Why I’m skeptical of some puzzling polls

  • @kescusay@lemmy.world
    link
    fedilink
    English
    319 months ago

    Yep. For the last half a dozen elections, the polls have overestimated Republican strength, and underestimated Democratic strength. I think it’s in large part because the pollsters still haven’t managed to figure out how to poll people who simply will not answer unknown numbers.

    • @ryathal@sh.itjust.works
      link
      fedilink
      39 months ago

      Polls were really bad in 2016, but the seem to have largely corrected that. 2018 polling just before the election was accurate. 2020 was projected as close and it was, there were a few problem states, but nothing like 2016. 2022 was again very accurate.

      Polling around the presidential election or maybe just Trump is less accurate, but it’s been getting better since 2016.

      • @kescusay@lemmy.world
        link
        fedilink
        English
        39 months ago

        Not quite. Polls underestimated Democratic support in 2018, 2019’s special elections, 2021’s special elections, the 2022 midterms, and last year’s elections. It’s been remarkably consistent how far off they’ve been.

        Do I prefer that to the other way around, as happened in 2016? Sure. But they’ve clearly over-corrected, and are having significant trouble getting back on track due to the difficulty of polling young people.