Not So Fast: Another Poll Shows Franken Trailing
A poll out the other day gave hope to national Dems, because it seemed to suggest that Al Franken had taken an improbable lead over GOP incumbent Senator Norm Coleman. But not so fast: Another survey out finds the opposite result.
The numbers from SurveyUSA: Coleman 52%, Franken 39%, not significantly changed since a month ago. Somewhat counterintuitively, the poll's internals show Coleman ahead 58%-31% among voters age 18-34.
Yesterday's Rasmussen poll gave Franken a two-point lead, after trailing 48%-45% a month ago.
Both pollsters have good reputations, so it's really not clear what the status quo is at this point -- but a one-time comedian might just have a real shot at knocking off a GOP incumbent in a real upset.
Advertisement
Hey, how come you don't have the Quinnipiac poll up? Oh wait, it shows a good lead for Obama, hmm...
July 15, 2008 11:11 AM | Reply | Permalink
we do have it up. it's the lead item in the morning roundup, which is linked on the front page of TPM.
but we deliberately didn't give it a separate post because we're secretly pro-McCain!
July 15, 2008 11:13 AM | Reply | Permalink
I knew it!!
July 15, 2008 11:16 AM | Reply | Permalink
:-) Thanks, but why isn't it anywhere on the front pages, as new polls are normally?
July 15, 2008 11:17 AM | Reply | Permalink
It is on the front page.
July 15, 2008 11:27 AM | Reply | Permalink
I thought it was that you own stock in Newsweek since that was your prominent "FISA cave" redux post for the past weekend...
July 15, 2008 11:17 AM | Reply | Permalink
Close, but no cigar. He owns stock in Associated Press and The New Yorker.
Seriously, it's a damn shame that headline wasn't four millimeters higher and half a point-size bigger. An obvious McSame mole.
July 15, 2008 1:59 PM | Reply | Permalink
This might be worthy of a post, no?
http://www.thecarpetbaggerreport.com/archives/16199.html
July 15, 2008 11:19 AM | Reply | Permalink
And what about all those young voters who exclusively use cell phones thus do not participate in polls because pollsters don't want to pay the additional expense to manually dial a cell phone?
I think all the polls are missing a large demographic which will skew the results
July 15, 2008 11:15 AM | Reply | Permalink
survey USA is really stinking it up lately.
just look at the internals of the poll.. Dems and republicans are tied... ummmm... no way. Dems have a 7-10% advantage in Minnesota, if not more.
they are supposed to make us believe the young support Coleman almost 2 to 1? LMAO.
bad poll. Survey USA i have no respect for lately. their polls are BAD.
July 15, 2008 11:16 AM | Reply | Permalink
thanks for this analysis because I have been working from the assumption that SurveyUSA was one of the best polls. They seemed to do very, very well in the Dem primary. Maybe they don't have their party split assumptions correct and that is skewing their results.
July 15, 2008 11:20 AM | Reply | Permalink
i watch polls like a hawk and I dont understand why they have such a good rep. they are awful.
there have been many instances where they even have McCain beating Obama among blacks in some states. LMAO. i mean its a joke.
what gets them in trouble IMO, is that they dont survey many people in each demographic.. so if you survey 4 blacks.. and 3 support McCain.. they will put out a poll showing 75% of blacks support McCain, and the final results will skew that. you have to poll 50 blacks to get a good level of idea how they will vote or you end up with a huge margin of error.
July 15, 2008 11:24 AM | Reply | Permalink
Yet another chance for all of us to learn how the margin of error works.
Now, who can tell me why both of the recent Minnesota Senate polls are consistent and don't contradict?
July 15, 2008 11:24 AM | Reply | Permalink
I'm very interested in these internals. As an 18-34 year old in Minnesota, I find it extremely questionable that we break for Coleman 58-31.
Maybe they worked their poll on a listingf rom FOX Radio?
July 15, 2008 11:25 AM | Reply | Permalink
This is the fourth poll in a row that has Coleman up double-digits in Minnesota. And they are the ones who are in line with other polls of the state. The Quinnipiac poll had Coleman up 10% just last week.
July 15, 2008 11:27 AM | Reply | Permalink
Rasmussen has Franken up by 2%.
July 15, 2008 11:43 AM | Reply | Permalink
Also, it's unfair to use one internal to question the whole poll given the HUGE MoE on internal polls.
July 15, 2008 11:27 AM | Reply | Permalink
so you are saying that republicans and dems make up the same # of the electorate in Minnesota?
July 15, 2008 11:42 AM | Reply | Permalink
It's the middle of July. The best we can expect from these serial polls is to see a trend one way or the other over an extended period. As is often said, any one of these polls is no more than a 'snapshot in time'.
July 15, 2008 11:48 AM | Reply | Permalink
Eh, they're not polling Independents either, which is a major problem for pollsters in MN, since we have a large number of independents of varied parties and no party. No polls done in MN will be truly accurate unless they take the independents into account.
July 15, 2008 3:20 PM | Reply | Permalink
What Dem primary? The primary isn't until September, and it looks like Franken has just token opposition.
That huge Coleman lead in the youth vote looks bizarre. Why has no other poll picked that up if it's right? I haven't seen an age breakdown in other polls. Anyway, even if Franken was supposed to be up big, it's still just mid-July. I wouldn't be asking where Franken's senate office will be located just yet.
July 15, 2008 3:32 PM | Reply | Permalink
Regardless of the polls, how is Franken doing on the stump? A friend of mine in St. Paul, very liberal, saw him speak a couple months ago and said he was terrible--very wooden, uninspiring, etc. This certainly surprised him, and me, given Franken's years of being in the spotlight. How's he doing now?
July 15, 2008 4:40 PM | Reply | Permalink
"a one-time comedian"???
July 15, 2008 5:24 PM | Reply | Permalink
SurveyUSA, MasonDixon and Quinnipiac tend to overestimate Republicans in most polls. Rasmussen is better, but also tends to have above average numbers for Republicans. CBS and Gallup tend to be at the opposite end. Others are all over the place.
It is also important that sampling is different for each organization between the national and regional/state polls. It's not entirely clear what methodology would make polls more accurate because there are so many variables. And the weighting on top of that could also skew results if it relies on old demographics figures.
My guess is that SUSA underestimates the potential Democratic vote based on past elections. However, doing so consistently is also important, as stable numbers even in badly skewed polls show that the race is not changing overall. Trouble may arise if demographics somehow change between polls, e.g., one party constituents have a surge in voter registration or suddenly become more likely to show up to vote. But other comments are right on one point--it makes no sense that SUSA internals show younger voters going for Coleman. Given a number of obvious flaws in SUSA splits and weights, the actual numbers are badly off and Franken is likely closer than SUSA shows. But he has not made a significant move over time and he clearly started behind. If his numbers don't start moving by convention time, Franken will be a hard sell the rest of the way. Coleman's best strategy right now is to stay out of the news.
July 15, 2008 9:30 PM | Reply | Permalink