Poll shows Hillary landslide in Tennessee; Obama gaining ground nationally

Some more new polls today, and the news is mixed on the Democratic side. In Georgia, which is supposed to be solid Obama country, InsiderAdvantage shows the Illinois senator with a big lead, 52-36. Obama gets 73% of blacks, 54% of Hispanics, and 33% of whites.

But here in neighboring Tennessee, which is supposed to be a hotly contested tossup/lean-Hillary state, the same firm’s polling shows Clinton with a huge lead, 59-26. The poll was taken yesterday, and thus may reflect a shift of Edwards voters into the Clinton camp (though 7% still prefer "other"). But what’s really surprising is that Clinton isn’t just winning the white vote, 64-19; she’s also winning the black vote, 46-43!

Moreover, Clinton has almost identical margins among men and women; she wins big in all age groups (with her biggest margin of all, 81-14, among 18-to-29-year-olds!!); and she is favored by self-identified Democrats, Republicans and Independents. So either this is a screwy poll, or Obama-mania really just hasn’t caught on at all here in the Volunteer State. If Obama’s internal numbers are showing a similar trend, maybe that explains his conspicuous absence — this supposed battleground state may be, for whatever reason, a lost cause for him.

InsiderAdvantage also released a poll on the Republican race here in Tennessee, and it’s much tighter, with McCain leading Huckabee and Romney 33 to 25 to 18. Thirteen percent are undecided.

UPDATE: On the bright side for Obama, he’s setting fundraising records, and is already buying ads in post-Super Tuesday states. Barring an enormous Hillary sweep on Tuesday that re-establishes her "inevitability" and thus causes a paradigm shift in the media storyline, this race will go on for a while. As this chart shows (context here), there are a ton of delegates at stake on Tuesday, but a ton more after Tuesday, too.

Meanwhile, the very early returns from Rasmussen Reports suggest that Edwards’s departure is helping Obama nationally:

In the race for the Democratic Presidential Nomination, it’s now Hillary Clinton 42% and Barack Obama 35% [in the three-day average]. Last night was the first night of interviews without John Edwards in the race. For last night’s data alone, Clinton and Obama were essentially even. Samples for individual nights are very small and results should be interpreted with caution.

The daily history shows that it was 41%-32% Clinton (a 9-point lead) from January 28-30; now it’s 42%-35% (a 7-point lead) from January 29-31. The big question is what happens tomorrow and the next day, when we’ll see a three-day average that is entirely after Edwards’s withdrawal. Gallup‘s tracking poll will also be worth watching closely; they had the race slightly closer (6 points) even before Edwards bowed out.

UPDATE: Gallup’s new data is out. It shows the Clinton-Obama race narrowing further, from 42-36 in yesterday’s three-day average to 43-39 in today’s three-day average — just a 4-point lead for Hillary!! However, Gallup’s write-up says that "Wednesday night’s numbers (the first with Edwards excluded from the ballot) show no clear indication that either candidate is benefiting disproportionately." I guess that means Obama’s 2-point gain between Jan. 27-29 and Jan. 28-30 is an indication that Jan. 27 was a good day for Hillary, rather than an indication that Obama cleaned up among former Edwards voters on Jan. 30.

24 Responses to “Poll shows Hillary landslide in Tennessee; Obama gaining ground nationally”

  1. Angrier and Angrier says:

    I think the Tennessee poll is a bad harbinger for Obama. Sounds like in states with older white populations, those folks are going to Hillary.

    It was good while it lasted.

  2. Derek says:

    First, Tennessee has the fourth-highest African-American population among the Super Tuesday states. If the black-white voter trend continues, it’s difficult to see how Obama can pull in enough delegates (i.e., if he can’t win in Tennessee, how can he pull off a nationwide victory). Ditto for Alabama.

    Second, Hispanics have “underperformed” so far in the primaries. In Nevada, Hispanics are almost 20% of the population, and blacks are 6.6%, but each was 15% of the voter turnout in the caucus. In Florida, Hispanics make up about 20% of the population and blacks are 17%, but 19% of the voters were black and only 12% Hispanic. If Clinton expects the edge among Hispanics (who, weighted for delegates, account for about 14% of Super Tuesday voters, as opposed to about 11%, also weighted, for black voters), the Hispanic turnout must be higher. Otherwise, it’s much-needed good news for Obama.

