Tuesday’s political round-up

December 18, 2007

Today’s installment of campaign-related news items that wouldn’t generate a post of their own, but may be of interest to political observers:

* Bill Clinton went a little off message yesterday, telling supporters of his wife’s presidential campaign yesterday, “Well, the first thing she intends to do, because you can do this without passing a bill, the first thing she intends to do is to send me and former President Bush and a number of other people around the world to tell them that America is open for business and cooperation again.” George W. Bush’s father would travel with Bill Clinton to improve some of the diplomatic failures of his own son? It seems unlikely.

* John Edwards got a boost in Iowa yesterday when the state’s First Lady, Mari Culver, endorsed the former senator’s campaign. While her husband, Democratic Gov. Chet Culver, remains neutral, Mari is backing Edwards, citing her affection for Elizabeth Edwards, his anti-poverty campaign, and Edwards’ electability.

* Hillary Clinton’s campaign, in an apparent effort to help demonstrate the senator’s warm side, unveiled a new project yesterday called, “The Hillary I Know,” featuring testimonials from people “whose lives have been directly touched by Hillary.”

* A new national poll from USAT/Gallup doesn’t show too much change from November. Clinton still leads the Dems with 45%, followed by Obama at 27%, and Edwards with 15%. Giuliani leads the GOP field with 27%, followed by Huckabee with 16%, and Romney, McCain, and Thompson, each with 14%. No other candidate, in either party, had more than 3% support.

* Romney is taking on Huckabee with a new TV in Iowa, hitting the former Arkansas governor for being soft on crime: “The ad, entitled ‘Choice: Judgment,’ points out that Romney did not issue a single pardon during his time as governor, compared to the 1,033 pardons and commutations Huckabee granted during his tenure…. While Romney’s ad doesn’t specifically mention the DuMond episode, it does clearly try to make the case that Romney is strong on crime and Huckabee is not.”

* Ron Paul told the AP he now has the resources to hang on through the de facto national primary on Feb. 5, so he’ll be in the race through then no matter how poorly he fares in the early contests: “‘Nobody would understand if I faded out before Feb. 5,’ Paul said. The date makes sense as a time for Paul’s candidacy to end — as the Republican nominee will probably be decided by then, which means there won’t be any more debates or opportunities for Paul to tweak the GOP establishment from within.”

* Speaking of Paul, he held a press conference yesterday to talk about his new fundraising success. Reporters didn’t exactly show up in droves.

* Have you heard the Goonies’ theory behind Bob Kerrey’s provocative Obama comments? Clever.

* And on the heels of his Steve King endorsement, Fred Thompson is apparently going all out in Iowa (he’s now apparently even willing to campaign for votes): “Thompson began a 15-day tour of Iowa yesterday, which will go all the way through the caucuses and take a break only for Christmas. For some context: Since September, when he officially declared his candidacy, Thompson had only spent a total of 14 days in Iowa — meaning that this tour will more than double his previous time on the ground here.”

George W. Bush’s father would travel with Bill Clinton to improve some of the diplomatic failures of his own son?

And Bush Sr. will essentially go to work for the hated Hillary, right after she beats the Republican nominee?

I think maybe Bill’s gone a little soft in the head.

  • you know, i could actually see GHWB doing that. First, he and Bill really do seem to get along. Second, think of it not as doing any favor for the Clintons, but ask yourself what HW would do to try and clean up the family name – to put his theoretical success out there instead of what he has to know is the royal screw up that his namesake son has been.

    On a more serious note, I think Huck is rightly vulnerable on Dumond, but it seems in American Politics we always overreact instead of finding a middle. Do we really want the kind of country and society where “point[ing] out that Romney did not issue a single pardon” is considered a big selling point? That one has no mercy, no matter how carefully doled out?

  • Fred Thompson on campaigning: When’s the hiatus?

    Maybe Fred quit campaigning because of the Writer’s strike. No one to prepare his scripts.

    Fred thinks the Presidency would have an off-season like Westwing.

    Fred thinks he only has to handle the Order part of the job. The other cast does the Law..

    He will be glad to know that the Presidency does pay residuals.

    Congratulations to Senator Dodd. He’s got my vote now for actually doing something to resist the Republicans and their Democratic enablers.

