27 thoughts on “An Idle Thought On The Polling”

  1. Here in CO, we get lots of live and robo-call pollsters. I hang up on all of them. (My wife is willing to answer the polls).

    I suspect curmudgeons like me break more for one party than another (which would potentially skew the result), but it’s hard to know. We’ll find out soon enough!

  2. Another quick thought is that teens and early twenty-somethings used to be almost impossible to get on the phone because they were always out running around. Now there’s hardly a one of them that won’t answer the phone on the first ring, and all of them are probably willing, even thrilled, to answer polling questions.

  3. I live in PA, and several months ago I got so sick and tired of telemarketing calls that I started leaving my answering machine on all the time. Now I get 5-6 calls per day, and most of them are hangups. I’m sure some of them are pollsters.

  6. Would there be so much interest here in the failings of pollsters if Romney was up by 5%?

    There are lots of practical obstacles to taking accurate polls, but nonetheless the polls are usually right.

    2. Actually, the fact that Obama isn’t polling in single digits is quite disheartening. It doesn’t matter who he’s running against. This is an indication of severe problems with this country.

    3. Rasmussen is the only Presidential Poll that can accurately claim to be usually right.

      Anybody here remember how badly Scott Walker lost? The polling wasn’t ‘usually right’ there!

        1. So at the end of October, we can expect decent RCP averages. But this isn’t the end of October.

          1. RCP averages are computed the same way in late October as they are in early October. The thing that (sometimes) changes in between is the voters, not the pollsters.

          2. FWIW, RCP had Walker up by 6 three weeks before the recall vote. The voters didn’t move very much at the end.

    4. Jim writes:

      “Would there be so much interest here in the failings of pollsters if Romney was up by 5%?”

      Would there be much interest in the media in getting Holder’s resignation over Fast and Furious if he were a Republican?

      Would there be much interest in the media over an unemployment level of over 8% in the last 40+ months…………

      Or a $16 trillion dollar debt….

      ….or more debt racked up in 3.5 years than in the entire history of the nation…..

      …or the incredible arrogance of “I won”…..

      Or millions wasted on cozy chummy Solyndra………

      if Obama was a Republican?

      Would there be much media interest in Sebelius’ lawmaking by decree if she were a Republican?

      Would there be much media interest in Obama’s comment to Medvedev that he’ll have more flexibility after the election of Obama were Republican?

      Would there be much media interest in Holder’s not prosecuting the Black Panther’s intimidating people at a voting site….

      ..or Holder calling the nation’s people cowards because of an alleged disinterest in talking about race?

      …..if Holder was a Republican?

      Would there be more media interest in Lie-awatha’s alleged Cherokee heritage and the possible use of that as an entre into Harvard if she were a Republican?

      I could go on and on and on but I have better things to do that recount 4 years of graft, lies, un-constitutional usurpations of power that the Democrats have indulged in and which the media totally ignore because they have thick kneepads.

  7. “There are lots of practical obstacles to taking accurate polls, but nonetheless the polls are usually right.”

    Jim , are you contradicting yourself?!

    1. No. There are obstacles, and pollsters try to correct for them. Evidently the remaining errors mostly cancel themselves out, at least when you take an average of polls. The end result is that polling is usually accurate.

      1. Mabey at the end Jim but polls tend to be noisy this far out. I also suspect Pollsters put more money into getting the last few right as thoses are the ones that will be remembered.

        Why spend big bucks on polls six weeks out when the cheaper ones sell to the media for the less money and more profit?

        1. Polls are noisy at the end too, but an average of polls performs very well.

          Reputable pollsters use the same procedures whether the election is months away or days away. Polling is just as accurate now as it will be the day before the election.

  8. I stopped answering our “landline” phone years ago because despite the Do Not Call registry the overwhelming majority of calls were from telemarketers.

    Now the landline almost never rings but even when it does I only answer it if I know who it is. Anyone who really needs to talk to me either has my cell phone number, or can damn well leave a message.

    Pollsters, like telemarketers, don’t ever seem to leave messages.

  9. One thing about polling accuracy; there is one, and only one, way to truly measure it, and that’s an election.

    Therefor, while we can look back at the past and say pollster X was more accurate than pollster Y, what we are really doing is to look and see how accurate they were in the final few days. That has little to no bearing as to how accurate they were weeks or months out.

    1. Of course it has bearing. The polls taken November 5 will use the exact same procedures as those being taken now. There’s no reason to think that polls a month before an election are any less accurate than the same polls the day before the vote.

      Polls are taking a snapshot, measuring what people think at one point in time. People can change their minds. So today’s polls don’t tell you how the election will turn out, but they tell you where it stands, and how much public opinion would have to change for the outcome to change. A series of polls taken over time tells you which way public opinion is moving.

      This year’s polls have shown Obama with a small, consistent lead, growing a bit larger after the DNC and the 47% video. They show movement towards Democratic Senate candidates. That’s the state of the election right now.

  10. Now they are reporting a response rate of only 9%. I wonder if the wheels are coming off of traditional polling the way they did the once-infallable exit polls?

    Mabey pollsters are going to have to go back to knocking door-to-door like in the olden days. I have found some of the things that purport to be phone polls are just lead-in’s to sales calls. That makes me want to hang up immediately when I get a call that has a delayed greeting from an unknown number.

    1. Neal Stevenson, writing under his pseudonum of Steven Bury, wrote a novel called Interface (a href=http://www.amazon.com/Interface-Stephen-Bury/dp/0553572407>link ) in which a company gave several hundred carefully chosen people a simple electronic device so their moods could be tracked for ad campaigns, a bit like a higher-tech version of being a Neilsen Family. Since the same people were used for all the data, there wasn’t a sampling problem, and the individuals could be weighted to correct for detected differences from the overall population.

      Then they wired the output of the polling data into the brain of a Presidential candidate, and sci-fi political fun ensues.

  11. I’ll never forget Election night 2004. Was watching Fox and the panel had these hangdog expressions as it seemed as if Bush would lose….

    Susan Estrich was croaking that Kerry had won…

    People were ready to unofficially declare Kerry the victor when….

    Michael Barone said, “Wait just a minute…….”

  12. I have Caller ID, and I don’t talk to pollsters. After you’ve been push-polled a few times, you want to wish an eternity of pain to those bastards.

Comments are closed.