:

AlanSnipes: - Correction. Dems have won 3 of the last 5 pres elections. 92,96,2008. GOP won 2, 2000,2004.

But its a meanginless stat, since the country goes back and forth between parties all the time. We swing to the right, we swing to the left...back and forth we go.

____________________

Paleo:

Interesting. Because I see a lot of the righties on here citing Silver when it suits their purposes.

____________________

Steve Miller:

"Correction. Dems have won 3 of the last 5 pres elections. 92,96,2008. GOP won 2, 2000,2004"

Actually, the vote count was

Al Gore 50,999,897
George W Bush 50,456,002

Gore won the election; Bush won the presidency

____________________

Xenobion:

lol most Democrats would realize that after looking at this Rasmussen poll and its "house effect" you'd place a "Rassmussen effect" and place the poll at Boxer 56% Fiornia 36%.

____________________

Field Marshal:

Silver is a Kos loon with little integrity and standing.

I like how AlanSnipes says "You guys never let facts get in the way of your opinions." when his facts are completely wrong two sentences earlier. Hysterical and typical of the left.

And yes, that is a meaningless stat. Heck, go to the last 8 and the GOP has won 5 of them! WOW!

____________________

Stillow:

Steve Miller: - You do not win the presidnecy by the popular vote. You win by getting the most electoral votes. Therefore, 2000 Bush won the elction and the presidency. so the correction stands, 3 of the past 5 won by Dems, 2 won by GOP. Its a very simple concept and the law on how a presidnet wins is clear.

____________________

Louis:

It is not a question for me of good and bad pollers. First Rasmussen has the highest Republican house bias of any pollster (see article published on Pollster.com). This means when looking at a Rasmussen poll you need to adjust Republicans down and Democrats up by about 2% for a Senate race. Second is the question of likly voter screens. this is a very debatable issue but at best I think one has to question using more than 30 days before an election. Third the likly voter screen is a joke when applied to Presidential approval and Rasmussen seems to be the only one who uses it here making his numbers not comparable with any others.
I think he has a definte personal political baises which occasionaly show in pharseing of his polls but not generally. On the whole I think when adjusting for the "house bias" his polls are resonably accurate.
I think he has made a terrible mistake in applying a likly voter screen to prsidential approval. As the point of such polls is to find out how people feel the president is doing his job and not how they are going to vote in an election. The only time such a screen might be meaningful is close to a presidential election when running the number with and without the screen might add some information on the potential election results.

____________________

Xenobion:

"I think he has made a terrible mistake in applying a likly voter screen to prsidential approval."

I completely agree Louis. Why would you place approval only on the shoulders of a Likely Voter model? I've said over and over and over again that approval does not = votes. Approval should be a pulse of national/state sentiment toward a candidate.

Its like gauging popularity using only the votes of popular people.

____________________

Stillow:

Louis - then how do you explain Rass haivng Barry's approval right along with everyone else? in fact for the past tweeks or so, Rass has had a higher approval # for Barry than gallup....Rass was up around 49/49 and Gallup was down around 44/45 for a while? how do you explain this based on your claims of Rass's bias?

If Dems are correct, then rass should always have Barry at the lowest approval rating. yet his approval #'s are right i nthe mix with everyone else....one difference is his disapprove, but that is his polling method, he does not allow for an I don't know, he makes you choose on, which means he makes leaners each way actually give their opinion at the time.

But this is not about rass...its about polling in general. Dems on this site whine about all pollsters who give bad #'s. One day they love Gallup, the next they hate them, one day rass is the devil, then Rass is cooler than santa clause.

____________________

Stillow:

X - Because X, politicians do not care what non voters think. If your not going to vote then elected officials could care less what you think. its important to know what a presidnet sapproval is among those likely to vote. Because they retain or lose their power based on those who actually vote, not those who do not. Its actually a pretty simple thing...not sure why you intelectually stimulated libs do not get it. I mean, libs pretending not to be libs who really are libs, but don't..........ahh hell.

