Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
×
Cellphones Communications Democrats Republicans Politics

'Cellphone Effect' Could Skew Polling Predictions 836

Ponca City writes "A good deal of polling data suggest that Republicans may win the House of Representatives in today's mid-term elections. However, Nate Silver writes in the NY Times that there are several factors that could skew the election, allowing Democrats to outperform their polls and beat consensus expectations. Most prominent is the 'cellphone effect.' In 2003, just 3.2% of households were cell-only, while in the 2010 election one-quarter of American adults have ditched their landlines and rely exclusively on their mobile phones, and a lot of pollsters don't call mobile phones. Cellphone-only voters tend to be younger, more urban, and less white — all Democratic demographics — and a study by Pew Research suggests that the failure to include them might bias the polls by about 4 points against Democrats, even after demographic weighting is applied. Another factor that could skew results is the Robopoll effect, where there are significant differences between the results shown by automated surveys and those which use live human interviewers — the 'robopolls' being 3 or 4 points more favorable to Republicans over all. It may be that only adults who are extremely engaged by politics (who are more likely to be Republican, especially this year) bother to respond to robocalls. Still, when all is said and done, 'more likely than not, Republicans will indeed win the House, and will do so by a significant margin,' writes Silver. 'But just as Republicans could beat the consensus, Democrats could too, and nobody should be particularly shocked if they do.'"
This discussion has been archived. No new comments can be posted.

'Cellphone Effect' Could Skew Polling Predictions

Comments Filter:
  • by Pojut ( 1027544 ) on Tuesday November 02, 2010 @11:30AM (#34102034) Homepage

    I'm sitting this one out, and possibly 2012 as well. Voting for the guy or gal that lies the least still means I'm supporting a liar. The very nature of politics nowadays automatically means someone with enough clout to run for election is unfit to serve...

    • by Archangel Michael ( 180766 ) on Tuesday November 02, 2010 @11:36AM (#34102120) Journal

      You should vote, if only to vote for a write-in or third party candidate. This election is as much about the "Two Parties" screwing things for "Joe Sixpack" in favor of their corporate overlords. The problem is that we don't have much of a choice from the two major parties.

      So, vote, but send a message. If third parties get more than 20% combined, there can be no call for "mandate" from either of the two parties.

      Voting for the lessor of two evils is a logical fallacy. There are more than two evils running for most posts.

      • by Nadaka ( 224565 ) on Tuesday November 02, 2010 @11:51AM (#34102364)

        In many cases, I agree. In 2000 I voted green even though I didn't agree with half their agenda, I've been disenfranchised by moving around for several years since then. But I just voted strait democrat in this election because the republicans in my area decided to go with comic book villain style candidates.

        Rick Scott (R, FL gov) = Lex Luthor

      • by DragonWriter ( 970822 ) on Tuesday November 02, 2010 @12:04PM (#34102604)

        The problem is that we don't have much of a choice from the two major parties.

        Voting for third-party candidates (unless you can rally enough votes to actually win, which is structurally unlikely without changing the electoral system first) is unlikely to change that.

        So, vote, but send a message. If third parties get more than 20% combined, there can be no call for "mandate" from either of the two parties.

        This is rather well demonstrated to be false from the fact that, in the rather rare individual elections in which third party candidates have won more than 20% of the vote, the winning major party candidates have still claimed mandates.

        For the most part, the whole point of negative campaigning is to get people who might otherwise vote for the other major party candidate to, in rough order of preference, vote for the candidate on whose behalf the negative ad is prevent, not vote at all, or vote for a third-party candidate. There's a reason why major parties often are found channeling support to "independent" or third-party candidates whose natural appeal overlaps that of their major-party opponent.

        Voting for a third-party candidate doesn't "send a message" to the major parties, except the message that their negative campaigning against eachother is working exactly as designed.

        Voting for the lessor of two evils is a logical fallacy.

        No, its not. It may or may not be good tactics, but its certainly not a logical fallacy.

        • by King_TJ ( 85913 ) on Tuesday November 02, 2010 @02:54PM (#34104810) Journal

          You make several good points, but ultimately, I still find I disagree with you on some of it.