    Finally, though it’s a “delegate hunt,” the media will focus on the “sexy state hunt,” as in, which candidate wins each state. Voters will be watching the faux electoral map to see which states turn Clinton’s color or Obama’s color. Though the delegate race so far suggests a(n approximately) 50- or 75-delegate win for Clinton on Tuesday, the map may tell a different tale. With polls showing “winners” so far, and based on “sex appeal” (i.e., size)… Clinton may win by a closer margin, and Obama will certainly pick up delegates nearly everywhere, but those “won” states may leave an indelible impression among most voters for post-Super Tuesday.

    <> (single or double carat indicates strength of current lead)

    <>

    <>

    < ND>

  3. Derek says:

    ((Clinton Obama)) (single or double parantheses indicates strength of current lead)

    ((CA

    ((NY

    IL))

    ((NJ

    (MA

    GA))

    (MN

    (MO

    ((TN

    ((AZ

    CO)

    ((AL

    (CT

    ((OK

    (AR

    KS)

    (NM

    (UT

    ID)

    (DE

    AK)

    ND)

  4. Derek says:

    Sorry, the carats were a mistake.

    A further musing: if McCain becomes inevitable, doesn’t that help the “independent” voters go to the polls for Obama after Super Tuesday? Indeed, Brendan’s call for “CHAOS!” on the GOP side might be the only thing keeping Obama down: many independents are assisting McCain, who needs it, rather than voting on the Dem side after the GOP race is sewn up.

    (Of course, Obama has fared far worse in closed primaries than open ones, suffering a McCain-lite syndrome.)

  5. yea says:

    it looks like obama is spending a significant amount of money on post super tuesday states. even if hillary is a huge favorite, it looks like this primary will be contested for longer than any primary i can remember in my short life.

  6. yea says:

    derek,

    isnt there a new poll out that has hillary leading california by only 3%? even if that poll is an outlier is a “((” edge really accurate at this point?

  7. Derek says:

    yea, I suppose if you’d like to nitpick on that one, it’s a plausible argument. There have been four polls this week in CA: +12, +11, +17, and +3. One survey indicated that 20% of voters have already cast a ballot. Another indicates that 23% have, and they cut 56-32 for Clinton. Finally, Rasmussen has been slightly more erratic than most other polls (ARG and Zogby generally performing even worse). For those reasons, I think it’s still a “substantial” Clinton lead. Any Obamentum must make up for those early voters; and if there’s a second poll showing a single-digit Clinton lead, I would be inclined to say that it’s a “slight,” rather than “substantial,” lead.

  8. Joe Loy says:

    Per the Courant, Clinton and Obama are both to campaign in CT on Monday, Pooper Duper Eve :}. Barack in Hartford, Hillary it doesn’t say Where yet (she did a Town Hall in Hartford just this last Monday). // Were it up to me I’d direct her next to Waterbury & environs: the storied Naugatuck Valley, ancestral seat of our bluecollar middleclass Reagan Democrats (which Reminds me, poor Johnny Edwards is gonna need Healthcare to have 2 pairs of Lips excised from his Butt, which I hope has some Coverage :). / OTOH I’d steer Hillary clear of venues such as Simsbury and Greenwich, whose whitewinesipping graduate-degreed betterclass of Democrats, being enthralled with New Camelot :}, are probably beyond Reach ;>.

    Meanwhile General John “Old Blood & Guts” McCain ;> rolls into Fairfield aboard the Inevitability Express on Sunday ;}.

    So, the humble Nutmeg State is on the Map. We must Savor all this while we can. / Yes, in years to come I’ll sing to my darling granddaughter of how for One brief shining moment, there was Connecticut. :)

  9. David K. says:

    Nothing about MoE? Or keeping in mind how the polls have not been as reliable in recent years, or even this year?

  10. Derek says:

    Hi, David K.,

    First, the polls have been overwhelmingly accurate this year, with the singular and notable exception of NH-D. That is, if you identify the more accurate pollsters (setting aside, say, ARG, and perhaps occasionally Zogby and Rasmussen).

    Second, the lay-prestige of “MoE” is often a statistically-incorrect use to try to write off most leads as a “statistical tie,” when, in reality, that’s not what the MoE actually means. (I’ll refrain from delving into mathematical formulae here, but most local libraries carry good books on statistics, if you’re truly interested.)

  11. Brendan Loy says:

    Clinton and Obama are both to campaign in CT on Monday … Meanwhile General John “Old Blood & Guts” McCain ;> rolls into Fairfield

    Harumph. The candidates appear hell-bent on campaigning in or near practically every single place I used to live — Phoenix, L.A., NYC, Greater Hartford (I think the only one they haven’t hit is South Bend) — as well as a place I may someday live, Denver… but NOT the place I currently live, Knoxville, even though TN is a Feb. 5 state too.

