Reviewing Ford County’s Ballot Reconciliation Process During the 2024 General Election (Limited-Scope)

Introduction

The Legislative Post Audit Committee requested this limited-scope audit, which was authorized by the Legislative Post Audit Committee at its December 16, 2024 meeting.

Objectives, Scope, & Methodology

Our audit objective was to answer the following question:

1. Did Ford County have an adequate process for ensuring the number of voted ballots reconciled with the number of voters in the 2024 general election?

The scope of our work included Ford County Election Office’s policies and practices for ballot reconciliation during the 2024 general election. This included reviewing documents from polling sites such as poll book records, election night ballot reconciliation sheets, and election night result tapes from tabulation equipment. We also interviewed officials from the Secretary of State’s office and reviewed state laws about ballot reconciliation and the post-election audit.

More specific details about the scope of our work and the methods we used are included throughout the report as appropriate.

Important Disclosures

We conducted this performance audit in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. Overall, we believe the evidence obtained provides a reasonable basis for our findings and conclusions based on those audit objectives.

Audit standards require us to report our work on internal controls relevant to our audit objectives. They also require us to report deficiencies we identified through this work. In this audit, we reviewed Ford County’s practices for reconciling ballots with poll book signatures. We did not identify deficiencies with those processes. However, we identified 2 deficiencies that may indicate control weaknesses with other processes (e.g., obtaining voters’ signatures and sealing ballots). We did not evaluate the county’s controls in these areas because they were outside the scope of this audit, but we describe both issues in more detail later in the report.

Ford County’s ballot reconciliation process ensured the number of voted ballots reconciled with the number of voters in the 2024 general election, but the county did not ensure all poll books were signed and ballots were sealed as state law requires.

Background

County election offices use the Election Voter Information System and poll books to determine whether voters are eligible to vote.

• The main objective of the county election officer in any election is to ensure that every valid vote is counted. To do this, county election officers rely on a variety of systems and processes to ensure valid votes are accurately counted and complete. This includes the voter registration database, poll books, and tabulation machines.  

• The Election Voter Information System (ELVIS) is a database maintained by the Secretary of State’s office that stores information about registered voters in Kansas. ELVIS stores information such as birth dates, addresses, registration dates, party affiliations, and voter histories.

• County election offices use the ELVIS database to update their poll books before the election and to determine voter eligibility.

• Poll books are essentially a list of registered voters that poll workers use at polling sites on election day to determine whether a voter is eligible to vote at that location. Voters present their identification to election office staff or poll workers who look up their information in the poll book.

• Eligible voters must sign the poll book before they may vote. This helps ensure that each voter only casts 1 ballot. In Ford County, eligible in-person voters sign electronic poll books and in-person provisional voters sign paper poll books. Voters who vote by mail sign their ballot envelope in place of signing a poll book.
  • If an in-person voter isn’t eligible to vote based on the poll book, they must sign the provisional poll book before they vote a provisional ballot.
  • A provisional ballot is only counted after the county election office researches the voter and determines they are eligible to vote. Otherwise, provisional ballots are not counted.

Voters in Ford County have a few options for voting and casting their ballots.

• One way eligible voters in Ford County can vote is by advanced mail-in paper ballot. Voters apply for a mail-in ballot with the county election office. Once approved, voters receive and vote their ballot, place it in the supplied ballot envelope, and sign the ballot envelope. They then return their ballot to the election office or a polling site to be counted for the election. Mail-in ballots in unsigned envelopes aren’t valid unless the election office can contact the voter, and the voter signs the ballot envelope.

• Eligible voters can also vote in-person during early voting or on election day using 1 of 2 options.
  • Voters may use a ballot marking device to vote their ballots. The ballot marking device is a touch screen computer used to mark and print ballots at the polling site.
  • Voters may request a pre-printed paper ballot that they mark by hand at the polling site.

• In most cases, voters cast their ballots by inserting them into a tabulation machine. The tabulation machine counts the ballots using an optical scanner. If a tabulation machine can’t read a ballot, it’s counted by hand at the county election office.

• Ford County has 5 polling sites. 1 polling site doesn’t use a tabulation machine because the site doesn’t have many voters. Ford County’s election officer explained this site uses a ballot box instead. Voters at this site have the same options for voting as other polling sites, but they place their ballots into a locked ballot box. The ballots are then counted by a tabulation machine at the election office.

• Provisional voters vote provisional paper ballots. These are kept separate and not sent through the polling site tabulation machines or placed in a ballot box for valid ballots. They are counted only if the voter is determined to be eligible to vote.

The number of ballots cast must reconcile with the number of poll book signatures as soon as the polls close.

• The number of ballots cast is the sum of the tabulation machine counts for non-provisional voters. Provisional ballots are counted separately.

• The poll book signatures for each type of poll book (regular poll books and provisional poll books) are also counted and totaled when the polls close.

• Ballot reconciliation is ensuring that the number of ballots cast matches the number of poll book signatures plus canvassed mail-in ballots.

• State law (K.S.A. 25-4412 and K.S.A. 25-4611) requires that as soon as the polls close, the supervising poll worker of the polling site must count the number of ballots cast and the number of ballot containing envelopes. These counts must balance with the number of voters who signed the poll book or provisional poll book. The supervising poll worker must report any discrepancies in writing to the county election officer with the reason for the discrepancy if known.

We previously audited Ford County’s ballot reconciliation process for the 2024 primary election and found poll workers at 1 polling site did not follow the proper processes.

• In October 2024, we completed an audit reviewing whether the number of voted ballots reconciled with the number of voters in Ford County’s 2024 primary election.

• Our work showed the poll book signatures were 1 less than the total number of ballots counted at 1 polling site. This means there appeared to be 1 more ballot cast than there were voters who checked in.

• The election officer was not aware of the discrepancy until we identified it through our audit work. According to the Ford County election officer, it was a poll worker who voted without signing the poll book. The election officer said they told her that the poll book malfunctioned and didn’t record their signature when they voted.

• This audit is a follow up to the October 2024 audit. It is focused on the 2024 general election rather than the primary election.

Ford County’s Ballot Reconciliation

We reviewed Ford County’s election records from the 2024 general election to evaluate Ford County’s ballot reconciliation process.

• We went on-site to Ford County’s election office to review their ballot reconciliation process. We reviewed result tapes printed from tabulation machines for all polling sites, advanced ballots, and mail ballots. Election office staff and poll workers print result tapes at the close of polling. The tapes show the results for each race and the total number of ballots counted by the tabulation machine.

• We compared these totals against poll book signature records for each polling site and mail-in ballot records. We also checked ballot reconciliation sheets for each polling site for accuracy. Ballot reconciliation sheets are forms poll workers use to document the total number of poll book sign-ins and ballots cast and to identify any discrepancies to the county election officer. This helped us determine if Ford County’s ballot reconciliation process was adequate for ensuring the number of voted ballots reconciled with the number of voters in the 2024 general election.

• We also reviewed Ford County’s official election results report.

Ford County’s poll book logs, ballot reconciliation sheets, and result tapes reconciled with each other for regular voters and provisional voters in the 2024 general election.

• Ford County’s documentation showed poll book signatures and voted ballots balanced at all polling sites for the 2024 general election.
  • Ford County polling sites reported 8,300 total ballots counted for the 2024 general election. This was for election-day voting, early in-person voting, mail-in voting, and canvassed provisional ballots.
  • Poll book records and other documentation showed 8,476 voters signed a poll book or mailed a ballot.
  • The difference between those two numbers is 176 signatures. These included 174 provisional ballots that weren’t counted and 2 voters who poll workers said signed a poll book but didn’t cast a ballot.
  • According to Ford County’s election officer, the most common reason for not counting a provisional ballot was because the provisional voters were not registered to vote before the registration deadline. She didn’t know why 2 voters signed a poll book but didn’t cast a ballot. She said the voters may have used a ballot marking device and mistakenly assumed their ballot was cast instead of running the ballot printout through the tabulation machine. We couldn’t evaluate how this happened because we weren’t present on election day.

• We saw the ballot reconciliation sheets for each polling site were accurate and included the necessary information. For example, we saw that the information on the ballot reconciliation sheets matched the counts in the tabulation machine result tapes and poll book signature records.

But 2 provisional voters did not sign the provisional poll book before voting as state law requires.

• State law (K.S.A. 25-2908) requires voters to sign the poll book before voting. But during our audit work, we reviewed documentation that showed 2 voters did not sign the provisional poll book before voting their provisional ballots. The election results reconciled because election staff noted the errors in the provisional poll book. They still counted both voters as having signed in even though they didn’t sign the provisional poll book.

• The Ford County election officer said this was a mistake that was likely caused by human error. She said the mistake didn’t affect her ability to determine whether the 2 voters were eligible to vote. She said provisional voters fill out their information on a voter registration application that’s attached to the provisional ballot envelope. The application and envelope include spaces for voters’ signatures. The county election officer then uses this information to research their voter registration and the voter’s signature. She said if a provisional ballot envelope isn’t filled out correctly, then the ballots aren’t counted. We didn’t evaluate whether this process was followed during the 2024 general election for this audit.

• We found a similar issue during our previous audit of Ford County’s ballot reconciliation process for the 2024 primary election. For that audit, we recommended that the Ford County election officer should ensure that poll workers follow procedures for reconciling the number of ballots cast with poll book signatures on election day in accordance with statute. The Ford County election officer responded that she would go over the poll books with her poll workers during training.

We also found a voted provisional ballot that wasn’t sealed as statutorily required.

• State law (K.S.A. 25-2708) requires election officers to seal voted ballots from presidential elections in envelopes after they are canvassed. The canvassing process occurs after election day and confirms all valid ballots are accurately counted and included in the election results. The election officer and 2 other people must then destroy the ballots after 22 months without opening any of the containers the ballots are sealed in.

• We found a voted ballot for the 2024 general election that was stuck in a tabulation machine and therefore not properly sealed with other voted ballots. The Ford County election officer said it was a provisional ballot based on the mark in the upper corner of the ballot. She said it likely got stuck in one of the bins inside the tabulation machine during the final canvass of provisional ballots.

• Because we were able to reconcile ballots cast with poll book signatures and ballot reconciliation sheets, we think this ballot was counted in the election results. However, we didn’t perform additional work to evaluate Ford County’s processes related to this issue because it was outside the scope of this limited-scope audit.

Other Findings

Ford County audited one precinct for their post-election audit as required by state law.

• As part of this audit, we were also asked to evaluate whether Ford County completed the required post-election audit for the 2024 general election and to compare the number of voters who participated in the election to the total voting age population in Ford County.

• State law (K.S.A. 25-3009) requires county election officers to conduct a manual audit or tally of each vote cast in 1% of all precincts in the county before final canvass or at least 1 precinct for smaller counties. Ford County had 42 precincts for the 2024 general election. This means Ford County’s election officer had to audit 1 precinct to comply with state law.

• The law also requires the audit to include certain races. In presidential election years like 2024, the audit must include a federal, state legislative, and county race. It’s also required to include a constitutional amendment question if applicable.

• We reviewed documentation showing that Ford County completed the post-election audit according to state law by auditing 1 precinct. The audited precinct was one of Ford County’s largest precincts with over 1,000 ballots cast.

• The documentation also showed that Ford County audited the required races during the post-election audit. The audit included 1 federal race, 1 state legislative race, and 1 county race. There was no constitutional amendment question to audit for the 2024 general election.

• Based on this work, we identified a potential issue with how state law on the post-election audit is written.
  • State law (K.S.A. 25-3009) requires counties to audit a minimum of 1 precinct, but it doesn’t require the precinct to have a minimum number of voters.
  • This means counties may comply with statute by auditing a single ballot. For example, during the 2024 general election, 3 of Ford County’s smallest precincts had only 1 ballot cast each. If Ford County had randomly selected one of those precincts for its audit, it appears they’d comply with state law. The Ford County election officer told us she only included precincts with more than 25 voters to avoid this issue even though it’s not required by law.

Most of Ford County’s voting age population are registered to vote and about half of registered voters in Ford County voted in the 2024 general election.

• The U.S. Census Bureau estimated that Ford County had a little more than 23,600 residents who were 18 years old or over in 2023. The Census Bureau’s data did not identify the part of that population that’s eligible to vote.

• Ford County reported about 16,650 voters (or about 70% of its voting age residents) registered for the 2024 general election. They also reported a total of 8,300 ballots cast during that election. Their voter turnout was 50%.

• The Ford County election officer said the registered voter total includes inactive and suspended voters. She said if those voters were excluded, the percentage of registered voters would be slightly lower. Officials from the Secretary of State’s office said they think 70% is typical for most Kansas counties.

Recommendations

1. The Ford County election officer should strengthen poll worker training about the importance of obtaining all voters’ signatures in the poll books.
   • Agency Response: I will come up with a checklist for checking people in for provisional ballots to make sure the proper procedure is followed.
2. The Ford County election officer should develop a process or checklist to ensure election workers check tabulation machines for ballots after election day and the final canvass.
   • Agency Response: I will make a checklist for poll workers to use to see that all the tabulator bins are cleared of all ballots election night and before sealing.
3. The Legislature should consider if state law (K.S.A. 25-3009) should be amended to require the precinct(s) selected for the post-election audit to have a minimum threshold of voters.

Agency Response

On February 12, 2025 we provided the draft audit report to the Ford County clerk’s office. Its response is below. Agency officials generally agreed with our findings, conclusions, and recommendations.

Ford County Clerk’s Office Response

The mistakes that happen are human error. We will always have human error.

I have board workers that work the in-office voting. I have always started in-office voting 19 to 20 days before Election Day and have 2 evenings and a Saturday voting times. For the General Election 2024 we had over 200 voters a day vote early. Board workers that work in-office voting attend training before they work in office. Most will also attend another training before Election Day.

I will continue to train my board workers, and I will continue to attend all trainings available to me.

All I can do is keep training and learning. I take the audit as a learning tool to see where the weaknesses are and to improve upon them.