Why does the "unreliable" attack deserve an eyeroll???......

I don't trust this poll at face value myself.  And I'm rooting for Halter!  (I don't really have anything against Lincoln, I just see her as having NO chance in November, while Halter has at least an outside chance.)

This Talk Business-commissioned poll has really weird crosstabs that I can't decipher, with columns "D1" through "D4" without any guide as to what those categories mean.

The narrative explanation says there's no weighting for anything, and that interested parties can do their own weighting, but there are no decipherable crosstabs provided to do that!

On balance I trust Lincoln's internal poll more than Halter's.  Whatever support she's lost, she still gets majority job approval from base Democrats in Arkansas, who have been voting for her for a long time while recognizing they're in a conservative state in a tough year.  Halter is going to have a very tough time actually winning the primary, he needs to somehow establish himself as clearly more viable than Lincoln in November to get over the hump with primary voters.  I already think that's true, but a lot of people don't, at least not yet.

43, male, Indian-American, Democrat, VA-10

---

by: DCCyclone @ Thu Apr 15, 2010 at 01:05:50 AM UTC

I think the eyeroll

was for the oft-repeated and verifiably false premise that automated polls are inherently unreliable.

34, WM, Democrat, FL-11

---

by: spiderdem @ Thu Apr 15, 2010 at 01:09:09 AM UTC

[ Parent ]

Sorry

I should reword that sentence to make it more explicit that this is indeed the case.

---

by: DavidNYC @ Thu Apr 15, 2010 at 02:35:05 AM UTC

[ Parent ]

D1=District 1? D2=District 2?

It would make sense seeing as Arkansas has 4 districts.  

---

by: JeremiahTheMessiah @ Thu Apr 15, 2010 at 01:13:13 AM UTC

[ Parent ]

Thanks,I bet you're right! BUT...

...why didn't they come out and say that, and help the reader?

And no other crosstabs!  Race, gender, ideological self-identification, and age would be nice!

43, male, Indian-American, Democrat, VA-10

---

by: DCCyclone @ Thu Apr 15, 2010 at 01:21:00 AM UTC

[ Parent ]

This is the problem

"I don't fault Roby Brock for conducting an automated poll because they are very inexpensive,"  Lincoln campaign manager Steve Patterson said.  "However, they are also very unreliable. [...]"

Team Lincoln is insulting the method, not necessarily (or only) the pollster.

---

by: James L. @ Thu Apr 15, 2010 at 01:51:42 AM UTC

[ Parent ]

OK, now I get you. But here's a valid counterpoint......

Chuck Todd, himself skeptical of robopolls, pointed out awhile back that if robocall polling is just as good or better than live callers, then candidates and campaigns would use them.  But they don't...even though they're cheaper!  OK, well, PPP is a partisan outfit that presumably does political polling for paid clients......but do we know that their clients include people running for office (as opposed to, say, trade organizations or other non-candidate and non-party organizations), and even if so, might they be pretty low-level candidates with a tight budget?  I really don't know.

Now, I understand it's pretty solid at this point that, Rasmussen's constant shark-jumping notwithstanding, robopolling can be just as accurate as live calling, when comparing the public pollsters who poll frequently enough to compare them.

And, maybe, just maybe, candidates and campaigns are mistaken to remain leery of robopolling.

But I tend to think the people who have to care about accuracy and cost-effectiveness in polling far more than anyone else, the candidates and parties, are not that far behind the curve.  They must know something about reliability that might not show up when comparing, say, PPP with Quinnipiac.

I do think PPP actually has exposed, unwittingly, some flaws in robopolling this cycle.  There's been this notion by robopoll defenders that, usually brought up regarding willingness to vote for a black candidate, "people are more likely to lie to live callers" than to automated questions.  But PPP has floated some outlandish questions in its polls the past year with results that suggest the opposite:  people are willing to say just about anything to an automated recording.  I don't believe for one second that up to one-quarter or one-third of Republicans believe Obama is the anti-Christ, or that they believe some of the other crazy things PPP's polling results suggest they believe.  I think those results show that people who dislike Obama are willing to say just about anything disparaging when "talking" to a computer-generated voice that asks them only to "press 1 for yes."  It's just easier to be impulsive when interacting with a computer instead of a human being.  I bet these same respondents would respond very differently if asked the same questions by live callers......and contrary to what some robopoll advocates might say, I think what they'd say to live callers would be more truthful than what they tell PPP or Rasmussen on some of those questions, because responses to live callers are more thoughtful and less impulsive.

In the meta-picture, I think robopolls are at least as accurate as live call poll results on questions respondents already consider on their own.  People do, indeed, think on their own about their opinion of the President, who to vote for in a fast-approaching election, and maybe a few other things.  But they don't spend time contemplating hypothetical Presidential contests that are years away, or whether the President is the anti-Christ or a socialist.  Every poll is necessarily guilty of catching every respondent off guard with the phone call in the first place, always interrupting the respondent in the middle of something else.  When you add to that questions that themselves catch the respondent off-guard because they raise an issue that the respondent has never considered, and especially if the question itself is intrinsically biased or has biased wording or is just downright wacky, then you get very impulsive and unreliable responses.

43, male, Indian-American, Democrat, VA-10

---

by: DCCyclone @ Thu Apr 15, 2010 at 14:08:49 PM UTC

[ Parent ]

I think once Bill Clinton stumps for Lincoln, the Halter surge will stall

For daily political commentary, visit me at http://polibeast.blogspot.com/ and http://twitter.com/polibeast

---

by: andyroo312 @ Thu Apr 15, 2010 at 03:15:18 AM UTC

If Bill does that

He better plan on stumping for her A LOT going into the general election, because she's gonna need it.

