Early voting: what he tells us (and does not tell us)

I say we are arguing here and see where we agree and disagree on the early vote.

Eric:

Hi Harry. Let me start by admitting that many early votes do not tell us. We know how many people are voting, but we do not know who they are voting for, but only for the registration of their party. And we certainly do not know who the independents vote for. We do not know what the electorate will look like. Does the increase in voter turnout during the early days of early voting reflect a broader trend, or is it simply highly motivated people who would always vote? It's too early to say.

But for campaigns, early voting is tactically important. They use early voting data to adjust their ads and programs on the fly. Suppose, for example, that a candidate exceeds their goals across the country, but underperforms in a large county. This campaign would transfer more money into the important county. The same goes for a demographic group – like Latinos, which are a major concern this year for Democrats.

So, although the early days of early voting do not necessarily predict the outcome of a race, I think they give the campaigns (and us) big flashing warning signals about what might happen. Is it right?

Harry:

So, I think we come to something. There is a difference between how the public can use early voting and campaigns. In addition, there is a difference between painting a picture of the electorate and trying to predict how that voter will vote.

If I can, I just want to point out again a point or two that you said. Whatever we know about first voters, we do not really know how they vote. This is especially the case in the state without political affiliation. In these states, you do not even know if anyone is a registered Democrat or Republican. You only know the last primary of the party where they voted. And if this person does not vote in the primaries, you do not have anything.

Let me add another point on which I think you are referring. Early voting is a relatively new phenomenon. I think in a perfect world, we would be able to say if more Democrats or Republicans are voting earlier than expected. But what exactly is this base? More people are coming to vote than ever before.

It means three things.

1. Yes, the campaigns have an idea of ​​what is going on and can possibly adjust the strategies.

2. We do not really know if the participation rate will be higher. Campaigns may cannibalize people who used to vote on polling day.

3. This could mean that past trends, such as states in which early voters traditionally lean more in favor of a party than polling day voters, will not really be considered as trends. of the future.

One question I ask myself is this: why are people so interested in early voting? What is so magical?

Eric:

I think there are some answers to your question about why people are so interested in us. First, in some key states, almost everyone votes. In Arizona and Nevada, two states with significant seats in the Senate, about three out of four voters will have voted early. This, along with the adoption of the postal vote, shifts the timetable for everything. The "caravan" and the posted bombs arrive after a large number of people have already decided and voted.

I also think that the benefit of early voting this year is that after 2016, polls were "wrong". More specifically, these state polls, in particular, were a little out of the way of what was happening with certain subgroups. I think people are aware that interest in intermediate courses is higher than usual this year, and they are trying to avoid the same mistakes. Supporters worry about all these close races and they think / hope to watch key subgroups – Latinos, for example, for Democrats; Exurban white voters for Republicans – could be an indicator of new imperfect polls.

If you are considering the campaign from a campaign point of view, in addition to what we have already explained in tactics, there are two other factors. When you know that someone has voted early, you can remove it from your list of exit contacts from the vote – which makes this process more efficient. You can also redirect that person's energy toward volunteering to make calls, get others to vote, and so on.

Here's what I'm wondering: Is there a way to look at the polls – which I think are more useful than the anticipated voting data to predict outcomes – along with the early vote to learn something something new?

Harry:

I think you come to an interesting point about early voting, which is under-covered. It's not so much that we can "predict", but rather that the election season itself is a little shorter. From what I understand from the literature, those who vote early are usually pure supporters. As a result, events such as "caravan" will not have a disproportionate impact. Nevertheless, it should be noted that such events that could be forgotten on election day will be in the minds of those who will vote early.

I guess what motivates me is that people say that the polls were "wrong" in 2016 and that early voting can be a way to solve this problem. You do not do it here, even if it's something that definitely happens. I have not seen any evidence that early voting is more predictive than polls. I've certainly seen people who have no idea what they're doing while trying to believe that they can beat the ballot box. In the words of Michael Jordan, "stop that".

Here is what I think we can learn from the ballot along with the advance vote. We can know if the polls correctly represent the anticipated vote. In states like Nevada and Florida, we know who votes early. We can see if the polls reflect exactly that. If they are not, polls may be wrong.

Are there specific elements that you are looking for in advance voting?

Eric:

Let me quickly come back to something you just said: "We can know if polls correctly represent early voting."

What do you mean? Are you saying that the demographics or party registration numbers of those who voted early are predictive and therefore have to roughly match the poll results?

Harry:

I mean, the investigators will ask if someone voted early. If they do it and the polls do not roughly represent this vote demographically, the registration of parties, etc., this could imply that these polls are inactive.

Eric:

I have it.

One thing I am going to watch is whether the high turnout that we have seen – higher than the previous ones, but well below presidential year levels – lasts much longer. There are few trends, like that of registered Democrats exceeding those of Republicans enrolled in Washoe County (Nevada, as emphasized by Ralston), who do not seem sustainable.

I am also eager to see if there are signs of problems or controversy from the point of view of access to the vote, which is a great potential story this year. Early voting can be a kind of test of resistance. In North Dakota, for example, the Heidi Heitkamp campaign fears that a strict law on voter identification will prevent some members of Native American tribes from voting. Many eyes are also turned to Georgia.

And you? As a skeptic of early voting, is there anything you think you can learn?

Harry:

Strict laws on voter identification are interesting.

I mean, to be honest, I'm not looking for a lot.

I think if Ralston highlights something interesting in Nevada. Then I will listen to you. Same for Steve Schale in Florida. Otherwise, not really. I will certainly not integrate it into modeling.

Eric:

That's enough. I head west in a few days and cross Nevada – I can not wait to see how things are on the ground. Fun talk!