Hugh Hewitt Takes on the Pollsters

Radio host and conservative columnist Hugh Hewitt has done fantastic work this season interviewing pollsters and asking the tough questions looking for answers how allegedly reliable polls have such unrealistic internal make-ups.   In his Townhall column he lays out numerous arguments to challenge the data including two objective data points that make the Ohio sampling more obviously incorrect:

There are plenty of data points to encourage Republicans, and these are genuine data points as opposed to the junk food offered up by Quinnipiac and Marist, which derived their predictions from samples that included enormous Democratic voter margins in key states, pro-Democratic turnout margins that were even greater than those achieved in Obama’s blowout year of 2008..

Two data points that warm GOP hearts and undermine the junk polls: (1) Absentee requests in Ohio by Democrats are trailing their 2008 totals –often by a lot in key Democratic counties like Cuyahoga County; and (2) overall voter registration for Democrats in the Buckeye State is down dramatically from 2008.

These two bits of info undermine the credibility of the Obama booster polls, as did the interviews I conducted with key leadership from both polls and with other informed observers.

In addition to doing the media’s job actually finding the data to challenge the assumptions, Hewitt has used his radio shows to go right to the sources on polling and how we should interpret the data.  After numerous interviews Hewitt provides five major takeaways:

  • The pro-Obama pollsters don’t have answers as to why their skewed samples are trustworthy beyond the fact that they think their approach to randomness is a guarantee of fairness, and they seem to resent greatly that the questions are even asked. Like [Convicted fraud Bernie] Madoff would have resented questions about his stunning rate of return.
  • Barone notes that percentage turnout by party in a presidential year hasn’t been much greater for the president’s party than it was in the preceding off-year, which makes samples outstripping even the 2008 model of Democratic participation “inherently suspicious.”.
  • Cost notes that Romney is winning the independent vote in every poll, which also makes big Obama leads suspect.
  • And my conversation with Mr. Shepard, whose employer National Journal has a reputation for the best non-partisan work inside the Beltway, didn’t find any academic, disinterested support for the proposition that party identification cannot be weighted because of the inherent instability of the marker.
  • The biggest unanswered question of all: If party ID is so subject to change that it should not be weighted according to an estimate of turnout, why ask about it at all? And if it is for the purpose of detecting big moves, as Mr. Shepard argued, why not report that “big move” in the stories that depend upon the polling?

There are a number of reasons polling organizations could offer for their curious sampling but they offer no defense of these results other than it is consistent with the prior election which fails to take into account that admitted notion that party identification changes every election and the current samplings do not reflect the reality of today’s electorate.

44 Comments

  1. Posted September 29, 2012 at 1:56 pm | Permalink | Reply

    And the conspiracy continues! Gallup has Obama up by 6 for the last 4 days. Remember when Gallup was what the right wingers claimed as the best source?

    And Rasmussen has Obama up by 2! Which given Rasmussen’s pro R bias means Obama is really up by 6.

    • Waingro
      Posted September 29, 2012 at 2:01 pm | Permalink | Reply

      Barry also lost two in approval and gained two in disapproval in Gallup. He will be coming back to Earth there soon, especially when they change to “likely voter” next week.

    • No Tribe
      Posted September 29, 2012 at 2:10 pm | Permalink | Reply

      Peter, what’s the comeback for the lack of turnout among Democrats in Ohio thus far?

    • Posted September 29, 2012 at 2:22 pm | Permalink | Reply

      Oh noes! Yet more of the “if I look at this little variable here I can safely ignore that all the polls at against my candidate” routine.

      • Posted September 29, 2012 at 2:25 pm | Permalink

        Peter if you’re going to only offer nonsensical comments to stir up trouble I’m going to send you back to the Kos Kids. I’ve left you here for awhile but you offer fewer and fewer worthwhile posts and more and more trash littering up the comments section. That won’t last for much longer.

