Skip to content

How the election might have looked in a world without polls

On the radio this morning it was all about how Biden’s in the lead but Trump outperformed the polls just about everywhere.

What if there had been no trial-heat polls? Then maybe the reporting would be how Biden outperformed Clinton almost everywhere, but given all the problems with the economy it’s surprising Trump kept it so close. I think that even without polls, Trump would’ve been the underdog, given that he lost the popular vote in the previous election and it’s been a tough year for everybody, especially for the olds. Then again, Trump has the advantage of incumbency. There are also the complications of voter registration and in-person voting with the coronavirus, given that it seems that Biden and Trump voters have different perspectives on the risks of exposure.

Another interesting factor is that neither candidate seemed to make much of a play for the center during the campaign. It seemed to me (just as a casual observer, without having made a study of the candidates’ policy positions and statements) that in 2016 both candidates moved to the center on economic issues. But in 2020 it seemed that Trump and Biden were staying firmly to the right and left, respectively. I guess that’s what you do when you think the voters are polarized and it’s all about turnout. Anyway, that was just something I noticed.

All the polls in the news can distract us from other aspects of an election, so it can be helpful to step back and look at what else is there.

P.S. Despite the above title, this is not a counterfactual imagination of a world without polls. If there were no polls, the whole campaign would look different. So really it’s more an imagination of ways we could think about the campaign if the candidates were looking at polls but we weren’t.


  1. Tom Passin says:

    I remember a science fiction story from long ago where the election forecasts had become extremely accurate, to the point that they needed just one (highly representative) voter to account for the remaining uncertainty. So picking that one voter became critical, and everyone held their breath while he (nowadays he/she, I suppose) voted.

  2. Joshua says:

    A thought. Yes, history in past elections is an input,but…

    I think we tend to underestimate a rather simple dynamic…people will tend to stay in the same place in this election because changing their vote from last time would mean they were wrong last time and people don’t like admitting they’re wrong.

    Protecting a previous decision should be the overriding default. It takes big energy to overcome that inertia. Do we tend to think that the slate is wiped clean and that people are deciding for this election as an almost independent process of evaluating things like the economy or the pandemic?

    • Andrew says:


      That’s easy for you to say now, given that the outcome of the 2020 election seems mostly like 2016 with a small shift to the Democrats, but . . . there are lots of contrary examples. Consider what happened in 1980.

      Regarding the question in your last sentence: We all agree that the vast majority of voters are going with their party and voting as they have before. And there are fewer voters in the center than there were in 1980, say. The issue is that with the electoral college where it is right now, there’s a big difference between a 0% shift (Clinton won the popular vote but lost the electoral college), a 1% shift (could be enough for Biden to pull off the electoral college win), or a 3% shift and a resounding Biden win. Small differences in the vote, big differences in that colored map.

      • Joahua says:

        Andrew –

        > , but . . . there are lots of contrary examples

        Sure. Good example. Maybe I’m over-evaluating the impact incumbency. But one frame here is that there were two basic camps in 2020: those who were really excited about Trump and those who who were just voting against Trump but not really for something concrete. The same narrative applied to Clinton in 2016 – where her message was muddled. It would be hard for a vote for Biden to be “for” something because his candidacy had to cover a pretty wide range spanning from pretty hard left to moderate to win (and he doesn’t have the personality draw of an Obama or Trump).

        Are those dissimilarities that distinguish 1980 and 2020?

        • Andrew says:


          1980 was different. Back in 1980 it was my impression that everybody hated Jimmy Carter. OK, not everybody, but lots and lots of people. Even lots of Democrats hated him. As for Reagan, many people were enthusiastic about him but many people were afraid of him too: he was so extreme and such an unknown. So, from the standpoint of the campaign, incumbent Carter vs. challenger Reagan was a much different story that incumbent Trump vs. challenger Biden. On the other hand, our model right now is pretty much just based on partisanship, and none of these factors directly come into play.

    • jim says:

      If you’re looking for a Trump effect, here are some candidates:

      Lots of people reject liberal policy and want an alternative. But Trump is such an incredible ass that it’s really difficult to vote for him, no matter how much you agree with his policies and no matter how strongly you object to liberal policy. So lots of people don’t make up their minds *really* until the last minute.

      I suspect that many of Trump’s last minute policy moves also pulled back some people that had thought to vote Biden. His removal of wolves from Federal protection was a big reminder to people that Biden will try to deliver plenty for his environmental constituency – and that might cost them their jobs and businesses.

      Also seems like his SCOTUS pick played far better than Dems would ever have dreamed – and talk of expanding the court played badly for Dems.

      • Joshua says:

        Those would be reasons why Trump did well, but not why the polls got it wrong – except to the extent that these were late-breaking influences that the earlier polls necessarily missed. But I would expect that would be suggested by trends in the later polls relatively speaking and that doesn’t seem evident.

        Trump’s improvement with minority voters would be another factor in his performance that wasn’t necessarily expected – but again, it’s not clear how that would show up as polling error unless that shift was late-breaking. There were some events with black celebrities that were heavily pushed by pubz late in the game, but is there evidence that there was a late trend there?

        Personally, I think that the differential impact of social media is generally stronger than people recognize, but how does that reflect in polling errors? The Bingerhunter Biden laptop brouhaha was a late-breaking event – was the impact under-estimated?

        • jim says:

          “except to the extent that these were late-breaking influences”

          That’s one part of the point.

          “but how does that reflect in polling errors? ”

          Was there a polling error? Trump’s approval ratings have been much lower than even his poll numbers. The polls reflect that. The simple fact is that a solid majority of the country *really really doesn’t like him*, and for good reason.

          But when it comes to voting a person is faced not just with the candidate but with the alternative. And once you get in the voting booth, or sit down with that mail in ballot, that’s the only decision that matters. You’re free to reflect and change your mind.

