Skip Navigation
Archive

Study: Design Flaws Contribute to Hundreds of Thousands of Lost Votes in Recent Elections

Design defects in ballots, voter instructions, and voting machines contributed to the loss of several hundred thousand votes in the most recent national elections, a new Brennan Center for Justice study found.

July 31, 2012

Report Details Major Ballot Design Prob­lems, Proposes Non-Partisan Solu­tions

Contact: Erik Opsal, erik.opsal@nyu.edu, 646–292–8356

New York, NY – Design defects in ballots, voter instruc­tions, and voting machines contrib­uted to the loss of several hundred thou­sand votes in the most recent national elec­tions, a new Bren­nan Center for Justice study found

In addi­tion, the report notes that in the 2008 and 2010 general elec­tions combined, as many as 400,000 people had their absentee or provi­sional ballot rejec­ted because they made tech­nical mistakes complet­ing forms or prepar­ing and return­ing the envel­ope. Poor design increases the risk of lost or misrecor­ded votes among all voters, but the risk is even greater for partic­u­lar groups, includ­ing low-income voters, and the elderly.

The compre­hens­ive study outlines simple meas­ures elec­tion offi­cials can take before Novem­ber to cure design defects and ensure every voter can cast a ballot that counts. View a slideshow of design flaws and solu­tions in recent national elec­tions.

“In the age of smart­phones and tablets, many have real­ized  the import­ance of good design and usab­il­ity, but Amer­ican elec­tions are still marred by major design prob­lems, ” said Lawrence Norden, deputy director of the Center’s Demo­cracy Program and co-author of Better Design, Better Elec­tions. “The rise of absentee and provi­sional voting since 2000 has made ballot design in our elec­tions even more import­ant. If a voter takes the respons­ib­il­ity to vote, elec­tion offi­cials must do everything in their power to make sure that vote counts.”

The Bren­nan Center’s report details four design and usab­il­ity prob­lems in 2008 and 2010. Here are a few select examples:

Prob­lem 1: Ballot Layouts that Invite Over­votes or Under­votes

  • In East St. Louis, IL in 2008, the ballot design led 1 in 10 voters to skip the U.S. Senate contest by mistake because of an inad­equate header identi­fy­ing the race. More than twice as many votes were lost in East St. Louis than the rest of the state. The Bren­nan Center’s revised ballot (page 17) could have saved many hundred votes.

Prob­lem 2: Poor Voter Instruc­tions

  • In the governor’s contest in Ohio in 2010, several counties repor­ted unusu­ally high numbers of voters select­ing more than one candid­ate. The culprit appears to be the instruc­tions, which state “select the set of joint candid­ates of your choice.” In Cuyahoga County alone, more than 2,000 voters did not have their vote for governor coun­ted because they selec­ted more than one gubernat­orial candid­ate. The Bren­nan Center’s sugges­tion for revis­ing the instruc­tion appears on page 25.

Prob­lem 3: Unclear Voting Machine Messages

  • Tens of thou­sands of votes were not coun­ted in 13 Flor­ida counties in 2008 and in New York State in 2010 because of inef­fect­ive over­vote warn­ings. If a voter selec­ted too many candid­ates in a race, a confus­ing error message appeared. If the voter pressed the green “Accept” button, marked with a check, the ballot would be cast with the over­vote, and the vote would be lost. The Bren­nan Center’s sugges­ted fixes appear on pages 27 and 28.

Prob­lem 4: Diffi­cult Absentee and Provi­sional Ballot Envel­opes

  • In Minnesota in 2008, nearly 4,000 absentee ballots were not coun­ted because the envel­ope was not signed. Recog­niz­ing the prob­lem, the Minnesota Secret­ary of State’s office worked with design, usab­il­ity, and plain language experts in 2009 and 2011 to improve the ballot envel­ope.  The changes made to the envel­ope can be found on pages 31 and 33.

“The design flaws that this report docu­ments are not diffi­cult or unknown prob­lems,” said Whit­ney Quesen­bery, co-author of the report and a user exper­i­ence researcher. “I hope that this stark evid­ence of lost votes inspires every elec­tion offi­cial to follow good design prin­ciples, and test their work to be sure that voters under­stand how to fill out forms and mark their ballots so their votes will be coun­ted."

As elec­tion offi­cials final­ize ballots and other elec­tion forms in the next several weeks, the Bren­nan Center’s report recom­mends several simple meas­ures that can be taken to ensure votes are coun­ted accur­ately. Elec­tion offi­cials should:

  1. Review data on lost votes to determ­ine what prob­lems they may encounter in Novem­ber.
  2. Create a check­list of design best prac­tices to make ballots and other elec­tion mater­i­als better organ­ized and easily compre­hens­ible.
  3. Conduct usab­il­ity test­ing to uncover poten­tial prob­lems that may arise.
  4. Make voters aware of poten­tial prob­lems if those issues cannot be addressed before the elec­tion.

The Center’s study provides four case stud­ies that demon­strate the power­ful impact usab­il­ity test­ing, voter educa­tion, and other correct­ive action before an elec­tion can have in redu­cing voter error in elec­tions (begin­ning on page 36).

For all the latest voting rights news, view the Bren­nan Center’s Elec­tion 2012 page.

To set up an inter­view, please contact Erik Opsal at erik.opsal@nyu.edu or 646–292–8356.