New Briefing on Provisional Voting
"Solution or Problem? Provisional ballots in 2004"

By: electionline.org
Published: Fri, 18 Mar 2005
 Whether a provisional ballot was counted in last year's election depended nearly as much on where you lived as your status as a voter. A new report by the nonpartisan and non-advocacy electionline.org found that differing practices around the country in issuing and counting provisional ballots led to wide variations in the numbers fail-safe of ballots distributed and counted.

Election Reform Briefing 10, "Solution or Problem? Provisional Ballots in 2004," asserted that pre-election predictions that provisional ballots would become "the hanging chad" of last year's election never came to pass ? but only because of the final margin of victory in the presidential race. However, the report found enormous gaps in provisional ballot counting, usage and rules that suggest Congress' cure for the ills of the 2000 election in Florida ? eligible voters being turned away without recourse ? might itself pose problems for voters and election officials nationwide.

The Help America Vote Act (HAVA) mandated the use of provisional ballots as a safeguard for voters who believed they were registered but whose names were not on voter rolls. While more than two-thirds of states had some safeguard in place for those voters prior to HAVA, November 2004 marked the first time every state and U.S. territory (except those with election-day registration or no voter registration) had provisional ballots available as well as methods by which voters could find out if their ballots were counted.

"The national mandate for provisional voting did not mean national uniformity," said Doug Chapin, electionline's director. "In fact, we found that whether a voter had his or her provisional ballot counted relied nearly as much on where they cast it as their actual registration status. Had the election been closer, this would have been a flashpoint for controversy. Our findings, however, suggest that the debate over provisional ballots is far from over."

"If HAVA's aim was to ensure no qualified and registered voter would ever again be turned away from a polling place without having the opportunity to vote," Chapin continued, "this report shows that the patchwork of rules and practices on provisional voting might actually be frustrating rather than advancing the aims of HAVA."

Among the findings:

* Provisional vote-counting varied widely among states, from a national high of 97 percent counted in Alaska to a low of 6 percent counted in Delaware. The report found a national average of 68 percent counted. While lacking any concrete data of the reasons for accepting or declining the votes, anecdotal reports from state and local election officials indicate some people were confused into thinking that provisional ballots could be used in lieu of registration. Those ballots were not counted.

* State rules had an impact on the percentage of provisional ballots counted. States that allowed voters to cast ballots outside of their correct precinct but inside of their jurisdiction counted an average of 70 percent of provisional ballots. States that disqualified any provisional ballots counted outside of a voter's correct precinct counted 60 percent of the ballots.

* Statewide voter registration databases led to fewer provisional ballots cast, presumably because each voter's registration record was tracked better. However, of those that still needed to cast provisional ballots, there is no evidence to suggest that a fewer percentage of provisional ballots were counted in states with statewide voter registration databases than in states without them.

* Inconsistent application of rules and procedures might have had an impact within states. Some counties sought out provisional voters to have them fix problems with their registration forms that kept them off of precinct lists.

Despite the disparities, provisional voting could nonetheless be considered a qualified success. More than 1.6 million fail-safe ballots were cast, with nearly 1.1 million (68 percent) counted. In Florida and Ohio alone, this meant nearly 200,000 voters who would have been turned away from polling places if their names could not be found in 2000 had a chance to cast a ballot in 2004.

The Briefing, the first extensive analysis of provisional ballot usage, counting and practice across the country, is available at electionline.org.