The thought of the former president killing our best chance at a seat retention is pretty amazing.

23, Male, Democrat, OH-13

---

by: StephenCLE @ Thu Apr 15, 2010 at 03:23:50 AM UTC

[ Parent ]

I don't know

I'm pretty sure that the seat is gone and isn't coming back, to intentionally misquote Rick Pitino, "William Fulbright is not walking through that door, fans. Dale Bumpers isn't walking through that door, and John Mclellan isn't walking through that door." The days of Democratic senators serving out long careers in Arkansas is fading away.
Also I haven't seen any polling data to suggest that Halter has a heck of a lot better chance versus any of the Republicans in the general. If it is out there, and not done by Kos or an internal, I would like to see it.  

---

by: Jfindl1 @ Thu Apr 15, 2010 at 04:09:51 AM UTC

[ Parent ]

Out of curiosity

Why do you view a Research 2000 poll for DailyKos with a jaundiced eye?

---

by: James L. @ Thu Apr 15, 2010 at 04:42:20 AM UTC

[ Parent ]

I think R2K has a very slight Democratic "house effect"......

It's only very slight, not enough to discredit the pollster, but Nate Silver measures R2K as having it, and my own personal sense is the same just from being an obsessive consumer of polls.

That said, DailyKos-commissioned R2K polls are no more biased than R2K polls commissioned by other clients, so it's nonsense to slam "the source."  The source, where it's a client and not the pollster, matters only to the extent one finds problems in the questions themselves, since that's the only part of the methodology the client controls.

43, male, Indian-American, Democrat, VA-10

---

by: DCCyclone @ Thu Apr 15, 2010 at 14:23:16 PM UTC

[ Parent ]

for most races I wouldn't be suspicious

Sure they probably have a Dem House effect but overall no big deal. However, in the Dem primary in Arkansas Kos is backing Halter big time. So I would take data from them showing a tight race there with a grain of salt.  

---

by: Jfindl1 @ Thu Apr 15, 2010 at 16:09:55 PM UTC

[ Parent ]

I think really Jfindl1 and StephenCLE are BOTH right......

I think it's pretty clear at this point that Halter has a better chance in November than Lincoln.  Lincoln is in the same position as Dodd, she's unpopular and a majority of voters have made up their minds that they won't vote for her against either Boozman or Baker (and the GOP nominee will be one of them, not any of the others).

The general trial heats with Halter clearly show substantially more undecideds, and the Republicans substantially below 50% against him, meaning Halter just has more room to grow than Lincoln.  There is nothing Lincoln can do to bring down Baker's or Boozman's image that Halter can't do if his campaign is merely competent, so one can't say "but Lincoln could tear down her opposition and have a chance" as something that somehow distinguishes her from Halter.  The only electability arguments in Lincoln's favor would have to be that she'll run a better quality campaign than Halter which is not an argument I've seen even attempted, or that Halter ends up performing at least as badly for being "too liberal" which is purely speculative because how the voters frame the choice between the candidates, ideologically or some other way, is really up for grabs.  One could have speculatively argued "money" in Lincoln's favor before Halter released his numbers, but his numbers were too strong to make that argument now.

But the flip side is that, yes, this seat and Arkansas more broadly probably are GONE from federal Democrats' reach due to ideological reasons that pushed the rest of the South to the GOP the last century, with the same effect delayed in Arkansas only because of Bill Clinton's Presidency.  That metaforce likely ends up giving Republicans this seat in November no matter what, so it's no slam dunk that Halter is substantially better than Lincoln.

43, male, Indian-American, Democrat, VA-10

---

by: DCCyclone @ Thu Apr 15, 2010 at 14:36:04 PM UTC

[ Parent ]

Prelininary comclusions

(from not especially detailed polls)

1. Democratic primary electorate in Arlaksas is substantially to the right of those in most states (including such Southern states as Virginia, North Carolina, Florida or Texas)

2. Lincoln is still favored in primary (may be not by much, but still)

3. Halter popularity among netroots gave him money, but didn't substantially strengthened his position as far as "electability" is considered - he still loses to main Republican candidates and, in most cases - by only few points less then Lincoln

4. The race is now Republican's to lose in any case, but, given the tendency of Arkansas republicans to put up "strange" or extreme candidates - that's quite posiible.

---

by: smoltchanov @ Thu Apr 15, 2010 at 05:39:53 AM UTC

Good summary, but I don't think your qualifier on no. 4 applies here......

I think Boozman or Baker, both mainstream enough for this state, is the Republican nominee, no oxygen for the others, who truly are loonies as you describe.

So it will be a very tough fight, yes.

43, male, Indian-American, Democrat, VA-10

---

by: DCCyclone @ Thu Apr 15, 2010 at 14:38:46 PM UTC

[ Parent ]

Probably yes

But with "teabagger-style" candidates appearing  from almost nowhere (like Hoffman in NY-23 last year) in many states and districts (and being competitive in many cases) i still can imagine some unwxpected results on Republican side (not Baker and not Boozman i.e.)

---

by: smoltchanov @ Thu Apr 15, 2010 at 14:44:30 PM UTC

[ Parent ]

How do AR primaries work?

Are there runoffs if no one clears 40% or what? Because if it's a free-for-all, I could see Baker & Boozman splitting "establishment" votes while one of the crazies slips past them.  

Kansan by birth, Californian by choice, and Gay by the grace of God.

---

by: arealmlc @ Thu Apr 15, 2010 at 15:39:35 PM UTC

[ Parent ]

Runoff if no one gets majority

if i remember correctly...

---

by: smoltchanov @ Thu Apr 15, 2010 at 15:48:11 PM UTC

[ Parent ]