Election Data Analysis

Detecting Irregular Patterns of Election Results

Examination of limited election data shows irregular patterns of vote counts in Florida, Ohio, Washington, New Mexico, and other states. We cannot know the full extent of problems until we collect and analyze the data.

Analysis of Undervotes in New Mexico’s 2004 Presidential Ballots, January 03, 2005

, 2004

Detecting Voter Disenfranchisement & Voter Service Levels

Election Data can be used to monitor vote count integrity; voter service levels; and voter disenfranchisement.

Reports released by election jurisdictions must include the number of registered voters assigned to that polling location, whether polling locations represent one or more precincts (if more, how many), machine allocation, equipment failure and breakdown, under-vote, over-vote, and uncounted ballot rates, absentee and provisional ballot & voter registration handling, and other crucial measures of voter services. Also, after reasonable time frames, include reports on whether supplies ran out, and maximum reported wait time for voting. The goal is to have voter service reports submitted prior to the deadline to contest an election and prior to swearing in.

US EAC 2006 Election Administration & Voting Survey

EAC Election Administration & Voting Surveys

Checking Vote Count Accuracy

In any field, the audited data must be committed first. I.e. An auditable report must be released publicly prior to the random selection of vote counts to audit so that the public can verify the audit.

“Auditable report” means a report of detailed vote counts and ballots cast on each vote counting device in each precinct, for each election office, for each candidate and ballot contest, separately for each vote-type including Election Day, early, provisional, absentee, mail-in, military, etc.

All ballot types must be tracked separately for that jurisdiction, from provisional to absentee to polling place electronic to polling place paper,... for each vote counting device for each race.