The Exit Poll Blackout Strikes Back

Avatar

Will Truman

Will Truman is the Editor-in-Chief of Ordinary Times. He is also on Twitter.

Related Post Roulette

23 Responses

  1. Avatar Alan Scott says:

    I’m reminded of the proposition 8 polls, where someone pulled exit poll data from one precinct in LA and used it to show that Black Voters supported prop 8 by some absurdly high margin. Trying to make a story out of numbers is never going to work if you don’t make an effort to get good numbers.Report

  2. Avatar Damon says:

    This just sounds like sloppy journalism.Report

    • Avatar Vikram Bath says:

      Is it just me or are we seeing more of this sort of thing? I can feel some of the desperation of old media trying to keep up with new media antics in some of these mistakes.Report

  3. Avatar Burt Likko says:

    IIRC, the objection to exit polling was not the gathering of information, but the use of that information when deployed on the broadcast media before analysts really had time to understand the sudden rush of data. The solution to that problem is not to foreclose the gathering of data, but rather maturity in its use once gathered.Report

    • Avatar Will Truman says:

      The reason they stopped gathering the information (in most states) was a financial one. “Who needs the data for North and South Dakota? We know which way they’re going to go.” (My initial criticism being “But Vermont and Maine are important?!” but those two were also dropped later, apparently.)

      Which is true, as far as that goes. And I understand the decision. But there’s still value in collecting the data, even for useless states. Which is why I hope somebody (state university political science departments?) steps up in 2016.Report

    • Avatar Kolohe says:

      The embargo on exit poll info isn’t because of insufficient analysis; it’s because telling people the results of an election before voting has ended may influence election participation and thus results.Report

      • Avatar Glyph says:

        Yeah, this. We already know people are prone to peer pressure and tend to “like” what other people like.

        Also, ppl might look at exit polls and falsely think it’s already in the bag for their guy and stay home, or panic because they think their guy is getting pasted and rush to the polls.Report

      • Avatar Will Truman says:

        The former is more common than the latter. Which is part of why I think the treating of pre-election polls as holy writ (or holy science, and only knuckl-dragging science-hating heathens are skeptical of them) is dangerous. Bad polls could distort elections, instead of just misreporting them.Report

      • Avatar Mike Schilling says:

        That’s why Romney lost. Millions of his supporters didn’t bother to vote because they read about the unskewed poll results and thought his election was in the bag.Report

      • Avatar Stillwater says:

        Pre-election polls and exit polls were incredibly reliable until … what was it again? Bush’s first run? Since then they’ve been politicized by the GOP and conservatives just like everything else has been.

        I don’t wanna go on record saying that there were election shenanigans that year. And for a few cycles after. But my conscience forces me too.Report

      • Avatar Damon says:

        There are always election shenanigans.Report

      • Avatar Will Truman says:

        Among other things, an election needs to be closer than the 2012 one was for polls to change an outcome.Report

  4. Avatar Michael Cain says:

    Exit polling is becoming less useful as any sort of indicator anyway. Colorado has switched to sending a mail-in ballot to every registered voter, and IIRC now has something >90% of all votes cast being cast by mail (or early drop off). Arizona, by virtue of a permanent no-excuse absentee ballot list, had >60% of all votes cast being mail-in this last time. California had >50%. I expect both of those states to adopt the practice of sending a mail-in ballot to every registered voter in the next few years (if the legislatures drag their feet, both have easy citizen initiative processes).

    If/when those two states adopt universal mail ballots, you would have something over 75% of registered voters in the West (as defined by the Census Bureau) getting a mail-in ballot.Report

    • Avatar Will Truman says:

      It’s true that going forward they may need to find another way. (Day before election poll + day after + exit polls + results?)

      I can’t find validation, but I think the National exits weren’t technically exits this year. But I could be wrong.Report

      • Avatar Michael Cain says:

        As the County Recorders have all of the relevant data, I anticipate a profitable new line of business for them, selling breakdowns of the vote by gender, race, etc. Not as useful as exit polls for same-day news, but probably better for post-election analysis.Report

      • Avatar Will Truman says:

        Except who precisely voted for whom, I assume.Report

      • Avatar Michael Cain says:

        I expect to live long enough to see some County Recorder — not necessarily one where it’s all mail-in ballots, although that does seem more likely — either accidentally send individual votes out due to insufficient quality control, or have that information be extracted by hackers. Hell, if the health insurance companies, who are presumably going to face stiff HIPAA penalties, are going to leak customer information 80M people at a time, some poor schmuck of a County Recorder is eventually going to have actual ballots with names leak.Report

      • Avatar Will Truman says:

        See?! Mail-in ballots are totalitarian!Report

    • Avatar Will Truman says:

      Also, yet another reason to mutter at the increased availability and utilization of alternative voting methods. Mutter mutter.Report

  5. Avatar Mike Schilling says:

    Here’s my map, based on voters I drove home from the polling place with:

    Report

  6. Avatar ???? says:

    租赁合同主体问题的认定基本案情:2007年4月27日,二被上诉人就宇鑫公司所有的位于麦积区一马路东天美花园2号楼一层临街铺德阳婚姻调查公司&#38Report