____________________

Xenobion:

Rassmussen has Obama's approval for the lowest vs. every pollster that doesn't use undecideds in their voter models. Rasmussen also hold higher Disapproval vs. all pollsters. Rasmussen has so much weight over the graph on the front page, if you take them out. Obama's Disapproval goes down 2%.

____________________

Xenobion:

Uhhh then polling approval is dumb in this scenario. If non-voters are not going to be polled for this and you make justifications about the "pulse" of America and what it wants... Then you can really never make that justification again using these types of models.

And why wouldn't you use registered voters? They are people that are positioned to vote, the standard error of a likely voter model seems to go way up even allowing the word "likely" to show up.

____________________

Stillow:

Again, Rass uses those most likely to vote. Why are you libs not whining about gallup polling RV's? Why not adults? your arguments are thin and they stem from a simple matter of not liking the results.

Otherwise there is no way Rass would ever have a higher approval for Barry than gallup or others, which he does all the time. For a while now Rass has been up around 48, many others down lower.

Your arguments are bogus and the #'s show it.

____________________

Farleftandproud:

I am not going to say Rasmussen is a good pollster in general, but I will say that common sense told me the poll yesterday was poorly conducted in CA. They may be right with Meg Whitman, but CA voters who are Independents don't usually vote for extreme candidates like Fiorina. IF this were PA or Ohio, and she were running, I would have likely believed it, but I think the GOP was worried about their future of winning statewide elections, that is why they planned Arnold Schwartzneggar's "Total Recall". It took someone like him to get the younger and Independent voters to get a Republican elected in CA, statewide I mean.

____________________

Stillow:

Then why not just always use adults? they are a simple form away from being able to vote....

You libs cannot grasp the simplest of concepts. its good that we get polls of LV, RV and adults. We need data from all 3 groups to help get an accurate reading. The approval rating from a voter who votes every single time should carry more weight than someon who is regsitered but has never voted before.

We need data sets from all these groups, then we can average them out to try and find good #'s. Seems to be obvious to most of us.

Both gallup and Rass give us good data from there respective models. Not sure what you libs have a beef with other than you simply don't like the #'s Rass always gives you.

____________________

Paleo:

"Rass uses those most likely to vote. Why are you libs not whining about gallup polling RV's? Why not adults?"

Because those "most likely to vote" creates a degree of subjectivity inherent in the voting screen not present with registered voters. You're either a registered voter or you're not. Whether you are a likely voter involves a degree of variables.

____________________

Xenobion:

Stillow you base yourself at the lowest common denominator and can't actually make arguments other than accusing me of being angry with the results, and somehow biased because I don't cherry pick every poll, only Rasmussen ones. Sorry I don't have the time to go over every pollster only the most prolific one, Rasmussen, and they functionally have a flawed methodology. Call me an idiot, a liberal, or whatever since you make hackneyed ad-hominem attacks at everyone at Pollster because you can't come up with a cogent argument yourself, but let the people that know a little about statistical science talk.

My argument is you can't justify national sentiment with this polling methodology. You can't say a majority this, or a majority that in reference to national sentiment on a likely voter model. The numbers further get scewed especially after the 2008 election where likely voter models reflect more Republican likely voters than Democrat likely voters, something 2008 shattered, yet Rasmussen continues to keep, and keep the methodology hidden away in some safe 20 miles underground. The lack of transparency for this methodology makes it hard to justify. Its like saying, "Eat this, its good for you." Well why? Rasmussen continues to only let people that pay for cross-tabs of their data sets see the results. Their is no real transparency to their "science."

____________________

Stillow:

paleo - Do you know how many RV's are out there who have never voted in a single elction? Millions. It makes perfect sense to gather data from both RV's and LV's. If I am age 50 and never voted before, then my approval or disapproval of Barry should mean a littl less than a guy at age 50 who votes in every single election.