          Although major party candidates of the "2 party system" may in fact use a 3rd. party candidate as "leverage" to get more votes (channeling support to them so votes for them siphon them away from their direct competitor), I'm not sure that should be viewed as a "problem"? If you happen to believe that neither a Republican or a Democrat that's running for a given position is right for the job, you're stuck with the following options:

          1. Refuse to vote. (Useless, because by sitting it out, you're ranked among the apathetic. Everything goes on without your input.)
          2. Go to the polls and vote for the "lesser of 2 evils" of the Republicans/Democrats in question. (That means your vote just counted the same as the next guy who was in FULL SUPPORT of the candidate you disliked, but only voted for because you hoped they were slightly better than the alternative.)
          3. Vote for a 3rd. party candidate who is closer to your own beliefs than the others in the running.
          4. Vote for a write-in candidate. (Practically-speaking, this option seems to accomplish nothing except in some oddball case where you knew the majority agreed to go with a write-in. Otherwise, it's pretty much a statistical impossibility your random write-in candidate is going to be selected over people actually named on the ballot as choice, who spent money actively campaigning.)

          So optimally, I think your options really boils down to either 2 or 3 here, if you're wanting to accomplish anything at all? And the way I look at it? If nobody cast any votes for the 3rd. party guys or gals on the ballot, they'd eventually just go away completely. Sure, their chances of actually winning might be slim to none, but your vote for them helps legitimize what they're doing. For example, the Libertarian party has pretty much always managed to get somebody onto the ballot in every election I've ever voted in. They may only get 3% of the vote in the end -- but that's enough so they know some people out there are listening to what they're saying. In turn, they may influence some of the Democratic or Republican voters to demand more Libertarian-minded solutions from their candidates down the road.

      • by dkleinsc ( 563838 ) on Tuesday November 02, 2010 @12:14PM (#34102780) Homepage

        Voting for the lesser of two evils is not necessarily a bad strategy.

        Let's say you have an election between candidates A, B, and C. You really want A to win, you think B is somewhat evil but much better than C, and you really really don't want C. Polls show A at 2% of the vote, B at 49% of the vote, and C at 49% of the vote. Now, who do you vote for? No question that C is out. But the choice between A and B is tougher - if you vote for A, you increase the chance C will win. If you vote for B, A can never get the support they need. As an individual voter, you're in a bind - voting for A will help in the long run, but voting for B will be an improvement right now.

        It also matters a lot how bad the various evils are. If, in the above situation, you'd rate A at +100, B at -10, and C at -10000, B is probably the better choice. If you'd rate A at +100, B at -100, and C at -150, then A is probably better.

        • by Nadaka ( 224565 ) on Tuesday November 02, 2010 @02:05PM (#34104232)

          This is why some form of range voting is ideal.

          I favor negative voting as it also preserves one person 1 vote. Basically you can vote for someone as normal, or against someone and that negates one of their for votes. A candidates final count is the difference of their for and against votes.

          Other forms of range voting offer more subtle gradation at a slightly increased level of complication.

      • Local is pretty important. The small suburb I live in was briefly taken over by social conservatives when most everyone except their base stayed home. Probably the corruption of the incumbents, involving a land deal and favorable zoning changes to bring the worlds tallest building (!) to town. The deal collapsed, but not before the mayor made a bundle off the land where the building was to go. After they got the boot, the social conservatives proceeded to screw up big time. Went on a holy crusade again

    • by fuzzyfuzzyfungus ( 1223518 ) on Tuesday November 02, 2010 @11:39AM (#34102168) Journal
      So you've essentially been avoiding democracy since classical Athens?
      • Re: (Score:3, Informative)

        by drinkypoo ( 153816 )

        Athens was an oligarchy, you had to be racially and economically privilged to get a vote.

    • by CraftyJack ( 1031736 ) on Tuesday November 02, 2010 @11:40AM (#34102180)
      You don't get to sit out the result, so you might as well toss a vote to whoever you find less abhorrent.
    • by tbannist ( 230135 ) on Tuesday November 02, 2010 @11:43AM (#34102232)

      Hmm. Taking your complaint seriously:

      I'm not sure you conclusion is supportable. If you are voting for whomever "lies the least" then you're actually supporting honesty (assuming you actually can telling more lies). If other people vote the same way then you could counteract the effect of people voting for whomever tells them what they want to hear. Looking at it from a macro point of view, voting for the least dishonest person increases the value of honesty in campaigns. Failing to vote at all on that basis does the opposite of what you want, it actually encourages more dishonest behavior because it increases the relative value of the votes of the gullible (by making the votes of skeptical irrelevant).