      • Posted September 29, 2012 at 2:37 pm | Permalink

        It’s your blog, do as you wish. But if you are going to open your posts for comments you are going to have to deal with people that disagree with your analysis. This whole idea that polls are wrong comes up every election cycle, every single one. Each time the losing side finds some variable or way of looking at the numbers that proves, just proves that their guys is really winning. Happened in 2004 and 2010 to Democrats.

        But in the end it’s always a fools errand. Polls are by no means perfect but the average of polls tuns out to have great predictive power. The same will happen in November.

        This whole debate about D and R samples is based on confusing party ID and party registration, which are two different things. It ignores the ample historical evidence that people will report different party ID across time and across pollsters. So you can’t look at a party ID of the NYT poll, for example, and compare it with the party ID of the past election, two different things.

        The polls are clear, Obama is ahead. There is a normal distribution of polls so you will find some with Obama up by 8 others up by 2 and an average of around 4.

      • jeff
        Posted September 29, 2012 at 3:10 pm | Permalink

        Face it. BO has peaked. Once gallups outlier is out and switches to LV the numbers will start to agree with rasmussen. The other msm polls will start to show the true state of the race in a couple of weeks in order to save face and maintain a semblance of credibility. BO is in trouble. Hes polling less than 50 in PA and OH and in fact there are several credible polls.
        that has Romney within 2 In PA. AND
        the absentees are favoring Romney in OH by a pretty good margin. Tell me. Do you really believe that BO is leading Romney by 9 in a traditionally red state like Fl like some of those msm polls show. I live in Fl and theres no way in hell that BO is going to win Fl.

        the.absentees are favoring Republicans in a substantial way. I believe the folks are starting to panic

      • No Tribe
        Posted September 29, 2012 at 5:39 pm | Permalink

        That’s a pretty weak comeback. I’d like to have you here contributing, but there’s no reason why you can’t leave your baggage at the door.

        I don’t think anyone here has posted that there isn’t a difference between party ID and party registration, or that party ID fluctuates in the course of an election cycle. That’s really not the argument at all. This has been an ongoing issue on both sides for the past 3 cycles and this election. It’s why many pollsters missed ’10 and ’06 wildly a month or two out from the election. The point in to do polling not as a snapshot but as a prediction of the election. And if you are going to do RDI as a snapshot, make that the narrative, and not pretend it is a predictive polling model.

    • WillBest
      Posted September 29, 2012 at 2:27 pm | Permalink | Reply

      3 days ago Gallup had an outlier poll day (it happens about twice a month statistically speaking) and rolled something like 60% Obama, 40% Romney. That result is still in their poll. The key thing is that gallup has stayed at 6% meaning that the new data entering is nearly identical to the data leaving it. As such Gallup has been polling about 2% for Obama the last 2 days. Relax on Gallup.

      • WillBest
        Posted September 29, 2012 at 2:29 pm | Permalink

        To clarify. Gallup uses a 7 day rolling average. So todays data replaces last saturday’s data. If the poll doesn’t move then the data it collects today is essentially the same as last weeks.

        Now Gallup is supposed to move to LV this week, but if it didn’t you would see a day where Obama’s lead dropped 3 points.

    • shane
      Posted September 29, 2012 at 2:30 pm | Permalink | Reply

      do you actually look at Ras numbers? He polled most of the summer D+3. Based on his previous three week weight on stated party id I.think his current break based on what people are actually telling pollsters is R+1.5.

      Funny how libs rail on Ras as bias but refuse to say Amy other poll has a bias despite Ras bein undisputedly (media, dem pundits, etc) one of the top 3 most accurate since about 04.

      As for Gallup it is used as reference as the longest running poll but in the last decade they have not been very accurate in the final tallies.

      • Posted September 29, 2012 at 2:41 pm | Permalink

        See, this is the problem. People create an alternative reality, in this case that Ras is one of the most accurate pollsters despite the evidence that he has a very strong R bias. In 2010 Ras predicted the Congressional vote would be R+12, and it was really R+6.8, a five point R bias. Yet you see in conservative websites repeated over and over that Ras was the most accurate in 2010. That’s just making stuff up.