          So my point is not that Trump voters are shy. Many may have intended to vote Biden right up until the last minute. Trump’s approval ratings reflect that. But in the end they voted not for the man but for the agenda. People just changed their minds at the last minute.

          • Andrew says:


            I’m skeptical about this explanation given that Republicans also outperformed the polls in the congressional races. I’m more inclined to attribute the discrepancy to some combination of differential nonresponse, differential turnout, and discarded ballots. The turnout thing could be a big deal considering how the two parties differed in their on-the-ground voter registration.

            • Joshua says:

              Andrew –

              > I’m skeptical about this explanation given that Republicans also outperformed the polls in the congressional races.

              IIRC, pub congressional candidates didn’t overperform polling in 2018, but pub Senate candidates did in 2016.

              Would differential non-response lolely be explanatory in 2016 and 2020 but not
              2018? Does Trump in the ticket drive turnout from voters pollsters deem not likely to vote?

              I dunno. I wonder if there are just idiosyncratic factors that make individual elections hard to predict. Sometimes.

      • Ben says:

        > His removal of wolves from Federal protection was a big reminder to people that Biden will try to deliver plenty for his environmental constituency

        This is a really weird use of environmental constituency. I Googled this and found:

        Like why not say hunters and ranchers? I don’t think Donald Trump reads the blog, so we don’t really need to rebrand things. Obviously hunters and ranchers interact with the environment and it’s a big part of what they do. I don’t know their concern with grey wolves, but that article claims it was enough to motivate this political move; and I doubt hunters and ranchers would be mad about being called hunters and ranchers!

        But the environment is also a big concern to lots of people. Like is oil and gas part of the environmental constituency now? I can see how if you make a bunch of perspective assumptions you could get to this, but it’s just a really unclear way to write something.

        • jim says:

          You seem to have misunderstood my comment.

          The reminder is to industrial and resource businesses and workers that Biden will ramp up environmental attacks on their livelihoods. They understand quite well what’s going on with the ESA.

          Trump doesn’t give a rats ass about what the public thinks about the environment. He’s interested in saving the economy, not the environment.

  3. Ken Carlson says:

    I think for a lot of Democratic primary voters, Biden WAS the play to the center. Most of the field was to his left. His actual positions were pretty much irrelevant to the GOP attack, which seemed to be about keeping the country safe from AOC and Hunter Biden.

    • Choosing Kamala Harris as a running mate was a strong move to the Left. And, let’s be honest — Biden won’t serve out his term. Either the 25th Amendment will be invoked, or the corruption scandal will cause him to step down. Which means that a vote for Biden was a vote for Harris.

      • Joshua says:

        > Choosing Kamala Harris as a running mate was a strong move to the Left.

        I disagree. You state that as if it’s a fact rather than an opinion. There were obviously potential, viable VP candidates who were significantly to the left of Harris.

        So how do you support your determination of fact?

        • I consider choosing Harris a move to the Left based on her voting record as a Senator. Newsweek ranks her as the most left-wing U.S. Senator, while CNN says, “Harris’ voting record in the Senate is certainly one of the most liberal”.

          • Howard Edwards says:

            That’s interesting. During the Democratic primaries it was Sanders and Warren who were painted as the most leftwing socialist candidates. I don’t recall Harris’ name coming up in this regard, and she ended up endorsing Biden.

            My view may be an outlier as it was from outside the US and relying on CNN for coverage of the debates.

            • Howard Edwards says:

              Ah, found it – it’s from a news item on Fox. Seems she voted in favour of gun control and universal healthcare among other other things. I can only guess that becoming Veep qualifies you for the “most left wing” title ahead of Sanders and Warren according to Fox.

          • Joshua says:

            What Howard says. I don’t know any lefties who consider Harris “left” within the dewoek of the party. Many other demz, like Warren or Sanders or many of the other potential VP picks. Her candidacy for president was pretty middle of the pack.

            Which part of her platform did you think was particularly on the left side as opposed to the middle of the dem spectrum?

  4. David Marcus says:

    As a candidate, does it really help to try to position yourself in the center (whatever the “center” means)? What is in between reducing the use of fossils fuels and not reducing them? Asking people to reduce, but not insisting? Maybe Lincoln is an example: he was the middle candidate on the issues, I believe. But, that was a four-person race.

    • Paul Hayes says:

      What indeed. Political ideology space is an affine space and failure to recognise that often leads to nonsense. For example this complaint that “In recent years British universities have drifted way to the Left”.

    • jim says:

      “What is in between reducing the use of fossils fuels and not reducing them?”

      Not having a fossil fuel doctrine, for example? What about just focusing on the economic benefits of energy, regardless of how it’s obtained?

      • Robert says:

        Well the central question is evaluating what the net economic benefits really are when there is some uncertain degree of future damage at stake. It seems odd to call positions on this “doctrines”, since we do actually have to make an estimate before the numbers come in whether we like it or not. If we assume that “focusing on the economic benefits” refers to something unambiguous then it seems we’ve jumped past the main question. Everyone involved believes that they are doing this.

        I guess this sounds to me like saying “What is in between policy A and policy B? Deciding I’m going to do the best of A or B situationally, whichever that turns out to be”. It’s easy to take a high road if we punt all the actual choices into the future.

        • jim says:

          “Everyone involved believes that they are doing this.”

          I don’t agree with that at all.

          Climate change policy advocates couldn’t care less about economic benefits. They’re happy to jigger up some kind of economic angle that appears to justify the policy they want if that’s what they need to do, but they don’t give a hoot if it’s true or not. It’s just a tool to get what they want.

          As a general rule the environmental movement is anti-growth, anti-consumption and, for the most part, anti-human.

Leave a Reply