I thought you libs were all about getting as much data as possible on everything....no?

____________________

Stillow:

X - That as the typical reply from someone who has lost the debate. all that si left is for you to call me a racist.

Check and mate. The lurkers can decide for themsleves. Do we want to maximize our data from all sources, or only use sources from polls that liberals approve of and give better #'s for you guys.................

Way to easy today..............

____________________

Paleo:

"I thought you libs were all about getting as much data as possible on everything....no?"

Accurate data. The time for likely voter screens is after Labor Day. Some registered voters tune in later than others. Because each pollster has a different screen, you have both the chance for more accurate results and less accurate results with LVs than you do with RVs.

____________________

Stillow:

paleo - if that is ture then rass, PPP and some other LV pollsters will reflect the changes after Labor day. I do not understand the beef you libs have with LV pollsters using there LV screen now as opposed to later. All it does it give us another datapoint to work from. A LV now will probably be a LV in November....so it gives us sort of a baseline to work with. then come september when more RV's might start paying attention and beocming LV's....it serves us well to see the alteration in LV #'s that occur.

I'm an all the above guy, I want Adults, RV and LV polling all year long. You lefties may like to pick and choose what and when you ant soemthing, but I like getting the most amount of data possible....as it gives us the best chance to get the end results correct.

____________________

boomer40:

89% can offer an opinion on Fiorina? Huh?

____________________

Gopherguy:

Don't forget about the article posted yesterday regarding Rasmussen. It's not only Silver who believes Rasmussen's numbers are off.

Also, people assume that this poll is an anomaly. What if this poll still has Rasmussen's house effect? That would mean Boxer is up by 10. What if the house effect has now changed? What if this is actually an anomaly? I find the first question to be quite intriguing.

____________________

Paleo:

"I do not understand the beef you libs have with LV pollsters using there LV screen now as opposed to later."

Because an accurate estimation of LV only can take place the closer you get to an election. It's meaningless well removed from an election. And worse, can give a far more distorted picture of the likely electorate, so that a RV sample is more reliable.

____________________

Stillow:

Paleo - That is subjective. I recall in NJ last year LV polls all year showed christie leading corzine. the argument from the left was exactly what your saying. The lefties would say these are LV and its so early. corzine hasn't started advertising yet, etc, etc.

But the early LV samples gave us a very solid baseline to work with and all year they shoed christie at a cap of 45 percent....Christie bounced around between 48 and 52. It ended 49-45....

Based on your argument, polling RV's would be just as flawed, because there are tens of million of people who are registered, but do not vote. So by polling RV's only, your polling many many people who are not even going to vote.

The correct thing to do would be to get pollsters doing polls on adults, RV and LV's....that way we maximize our data and thus give us a bigger picture in which to make comments on.

Narrowing your datastream to only RV's is flawed. I think you guys know it, but at this point are unwilling to admit it.

____________________

real_american:

"Gore won the election; Bush won the presidency"

Actually, we have this thing called the consititution and within it we have this thing called the electoral college so who has a bigger overall popular vote has as much effect on an election as a butterfly farting in China.

Sorry 'bout that dems, that nasty old constitution stood in your way again. No wonder Obama hates it so much and keeps nominating supreme court justices who want to rewrite it.

____________________

dpearl:

I agree with Stillow's point that more data is better here. It would be nice to see Rasmussen's RV numbers (i.e. before their weighting to produce LV) to make it easier to compare between pollsters. I also think that LV is the proper population for any ELECTORAL issue (Presidential approval is not an electoral issue in my mind since he is supposed to be the President for all Americans). Just because you don't have perfect information 3 or 4 months out doesn't mean you don't have any information at all - weighting to better predict the election is the correct thing to do. As the election draws near your LV model will get better as it should if you are doing your job. When we get very close to the election I think pollsters need to track three categories: Registered to vote, Likely to vote, and Already voted. The difference between RV and LV combined with a knowledge of what % of the electorate has not yet voted can tell you about the potential benefit of get out the vote election day organization. Of course, the election prediction itself should be based on % from Already voted + % estimated from LV's who have not yet voted.