  • And Bush Sr. will essentially go to work for the hated Hillary, right after she beats the Republican nominee? -RacerX

    Why not? They’re all one big happy family now with the single minded goal to keep fooling Americans into alternately putting one of them in the White House. Gotta make sure it’s ready for Jeb after Hillary’s had her turn.

    Someone should sail across an ocean and form a new country to avoid stuff like this.

  • While I don’t have a hell of a lot of love for Bush Sr, hasn’t he gone out on goodwill tours with Clinton over the past few years? And aren’t there a lot of Bush Sr. supporters who have gotten completely fed up with Dubya? And hasn’t it been more or less verified that Dubya’s more of a momma’s boy, & Sr. always thought Jeb would be next in line for the White House? And that while Sr wouldn’t denigrate his son in public, he’s also made it gently known that he would’ve done some things differently? And on the flip side, hasn’t Dubya publicly state he wouldn’t stoop so low as to go to his dad for advice on how to handle the job or facets of it? (after all, Daddy’s just a one-termer) And wouldn’t that stick in the craw of a father who is also a former President, someone in a unique position to understand the stresses and pitfalls of the job, and his own son won’t bother to talk to him?

    Maybe just maybe, Bill knows more of what he’s talking about than we do.

  • Bush, Clinton, Bush, Clinton. Our “democracy” RULES!!!

    Too bad neither couple had a son–we could Unite the Kingdom. Jenna’s getting married, but maybe Not-Jenna or Chelsea could be persuaded to go transgender. If it’s Chelsea, I’m sure Bill would be willing to provide the genetic material to produce an heir.

    (Yes, this is satirical. Today.)

  • Big Dog went a little off-message yesterday, telling supporters: I’M A PANDERBEAR!!!

    I’ll wager that there aren’t very many Hillary testimonials from all the guys who’ve died in Iraq so far.

    Romney is taking on Huckabee with a new TV? Would that define television as a weapon of mass destruction?

    Hey—Napoleon didn’t quit before Waterloo—why should Paul be any different? Let him keep sucking up all those fringe, neocon-in-a-very-clever-disguise profiteer dollars. Every penny he burns is a penny less that can be spent next fall by the GOPers.

    And—we sent Chris Dodd a nice Christmas present today (a check to his campaign and a card with the family Christmas photo to his home). He’s done a great deed for the Republic, that’s for sure….

  • According to the USA Today/Gallup Poll, 399 “Republicans or Republican leaners” and 513 “Democrats or Democratic leaners” accurately reflect the consensus of the country regarding the 2008 Presidential race.

    Americans are rather non-descript, after all.

    I especially appreciate that they didn’t bother linking to a detailed analysis of the poll or a detailed explanation of the polling methodology used in the article. No, the American Public should just put their full faith and confidence in the Corporate Military Industrial Media which is doing a fine job of reporting unbiased, unfiltered information.

    Scrutiny of polling methodologies is completely unnecessary in a representative democracy when you can trust five corporations to deliver the news to the voting (and non-voting) public.

  • * Hillary Clinton’s campaign, in an apparent effort to help demonstrate the senator’s warm side, unveiled a new project yesterday called, “The Hillary I Know,” featuring testimonials from people “whose lives have been directly touched by Hillary.”

    The online ad for this isn’t that cheery though. It shows pictures of four people, and three of them have blank or frowning expressions on their faces.

    More bad work by Hillary’s own people.

  • Yes, JKap, an unscupulous polling firm could manipulate the results, or an unscrupulous media could simply lie about those results. You have no evidence of it, but as suspicions go, at least it is possible (although in either case there would be some longer-term economic risk — a polling firm whose results are inaccurate wont get as much repeat business, and if the media lied and the polling firm came clean, the media would lose credibility and, in theory, readership).

    But what one cannot help but see in your frequent posts on this is an attack on sample size, as if the only way to obtain accuracy is to have the entire election. At one level, of course that is true: we do hold elections, we don’t just look at the last poll and declare a winner. Sometimes, particularly in close elections, the results defy the polls.

    At another level, however, I think your concerns are essentially a complaint about the principles of statistics – which is about like arguing against the law of gravity. Statistics are not political. There are formulae that have been tested in hundreds of contexts for hundreds of years that, whether you like the result or not, and whether it feels democratic or not, say that yes, you really can use a sample of 399 to extrapolate what millions of people think or will do (within a known probability of accuracy, of course, but any reliable poll will give you the margin of error, and some will give you the “confidence level.”) The attack on sample size just makes no sense, and undermines the credibility of your larger complaint about polls and politics.