    • in your words, is the perfect cattle of an authoritarian country, the perfect double plus good citizen

      the simple truth of the matter is, if you wait for your perfect candidate, you will never vote. and even then you will find something wrong with them. every election, ever held, and will ever be held, will simply be a choice between the lesser of two evils. no one is pure, no one doesn't have lies spread about them

      the real criminal is you: you who hold your candidates to impossible standards, and then complain no one meets those standards

      what you are really doing is rationalizing your desire to absolve yourself of responsibility for the society you live in. you are detaching yourself from any crimes that happens in your society, absolving yourself of guilt: "i didn't choose our leaders"

      and in a country composed of people who think like you, sits the happiest tyrant

      go to work slave. don't ever complain again. even when they increase your workhours and decrease your salary. not your fault, right?

      you, all by yourself, no one else to blame, have given up the right to complain, by choosing not to do the ONE TINY THING that guarantees that you live in a free country: VOTE

    • by cgenman ( 325138 ) on Tuesday November 02, 2010 @12:08PM (#34102656) Homepage

      While the high-profile election seats may be that way, local elections actually generally have people that will effect your daily life. Should the city revise your street to be more cyclist friendly, at the expense of parking? Will they approve of installing billboards in front of the local lake? You have a pretty solid voice in deciding who makes those decisions. And people at the local level tend to be genuine and earnest.

      Similarly, if your state does ballot propositions, they can be incredibly powerful. California might end the war on pot. Massachusetts might kill affordable housing. These are important things which are up for a yes-or-no vote.

      I once thought like you do. In 2000, I thought "These guys are both sellout corporate tools who are only interested in money." "They both must be equally bad," I thought. OMFG did Bush prove me wrong.
      The lesser of two evils might still be evil, but damn can the greater of two evils get us into some huge intractable problems.

      • by Bigjeff5 ( 1143585 ) on Tuesday November 02, 2010 @12:58PM (#34103438)

        While the high-profile election seats may be that way, local elections actually generally have people that will effect your daily life.

        More importantly high-profile candidates almost always come from these positions first. If you shut them down before they become "somebody" you've prevented that future bad candidate from being a possibility.

        The reason we have such a huge problem with representatives is because politicians are practically ignored at the time when they are most easily influenced.

        For example, Obama started out as a state senator (some funny business there, of course, but that's Chicago politics), became a US Senator for the state of Illinois 8 years later, and is now the president. He could have very easily been shut down at the state level by a competent opponent and a few thousand votes.

        Don't tell me your vote doesn't matter. It doesn't have great influence when you think it should, but it certainly has a huge impact when you aren't really paying attention.

  • So... (Score:3, Interesting)

    by WillyWanker ( 1502057 ) on Tuesday November 02, 2010 @11:32AM (#34102074)
    Basically you're telling us what we've already known for decades... that polling is retarded and highly inaccurate.
    • Re: (Score:3, Insightful)

      by LWATCDR ( 28044 )

      Actually I would like to see the polling banned.
      It introduces way to much bias into the process. People tend to not want to throw away their vote so once a canidate is in the lead people tend to want for them or not for them instead of the person that they think is the right one.

      That and they should keep primary results a secret until every state votes.

      It is funny but I had a long drawn out discussion about the value of randomizing ballots and bias. This bias is probably a million times greater than who is

      • Re: (Score:3, Insightful)

        by tbannist ( 230135 )

        That's not an issue with polling, it's an issue with using a terribly flawed voting system (first past the post). Fix the system and it would fix quite a few political problems. For example, preferential voting eliminates the need for strategic voting as you've described above.

      • Re:So... (Score:4, Informative)

        by stoolpigeon ( 454276 ) * <bittercode@gmail> on Tuesday November 02, 2010 @12:00PM (#34102506) Homepage Journal

        I understand your sentiment but polling is free speech. Someone should be free to ask, and others should be free to answer, or not. If someone does ask a bunch of people a question, they should be free to share those results. If there is a chunk of the population that is too stupid to filter those results or understand what they may or may not mean - then the solution is education not the restriction of speech.

    • Re:So... (Score:4, Insightful)

      by interkin3tic ( 1469267 ) on Tuesday November 02, 2010 @11:41AM (#34102204)

      And even if it weren't WHY THE FUCK IS POLLING ALL THEY TALK ABOUT? Paying attention to the news will tell you 1. Who is running 2. How likely they are to get elected 3. If they are having sex with someone who isn't their spouse 4. What their opponents are saying about them, in order of most to least information.

      Not on there: their history or what they will actually do (if anything) when elected. Who do I vote for, the guy who's likely to win? Because that's about the only thing you'll get from the news.

      How a candidate is polling is of interest to the candidate and his staff, and to people who already know who they are voting for to either say "Ha ha, we're going to win!" or "Damnit, we're going to lose!" To everyone else, it should be trivial information.

