How to avoid misreading the early voting numbers

One week out from Election Day, who is casting early votes is getting increased attention. NBC News’ tracker of early ballot returns is one of several presenting information on who has already cast ballots state by state — tantalizing hard numbers after a campaign full of polling suggesting a close race but nothing more definitive than that.

But while early voting is a useful tool for understanding what has transpired to date, it is easy to read more than is appropriate into these patterns and what they mean for the 2024 election, given what we know about when different types of people are likely to vote. In particular, we know that younger people disproportionately wait until closer to Election Day to cast their early votes.

Our main takeaway: Even with weeks of early voting already complete in some places and only days left before it’s all over, it’s still dangerous to read too much into partisan breakouts of who has voted so far. There is still time for those patterns to shift — and we’ve already seen them shift in twists and turns through recent weeks of early voting.

About 10 days ago, NBC News explored some initial patterns in Virginia’s early voting, noting that as a share of total turnout from 2020, voters in parts of the state that generally support Republican candidates had cast more ballots than in parts of the state that generally support Democratic candidates.

Because we speculated that these differences might reduce once more of Virginia’s satellite voting locations opened, we thought it was premature to conclude based on these patterns that Republican-learning areas were headed toward higher turnout than Democratic-learning areas. Consistent with that hypothesis, the below table of early voting through Oct. 26 shows substantial leveling in the usage of early voting over different areas.

However, it’s still the case that substantially more ballots have been cast in areas that generally lean Republican than other parts of the state. 

Even though most satellite voting locations are now open, though, these patterns are not clear evidence that total turnout will be higher in areas that lean Republican than in the rest of the state.

One reason caution is still in order is because the timing of when votes are cast can vary across a state.

Previous research shows that older and more experienced voters disproportionately cast their absentee ballots long before Election Day, with younger and less experienced voters waiting until closer to Election Day to cast theirs. Thus, we expect areas that have more older voters who regularly participate in elections to cast their absentee ballots earlier than areas that have more younger voters who participate less often.

This is certainly true in Virginia right now: According to our NBC News’ tracker of absentee ballot returns in Virginia, about 44% of the absentee ballots cast to date have been cast by people over the age of 65 compared to only 16% by people under 40. As a point of contrast, in the 2020 election, TargetSmart’s voter file shows that 30% of voters were under 40 and 24% were over 65.  

The get-out-the-vote operations active in an area also likely play a role in determining when people vote. For example, an experiment during the 2020 election showed that people returned mail ballots sooner when they were randomly assigned to be sent an informative postcard about the process of mail balloting, even though the postcard made no mention of timing. 

And given that younger voters are disproportionately Democrat and older voters are disproportionately Republican, there are reasons to expect that, all other things being equal, areas that tend to support Democrats will see more of their absentee ballots cast just before Election Day than areas that tend to support Republicans. This was the case during Virginia’s 2022 general election: The parts of the state most supportive of Democrats cast about 57% of their absentee ballots within 10 days of the election, while the parts of the state that tend to support Republicans returned only 40% of their absentee ballots during that final-stretch time frame.

Will absentee voting follow similar patterns in Virginia in 2024 as in 2022? It is really hard to say. One of the broader challenges of interpreting early vote totals in most states is the lack of sufficient historical data to benchmark against the current returns.

Because 2020 ushered in the use of more absentee voting in many states, we don’t trust that data on the timing of absentee votes from before 2020 is necessarily indicative of the patterns we should expect to observe now. And because so many more ballots were cast by mail in 2020 than will be cast by mail this time, we don’t think that it is wise to benchmark relative to 2020.

Benchmarking to 2022 is an option in some states, but there are still reasons to suspect that people may vote at different times during a presidential election than a midterm election. Thus, while we anticipate that there will be more absentee ballots cast over the next 10 days in Virginia in Democratic areas than Republican areas, we don’t think that we have enough information to be able to say anything concrete on exactly how many. 

Another reason why we aren’t concluding that there is likely to be disproportionate turnout in those areas that lean Republican comes from looking at the vote history of people who have already voted in Virginia. The below table shows the number of recent elections in which this year’s Virginia early voters participated. Just as previous research suggests, an overwhelming number of the absentee ballots cast so far have been cast by consistent voters.

If an area is going to experience higher than normal turnout, we’d expect to observe a greater number of less consistent voters casting ballots. But if anything, the below table shows that the people who have voted in the areas that lean Republican are more consistent voters than the average person who has cast an absentee ballot in Virginia so far.

These patterns in Virginia have implications for how you should think about the early vote data being reported in any state.

In particular, it says you need to be very cautious about how you interpret any partisan breakouts of the early vote reported by NBC News’ tracker of absentee ballot returns or any other tracker. In particular, you need to consider the possibility that absentee voters from one party may be overrepresented in the current returned ballots compared to what the final results will show because of differences in when voters cast their absentee ballots.

This is especially true in the battleground states, where campaign activity is also likely shaping when people are voting.

And while it can be helpful to compare a state’s patterns this year to historical markers showing when votes were cast in the state in 2022, nothing guarantees that these return patterns are going to be similar over election cycles.

Finally, we should keep in mind that a vote counts the same, whether it is cast on the first day of early voting or as polls close on Election Day. And thus we should be careful about assuming too much before all ballots have been cast.

This story first appeared on NBCNews.com. More from NBC News:

• 9 congressional sleeper races to watch on Election Night
• 'Ready for anything': Democrats tout post-election legal preparations
• New York delivered Republicans the House in 2022. Will the blue state help them keep it in 2024?