  • Actually, Z, I think that “the problem” isn’t so much with the sample size as it is with the polling results not saying what the listener wants to hear….

  • Ok, Zeitgeist, I’m game.

    Can you indulge me with your speculation as to who the vague “Republicans or Republican leaners” and “Democrats or Democratic leaners” (who answered their landline telephones and chose to participate in the poll) are since USA Today neglected to describe their methods?

    I understand that you trust the Old-Stream Media as “credible” is this instance –that is where you and I differ. I believe that the credibility of the Old-Stream Media is the exception today and not the rule.

    One point that I am making is that I have never seen the Old-Stream Media (or the Carpetbagger for that matter) scrutinize polling methodologies. Have you?

    Are not polling methodologies important to our representative democracy?

    When I mention in these posts that cell phone-only households are a greater percentage than landline-only households today and that this fact calls into question polling methodologies that omit this growing demographic of cell-phone only households (largely under the age of 34), I rarely see an intelligent discussion about it here. Instead, some left wing authoritarians poop their pants and hurl insults.

    Yet, the “conventional wisdom” states that great faith is to be placed in these polls. One argument is that the polls often predict the winners of these contests.

    However, rarely is there “serious” discussion of the effects of polls on the outcome of these contests.

    One meme that is often perpetuated is this idea that a given candidate is not “credible” if they are behind in the polls (“second or third-tier,” etc.). To me, this amounts to a self-serving, self-fulfilling prophecy.

    It should be no wonder that 50% of Americans do not vote. After all, those 50% of Americans are effectively locked out of the polling process that largely predicates Old-Stream Media coverage. And current polling methodologies are sure to perpetuate that apathy of 50% of Americans.

  • Why would Clinton and Bush traipsing around the world do any good?

    People neglect the fact that Al Queda’s hatred of america, and their strength, grew on Bill Clinton’s watch.

    9/11 happened shortly after Bill left office.

    The animosity against the US and all the planning of terrorist acts happened when BILL was at the wheel. Oh, excuse me, when Bill and HILLARY (co-president in case you didn’t know) were at the wheel.

  • Again, JKap, you are deflecting. What I was defending was the idea that the small sample size by itself is not a problem.

    Not once did I discuss the methodological use (or lack of use) of “leaners,” nor did I discuss wired versus wireless phones.

    And not once did you try and defend your apparent attack on statistics – i.e. on the sample size itself.

    Nonetheless, let me address the wired versus wireless issue. To say that the proportion of wired only homes and wireless only homes is equal is irrelevant. That omits the truly relevant information: an overwhelming majority of homes have both. For those homes — again, a huge majority — it doesn’t matter how you reach them; wireline or wireless works the same (presumably they don’t prefer Biden when they answer their cell phone but Tancredo when they pick up the wired phone).

    It makes sense to use random selection from a relevant wireline number pool to do phone polling because your coverage is better: more wireline numbers are available than wireless, and people are more willing to answer because wireline is “calling party pays,” where wireless the receiving party burns minutes. This also captures that small percentage that is wireline only.

    As a result, the only segment “left out” is wireless only. What is relevant is not whether that segment is bigger or smaller than other segments; all that matters (a) its absolute size and (b) whether it differs significantly from the rest of the population. Its impact on polling accuracy will be a factor of (a) and (b) — the more similar it is to the general population, the larger it can be without impacting polling accuracy.

    I know you like to say the wireless only number is in double digits, but that is not the most credible research. The most credible research still has the number of adult households using solely wireless telephony in the single digits. Add to that two considerations. First, as discussed above, some number of these people will act just like the rest of the population. Whatever number that is, they do not adversely impact the accuracy of the wired phone poll. Second, a disproportionate number of “cord cutters” are very young — which also happens to be the group with a long and well-established history of voting in smaller percentages than the rest of the population. These factors mean that even the already small percentage represented by wireless-only homes can be further discounted.

    Mathematically, those remaining in this group would have to be virtually another species from another planet to be different enough to meaningfully skew the poll results beyond the margin of error that is already acknowledged.

    Wireless-only bias will, someday, be a meaningful polling phenomenon. But wireless only adult households will have to be a 20% and they will have to be distributed to include households of an age that actually votes before that is the case.

