POLLS ARE STILL STUPID

The media's collective insistence on running an Obama-McCain horse race poll every 45 minutes is enough to make me want to adopt a homeless dog and punch it.

online pharmacy buy fluoxetine online no prescription pharmacy

Or go whaling. One of the two, just out of spite.

I struggle to think of things that matter less than mid-summer general election polling. Voters fall into two categories at this point in the campaign season: either they have made up their minds (which a lot of us, frankly, did in 2001) or they have almost zero political information on account of paying no attention whatsoever to the race. The media's mythical fence-sitter, torn between two equally appealing candidates, is a rarity. The inability to express a preference speaks more clearly to ignorance than ideological ambiguity/ambivalence.

online pharmacy buy neurontin online no prescription pharmacy

Let's take a Pew Research poll, conducted from July 31-August 10. It polled 2,414 adult likely voters nationwide and collapsed the leaners (MoE 2.5%):

McCain 43%
Obama 46%
Unsure/other 11%

Obama's ahead, but McCain's closing in! Right? Aside from the fact that the margin of error overlaps (McCain: 40.5% – 45.5%, Obama: 43.5% – 48.5%) how can anyone put the slightest bit of stock in a poll – of a two-way race – with 11% undecided? Enjoy the many levels on which a single poll can negate itself.

"I don't know" or "undecided" in a high-profile race with saturation media coverage means "I do not want to embarass myself by telling the survey guy that I have no fucking clue." Many such individuals will not vote ("likely voting" is another wildly inflated aspect of polling) but many will.
buy wellbutrin online buy wellbutrin no prescription

They will do one of two things. They could remain clueless and literally vote at absolute random.
buy symbicort online buy symbicort no prescription

Such voters would have no impact on the race because, as the Marquis de Condorcet discovered a few hundred years ago, true randomness cancels out. Second, they can start paying attention to the race at the last minute and make some sort of non-random decision. Some voters parcel out their tiny attention spans strategically. They ignore the race, giving "I'm clueless" answers to pollsters, until the last minute. So they have a real preference, they just don't realize it yet.

Imagine yourself on a playground, having evenly divided your votes between kickball and freeze tag, waiting for the "slow" kids to stop picking their noses and eating glue long enough to break the tie. Yes, the 11% of individuals who, for one reason or another, can't give an answer in polls such as this one will ultimately decide the outcome in a close two-way race. How's that for depressing? The people who pay attention have essentially come to a split decision. Now we wait for the lame, the halt, and the ugly to furrow their brows and cast their votes – votes which may be entirely random. That's all these polls tell us.

2 thoughts on “POLLS ARE STILL STUPID”

  • The mainstream media outlets specialize in creating a story where there is nothing to tell. Polls like the one you describe count on the public's statistical ignorance in order to manufacture a "close race" narrative. It's more exciting than saying that Obama will more than likely beat McCain handily. It's also a lot easier to say that it's a close race rather then risk ticking off as much as half of your viewership/readership by claiming that either candidate has an edge…even when it is obvious to everyone that that is the case.

    Stated bluntly, the mainstream media will lie through its collective teeth to make a contest seem more competitive than it realistically is because they have to justify 24-hour coverage of what are essentially non-events and maintain a "balanced" view to appease everyone.

    In short, the mainstream media is far more interested in selling stories than in actual reporting. The media is in a race for the bottom…they're vying for the "nose-pickers" as you call them–those poor folks who can hardly maintain a thought without assistance and reassurance from "their" news outlet…(Fox viewers, I'm referring to you specifically.)

    When the news is created with the intent to entertain rather than inform, good reporting (and thoughtful analysis of polls) gets edged out until the news looks more and more like an episode of Jerry Springer and reads more and more like an Ann Coulter screed.

  • Ed –

    Great post. At some level I belive polling was invented by statisticians who weren't able to make it in academia or who wanted to make money off of people who didn't take a math class past their junior year og high school.

    I do have a question for you though… What about those respondents who lie to pollsters?

    I was at a seminar for lobbyists recently and learned the following (which probably isn't all that newsy, I just don't pay attention to polling questions) from a group that does a significant amount of polling.

    The presenter indicated they will ask a question along the lines of, "Would you vote for Barak Obama for Preisdent?" and X% will say yes. However, when they ask the same respondent, "DO you think your friends and neighbors would vote for Obama?" the answer comes back substantially lower than X%.

    Do you think this is a function of people lying to pollsters about their own prejudices (i.e. I don't want this person to know I'm racist) or more closely related to them not thinking very highly of theor neighbors??

Comments are closed.