  • by xxxJonBoyxxx ( 565205 ) on Tuesday November 02, 2010 @11:33AM (#34102080)

    Cellphone-only voters tend to be younger, more urban, and less white — all Democratic demographics — and a study by Pew Research suggests that the failure to include them might bias the polls by about 4 points against Democrats, even after demographic weighting is applied.

    Umm...isn't the point of demographic weighting to factor in "unweighted" demographics like this?

  • by eldavojohn ( 898314 ) * <eldavojohn@gm a i l . com> on Tuesday November 02, 2010 @11:35AM (#34102112) Journal

    It may be that only adults who are extremely engaged by politics (who are more likely to be Republican, especially this year) bother to respond to robocalls.

    Poll Phone Operator: Excuse me, sir or ma'am, do you have a free minute to answer a few simple questions anonymously about who you plan to vote for?
    Phone Respondent One: Well, let's see, what would Jesus do?
    *Poll Phone Operator hangs up the phone and puts a check mark next to the Republican candidate*

    Poll Phone Operator: Excuse me, sir or ma'am, do you have a free minute to answer a few simple questions anonymously about who you plan to vote for?
    Phone Respondent Two: Sorry, what did you say? It's cloudy and my solar powered phone is cutting in and out.
    *Poll Phone Operator hangs up the phone and puts a check mark next to the Democratic candidate*

    Poll Phone Operator: Excuse me, sir or ma'am, do you have a free minute to answer a few simple questions anonymously about who you plan to vote for?
    Phone Respondent Three: Yes I do, just let me put NASCAR on mute, I can talk and watch at the same time.
    *Poll Phone Operator hangs up the phone and puts a check mark next to the Republican candidate*

    Poll Phone Operator: Excuse me, sir or ma'am, do you have a free minute to answer a few simple questions anonymously about who you plan to vote for?
    Phone Respondent Four: I'm so sorry but I just put on a 180 gram vinyl Arcade Fire album and I fear that if I remove the needle prematurely I would ...
    *Poll Phone Operator hangs up the phone and puts a check mark next to the Democratic candidate*

    Poll Phone Operator: Excuse me, sir or ma'am, do you have a free minute to answer a few simple questions anonymously about who you plan to vote for?
    Phone Respondent Five: Fuck you and fuck the establishment you rode in on.
    *Poll Phone Operator hangs up the phone and puts a check mark next to the Independent candidate*

  • Lopsided summary... (Score:5, Informative)

    by jusdisgi ( 617863 ) on Tuesday November 02, 2010 @11:36AM (#34102126)
    Some background is in order here; this is not a typical piece for Silver. He did a companion [nytimes.com] to it a couple days ago, giving the reasons the GOP could overperform. These are just "what if" stories, designed to flesh out the message he's been driving for some time now, which is that this election has unusually high uncertainty. He isn't engaging in hackery and claiming everything will be fine for Democrats...
  • by Anonymous Coward on Tuesday November 02, 2010 @11:37AM (#34102144)

    American public: "Wow, those Republicans sure fucked everything up. Better vote Democrat this time."
    T+4 years: "Wow, those Democrats sure fucked everything up. Better vote Republican!"
    T+8 years: "Wow, those Republicans sure fucked everything up. Better vote Democrat this time."

    Umm, people? We have other choices, you know. The extremes of *any* party are going to be nut-jobs, but we can probably do a lot better to let the D's and R's set a few rounds out.

    But we won't, will we. Because voting is supposed to be about thinking with other people's brains and voting with the flock.

    • by fiannaFailMan ( 702447 ) on Tuesday November 02, 2010 @12:05PM (#34102608) Journal

      American public: "Wow, those Republicans sure fucked everything up. Better vote Democrat this time."
      T+4 years: "Wow, those Democrats sure fucked everything up. Better vote Republican!"
      T+8 years: "Wow, those Republicans sure fucked everything up. Better vote Democrat this time."

      Umm, people? We have other choices, you know. The extremes of *any* party are going to be nut-jobs, but we can probably do a lot better to let the D's and R's set a few rounds out.

      But we won't, will we. Because voting is supposed to be about thinking with other people's brains and voting with the flock.

      It sounds insane, but as long as you have this winner-take-all voting system then you're always going to have two dominant parties. If you feel strongly enough about fixing the broken system* then what you should be campaigning for is a Proportional Representation [wikipedia.org] voting system where you vote for as many candidates as you want in ranked choice. This would discourage a lot of the populist appealing to the extremes and bring politics closer to the center IMHO. You'd also see a lot more people registering as independents and better quality candidates defecting to the smaller parties and making them more credible, whereas right now anyone who's serious about getting elected has to pick their poison, R or D, and hold their nose and run on that ticket.