____________________

Xenobion:

Psh Stillow I do believe the one who starts name calling loses. And I can dissect your arguments into microscopic elements and your head will keep spouting off ad-homs like a child who dropped his icecream cone. I've posted an argument up there, one you cannot answer, nor was I really focusing you to answer it because its always white flags with you.

____________________

Louis:

On Gore. yes we have aconstitution and except for the Supreme courts interference Gore would now be President.
Fact the Florida Supreme Ct. declared there should be a state-wide recount and established the critera for a valid vote.
Fact The Supreme court despite it's long held position that state courts deserve deference in such matters refused to allow the recount to go on.
Fact the Miami Herald along with other papers latter undertook this recount and established that Gore had won Florida and hence the election.

But that is history and cannot be changed now.
Stillow,
If you will go to the Pollster.com chart and take the approval and disapproval rates with and without Rasmussen you will find that it 48.8 % disapproval and 45.9 % apprroval with Rasmussen and 47% Disapproval and 45.5% without him. The margin of Diapproval shrinks from 2.9% to 1.5%. the margin is almost double when rasmussen is included.
If you use my perfered method oe eliminating all party and internet polls and then going to the last 30 days including Rasmussen it is Disaprove 48% and 47.1 % approve. A .9 diapproval margin taking out Rasmussen it is 46.5% approve and 45.7 Disapprove a .8 margein Again the margin is doubled and in this case revesed when Rasmussen is included. I do not think this a bias of Rasmussen but of the choice to use a likely voter screen. Again even if by some chance they were to come out the same I do not believe polls using such different critera are comparable and Rasmussens numbers should be considered seperatly from the rest.

____________________

Stillow:

X - i picked apart your argument and the readers can see that. It was not that hard to do, you just choose not to read it.

Louis - Thanks for the good reply, its refreshign to ahve a lefty actually take o nthe issue unlike X.

You need to remember Rass forced an answer on approve / disapprove, not all pollsters o nthe chart do that. many allow for undecided. Plus most of the chart is made up of RV and adult pollsters. rass is LV and that would give him differetn numbers.

PPP for example today has Barry at 52 percent disapprove among RV. Are we now going to discount PPP because his disapprove his so high?

Unlike what X is trying to do, you cannot simply pick and choose the data that best suits your argument. We need all the data we can get from all sources and all modeling schemes. The left doesn't care of LV polling because it usually shows the GOP doing better than they do among RV or adults. that is because republicans tend to vote at a higher frequency than Dems, not always, but the majoirty of the time.

What you lefties are doing is simply adding another layer of wool over youe eyes. No one can stop you from picking and choosing the data you like, but it makes your commentary less useful and on election day it makes it bogus.

Polling is not an exact science, so it snecessary to get as much data as we can. To argue otherwise like x and paleo is naive at best.

____________________

Stillow:

Perhaps what is needed then is a 3 chart system. A rolling average of polls doing adults, one of RV and one of LV to give us a better idea of what is what. Then we can use subjective weighing to form a basis for opinion and prediction.

But each model is helpful in its own way...the picking and choosing method is simply bad form.

____________________

Another pollster.com posting with party affiliation omitted when the Dem is ahead. What gives?

____________________

seg:

AlanSnipes:
I haven't read much of this thread, so perhaps someone has already pointed this out: Al Gore lost to Bush.

I think that it was the Atlanta Constitution paper that put together a panel of reporters and had them re-count all of the disputed ballots, hanging chads and all.

Despite the fact that they were overwhelming liberals, Bush won easily by their count.

Perhaps you know of a House, Senate, or presidential election with an amazing surge of votes for the Republican after the "first" count. I can't think of any, but there are many such cases for Democrats.

