EAC Releases 2016 EAVS Data

[Screenshot image via EAC]

There is a lot happening in the election community right now – some of it encouraging, some less so – but last Friday, June 30 definitely included some very good news: the arrival of what has become one of the most highly-anticipated and valuable data sets available to election administrators and researchers nationwide, the Election Administration and Voting Survey (EAVS) collected and compiled by the U.S. Election Assistance Commission (EAC). Here’s the Commission’s release:

The U.S. Election Assistance Commission today released the 2016 Election Administration and Voting Survey (EAVS), the most comprehensive nationwide data about election administration in the United States. The survey is a deep dive into a wealth of election and voting data and includes findings such as an increase in registered voters since 2012, a growing number of Americans voting before Election Day and a larger number of jurisdictions using technologies such as e-poll books.
“In the face of unique challenges last year, election officials across the country administered fair and accurate elections,” said EAC Chairman Matthew V. Masterson. “Today’s report gives us a detailed look inside that process and provides data we can use to improve future elections and voter experience. The EAC looks forward to turning the 2016 EAVS responses into resources for state and local election officials and the American voters they serve.”

The 2016 EAVS included responses from all 50 states, the District of Columbia and four U.S. territories.  Data is collected at the state and local level by county or county equivalent. The EAVS was administered by Fors Marsh Group and in partnership with the Pentagon’s Federal Voting Assistance Program (FVAP).

Sean Greene, EAC Director of Research, leads the EAVS. He notes, “One of the most important results of the survey, which began a dozen years ago, is it allows us to see the impact of policy changes over time. For example when the EAVS was first administered, only one state allowed online voter registration (OVR). Now nearly three dozen states offer OVR and it was the second most common source of voter registration applications in the 2016 election cycle. The EAVS allows all stakeholders in the elections process access important data like this.”

The following data points are among this year’s primary findings:


  • Sixty-three percent of the U.S. civilian voting age population, 140,114,502 million Americans, voted in the 2016 Presidential Election.  
  • Five states – Colorado, Maine, Minnesota, New Hampshire and Oregon – had turnout rates exceeding 70 percent.  

Voter Registration 

  • States reported 214,109,360 million citizens as registered to vote. This represents a 6 percent increase in registered voters compared to the 2012 Presidential Election. Nationally, 86.7 percent of all registrants are considered active voters, and 8.7 percent are on an inactive voter registration list. 
  • States and territories reported processing 77,516,596 million voter registration applications, 83.4 percent of which were accepted by election officials. 
  • Online registration applications constituted only 6.5 percent of total registrations in the 2014 election, but accounted for 17.4 percent of registrations in the 2016 Presidential Election. Since its inception, registering to vote online has grown in popularity as it has been adopted by more states. The Department of Motor Vehicles still receives the most registration applications (32.7 percent), while other registration methods, like mail and in-person registration, have declined in use since the 2012 Presidential Election. 
  • From 2014 to 2016, 16,696,470 million citizens (8.8 percent of all registrants) were removed from state voter registration rolls. The number of registrants removed from registration rolls between 2014 and 2016 was 1.9 million greater than in the same period leading to the 2014 Federal Election (i.e., 2012 – 2014), a 12.8 percent increase. Most states and territories that provided information about the number of citizens removed from registration rolls reported removing between 5 and 10 percent of their registered voters. 

Pre-election Voting 

  • In 2016, more than 41 percent of all ballots were cast before Election Day. Of the total turnout, approximately 17 percent of ballots were cast using in-person early voting and nearly 24 percent were cast using by-mail absentee voting. 
  • By mail absentee voting rates vary dramatically across states. Nationally, 79.9 percent of absentee ballots transmitted to absentee voters were returned and 99 percent of returned ballots were counted. 

Military and Overseas Voting 

  • In 2016, 930,156 Uniformed and Overseas Citizens Absentee Voting Act (UOCAVA) ballots were transmitted and 68.1 percent — 633,592 UOCAVA ballots – were returned. Of the UOCAVA ballots returned by voters, 19,039, about 3 percent, were rejected. Of these rejected ballots, nearly half were rejected because they were not received by election offices on time. 
  • The number of ballots transmitted to overseas civilians increased by 23% from 2012 to 2016. Illinois, New Jersey and Washington are among the states that reported transmitting many more ballots to overseas civilians in 2016 than in 2012. Cumulatively, those three states accounted for an increase in about 40,000 ballots transmitted to overseas civilians. 

Precinct and Polling Places 

  • Administration of the November 8, 2016 General Election was a massive undertaking. Nationwide, there were 178,217 individual precincts (geographic voting areas to which individuals are assigned and that determine the ballot type voters receive) and 116,990 physical polling places (the locations where people can vote on Election Day). In addition, jurisdictions operated more than 8,500 early voting locations in the days leading up to the election. 

Poll Workers 

  • Recruiting poll workers continues to be a challenge for many jurisdictions; almost half reported they had a somewhat difficult or very difficult time recruiting poll workers. The poll worker population remains skewed toward older Americans, with 24 percent of poll workers aged 71 and older and another 32 percent aged 61-70. 

Provisional Voting 

  • There were 2.4 million provisional ballots cast in 2016, with almost half of those ballots cast in California. Of the provisional ballots cast, 71 percent were counted either partially or in full. 

Poll Book Technology 

  • Most jurisdictions across the U.S. (81.8 percent) use preprinted paper registration lists to check-in voters at the polls. From 2012 to 2016, however, there was a 75 percent increase nationally in the use of electronic poll books in elections. In 2012, 645 jurisdictions—7.9 percent of all jurisdictions nationally—reported using e-poll books to sign voters in. By 2016, 1,146 jurisdictions—17.7 percent of all jurisdictions—used e-poll books and 1,109 jurisdictions used them to sign in voters at the polls. Some of these jurisdictions used e-poll books to update voter history and to locate polling places.

The EAC conducts the EAVS to meet its Help America Vote Act of 2002 charge to serve as a national clearinghouse and resource for the compilation of information with respect to the administration of federal elections. Additionally, the EAVS fulfills EAC data collection requirements contained in both the National Voter Registration Act (NVRA) and the UOCAVA.

The full report is available here; you can also access a fact sheet on EAVS as well as a webpage with state-specific data and other resources.

It’s hard to overestimate the value of this data to the field – especially in the current environment where election administration systems across the nation are once again under what the Election Center’s Doug Lewis famously called called the “electron microscope”. To me, what’s exciting about the data (beyond the electiongeeky thrill of reams and reams of information) is the anticipation of what researchers like MIT’s Charles Stewart and others will do to analyze this comprehensive snapshot of the 2016 election. It may seem hard to believe, but I suspect we still have a lot to learn – and be surprised – about 2016 as a result of this data.

I also want to recognize that this is the first of what I hope will be many EAVS releases under the leadership of my longtime friend and colleague Sean Greene, who has made election research his life’s work, and who has gone from eager consumer of EAVS data to the person responsible for ensuring it gets shared with the election community. Congratulations, friend.

It’s also another reminder of the value of the EAC, who despite its central role on this and other key issues in the field, is still apparently (and mind-bogglingly, if that’s a word) threatened with elimination by another new piece of legislation on Capitol Hill.

The news brings lots to be frustrated by these days – which is why it’s so great to see (and count on) the release of the new EAVS data. Thanks to the EAC and all of the state and local election administrators who made this data set happen; I can’t wait to see what the field can do with it.

Stay tuned …

Be the first to comment on "EAC Releases 2016 EAVS Data"

Leave a comment

Your email address will not be published.