    Rather than just engaging in ad hominem suggestions that I shill for the status quo, I challenge you to actually point out where my analysis is wrong.

  • Please cite your “credible” research, Zeitgeist. Or am I to simply trust you as you would have me trust Old-Stream Media?

  • Sure. The most recent, completed FCC CMRS Competition Survey, which they are required to release each year, was the 2006 report. In that survey, staff reviews various studies that show around 8% of households and 12% of users use a wireless phone exclusively. (See para. 205)

    I have seen private consultant/financial analyst/expert witness studies that show more conservative figures of around 6%; there is a very recent survey from centers for disease control (of all things!) on wireless substitution that is touted by the CTIA (Cellular Telephone and Internet Association), whose interest is it in to have a high number, that puts the figure at 12%.

    I consider the FCC report the best information (a) because they are collecting and comparing numerous studies without an obvious preference for outcome and (b) as a cross-check, their result happens to be right at the average of the conservative private research and the liberal CTIA-perferred figures.

  • “Another survey from early 2006 found that 12 percent of cellphone users use cellphones as their only phone. 564”
    From footnote 564 of your citation:

    (citing a survey by the NPD Group). Another 42 percent said that they also had a landline phone, but they
    used their cellphones “most.” Only 43 percent said they still used their landline phones as the primary phone.

    Is this indeed the more “credible” research that you cited? If so, what makes NPD Group more credible than Medimark Research Inc. in what is more current research?

  • Here is an instance where sampling landline-only produces a clear bias:

    Adults living in poverty (21.6%) were more likely than higher income adults to be living in households with only wireless telephones.

    So, it stands to reason that by not sampling cell phones in OSM polls, adults living in poverty are less likely to be polled.

    Moreover, what I do not understand is why rational people would not favor polling of both landline telephones and cell phones when there is a clear trend underway that indicates that landline telephones will become obsolete or at least not the preferred calling device for the majority of Americans in the near future.

    Why not start now? Why not more inclusion as opposed to less?

    I have no problem with your opinions Zeitgeist, but it does not change my attitude about polling methodologies that are not reflective of reality in today’s world.

  • J, just out of curiosity—who “commissioned” the Mediamark research? I’ll grant you that you’re going to badmouth the NPD data because it was commissioned by FCC—and I think everyone here already knows how you (and your candidate) feel about various “government entities”—but you’re going to have an exceptionaly difficult time ranting about things like “Imperialist corporatism” if your cited research was bankrolled by the “corporate imperialism” of the cell-phone industry. It’ll fly about as well as the other day when you cited Mediamark via a link to a profoundly-neoconservative think tank—kind of like the glass house throwing great big rocks at itself, don’t you think?

  • Part of the problem, JKap, is how to sample wireless numbers. They are generally non-public numbers (going back to the “called party pays” issue), so you either have to (a) involuntarily violate their privacy, (b) randomly sample within an wireless NPA-NXX without knowing which numbers are active, which is horribly inefficient, or (c) have an opt-in system for polling, which being self-selected is highly biased. Wireless sampling is not particularly practical at present.

  • I guess that you’ve never heard of Halliburton, which could be characterized as an “Imperialist Corporation,” Steve.

    Or do you sympathize with Dick’s Private Empire.

  • Ok, Zeitgeist. Why am on the Carpetbagger Report and someone says “Wireless sampling is not particularly practical at present.”? I have never heard or read that alliteration prior to today.

    Why is this burgeoning lifestyle (that of cell phone-only) in America not relevant to the Old-Stream Media in its discussion of the 2008 Presidential race when so much weight is given by the five corporations who control the airwaves not scrutinized by the likes of rational people?

    The USA Today article about the USA Today/Gallup poll omits critical information, such as polling methodology. How arrogant must the Corporate Military Industrial Media be to presume that thinking people should just trust them in that respect?

    It is contemptuous of our representative democracy. It is manipulation. It is an instrument of autocracy.

  • Ok, JKap, let me try this one more time because you still are not addressing what was originally my issue and what has become two issues:

    (1) are you really arguing, contrary to established principles of statistics, that a small sample size cannot produce an accurate result? I still think the opening of your initial post attacks the polling for the small sample size itself. The “Corporate Military Industrial Media” did not invent the mathematical discipline of statistics. So throwing that perjorative around is not a valid argument.