    It’s a vast left/right-wing conspiracy to piss me off.

  12. yea says:

    derek,

    didn’t the majority of polls have obama ahead in nevada up until the end, and wasn’t obama’s win in sc much larger than the predicted margin by the concencus of the polls?

    i do agree with your point on people misunderstanding the margin of error. a 3pt differential with a 3.5pt margin of error != a statistical tie.

  13. Brendan Loy says:

    P.S. And, come to think of it, if the Dem race really does drag on, there’s an excellent chance they’ll eventually hit South Bend! Between Pennsylvania (April 22) and West Virginia (May 13), the only primaries are in Indiana and North Carolina, both on May 6. So there’s a two-week window for campaigning in just those two states. If Hillary and Barack are still going at it by then (probably unlikely, I admit, but certainly not impossible), they’ll have more than enough time to travel up and down the entirety of the great state of Indiana trolling for votes, stopping in every major and minor city along the way. So they’d certainly make it to South Bend, which is the biggest city in northern Indiana. Heck, forget South Bend, I bet they’d end up coming to Notre Dame itself, perhaps for a rally (or rallies) at the Joyce Center (as President Bush did on his Social Security Unplugged tour back in 2005).

    Again I say, harumph.

  14. Derek says:

    yea,

    In NV, I believe only 1 poll, ever, showed Obama with a lead.

    As for SC and “margin of victory,” given that all polls have a non-negligible number of “undecideds,” even up to the final day. (Exit polls in SC, for instance, showed that 21% of voters decided within the last three days.) Given that, I’m not as concerned with assorted, even dramatic, upward margins of victory for the winner in the end. (Think Romney in Michigan, or Huckabee in Iowa, or McCain in New Hampshire).

  15. David K. says:

    Derek, I have a B.A. in Applied Math, i understand that margin of error can be abused as well, but that doesn’t mean we shouldn’t include it in these poll posts, it gives a more accurate snapshot on how reliable the data is.

  16. Derek says:

    That’s just silly. What makes it a “more accurate snapshot” to use a term that is essentially, and solely, abused? And what makes it “more accurate” to use a margin of error without a confidence level, which you don’t ask for? And if we’re just looking for “more accurate,” why don’t we include the type of voters sampled (likely, registered, etc.), along with the underlying methodology as well? Or the sample size?

    I guess those aren’t as “trendy” to use as a margin of error. But putting the margin of error alone doesn’t give you a “more accurate” picture absent other data; and doing so is more likely, rather than less likely, to confuse the lay reader.

  17. Angrier and Angrier says:

    Speaking of Hillary, I would send Bill one of these, but I think he has the real one at home…

    http://www.hillarynutcracker.com/

  18. Joe Loy says:

    “Again I say, harumph.”

    Brendan: Waw haw haw! :]

    “Arrgh.”

    Jay Johnson: I know. It’s aggravating. Ah feel yore Pain :}. Looky here, I’ve got an Alternative for you to try. it’s Qualitatively no different but still, being event-Appropriate commentary, it might be easier to Take ;>

  19. David K. says:

    Just because someone might misuse the information or misunderstand it is not a sufficient reason nto to include it. Sorry but the more information on the polls we get the better we are able to make use of them. I figured getting the MoE would be an easy task for Brendan, where as things like male/female or racial splits might not be as readily available, hence why i asked for the MoE. I didn’t realize asking for that info would be so horribly controversial.

  20. Anonymous says:

    I continue to be amused that Applied Math tends to be an Arts degree in the US rather than a Science degree … (grin) …

    Still, in this day and age of the Cult of Global Warming, even ‘Climate Science’ seems to have become either an Arts degree or a Theology degree …

  21. Sandy Underpants says:

    So is Obama the Anti-Christ or Nostradamus’ “Mabus”? It’s an either/or question.

  22. David K. says:

    Actually anonymous, its a B.S. i just can’t type today (notice i also used nto instead of not in an above post) Thanks for pointing it out though.

  23. Alasdair says:

    David – I was/am the one amused by the BA in Applied Math … I’m not sure why the software didn’t show it as me … the ‘signature’ use of the ellipses should have been a clue, however …

    {/me is being *very* good right now, in not referring to David K’s obvious BS … (grin)}