I hope you found my early vote analyses informative. The election does not end for me, as over the next month I will post turnout rate estimates for the 2024 election. In some states I have a good sense we are close to the final number of votes to be counted. In others, there may be an unknown number of outstanding mail ballots or provisional ballots and I make an educated guess what the eventual turnout will be based on history and turnout in other states where we have more data. Of course, election officials will eventually report their final election results. I will update these numbers as states count their ballots and eventually certify the election.

These turnout rates are for the voting-eligible population or VEP, a phrase I coined about two decades ago. My VEP turnout rates are used widely by the media, textbooks, academics, and policy makers, and are considered by some to be the semi-official turnout rates for the United States.

The voting-eligible population is an estimate of everyone who is eligible to register and vote in an election. I calculate VEP by estimating the citizen voting-age population available from the Census Bureau and subtracting an estimate of the number of ineligible felons, depending on state law (I do not try to estimate ineligible felons that may still be ineligible after they serve their sentences because there is no good way to estimate these populations). Nationally, I also add in overseas citizens and overseas military (there is no good way to estimate these populations among the states).

VEP is different than registered voters, which many election officials report as the denominator their turnout rate. There are good reasons why election officials care about registered voters for their administration of elections. For example, they need to allocate polling equipment to precincts to ensure a well-run election.

The reason why scholars and many others prefer VEP turnout rates is that they provide a consistent way to compare turnout across states and time. To give an example, Oregon adopted a policy known as automatic registration whereby eligible persons who interact with a DMV office are automatically registered to vote unless they proactively state they do not want to be registered. This policy resulted in the registrations of tens of thousands of people. While some of these people did vote, many did not. Calculating turnout rates with voter registration as a denominator counter-intuitively resulted in a lower voter registration turnout rate after the implementation of a policy that empowered more people to vote. Using the VEP denominator, the turnout rate increased, as one might have expected.

More generally, states vary in how they administer voter registration. Do they have automatic registration? Online registration? Same day registration? A deadline to register long before the election? When do states remove registered voters? How aggressive are they at removing people? And so on. Like Oregon, these policies can change over time. These differences mean registration turnout rates may be higher or lower not based on campaign affects — like if a state is a battleground state — rather, they are influenced by the nature of how a state administers voter registration.

You can find my 2024 turnout rates here.