    (2) i still contend that the omission of wireless-only households, barring more radical changes in society, does nothing to meaningfully invalidate the polls. and let me try one more example and see if I can get you to actually respond to this one rather than throwing loaded phrases around and quibbling over sources.

    Operating assumptions:
    a) Assume a town of 100 citizens who are eligible to vote.
    b) We’ll split the difference between my 8% wireless-only and the new CDC survey (which historically runs high) – it says 12%. We’ll assume our town has 10% of the population in households with no wireline phone.
    c) Assume Ron Paul polls at 10% among the traditional wireline polling sample but — and I think this massively overstates the case in your direction — but lets say Ron Paul has 50% support among the “cord cutters.”

    Your theory appears to be that the absence of wireless-only people in the sample renders polls inaccurate (presumably you think huge amounts of Ron Paul support gets missed this way).

    Assuming the sample is statistically valid, we know who 90 people support (and 10% of them support Ron Paul) because they are similarly situated to the sample: they have wireline phones.

    But what about the 10% without (in this example, 10 people)?
    Half of them – 5 – are pro-Paul, but that true under-representation is the 50% minus the 10% (1 person) represented in the wireline survey, for an undercount of 4 citizens.

    But we also can assume that the cord-cutters are predominently young and, per your quote from the CDC survey, poor — both voting blocks that underperform in turnout. This means we need to reduce the impact of the undercount; lets lower it from 4 to 3.

    As a result, our total “misrepresentation” by omitting wireless only citizens is about 3% — less than the margin of error of almost any published poll. Which is to say, the published margin of error already accounts for the wireless only sampling error.

    Hence my contention that, unless you make a really wild assumption, like 20% wireless only and 90% of those support a candidate who only gets 5% in the rest of the population, there is no distortion of the results. The mathematical reality is that the block of wireline households is so large (90%) that the entire wireless only population would have to be “outliers” to throw off the accuracy. It is just very hard for 10% of something to “undo” 90%.

    Explain where that analysis is wrong (not undemocratic, not corrupt – where is it not logically sound?)

  • So J, you’re saying that you have credible research that was commissioned by Halliburton? Is that “for the record?”

    And a lot of the folks I know who have cellphones also have their numbers placed on the DNC registry. That—coupled with having their numbers held from list-publications—pretty much prevents pollsters from contacting them in the first place. In and of itself, that fact would negate any desire to poll people via wireless; it would, likewise, suggest that those wireless individuals participating in various surveys and polls are doing so because they sought out the poll, rather than the other way around. It also raises the spectre of the poll being tainted via “stuffing,” because only those who are aware of the poll prior to its taking—and agree with predetermined, sought-after outcome of the poll itself—will participate.

    But—perhaps Mediamark can “document” their ability to prevent such an occurence?

  • Why should I categorically respond to your opinions which I disagree with, Zeitgeist?

    You ignore my argument that, according to CDC, 20%+ of adults living in poverty are more likely to have cell phones only.

    As I said, this alone indicates a clear bias in polling methodologies, since adults living in poverty are then less likely to be polled by the OSM. But please, ignore that, maybe it will go away.

    I understand that it gives some people great comfort to place all of their faith in this system which brought us the Bush Disaster and which has aided and abetted fascism being enthroned in America. I am just not one of those people.

  • JKap, you are showing yourself to be as bad as any other fundamentalist in simply denying truths in favor of beliefs. Statistical concepts are not “my opinion” whether you agree or disagree. The real problem isn’t that you wont respond to my arguments, the problem is that you can’t respond to mathematics. It is what it is. It just happens to undermine your position. A good example: I never denied the CDCs finding that those in poverty are more likely to be wireless only (heck there are some programs to give them to the homeless). But you haven’t addressed an objectively verifiable truth: those in poverty vote at rates far below the general population. In other words, the more skewed non-sampling is towards the impoverished, the less inaccuracy it introduces in polls.

    Ultimately, this discussion is pointless because, as always, you resort to name calling and suggesting your counterpart is “aiding and abetting fascism.” This is not a matter of what I do or don’t take comfort in or aid and abet, and mathematical pricinples have no more values — they are no more pro-or-anti democracy — than gravity, thermodynamics or evolution (perhaps you disagree with those opinions as well? Of course you can no more opt out of statistics than you can gravity.)

    If you can’t actually respond to the points and have only recourse to dogma, you;ve already lost the debate.

    • Comments are closed.