  • Who can say (Score:4, Insightful)

    by alvinrod ( 889928 ) on Tuesday November 02, 2010 @11:57AM (#34102470)
    Who can really say. Counter-example:

    I'm doing graduate research involving monitoring students in computer science labs. Today the instructor asked how many students were planning to vote. Around 15% raised their hands. At least that many had a stunned look in their eyes as though they didn't even realize it was election day.

    Young people may be more likely to own only cell-phones and tend to be much more progressive, but it seems as though they may be a lot less likely to vote. Most of them probably live within a few blocks of where they can vote and it's a nice day out so there's not much of an excuse.

    I follow Silver's site as he often writes a lot about the statistics behind his model, which I usually find more interesting than the results or political commentary, but if these observations are true, why the hell aren't they built into his model? If these effects actually exist and skew polling results, why haven't they already been taken into consideration? Also, what effects exist that skew the results in the other direction and what evidence supports them?

    This article feels sloppy, especially when compared to the usual high quality from fivethirtyeight. Let's wait another twelve hours and then we'll have a pretty good idea about the actual outcome and can start speculating what might have caused it to deviate from the expected results so that the prediction model can be adjusted accordingly.
  • Exit Polls (Score:3, Interesting)

    by kenh ( 9056 ) on Tuesday November 02, 2010 @12:15PM (#34102790) Homepage Journal

    This reminds me of the great exit poll kerfuffle [washingtonpost.com] when John Kerry was seen to be leading in exit polls to a greater extent than the actual poll results bore out.

  • by jejones ( 115979 ) on Tuesday November 02, 2010 @12:23PM (#34102918) Journal

    Back in the days of ELIZA and people wondering whether computers really could be used to augment, if not replace, human therapists, I recall people describing their surprise at how readily people would confide in such software; perhaps people's inclination to post everything on Facebook is related. But:the existence of a difference between what people tell automated polls and human pollers doesn't, by itself, tell you which of those responses reflects what the people polled really think. For all I know, that ELIZA effect still holds, and people will tell the machine something that they wouldn't tell a human. A robocall won't turn on you and say "You favor tax cuts?! THINK OF THE CHILDREN!!!!" (at least not yet!) or give you a dirty look or inflection that indicates disapproval.

  • http://www.fec.gov/pubrec/2000presgeresults.htm [fec.gov]

    the results of the 2000 elections were decided by a razor slim margin. meaning those who chose not to vote had a real effect: they helped bush win

    and if you say "politicians are all the same": tell me with a straight face gore would have invaded iraq

    those who don't care, or don't want to be involved, are just as guilty as everyone else for the sorry state of the world, if not more so

    • Re: (Score:3, Interesting)

      meaning those who chose not to vote had a real effect: they helped bush win

      Nope. Bush won by just one, repeat just one, vote. 5 to 4.

      • that supreme court was put in place by a previous president. the vote would have been 5-4 for gore, if someone in 1980 or 194 or 1988 had gotten off their duffs and voted

        for example:

        by a vote of 5-4 in january of this year, our supreme court said it is basically ok that corporations spend freely on elections. this is a supreme court put in place by bush. bush barely won the 2000 election

        therefore, if the tiniest minority more had voted in 2000, gore would have won, we would not have invaded iraq, and the ridiculous pro-corporate dollars in elections decision in january would be 5-4 AGAINST

        meaning YOUR VOTE MATTERS, IT REALLY DOES

        when you don't vote, you are basically saying "i am completely happy with the way things are going, don't change a thing". if you think by not voting you are somehow being noble or acting principled, you are a complete and utter fool: corporations WANT you not to vote. an electorate that feels helpless and uninvolved is an electorate that can be raped

  • by jav1231 ( 539129 ) on Tuesday November 02, 2010 @12:46PM (#34103292)
    I firmly believe not everyone should vote. Stupid people, for instance. The king that are often interviewed in man-on-the-street interviews who can't tell you who the President is, Speaker of the House, and other basic information people should have before voting. Those people should stay home. That's why I always hope for rain on voting day, it keeps stupid and lazy people at home. Unfortunately, the rain called for my area is holding off.

Dennis Ritchie is twice as bright as Steve Jobs, and only half wrong. -- Jim Gettys

Working...