Republicans assume that they must win by at least 1,000 votes or democrat polls will stuff the boxes and change the outcome. The Gregoire and Franken farcical events (finding bags of ballots that were heavily democratic in car trunks, closets, etc.) were orchestrated by a group put together by Soros. The multiple re-counts when the previous one STILL showed Bush ahead were par for the course.

Ever notice that big, democratic cities always withhold their vote counts until last if the race is close? Notice that the laggards needing the re-counts in Florida were all 100% democrat controlled, though none were big cities? I am sure it was pure coincidence.

____________________

Stillow:

seg - Its commonly assumed these days that a republican needs to win close races by more than 1,000, in some cases more. Dems do have a habit of "finding" votes in close races. Many of us said that with Brown in MA, if he was going to win he had to do so by probably 5,000 votes....or votes would suddenly pop up for coakley in someone's trunk.

____________________

Bukama:

Louis

Sorry, but as seg notes above, the conclusion of all media reviews on the 2000 vote count in Florida is that, no matter what the Courts (FL or SCOTUS) might have done, the vote would have eventually gone to Bush (and thus the election). Here is a CNN report on the counts: http://archives.cnn.com/2001/ALLPOLITICS/04/04/florida.recount.01/

Now, you can try and argue that the election was flawed - and all sorts of people didn't vote that wanted to vote (felons, blocked by FL, and overseas military votes blocked by Gore). And some number of people voted for the wrong person due to flawed ballots developed by the Democrat controlled county bureaucrats. But that is all speculation. You can also speculate that had Richard Daley not been mayor of Chicago in 1960, Nixon would be have become President 8 years earlier.

As for the popular vote - that too is speculation. Considering how flawed the FL vote was, who is to say there were not swings of 10s of thounsands of votes in California, NY, TX, etc. - places where the vote wasn't close so nobody reeavuated the final number. This in fact is why we can't have a national popular vote - it would require 50 recounts every election cycle anytime the outcome was within a million votes difference.

By all measures, Bush won the election in 2000 fair and square, and it was the poor losership of Al Gore that delayed Bush from being able to set up his administration in as timely a manner as was normal, and that hobbled his administration with unwarrented questions of illigitamacy at least up till the 9/11 attack.

____________________

Xenobion:

You're really bad at this Stillow. I've said what I've said about the Likely Voter model, we don't have to agree, but you also don't have to always go into a tirade about Liberals or me representing liberals on an issue. Its a straw-man argument that has nothing to do with the subject at hand.

You disparage Daily Kos polls, perhaps rightly so, its easy to pick out the bad polls for flawed or very unrepresentative methodologies. How you cannot recognize this on the right shows your shortsightedness to defend your ideology over supporting the un-exact science of statistics.

____________________

dpearl:

With respect to the 2000 election - the most definitive recount was done by the Media Consortium Florida Ballot Project through the National Opinion Research Center. They found that Gore would have won by some criteria and that Bush won by others. Most interestingly it turned out that Gore was pushing the wrong issue in all of his campaign's legal arguments. Where he would have gained a substantial number of votes was in the "overvote" category where people checked his name on the ballot but also wrote his name in as a write-in vote (he had 885 more such votes than Bush) . However, at the time his legal team mistakenly thought the overvotes were either cases where people voted for two different candidates or that they came from areas that favored Bush so they argued against including them. Meanwhile the Bush legal team argued that not including the overvotes did not provide equal protection for those voters. Thus it appeared that Gore probably did win the election by the most objective criterion (i.e. full count of all ballots where voter intent was clear) but the Supreme Court did not stand in the way of his victory what-so-ever. If they had ruled in Gore's favor and ordered the type of recount he requested, he would have lost by more than he did.

____________________

Stillow:

X - seriously, ya gotta stop at this point. This thread has embarrassed you a great deal. Your just antoher cookie-cutter leftist on here who engages in Rass bashing on a daily basis....because you do not like his results. You are arguing in favor of selective data gathering, while I am on the side of getting as much data as possible. the readers can decide.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR