Votebeat is a nonprofit news organization reporting on voting access and election administration across the U.S. Sign up for Votebeat Pennsylvania’s free newsletter here.
Human error was the primary cause of a pollbook printing issue that forced thousands of voters in Chester County, Pennsylvania, to cast provisional ballots in the 2025 municipal election, according to a report by outside investigators.
Early on on Nov. 4, 2025, voters and election officials realized that the pollbooks — the printed voter lists used to check people in at polling places — didn’t include the names of Chester County’s more than 75,000 unaffiliated and third-party voters. Those voters had to either wait for supplemental pollbooks to be delivered or use a provisional ballot, an option used when there is some question about a voter’s eligibility.
According to the report — from West Chester-based law firm Fleck Eckert Klein McGarry LLC and released Thursday by the county — the “two inexperienced employees” who generated the pollbook files mistakenly selected the option to include “only voters from the major parties.” The report said the employees had “limited election experience, no formal SURE system training, and no direct supervision during a critical task performed under compressed statutory deadlines.” SURE, or the Statewide Uniform Registry of Electors, is Pennsylvania’s voter registration database.
As a result of the error, about 12,600 provisional ballots were cast in the county, accounting for roughly 6.4% of the county’s total vote, more than in any other election in the county’s recent history. The vast majority of the provisional ballots were counted after the county confirmed the voters’ eligibility.
Report cites other gaps in the system
The report said there was no evidence of intentional wrongdoing. While the report primarily blamed human error, it said that other factors also contributed to the issue
First, it noted that the SURE system allows a pollbook with only major-party voters — meant for use in the state’s closed partisan primaries — to be generated without any confirmation or warning screens. The report also said the state’s training did not sufficiently guide employees through the settings needed to generate a general-election pollbook versus one for primaries. Finally, the report said “no policy or procedure was in place” to check the files generated from SURE for accuracy before the pollbooks were printed.
Poll workers previously told Votebeat and Spotlight PA that in years past they had the ability to check pollbooks before Election Day, but the county recently began sealing those materials.
Asked if it would be updating the SURE system or its training materials, the Pennsylvania Department of State said in a statement that it “continues to monitor and update the system to meet election security and election administration needs” and was open to working with Chester County to improve the training process. It also highlighted the ongoing effort to replace SURE with a more modern voter management system.
“The new system will have validation and quality assurance built in around our business logic — meaning, for example, that a county user will not be able to select ‘only major parties’ for their general election poll books,” the statement said.
The report also said high staff turnover at the county’s election office in recent years “has increased operational risk.” The Philadelphia Inquirer reported in December that dozens of employees have left the office since its director, Karen Barsoum, took over in 2021. Some former employees described a toxic work environment under Barsoum.
While the pollbook report acknowledged complaints about Barsoum’s leadership and management style, it said the turnover trend was “not necessarily reflective of a cultural or managerial deficiency” and attributed it to recent high-pressure election cycles and better pay in surrounding counties.
What the investigators recommended
The report recommended several changes to avoid similar failures in the future. Among them, it said the county should ask the Department of State to implement safeguards in SURE that would alert users when they are about to export pollbooks that exclude some voters, as well as ask the state to revise its pollbook training guide.
The county itself should implement a procedure to double-check that pollbooks are accurate prior to Election Day, the report suggested.
“At least two trained staff members, including a supervisor, should independently verify SURE Poll Book generation selections prior to finalization,” the report said. “Before transmitting files to the print vendor, staff should also be required to complete a mandatory audit checklist certifying to the commission that randomly selected precincts have been spot-checked and that Poll Book files include all required data and all required voter categories.”
Allegheny County, in western Pennsylvania, told Votebeat and Spotlight PA in November that it performs such spot checks for accuracy.
The report said Chester County should also more thoroughly document its procedures, as well as who is responsible for performing what steps, and any action that could affect the entire electorate, such as printing pollbooks, should require signatures from a supervisor and administrator.
In a statement responding to the report, the county said that while it believes some sections could benefit from additional context, it otherwise “fully supports and plans to implement the report’s recommendations.”
The county said it has developed an action plan to address all the recommendations. The county board of elections will hold a meeting at the Chester County Judicial Center on Jan. 27 at 7 p.m. to discuss the report and the action plan. If the judicial center is closed that day because of weather, the meeting will be moved to Feb. 3 at 7 p.m.
Carter Walker is a reporter for Votebeat in partnership with Spotlight PA. Contact Carter at cwalker@votebeat.org.
