[Screenshot image courtesy of eac]

That sound you heard mid-day yesterday was hundreds (thousands?) of electiongeeks clapping their hands with glee. The reason? The U.S. Election Assistance Commission (EAC) released the summary report and datasets from the the 2014 Election Administration and Voting Survey (EAVS).

Here’s a summary of the findings from the EAC release (all emphasis in original):

Broad Highlights

• According to the data submitted by the States and territories, 81,133,122 individuals participated in the 2014 election.

• Of the more than 81 million people who turned out to vote in 2014, over 60% voted at the polls, 17.5% voted a domestic absentee ballot, and 10.7% voted early (prior to Election Day).

States reported counting 98.2% of the domestic absentee ballots submitted. The most common reason for absentee ballot rejection was a missed deadline for returning the ballot, followed by invalid signatures.

A total of 892,202 provisional ballots were submitted according to the States and territories. 80.3% of those ballots were counted in whole or in part. Of the 171,443 that were not counted, the most common reason was because the voter was not properly registered.

In 2014, States operated 178,636 precincts and over 114,000 physical polling places. States employed almost 731,000 poll workers on Election Day. The largest number of poll workers ranged from 41 to 70 years of age.

The report also covers data regarding the National Voter Registration Act (NVRA, or “motor voter”) and the Uniformed and Overseas Citizens Absentee Voting Act (UOCAVA). From the release:

NVRA: Key Report Findings

• There were approximately 190 million total registered voters reported for the November 2014 midterm election. The number of registered voters in 2014 represents an increase of approximately 3.2 million voters since the 2010 midterm election cycle.

• According to the responses to the survey and population estimates from the U.S. Census Bureau, 84.7% of American citizens of voting age were registered to vote in the 2014 elections.

33 States reported receiving voter registration applications over the Internet.

• Internet applications accounted for 6.5% of the total number received, up from 1.7% in 2010.

There were 16.6 million new applications in 2014, an increase in new applications as compared to the 2010 elections when there were 14.3 million new registrants.

States found invalid or otherwise rejected over 984,000 voter registration applications, a decrease from the 1.3 million voter registration applications that were rejected in 2010. In addition, nearly 3.5 million applications were duplicates of existing registrations. Altogether, 9.0% of registration applications were invalid or duplicates, a decrease from the 9.4% of registration applications that were invalid or duplicates in 2010.

States removed over 14.8 million voters from voter registration lists. The NVRA allows states to remove voters who have not voted in two consecutive federal general elections and failed to respond to a confirmation notice from an elections office. Other reasons include death, felony conviction, having moved from one jurisdiction to another, mental incompetence, or at the voter’s request …

UOCAVA: Key Report Findings

States transmitted 420,094 ballots to UOCAVA-covered voters for the 2014 election, with just over half (51.4%) going to civilian citizens living overseas. Another 46.0% went to uniformed services members.

Of the UOCAVA ballots transmitted, 34.6% (145,509) were returned and submitted for counting, which is an increase from the 30.2% in 2010.

States reported counting 137,683 UOCAVA ballots, or 94.6% of the total submitted for counting.

States reported rejecting 8,492 ballots. The most common reason given for rejecting a UOCAVA ballot was that the ballot was not received on time; States reported that 48.9% of rejected ballots were rejected for this reason.

States reported that 1,934 voters submitted a Federal Write-in Absentee Ballot, which is a sharp decrease from the 4,294 voters who submitted an FWAB in 2010.

Now with the data out, the analysts go to work; the University of Florida’s Michael McDonald spotted some reporting issues (Alabama with highest turnout? No – misreported voter registration as turnout), while Reed College/Appalachian State’s Paul Gronke checked totals and started producing maps. But the person I was most thinking about was EAVS maven and data jockey extraordinaire MIT’s Charles Stewart, who I imagined looking like this [image via armedrobbery]:


This release almost certainly heralds a future parade of analyses from these three and many, many others that both help us understand what happened in 2014 and what it might mean for 2016 and beyond.

Kudos to the EAC for “keepin’ on keepin’ on” with this work through the agency’s recent difficulties – and especially to the thousands of election officials who provide the data that fills in the color on this snapshot of the 2014 election.

Happy EAVS Day, everybody!