{"id":17402,"date":"2016-09-16T12:34:13","date_gmt":"2016-09-16T16:34:13","guid":{"rendered":"http:\/\/peterlevine.ws\/?p=17402"},"modified":"2016-09-16T12:37:19","modified_gmt":"2016-09-16T16:37:19","slug":"three-ways-of-thinking-about-fluctuations-in-polls","status":"publish","type":"post","link":"https:\/\/peterlevine.ws\/?p=17402","title":{"rendered":"three ways of thinking about fluctuations in polls"},"content":{"rendered":"

With the national\u00a0presidential polls suddenly looking very tight, here are three ways of looking at the state of the election.<\/p>\n

    \n
  1. An election is like\u00a0a race. As in a race, the contenders stand in some relation to each other at any given moment. They can increase or reduce their speeds, but it’s an\u00a0advantage to be in front, and more so as time passes.\u00a0If an\u00a0election is like a race, then it becomes increasingly important who’s ahead as the finish line approaches. A race\u00a0course may have features that favor one or the other contender at a given moment. For instance, each presidential candidate gets a burst of speed after her or his convention, and a debate offers a chance for\u00a0one of them\u00a0to speed up or stumble,\u00a0but the last stretch will be\u00a0pretty level and even. In that case,\u00a0it\u00a0is bad news for Clinton that her lead had dissipated as we’ve moved through September.\u00a0Much depends on whether that trend continues or reverses in the next few weeks, because\u00a0by mid-October,\u00a0a\u00a0candidate who trails has little time to make up the gap. (That conclusion follows from the race metaphor.) It\u00a0supports the idea that Trump has as much as a 40% chance of winning<\/a>.<\/li>\n
  2. An election is\u00a0an event that occurs at one moment (although kind of a stretched-out moment nowadays, thanks to early voting). Polls\u00a0ask people how they will<\/em> vote once the big\u00a0moment comes. It’s not clear when our\u00a0predictions are most accurate, and\u00a0accuracy may\u00a0not necessarily increase over time. Instead, we might think of each of the many hundreds of polls taken so far as a measure of how the public will vote once the actual election comes. The best estimate, from this Bayesian perspective, averages\u00a0all<\/em> the polls taken so far. It does so not only to maximize the sample size but also to negate the random variations in competitors’ standing due to recent events. As Sam Wang says<\/a>, “I still expect Clinton\u2019s lead to increase again, on the grounds that she has led all year. Previously, I noted that the national Clinton-vs.-Trump margin in 2016 has averaged 4.5 percentage points. The standard deviation is 2.2 points, comparable to the four Presidential elections from 2004 to 2012.\u00a0…\u00a0Today, conditions seem right for regression to the mean<\/a>.'”\u00a0There is no such thing as regression to the mean in a race<\/em>, where the leader accumulates\u00a0an increasing chance of winning. But this second way of thinking about the election avoids the race analogy. Wang<\/a>‘s own Bayesian prediction is a little more complicated, but it gives Trump only a 14% chance of winning.<\/li>\n
  3. An election is\u00a0an event that will happen at one moment in the future,\u00a0and each poll is\u00a0a prediction of what will happen when that moment comes–but the sample that responds to pollsters\u00a0varies<\/a> depending on recent events. Democrats, for instance, may have become marginally less likely to answer surveys\u00a0in the last two weeks because of some generalized discouragement–or Republicans who were going to vote for Trump all along may have become more willing to answer the pollsters’ calls.\u00a0If this theory applies, I think we should act as Wang recommends, because we should treat the variations in response rates as pretty random. But we might view\u00a0the real vote as similar to\u00a0a single poll and ask whether the experience of actually voting\u00a0will encourage or discourage the people who have been favorable to Clinton or to Trump all along. We cannot tell the answer to that question from poll data, but we might propose reasonable hypotheses about it.<\/li>\n<\/ol>\n

    Since I don’t know which of these theories\u00a0is true, I’m inclined to estimate the odds of a Clinton win somewhere between the Bayesian estimate (86% or so) and the horse race estimate of only about 60%.<\/p>\n","protected":false},"excerpt":{"rendered":"

    With the national\u00a0presidential polls suddenly looking very tight, here are three ways of looking at the state of the election. An election is like\u00a0a race. As in a race, the contenders stand in some relation to each other at any given moment. They can increase or reduce their speeds, but it’s an\u00a0advantage to be in […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[32],"tags":[],"class_list":["post-17402","post","type-post","status-publish","format-standard","hentry","category-2016-election"],"acf":[],"_links":{"self":[{"href":"https:\/\/peterlevine.ws\/index.php?rest_route=\/wp\/v2\/posts\/17402","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/peterlevine.ws\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/peterlevine.ws\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/peterlevine.ws\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/peterlevine.ws\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=17402"}],"version-history":[{"count":2,"href":"https:\/\/peterlevine.ws\/index.php?rest_route=\/wp\/v2\/posts\/17402\/revisions"}],"predecessor-version":[{"id":17404,"href":"https:\/\/peterlevine.ws\/index.php?rest_route=\/wp\/v2\/posts\/17402\/revisions\/17404"}],"wp:attachment":[{"href":"https:\/\/peterlevine.ws\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=17402"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/peterlevine.ws\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=17402"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/peterlevine.ws\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=17402"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}