What Early Voting Can (And Can't) Tell Us

Nov 5, 2018

Tomorrow is finally Election Day. After months of campaigning and dozens of polls, voters get the final say. Past elections tell us the vast majority of Texas voters cast their ballots during the state’s two-week early-voting period. With almost all those votes waiting to be counted, political analysts and pundits are doing their best to glean what they can from the turnout.

But what can we tell from those still-secret ballots?

Here's What We Know

You’ve seen stories on the record early-voting turnout: It’s double the last midterm and includes a large chunk of new voters. Candidates across the political spectrum will say it's helping them the most.

But how do they know? How does anyone even begin to guess what the early-voter turnout means?

RELATED | Planning To Vote? You Should. Here's Everything You Need To Know.

Let’s start with the data that are available.

Jim Henson is director of the UT Austin Texas Politics Project. He’s also a pollster. When you compare historic voter data to the state voter file, he says, you can find out if folks vote in primaries or just in general elections, or whether this is their very first election.  

Data miners use that information to tease out what the early vote could mean. There’s a good chance that a primary voter, especially one who has voted for the same party in several primaries, will support that party again.   

But then there are the people who vote only in general elections. How do you figure out that voter’s propensity at the ballot box? Henson says that’s where the dozens of other data points attached to most of us come into play.

“From our magazine subscriptions to our shopping habits to our income levels,” he says, “all these things ... go into trying to make an educated estimate, let’s call it, about what you’re going to do on Election Day.”

Here's What We Don't Know

As Henson says, it’s an estimate. He calls it voter propensity; others might call it an educated guess. It is taking the data available and trying to figure out what could happen in the future.

The early voting totals don’t tell us if a person who has voted in three straight Democratic primaries has had a change of heart. It can’t tell us if demographic changes in a county mean that a reliably red county is now voting much more purple. And it can’t tell us how new voters have voted in the past – for obvious reasons.

So when you read that 800,000 more Republicans than Democrats cast a ballot in Texas during early voting, know that there’s a bunch of data and work that went into that estimate.

Then go look up the definition of estimate.