      • shane
        Posted September 29, 2012 at 2:53 pm | Permalink

        and all libs can point to is 2010 and claim some massive bias. Most polls expected Repubs to take more than they did.

        But look at 04-06-08 and he and Pew (ABC news?) Have been the top three under a .5 to 1 point.

        Libs don’t want to admit to his consistency. Gallup is a highly respected pollster comp yet they have been very wildly wrong for many years…right winner way wrong numbers. Yet they are respected. ras has been right winner and damn near dead on numbers too but can’t get recognition because libs would have to admit a conservative is right.

        Every Lib wants to jump on the bare silver wagon because he works for Obama. Bottom line is a statistician can make any numbers say what they want. Silver is if the money ball persuasion who thinks stats can predict the future it is modern day palm reading. If that stuff worked then the As would have a shelf full of trophies. Stats can produce interesting items but they miss the vital human element..you can’t predict human action by looming at a number.

        The pollsters and media are majority left of center, they DESPERATELY want to believe Obama and his 08 turnout is the new norm. They cannot deal with the thought their creation has been the failure the numbers prove.

      • Posted September 29, 2012 at 3:13 pm | Permalink

        No Shane, Ras was not the best in 2008. He was Ok but so were several others, including Pew. Again, your side has created this alternative reality where Ras is better than other pollsters when in fact he has been wrong a lot.

      • jeff
        Posted September 29, 2012 at 3:35 pm | Permalink

        Gallup in fact was off by 6 in 2008 giving BO a 13 point advantage overMcain as their last tally. Ras almost hit the nail on the head and was off by 1

      • Posted September 29, 2012 at 7:01 pm | Permalink

        peter
        notice i said one of the top 3. Yes, as I stated, Pew and ABC news was it had it closer they were at like .3 off and he was .5 off.

        So no it is not our side that has made this up….look up the numbers. He has been, as I have stated, one of the top 3 most accurate each election. Pew has done well and then there seems to be a wild card. Gallup actually has pretty crappy record.

    • John Smith
      Posted September 29, 2012 at 2:58 pm | Permalink | Reply

      Here’s a nice dose of reality for you

      http://scottthong.wordpress.com/2008/10/22/ann-coulter-researches-the-pro-democrat-poll-bias-through-8-presidential-elections/

      • WillBest
        Posted September 29, 2012 at 3:54 pm | Permalink

        You can’t cite Ann Coulter to a liberal. They foam at the mouth and go full on stupid.

    • jpcapra1
      Posted September 29, 2012 at 4:44 pm | Permalink | Reply

      Rasmussen was the most accurate in the national polling in 08 and 04. His voter ID model was also extremely accurate, which is why his overall numbers were accurate. I would say obama +2 is about right. You may want to consider he won by seven last time, and has essentially conceeded three states he fought for last time(NC, IN, and MO). And he is having to defend two (MI and WI) that were uncontested last time. Before you say MI is not on his radar, you may want to know obama has more than doubled his offices there in the past month, hardly an indication he is not concerned at all about that state.

      • Posted September 29, 2012 at 5:00 pm | Permalink

        No, he wasn’t the most accurate in 2008. This is part of the right wing mythology. Just check the RCP numbers and you’ll see that several other pollsters had similar or even better numbers in 2008.

      • Posted September 29, 2012 at 5:01 pm | Permalink

        And no, MI is not a swing state at all.

      • WillBest
        Posted September 29, 2012 at 5:24 pm | Permalink

        If he isn’t worried about it, why would he up his ground game?

