Professor Alex Street Q & A with Senior Political Reporter Jonathon Ambarian of the Montana Television Network (MTN).
What are the challenges of doing a poll?
The core challenge in polling is to get a representative sample. That's more important than the number of people. To be representative you need to draw from the people who are going to vote on Election Day. It's hard because you don't really know, in advance, who will vote! Most of them will be people who have voted in previous elections. But the high-profile elections, when the Presidential ticket is on the ballot, tend to draw in some new people. And certain candidates draw in new people. For example, Donald Trump appealed to some people who had been feeling pretty cynical about the political system by saying he was an outsider who would shake things up. That is one reason polls have tended to underestimate his appeal. Those same cynical people tend to distrust the organizations that run polls, like the media or universities, so it is hard to get data on them.
Is it getting harder to poll these days? Is Montana a particularly hard place to poll well?
No, Montana is not especially hard to poll. It's hard everywhere. Nowadays polling is largely via the internet on your phone, or by calling phones. It has gotten harder now that fewer people have landlines, and the ones who do tend to be older so a landline-only poll won't be representative. But that's true across the country. Each state has some groups that can be harder to reach. In Montana that may include new residents, like remote workers who have moved to Montana, or students who recently moved from out of state, or poorer people who have less access to cell phones.
How should people interpret election poll results when they see them? What should they be thinking about when looking at polls?
Election polls often ask, "If the election were held tomorrow, how would you vote?" Even though the election isn't happening tomorrow! They do that because it reduces the number who say they're not sure. But it's only intended as a snapshot, not quite a prediction. The problem is that a lot of people follow politics the way they follow a sports team, and they like the idea of keeping up with the score. That's the wrong way to think about it. In general, if a poll says that a race now is within a few percentage points, say one candidate is polling at 45% and another at 51%, it is definitely close enough that it could go either way. If you care about the election you should vote and you should encourage your friends and family to vote. I think it's better to use polling to understand the issues that are important for people. This is one way that polls can really add something, beyond just the election results which we will see in a few weeks anyhow. For instance, it is helpful to know that some aspects of healthcare reforms from a few years ago, like bans on health insurance companies turning people away because they have a pre-existing condition, are now really popular. Or that there is very little public support for selling public land in Montana to private owners. This is why, every couple of years, the students in my class run an exit poll. We want to learn about the voters. It's a great way for students to apply what they are learning and try to understand American politics. For example, we have studied whether people in rural Montana think that urban areas in the state are unfairly getting more resources, which seems to be less true here than in some states.
We're going to be running an in-person exit poll on Election Day this year but we are also going to be sending out some surveys in the mail and we should start getting some of them back in late October. We will be able to give a little preview of what we are seeing in our initial results.
Stay tuned for a follow-up on the polling from Professor Street’s class.