      • jpcapra1
        Posted September 29, 2012 at 5:40 pm | Permalink

        He was off by one point in 08 and 0.5 in 04. Who was better than that? And you haven’t answered for why he has conceeded states he won like Indiana and North Carolina? And a state he only lost by a few thousand like Missouri, which is ALWAYS a swing state? Every candidate going back decades that has won re-election has increased the number of states they won:

        Bush picked up Iowa and New Mexico, lost New Hampshire
        Clinton picked up Arizona, Florida, lost Colorado
        Reagan picked up every state with the exception of Minnesota
        Nixon picked up every state with the exception of Massachussetts
        Johnson won every state except AZ, MS, and GA

        Not saying it can’t happen, but I will definitely say obama will not win by anywhere near the margin that he won by last time. What irks us is that if this was a Republican incumbant, the media would be 24/7 talking about how he has backtracked, given up on states he won, not competing in a perennial swing state like Missouri that goes with the winner in virtually every election, etc. They act like obama is up 15 points, when he is up 2, at most 4. And WillBest makes a good point, why would he double his offices in Michigan if it wasn’t on his radar?

  2. WillBest
    Posted September 29, 2012 at 2:24 pm | Permalink | Reply

    Its not party ID. Its race and sex. If you assume 51% female, 9% Hispanic, 14% Black, 5% other, 72% white, you will have a much different outcome than if you assume 49% female, 8% Hispanic, 11% Black, 4% other, and 77% white. You are literally talking about 4.25-5.25% swing from Obama to Romney from A to B.

    Whats the difference between Rass & Gallup (before it had its massive outlier day a couple days back) and The media polls? They are averaging 6% Obama lead to Rass & Gallups 1-2% Obama lead

    • shane
      Posted September 29, 2012 at 2:42 pm | Permalink | Reply

      no one on the Romney side with a brain is saying he is up. The point is this will not be a landslide as being portrayed and Romney can close.

      Long story short Romney/Ryan are nearly 180 degrees different than McCain/Palin. In 08 they had ZERO chance of winning Zero.

      ROMNEY may lose and Obama will have a few more entries in the history books. But our point is there is no way this is a landslide nor even as spread as 08. Really can’t be considering Obama isn’t picking off and 08 red states while it is HIGHLY probable Romney will at least get Indiana, North Carolina and a great chance at Iowa, Virginia and Wisc.

      Again Romney may lose but it will be 50-49 or something razor thin. Senate will stick, GOP will maintain or gain in house.

      • Posted September 29, 2012 at 3:13 pm | Permalink

        I think most conservatives like me are playing it conservative (pun intended) right now and saying it’s most likely Obama +1,2 max, or tied. Personally I think right now it’s really Romney up 1 or 2 and will end up Romney by 5 or 6 in the end when undecideds/independents break plus bigger turnout margins for GOP. The left is going to be in shock if they really think they are going to get above average turnout this election. No one will be surprised if there is above average turnout on the republican side.

      • WillBest
        Posted September 29, 2012 at 3:25 pm | Permalink

        I didn’t say he was up. I was just pointing out that nobody knows what the electorate is going to look like so using a minority model running from 22% to 28% is valid enough, but because whites away from Obama in the last 4 years the % of the voter base that is minority, specifically black, is highly determinate of Obama’s advantage. Whether it is Obama +1% or Obama +6%.

    • Posted September 29, 2012 at 2:43 pm | Permalink | Reply

      Gallup has Obama up by 6, that’s the same as Bloomberg, one more than Fox, and one less than the National Journal poll.

      • shane
        Posted September 29, 2012 at 2:57 pm | Permalink

        and they are still rolling registered which is a known dem pad. They others likely screen are still demographically mirrors to 08…not going to happen.

        Obama us leading no doubt but more closer to Ras’ 2-2.5 not 5+.

        Plus look at the Gallup tracking all year it us an EKG for Obama…up down up down. He swung up 5 to down 2 in a seven.day stretch not long ago. That is the maddening factor of tracking polls they show the real.time indecisivness of Americans

      • WillBest
        Posted September 29, 2012 at 3:16 pm | Permalink

        I explained why Gallup has a 6 point lead. Feel free to disprove the evidence.

      • Posted September 29, 2012 at 3:17 pm | Permalink

        Well, at least you recognize that Obama is winning.

  3. Posted September 29, 2012 at 2:43 pm | Permalink | Reply

    I think Gallup is just padding their stats for when the LV model kicks in, they don’t want to show Romney take a lead. Something about a lawsuit…….

    • jeff
      Posted September 29, 2012 at 3:14 pm | Permalink | Reply

      The true model will probably be 2010 Which the msm polls dont want to use because it would show Romney ahead or at least tied.

      • Mike DeVault
        Posted September 29, 2012 at 5:15 pm | Permalink

        I agree. Much has changed from 2008. Given the precipitous drop in early voting and absentee ballot requests in places like NC, OH, and IA,, you can “see” the changed electorate. It may not be 2010 but it will be closer to 2010 than 2008. I am very optimistic.

    • WillBest
      Posted September 29, 2012 at 3:29 pm | Permalink | Reply

      No. Gallup had an outlier polling day. They didn’t set out to have one. It just happens sometimes.

      • Posted September 29, 2012 at 4:08 pm | Permalink

        Sure it does and I know they had one and I thought I read they had two somewhere else, I just don’t believe in coincidences.

  4. jpcapra1
    Posted September 29, 2012 at 4:55 pm | Permalink | Reply

    obama may win, but not by 6. More likely 3, maybe 4. And he will lose FL, that I can promise.

  5. jpcapra1
    Posted September 29, 2012 at 5:44 pm | Permalink | Reply

    And Michigan may not be a “swing state” like Florida and Virginia, but my overall point is he is having to allocate resources in a tight race where his opponent has an equal amount of money, something he did not have to do last time and he would not be doing if he wasn’t a little concerned. He had ten field offices prior to the convention, he has twenty now, if not more.

  6. jpcapra1
    Posted September 29, 2012 at 5:48 pm | Permalink | Reply

    I will give you Rasmussen is not as accurate with state polling as he is nationally. My guess is his party weighting is not as effective on smaller levels.

  7. jpcapra1
    Posted September 29, 2012 at 5:50 pm | Permalink | Reply

    FYI, his party weighting in ’04 was +1D; it ended up even. It was +8D in ’08, it ended up +7D. My guess is he will have it even or +1D in November this year.

    • Harold Smith
      Posted September 29, 2012 at 6:48 pm | Permalink | Reply

      I went back and looked at Larry Sabatos site for 2010 and he changed his predictions often in October and always to the republicans side. I know it was a wave election but it seems to me they always give the democrats the benefit of the doubt.

      • No Tribe
        Posted September 29, 2012 at 8:53 pm | Permalink

        No, that’s not true. They just favor the last party to win. In ’06, it was just the opposite, that the predictions were often late, and changing to the Democratic side. That’s the main problem here– that we have a press that hurries to favor of those in power to have access or some other favor or just a standard laggard mentality that isn’t that capable of insight.

  8. truthseeker
    Posted October 3, 2012 at 6:14 pm | Permalink | Reply

    It’s the old put your fingers in your ears, jump and down and go “na,na,na” then you don’t hear the truth. Chris Christie said something about this poll whining thing – once you do that, shows you are losing. Plain and simple.

  9. Posted October 29, 2012 at 6:45 am | Permalink | Reply

    My ever expanding list of the growing Romneylanche – from polling data to flaccid Dem enthusiasm to the economy to dead consumer confidence.

    http://scottthong.wordpress.com/2012/10/23/why-obama-will-lose-november-2012/

One Trackback

  1. By Head Fake for the Ages | Tribble News on October 3, 2012 at 2:58 pm

    […] pronounce that Barack Obama has it in the bag and that Mitt Romney is done for. Every radio host (Hugh Hewitt included) has been out dissecting the bogus polling data that oversamples Democrats anywhere from 7 to 11 […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: