2022 Utah Election Audit

Report No.
2022-17
A Performance Audit of Utah’s
Election System
and Controls
Office of the Legislative

Auditor General
Report to the UTAH LEGISLATURE
The mission of the Office of the Legislative Auditor General is to serve the Utah
Legislature and the citizens of Utah by providing objective and credible information, in-
depth analysis, findings, and conclusions that help legislators and other decision
makers: Improve Programs, Reduce Costs, and Promote Accountability
Audit Subcommittee
President J. Stuart Adams, Co-Chair Speaker Brad R. Wilson, Co-Chair
President of the Senate Speaker of the House
Senator Evan J. Vickers Representative Mike Schultz

Senate Majority Leader House Majority Leader
Senator Karen Mayne Representative Brian S. King

Senate Minority Leader House Minority Leader
Audit Staff
Kade R. Minchey, Auditor General, CIA, CFE
Leah Blevins, Manager, CIA
Jake Dinsdale, Sr. Audit Supervisor, CIA
Ryan Thelin, Audit Supervisor
Jake Davis, Performance Auditor
Tanner Taguchi, Audit Intern
Tyson Cabulagan, Methodologist, CFE
Office of the Legislative Auditor General
olag.utah.gov
Office of the Legislative
Auditor General
Kade R. Minchey, Legislative Auditor General
W315 House Building State Capitol Complex | Salt Lake City, UT 84114 | Phone: 801.538.1033
Audit Subcommittee of the Legislative Management Committee

President J. Stuart Adams, Co-Chair | Speaker Brad R. Wilson, Co-Chair
Senator Karen Mayne | Senator Evan J. Vickers
Representative Brian S. King | Representative Mike Schultz
December 13, 2022

TO: THE UTAH STATE LEGISLATURE
Transmitted herewith is our report:

“A Performance Audit of Utah’s Election System and Controls” Report #2022-17.
An audit summary is found at the front of the report. The scope and objectives of the audit
are included in the audit summary. In addition, each chapter has a corresponding chapter
summary found at its beginning.
This audit was requested by the Legislative Audit Subcommittee.
We will be happy to meet with appropriate legislative committees, individual legislators,
and other state officials to discuss any item contained in the report in order to facilitate the
implementation of the recommendations.
Sincerely,
Kade R. Minchey, CIA, CFE

Auditor General
kminchey@le.utah.gov
a
AUDIT SUMMARY
REPORT #2022-17 | December 2022
Office of the Legislative Auditor General | Kade R. Minchey, Auditor General
PERFORMANCE ELECTION SYSTEMS AND CONTROLS

AUDIT
AUDIT REQUEST KEY FINDINGS

1.1 Utah’s election controls mitigate the risk of fraud as long as they
The Legislative Audit
Subcommittee requested an audit are used properly.
to provide assurance that Utah’s 2.1 Mistakes within the voter registration database highlight
election systems and processes opportunities for increased oversight.
continue to be secure, fair, and
trustworthy. 3.1 Canvass ballot totals from Utah’s 2022 primary election did not
match those recorded in the central voter database.
Specifically, we were asked to
3.2 Some counties’ chain-of-custody practices make it difficult to
examine the integrity of election
processes, the accuracy of voter account for all ballots.
rolls, and the security around 4.1 Utah lacks clear legal standards for election signature
voters’ ballots. verification
5.2 Adopting additional post-election audit methods could increase
BACKGROUND confidence in election processing and outcomes.
For our purposes, we define 6.1 Utah election code does not specify oversight and enforcement
election integrity as ensuring that roles.
eligible voters can vote, ineligible
voters cannot, and election
results reflect the will of the
electorate. KEY RECOMMENDATIONS
The Lieutenant Governor is 2.1 The Office of the Lieutenant Governor should finalize standards
designated as Utah’s chief for the frequency and use of key VISTA maintenance tools, and then
elections officer but does not monitor their implementation.
administer elections or directly
maintain voter registration 3.1 The Office of the Lieutenant Governor should create rules
records. requiring county clerks to publicly reconcile the number of ballots
tabulated with the number of voters given vote credit in VISTA.
Direct responsibility for voter
3.4 The Office of the Lieutenant Governor should finalize its chain-
registration and election
of-custody manual, including best practices for election staff in
administration falls instead on
Utah’s counties.
independently elected county
clerks throughout the state. 4.1 The Legislature should consider either including clearer
standards for signature review, acceptance, and rejection in Utah
Click or scan for interactive Code, or giving the Lieutenant Governor authority to establish these
election security website standards and instructions.
5.2 The Legislature should consider establishing a risk-limiting audit
pilot program to enhance Utah’s post-election audit methods, giving
the Office of the Lieutenant Governor rulemaking authority to
establish standards.
6.1 The Legislature should consider adding election standard
oversight and enforcement responsibilities and mechanisms to Utah
Code.
a
AUDIT SUMMARY
CONTINUED
REPORT
SUMMARY
1.1 Despite Opportunities to Improve, the Office of the Lieutenant Governor (LG’s
Utah’s Election Systems and Processes Office) playing a more active role.
Work Together to Guard Election Integrity
5.2 Adopting Additional Post-Election
Our team observed and tested election systems in Audit Methods Can Enhance Audit
every county during the 2022 primary election and Outcomes
found no evidence of systematic problems,
widespread errors, or significant fraud. This is Adopting risk-limiting audit (RLA) methods
primarily because multiple layers of process and could add to Utah’s election system by shifting
defense must be defeated to undermine election the purpose of the post-election audit to
integrity in Utah. Although we believe elections are validating that election outcomes are correct.
functioning well overall, there are several risk Many states are moving to RLAs as the new
areas—discussed throughout this report—that standard for post-election audits because they
provide opportunities to strengthen Utah’s election allow election officials to adjust the number of
system in important ways. examined ballots in an audit to provide
statistical confidence in election results.
2.1 Mistakes within Voter Registration Because there are various ways to conduct
Database Highlight Opportunities for RLAs, policymakers should deliberate on
Increased Oversight whether RLA methods could be beneficial in
Utah.
Maintaining accurate voter rolls ensures that
ballots are mailed to correct addresses, eligible 6.1 Utah Election Code Does Not Specify
voters can vote, and ineligible voters cannot. We Oversight and Enforcement Roles
found that county clerks are striving to maintain
accurate records and most records we reviewed To ensure elections are more uniformly
were accurate. However, we did identify important administered we believe Utah Code could
instances where county clerks can improve the clarify and define oversight and enforcement
accuracy of voter records. We believe the need to roles for elections as seen in other states. Utah
ensure accurate records is critical to a successful Code designates the Lieutenant Governor (LG)
election and the problems we identified warrant as the state’s chief election officer.
3.1 Canvass Ballot Totals from 2022

Primary Election Did Not Match Those attempted to reconcile these discrepancies
Recorded in the Central Voter Database and had some success, but the processes in
some counties do not allow for a
Our team analyzed ballot processing statistics reconciliation. Regardless of the exact
and found several counties that had cause, these discrepancies point to ballot
discrepancies between reported ballot numbers processing and recordkeeping problems
and recorded voter totals. We that should be understood and explained.
Table of Contents
Chapter 1
Utah’s Election Controls Mitigate the Risk of
Fraud as Long as They Are Used Properly .............................................................. 3
1.1 Despite Opportunities to Improve, Utah’s Election Systems and
Processes Work Together to Guard Election Integrity ............................................................... 3
1.2 Differences in Other States’ Election Laws Can Make Comparisons Difficult .......................7
1.3 Prosecution Is Rare Because Reported Intentional Election Crimes Are Rare ..................... 8
Chapter 2
Utah’s Voter Registration Process Is Strong but
Can Be Improved with Additional Oversight ......................................................... 11
2.1 Mistakes Within the Voter Registration Database
Highlight Opportunities for Increased Oversight ...................................................................... 11
2.2 The LG’s Office Did Not Ensure Deceased-Voter
Records Were Removed as Required by Statute ....................................................................... 20
2.3 The LG’s Office Should Analyze Staffing and Backup for Critical Positions ....................... 21
Chapter 3
Consistent Chain-of-Custody Practices Are Needed to
Accurately and Transparently Account for All Ballots ........................................... 25
3.1 Canvass Ballot Totals from Utah’s 2022 Primary Election
Did Not Match Those Recorded in the Central Voter Database ............................................... 25
3.2 Some Counties’ Chain-of-Custody Practices Make It
Difficult to Account for All Ballots ............................................................................................ 29
3.3 Counties Use Physical Security Controls Inconsistently..................................................... 32
Chapter 4
Standards, Training, and Audits Could
Improve Election Signature Verification.............................................................. 39
4.1 Utah Lacks Clear Legal Standards For Election Signature Verification ............................. 39
4.2 Training Standards and Requirements Should Be Consistent Across the State ................ 42
4.3 Signature Quality in VISTA Should Be Improved............................................................... 43
4.4 Mid-Election Performance Monitoring Helps Identify and Correct Problems .................. 44
4.5 The Legislature Should Consider Policies Related to the Cure-by-Phone Process............. 46
Chapter 5
The Legislature Should Consider
New Options for Post-Election Audits ................................................................... 51
5.1 The LG’s Office Has Recently Strengthened the
Post-Election Audit but Other States Have Gone Further ......................................................... 51
5.2 Adopting Additional Post-Election Audit
Methods Can Enhance Audit Outcomes ................................................................................... 54
Chapter 6
Oversight and Enforcement Roles
In Election Code Could Be Clarified ......................................................................65
6.1 Utah Election Code Does Not Specify Oversight and Enforcement Roles .......................... 65
6.2 County Election Workers Are Not Required to Participate in Election Training ............... 68
6.3 The LG’s Office Is Working on a Manual of Standards and
Guidance for Clerks to Prepare for and Administer Elections................................................... 71
Complete List of Audit Recommendations ............................................................ 79
Appendix ............................................................................................................. 83
A. Full Detail of Election Processes and Controls Highlighted in Figure 1.1 ............................ 85
B. Voter Status and Record Type .............................................................................................. 99
C. Examples of Washington State Reconciliation Reports ......................................................103
D. OLRGC Opinion.................................................................................................................... 111
Agency Response ................................................................................................. 117
CHAPTER 1 Summary
Fraud as Long as They are Used Properly
BACKGROUND
The Legislature asked us to examine the integrity of Utah’s elections. We define election integrity as
ensuring that eligible voters can vote, ineligible voters cannot, and election results reflect the will of
the electorate.
The Lieutenant Governor is designated as Utah’s chief elections officer but does not administer
elections or directly maintain voter registration records. Direct responsibility for voter registration
and election administration falls instead on independently elected county clerks throughout the state.
RECOMMENDATION
FINDING 1.1 Recommendations can be found in Chapters 2-6.
Despite opportunities to improve, Utah’s
election systems and processes work together to guard
election integrity.
NO RECOMMENDATION
FINDING 1.2
Differences in other states’ election laws can make
comparisons difficult.
NO RECOMMENDATION
FINDING 1.3
Prosecution is rare because reported
intentional election crimes are rare.
CONCLUSION
Our audit team observed and tested election systems in every county during the 2022 primary
election and found no evidence of systematic problems, widespread errors, or significant fraud. This
is primarily because multiple layers of process and defense must be defeated to undermine election
integrity in Utah. Although we believe elections are functioning well overall, there are several risk
areas—discussed throughout this report—that provide opportunities to strengthen Utah’s election
system in important ways.
1
Chapter 1
Fraud as Long as They Are Used Properly
1.1 Despite Opportunities to Improve, Utah’s

Election Systems and Processes Work Together to
Guard Election Integrity
Our team observed and tested election systems in every county during the 2022
primary election and found no evidence of
systematic problems, widespread errors, or
Our definition of
significant fraud. This is primarily because multiple election integrity is
layers of process and defense must be defeated to to ensure that
undermine election integrity in Utah. Although we eligible voters can
believe elections are functioning well overall, there vote, ineligible
are several risk areas—discussed throughout this voters cannot, and
report—that provide opportunities to strengthen election results
Utah’s election system in important ways. Of reflect the will of
the electorate.
particular importance is the need to bolster central
oversight and control through the Office of the
Lieutenant Governor (LG’s Office), as discussed in Chapter 6.
Election Systems Can Work Together to Mitigate Risk
For this audit, we did a risk-based assessment of Utah’s elections to determine
whether controls are in place to reasonably prevent process problems, human
error, and voter fraud. Although election laws and processes work well to reduce
risk and protect election integrity, there is no way to completely prevent
problems, errors, and fraud. While we do not believe the problems and risks
identified in this audit are at a level that would significantly compromise election
integrity, they are nevertheless concerning. We believe the information in this
report can help policymakers continue to assess election risks and potential
policy responses.
A Performance Audit of Utah’s Election System and Controls 3

Because elections are overseen by the LG’s Office and run by twenty-nine
different locally elected county clerks, our team traveled
Although we believe
the state to perform firsthand interviews and observations.
elections are On primary election day in June 2022, we placed auditors
functioning well from our office in every county in the state to carefully
overall, there are review the procedures and controls of key election
several risk areas— activities. These in-person observations and discussions
discussed were invaluable in understanding the similarities and
throughout this differences in election practices from county to county.
report—that provide
opportunities to In addition, we researched how elections are managed in
strengthen Utah’s other states and reviewed best practices. Throughout our
election system in audit, we built a list of the key election processes that
important ways.
support Utah’s election integrity. In this chapter, we
review parts of that list, touching on the process functions and strengths that led
us to conclude that Utah’s election processes are sound. This chapter also
previews the rest of this audit report by discussing the risks our team identified.
Each chapter then contains recommendations to mitigate those risks. Figure 1.1
shows a summary of the key election processes we believe are critical to reducing
risk and protecting election integrity.
Figure 1.1 Our Audit Work Identified Ten Elements or Processes We Believe Work
Together to Provide Critical Support for Utah’s Election Integrity. We also created a
detailed interactive website that explains the basics of each element and includes risks and
recommendations for improvement. Readers can access the website by scanning or clicking
the QR code on the next page. Appendix A shows the same detail as the interactive version
but in static format.
Source: Auditor generated; more detail is found in an online, interactive version of this figure, and in
Appendix A of this report
4 Office of the Legislative Auditor General

The controls listed in Figure 1.1 can combine in many ways to reduce the risk of
various types of election threats. For example, election administration in Utah is
structured at the county level (see element #1 in Figure 1.1). Experts say that this
localized structure subdivides process and power in a way that makes it extremely
difficult to compromise statewide and/or national races. This is because of the
complexity of coordinating and executing such widescale fraud against so many
independent offices. On the other hand, Utah’s county clerks use a centralized,
state-level voter registration system that enables voter management across
county lines and provides central controls over ballot security (see element #2 in
Figure 1.1). While this report identifies ways to further strengthen these controls,
this blend of state and local administration leverages the strengths of both,
resulting in local control with unified state-level data and technology resources.
Figure 1.1 provides a high-level view of election

processes in Utah. We created an interactive version of
this figure to provide more details on the basics of each
control, remaining risks, and recommendations for
improvement. This figure can be accessed by scanning or
clicking the QR code. Appendix A of this report shows the
same information in a static format.
An Example Highlights Election Controls in

Action When Used Properly
If the controls in Figure 1.1 are used according to best
practices, they can go a long way to mitigate risk in Layers of controls,
Utah’s election system. A hypothetical example of an when used
attempt to commit fraud highlights how these appropriately, work
controls might work. It also helps to show that if together to mitigate
the risk of election
controls are not well used, the risk of fraud will
problems, errors,
increase. and fraud.
Imagine a fraudster who wants to steal mail ballots,
forge the rightful voters’ signatures, and cast those votes in an attempt to sway an
election.
• The fraudster must first obtain mail ballots. If clerks are maintaining voter
records, only eligible, active voters will receive mail ballots. That can
reduce the number of ballots mailed to voters who should have been made
inactive or to wrong addresses where the ballots may not be missed if
stolen. If ballots are stolen from active voters, there is a higher chance
those voters would call the county clerk to locate their ballot and the
fraudster’s plan could be detected.
• If the fraudster is unable to steal authorized ballots, creating envelopes

and ballots from scratch is very difficult because each authorized envelope

is printed with a unique control number that is in turn associated with a
specific voter and election.
• If the fraudster were to find some mailed ballots, the next step would be to
forge signatures. Utah’s signature verification process compares voters’
signatures to validated signature examples held in the state voter
database. Although signature verification is somewhat subjective and not
foolproof, this process presents another hurdle the fraudster must
overcome.1
• It is also important to note that each instance of election forgery is a felony

offense, increasing the risk of criminal prosecution.
Our view, based on extensive observation and analysis, is that the election
controls described in this hypothetical scenario mitigate risk to a level where it is
reasonable to expect that the fraud attempt will be prevented. That is not to say
that every attempt will be prevented. The controls described in Figure 1.1 must be
well executed according to best practices to achieve the full effect of their risk
mitigation.
The controls in Figure 1.1 combine in similar ways to prevent or deter various
other election fraud methods like identity theft or double voting (i.e., voting by
mail and in person). The processes can also work together to detect human errors
and process problems, allowing election officials to catch mistakes and make
needed corrections in a timely manner. For further detail about these controls,
please visit our interactive website or Appendix A of this report.
The Remainder of this Report Highlights
Areas to Strengthen Election Controls
Although Utah’s election controls work well, we do not mean to suggest that
Utah’s election system is flawless. We found situations where some of the
controls described in Figure 1.1 were not used as well or as consistently as they
could have been, creating a higher risk for election errors, problems, and fraud.
Although the actual risk exposure in the instances we found was still small, each
small deficiency highlights areas where we believe systemic improvements are
possible.
The remaining chapters of this report highlight some of these deficiencies and
our recommendations to address them, such as:
• Chapter 2: Process gaps in voter registration The remainder of
and maintenance this report
highlights areas to
strengthen and add
to election controls.
1 The quality and strength of signature verification is, itself, an example of how multiple
election processes in Figure 1.1 overlap to mitigate overall risk. Chapter 4 discusses signature
verification in more detail, as well as opportunities to strengthen pertinent controls.

• Chapter 3: Significant concerning differences in vote total reports between
different information sources
• Chapter 4: A lack of clarity in standards and training for signature

verification
• Chapter 5: Opportunities to refine the purpose and results of post-election

audits
• Chapter 6: A need for clear oversight and enforcement powers in the LG’s
Office
1.2 Differences in Other States’ Election Laws Can

Make Comparisons Difficult
Within certain federal constraints, each state has a lot of latitude to craft its own
election laws. As we developed the audit findings in this report, we found great
value in examining other states’ election laws and practices. Many of the findings
and recommendations in this report stem from policy ideas from other states.
However, we also found areas of controversy and litigation in other states that,
because the laws and practices are so different, were all but impossible to
compare to Utah’s election practices.
For example, in 2020, as the COVID pandemic caused nursing home lockdowns,
the Wisconsin Elections Commission (WEC) voted to send absentee/mail ballots
to certain nursing home facilities. In the wake of that decision, allegations were
made that voter turnout in certain nursing homes was suspiciously high and that
mentally incapacitated elderly voters were improperly influenced, leading to
fraudulent votes.
Although these issues raise interesting questions, they flow from such drastically
different laws and practices that comparison with Utah laws becomes very
difficult. For example:
• Unlike Wisconsin, Utah does not have an elections commission that is
empowered to make election decisions.
• Instead of Wisconsin’s rigid legal procedure to send absentee/mail ballots

to nursing home residents, Utah simply sends mail ballots to all active
voters.2
2The WEC is required to send “special voting deputies” (SVDs) to assist certain nursing home
residents. If, after two attempts, the SVDs are unable to do so, election officials can then send
absentee/mail ballots to those voters. During the COVID pandemic in 2020, nursing homes
refused entry to SVDs for public health reasons. This and other factors led the WEC to vote twice
to bypass the statutory SVD requirement and send absentee ballots. Faced with poor options, the
WEC later cited a fear of disenfranchising nursing home voters as the unacceptable alternative.

• Wisconsin does not verify signatures on absentee ballots like Utah does,
which means that its absentee ballot-related decisions are made in a
different context.
While there are useful lessons we can learn by examining election practices in
other states, certain areas of comparison are not useful because the fundamental
practices are so different.
1.3 Prosecution Is Rare Because Reported

Intentional Election Crimes Are Rare
In addition to examining the election controls highlighted in this chapter, we
worked to understand prosecution as both a response and a deterrent to illegal
voting activity. Our audit work in this area was comprised primarily of
conversations with county attorneys and county clerks because other sources of
information were limited. We surveyed Utah’s clerks who, generally speaking,
said that election crimes should be referred to local law enforcement or the
county attorney for investigation and prosecution. Multiple election leaders
explained that there are scenarios in each election where family members sign
each other’s ballots, whether intentionally or inadvertently, and that doing so is
illegal. However, they report that the lack of criminal intent leads officials to treat
those as educational opportunities unless an individual has shown a pattern of
such behavior.
One of the county attorneys we contacted could not recall any election-related
criminal referrals in the eight years since he took office. He did recall one case
where two individuals were convicted of misdemeanor crimes related to
gathering petition signatures.
Another county attorney could recall only two election-related cases in his fifteen
years of experience, one of which involved an individual suspected of signing a
family member’s ballot.3 Upon further investigation, the county attorney
explained, it was determined that the individual was attempting to assist a family
member who could not sign the ballot on their own, and there was no clear
criminal intent. For those reasons, the case was not prosecuted.
Though election-related crimes do occur, statements made by clerks and county
attorneys show that they identify few scenarios they believe merit prosecution.
That said, county clerks and county attorneys should remain aware and vigilant
to prosecute any attempts to thwart elections in the future.
3 The other case was related to gathering petition signatures.

CHAPTER 2 Summary
Utah’s Voter Registration Process Is Strong, But
Can Be Improved with Additional Oversight
BACKGROUND
VISTA is Utah’s statewide voter registration system used by all counties to maintain accurate and up-to-date
voter data. The Office of the Lieutenant Governor (LG’s Office) manages VISTA and provides information
and processes to update records. The county clerks are responsible to maintain the accuracy of voter records
in their counties.
RECOMMENDATION 2.1
The Office of the Lieutenant Governor should finalize
standards for the frequency and use of key VISTA
maintenance tools, and then monitor their
implementation.
FINDING 2.1
Mistakes within voter registration database RECOMMENDATION 2.2
highlight opportunities for increased oversight. The Legislature should consider clarifying the
oversight role of the Office of the Lieutenant Governor
with regards to voter registration. This could include
authority for regular analysis of voter records and
rulemaking authority for minimum maintenance
standards.
RECOMMENDATION 2.3
FINDING 2.2
The Office of the Lieutenant Governor should monitor
The LG’s Office did not ensure deceased-voter
and ensure that the names of deceased voters are
records were removed as required by statute.
removed from voter rolls, as required in statute.
RECOMMENDATION 2.4
The Office of the Lieutenant Governor should review
its staffing plan to determine whether critical
FINDING 2.3
functions have sufficient staff. If, in this review, the
The LG’s Office should analyze staffing and backup
Office of the Lieutenant Governor determines that
for critical positions.
they need more resources for monitoring and
maintenance, we recommend that they report these
needs to the Legislature.
CONCLUSION
While we found that most county election records are accurate, we also identified important instances where
the process of maintaining voter records can be improved. We believe the need to ensure accurate records is
critical to a successful election, and the problems we identified warrant the LG’s Office playing a more active
role to provide oversight and guidance for voter registration activities.
9
Chapter 2
Utah’s Voter Registration Process Is Strong but
Can Be Improved with Additional Oversight
2.1 Mistakes Within the Voter Registration Database

Highlight Opportunities for Increased Oversight
Accurate and up-to-date voter registration records are the backbone of many of
the election controls outlined in Chapter 1. Maintaining accurate voter rolls
ensures that ballots are mailed to correct addresses, eligible voters can vote, and
ineligible voters cannot. Evaluating the integrity and accuracy of Utah’s voter
rolls was one of the key objectives of this audit request. We found that county
clerks are striving to maintain accurate records and most records we reviewed
were accurate. However, we did identify important instances where county clerks
can improve the accuracy of voter records. We believe the need to ensure
accurate records is critical to a successful election and the problems we identified
warrant the Office of the Lieutenant Governor (LG’s Office) playing a more active
role.
While we did not find that these concerns led to widespread fraud, they have
resulted in a relatively small number of inaccurate voter records. The problems
we found include inconsistency in clerks’ use of built-in maintenance tools,
mismatching information within voter files, and other isolated human errors.
Therefore, greater oversight is needed to ensure Utah has accurate voter
registration. We recognize that human error is, to some extent, unavoidable,
given the complexity of the work required to maintain voter records. However,
more formal standards and procedures for voter maintenance, coupled with
effective oversight efforts from the LG’s Office, could reduce both the likelihood
and impact of human error in future elections. We recommend that the LG’s
Office establish and implement a procedure to maintain the accuracy of the
statewide voter database and use of key VISTA controls, then monitor the
implementation of that procedure to ensure it is working as intended.
The Lieutenant Governor Has the Responsibility to Oversee
Voter Registration and Implement a Maintenance Procedure
The Lieutenant Governor (LG) is named in statute as the chief elections officer in
the state, with “direct authority over the conduct… for federal, state, and
multicounty” elections.4 This authority includes the duty to “oversee all of Utah’s
voter registration activities.”5 In addition, the LG is given responsibility to
“establish and implement a procedure to maintain the accuracy of the statewide
voter registration database…”6
4 Utah Code 67-1a-2(2)

5 Utah Code 20A-2-300.6
6 Utah Code 20A-2-109(2)

This code gives the LG’s Office general oversight for registration activities. The
Standards for Internal Control in the Federal Government (the Green Book),
defines an oversight body as follows:
An oversight body oversees the entity’s operations; provides
constructive criticism…; and where appropriate, makes oversight
decisions so that the entity achieves its objectives in alignment with
the entity’s integrity and ethical values.
As part of this oversight, the Green Book further emphasizes the need for regular
monitoring to ensure that controls are functioning as intended. As risks and laws
evolve, controls must be continually evaluated to ensure they address the
evolving risks. The Green Book lists two principles for monitoring:
Standards for Internal Control in the Federal Government

Management should establish and operate monitoring activities to
monitor the internal control system and evaluate the results.
Management should remediate identified internal control

deficiencies on a timely basis.
The LG’s Office has recently taken steps toward a

The LG’s Office has more active oversight role. In 2022, the Legislature
begun implementing created a requirement for the LG’s Office to conduct
the voter an annual voter registration audit. The LG’s Office
registration audit as reports that they have developed procedures and are
passed by the
currently in the process of completing the first audit.
Legislature.
The state director of elections said that their findings
echo some of those discussed in this chapter and will
inform future training. Though these voter registration audits have only just
begun, we believe they can continue to yield valuable insight into voter
registration maintenance and training opportunities for clerks.
The LG’s Office also reports that the following processes have been implemented
or are in process:
• Increased onsite visits with clerks
• Provided bi-weekly trainings on many topics, including voter registration
• Funded Olene Walker VOTE Certification training
• Working to finalize elections manual for all election officials
We believe the steps taken so far will improve voter registration across the state.
That said, this chapter highlights additional areas where the LG’s Office can
further formalize voter registration oversight and monitoring procedures.

Utah Has Many Processes to Update Voter Records
Although the statewide voter registration system—known as VISTA7—is managed
by the LG’s Office, the direct input and maintenance of voter records are the
duties of county clerks’ offices. Figures 2.1 and 2.2 show how the LG’s Office and
clerks interact with VISTA, voter records, and some of the key information
sources used to keep records accurate and up to date.
Figure 2.1 VISTA Is Maintained by the LG’s Office, but Voter Records Are Maintained
by County Clerks. Day-to-day activities are done by county clerks. The LG’s Office maintains
VISTA’s functionality and uploads data from other agencies.
ounty le s
LG s Office VISTA
aintains functionality
of VISTA and u loads o ides otential
otential atc es f o ote atc es f o ecei es info ation
a ious sou ces fo S L and in uts into VISTA
cle s to u date ote and it in VISTA
eco ds oes not edit
ote eco ds dates ote eco ds e
Syste used to aintain
ote eco ds e o al add ess c an e
na e c an e etc ased
on info ation f o LG s
Office o t e ote
Source: Auditor generated, based on Utah Code, VISTA, and discussions with county clerks and the LG’s
Office.
Note: DHHS (Department of Health and Human Services) provides data regarding deceased
individuals; data from DLD (Driver License Division) allows for identity verification; and UDC (Utah
Department of Corrections) provides a list of incarcerated individuals.
Maintenance of voter data is an ongoing task that requires constant vigilance as

voters move, die, get married, temporarily move away to school, etc. Our audit
found that these ongoing tasks are done inconsistently from county to county.
Improved oversight from the LG’s Office would help
counties maintain voter records more consistently.
Maintenance of
Additional oversight would involve developing clearer voter data is an
maintenance procedures as required in statute and ongoing task that
working to ensure that counties are adhering to those clerks conduct. We
procedures. found that these
tasks should be
The state’s voter registration system is more than a done more
warehouse of voter data. Election officials we consistently and
interviewed emphasized the functionality of VISTA as with more
a key safeguard of election integrity. In a structure oversight.
that balances state and local control, the LG’s Office is
required by law to upload multiple state data sources that clerks use to review
and maintain voter records. In addition to state-level data, VISTA has been
programmed with numerous tools to help county clerks maintain accurate voter
7 VISTA stands for Voter Information and State Tracking Application. It is a centralized
system, managed and maintained by the LG’s Office. The LG’s Office grants access rights to clerks
and their election staff who access VISTA remotely to input and maintain voter data from their
respective counties.

records. Figure 2.2 illustrates the main data sources that are fed into VISTA and
the system’s primary maintenance tools.
Figure 2.2 VISTA Is a Hub of Information Sources and Analytical Maintenance Tools.
The LG’s Office is required by law to upload multiple streams of data that enable clerks to maintain
voter records. Clerks can also run multiple maintenance tools to identify and correct voter records.
VIST Data Sources

loaded to VISTA y LG s Office ocessed y le s Offices
Ot e State Vote
eceased Vote ata Acti ity ata Inca ce ated
ta e a t ent of ealt and lect onic e ist ation Vote ata
u an Se ices Info ation ente I ta e a t ent of o ections
e e States
VIST Maintenance Tools

o a ed into VISTA sed y le s Offices
Annual aintenance Vote

u licate Vote tility
tility I Ve ification Tool
Includes iffe ent ue ies
t at Identify otential u licate Inacti ates Vote s o a e Ve ifies Vote ata a ainst ta
Vote eco ds ot a tici ated in ulti le i e License i ision and
ede al lections Social Secu ity Ad in ata ases
Source: Auditor generated based on VISTA and discussions with county clerks. This is not an exhaustive list
of maintenance resources in VISTA.
Note: ERIC is a non-profit governed and managed by member states, whose purpose is to improve the
accuracy of voter rolls through shared data.
To evaluate the accuracy of Utah’s voter database, including the use of the data
and tools depicted in Figure 2.2, our audit team did the following tests:
• Analyzed records to identify potential duplicate registrations or vote
credit.
• Evaluated compliance with laws that require timely removal of deceased

voters from eligible voting status. Our analysis included death records
received directly from the Department of Health and Human Services.
• Compared DLD and SSA data against a sample of voter records to check
accuracy and to determine whether clerks are using the voter ID
verification tools in VISTA.
• Compared voter history statistics against official canvass reports to

determine whether official vote totals matched the vote totals recorded in
VISTA.

These audit tests found issues that need to be
Audit tests found
addressed. We did not find any clear patterns that
issues in voter
would suggest that certain errors were more common records that need to
or that certain counties were consistently generating be addressed but
more errors than others were. Rather, the nature of most of the issues
the issues we discovered points to human error and appear to be caused
can be remedied with improved guidance and by human error.
oversight. Most records in Utah’s voter database are
accurate, and most of the controls in place are well used. Details about our
analysis and findings are included later in this chapter.
Built-in Voter Database Maintenance Tools
Are Not Being Used Consistently
VISTA offers numerous tools to help maintain an accurate database. These
include checks for duplicate voters, voters registered in another state, and
deceased voters. The tools are designed to keep voter rolls accurate. However, we
found that the usage of these tools varied from county to county.
Two Counties Have Not Performed Voter Status Updates for More
Than Three Years. Clerks should run the annual maintenance utility in VISTA
to update voter statuses. The utility accomplishes two tasks: (1) it changes voters
who have not voted in two consecutive federal elections to inactive status, and (2)
it changes voters who have not voted for four
consecutive elections (while in inactive status) to
Two counties have removable status.8 These are important distinctions,
not moved voters to because neither inactive nor removable voters are
inactive status,
mailed a ballot. The two counties that have not
likely leading to
ballots being mailed performed this task have likely been mailing ballots to
to individuals who individuals who should not be receiving them. In
may not be eligible. addition, individuals who should be categorized as
removable—and therefore ineligible to vote—are still
categorized as inactive, meaning they could go to a vote center and successfully
vote. Figure 2.3 shows how long it has been since counties used the annual
maintenance utility as of August 2022.
8 This annual maintenance utility was shown in Figure 2.2. For additional information on
voter statuses, see Appendix B.

Figure 2.3 Two Counties Have Not Used the Annual Maintenance Utility in More
Than Three Years. One county last used the annual maintenance utility in 2014, and the
other did so in 2018. More than half of the counties have performed the annual processing
within the last year.
16 17
14
12
Number of Counties
10
10
8
2
1 1
0
Within a year Within two years Over three years Over seven years
Year Since Maintenance was Last Performed
Source: Auditor analysis of reports within VISTA
Twenty-seven out of twenty-nine counties have used the annual maintenance

utility within the last two years. Some counties run the
process only after they complete federal elections. Of Without guidance,
greatest concern are the two counties (blue bars in expectations, and
Figure 2.3) that have not run the process in more than monitoring, counties
three years. One county last ran the process in 2014, may not complete
and the other did so in 2018. These are processes that important processes
have been built into VISTA to help ensure an up-to- to ensure accurate
date and accurate voter database. Thus, monitoring by voter records.
the LG’s Office would help ensure that counties
perform key processes as intended. Without guidance, clear expectations, and
monitoring, counties may not complete important processes to ensure an
accurate voter registration database.
Use of Voter Identity Verification Tool Needs Additional Guidance.
The identity verification tool was created to verify the identity of voters and
ensure the registrant is a real person. It is built into VISTA and can be used to
quickly match information. Election staff must manually perform the check and
verify that the information is accurate. For example, if a Utah driver license
number is provided as part of the voter registration application, information can
be matched to current information on file with the DLD. This matching tool can
also match to SSA data by using social security numbers.

To test the use of this tool, we sampled 870 records to compare the voter
registration information against data from the DLD and SSA.9 We found that the
identity of nearly 79 percent had been verified using the built-in tool. The
remaining 21 percent of records reviewed may have been verified using other
means. Among all the voter records we reviewed, we found that eight percent of
all records contained mismatching information, as summarized in Figure 2.4.
Figure 2.4 Our Sample of 870 Records Found That Eight Percent of Records Had
Information That Did Not Match. We found that 4 percent were not able to be matched
with SSA or DLD information.
No Match 4.0%
Wrong DL # 1.4%
Name Mismatch 1.4%
SSN Mismatch 1.0%
Citizenship Status
0.7%
Mismatch
DOB Mismatch 0.2%
0% 1% 2% 3% 4% 5%
Source: Auditor analysis of reports within VISTA
Note: Numbers will add to more than 8 percent because some records had more than one mismatch.
Citizenship status mismatches were records that returned a warning that the voter may not be a citizen.
These errors do not necessarily mean the registered voter was not a citizen, but they do highlight the
need for further investigation.
We found that thirty-five records (nearly 4 percent) within the sample could not
be matched at all, and an additional six (0.7 percent) returned an error warning
that the individual was not a citizen. Currently there is not a clear procedure to
guide clerks when information is mismatching.
Ensuring accurate information in voter records is critical to maintaining an up-
to-date voter file. If information is not accurate, removing duplicate and deceased
records is more difficult. For example, we found a potential case where the voter
record of a deceased individual was not removed from the voter file for seven
months after their death. We believe this occurred because the voter record did
not contain the individual’s legal name or other identifying data, making the
match with death records difficult.
There should be more oversight and instruction for clerks regarding steps to take
if information does not match. Additional guidance is needed to provide a clear
process for who should be verified, and when. While our sample of records shows
9 This was not a representative sample and cannot be used to extrapolate to the entire
population.

that a large percentage of voter records are accurate, and that the identification
verification tool is being used, there is room for improvement. The mismatches
do not necessarily mean the record is incorrect, but it may show a need for
further investigation. We believe improved guidance and oversight can enhance
the consistent use of the ID verification tool.
Most Counties Are Using Duplicate and Out-of-State Voter Checks In
VISTA. Voter data from other states is uploaded to VISTA to help clerks identify
voters who may be registered and voting outside of Utah. Registering to vote
elsewhere would be grounds to remove that voter from Utah’s voter file. There
are also tools within VISTA designed to help clerks review and consolidate
duplicate voter records. Figure 2.5 shows the number of counties, highlighted in
blue, that regularly use these tools.
Figure 2.5 Most Counties Appear to Regularly Use Maintenance Tools Within VISTA.
One county did not regularly check VISTA for duplicate records and nine counties did not
regularly check for voters registered in another state.
VISTA u licate Vote VISTA e iste ed in Anot e
e ist ation ec State ec
ounties Sufficiently ec ounties Sufficiently ec
Source: Auditor analysis of VISTA
As Figure 2.5 indicates, most counties are using VISTA’s maintenance tools.
However, one county did not regularly use the duplicate registrations check,
resulting in more than 100 potential duplicates (1.4 percent of active registered
voters in that county). Furthermore, seven counties did not appear to utilize the
check of voters registered in other states, which is concerning.10 Not regularly
using these tools increases the risk of voters being able to vote in multiple states
or having duplicate registrations, which could allow them to vote multiple times.
After the duplicate voter tool flags voters who are potentially registered more
than once, election officials must manually review each record and take action.
10 Utah is a member of Electronic Registration Information Center, Inc. (ERIC). ERIC is a
non-profit funded and managed by member states. The goal is to improve the accuracy of voter
rolls. One way this is accomplished is by sharing voter rolls to check for individuals registered in
multiple states; thirty-three states and Washington, DC are members.

This process can lead to human error. Our reviews found that a small percentage
of duplicate registrations were potentially active and had yet to be corrected.
Although the verification process largely appears to be working, there are still
duplicate registrations that are not identified. Currently there is no formal
procedure regarding how frequently counties should check for duplicate
registrations and out-of-state voters; nor is there any documented guidance on
how to use these tools. Establishing procedures and monitoring their
implementation would help ensure that key maintenance checks are done
properly.
Although the newly created annual voter registration audits will provide the LG’s
Office insight into how these tools are used, formalizing procedures would help
clarify expectations for clerks’ offices and help the LG’s Office better fulfill its
voter registration oversight duties.
RECOMMENDATION 2.1
The Office of the Lieutenant Governor should finalize standards for the
frequency and use of key VISTA maintenance tools, and then monitor their
implementation.
Additional Statewide Monitoring Could

Help Remedy Errors Auditors Found
In conjunction with formalized voter maintenance procedures, Utah needs better
statewide monitoring of VISTA, which is a key control for the election system. As
described in the previous section, our audit found inconsistent processes and
implementation in counties, leading to inaccurate voter records. We believe that
with enhanced monitoring, the LG’s Office should be able to identify and remedy
similar issues. In addition to inconsistent practices, Utah’s voter registration
utilizes manual processes that rely heavily on individual judgment which in turn
can lead to human error. While human error is not preventable, monitoring and
checks in the system can help catch, reduce, and correct errors. The following is a
summary of additional errors auditors found while reviewing VISTA:
• One county’s interpretation of code led officials not to remove the names
of deceased voters during an election. This led to a cast vote that was
potentially fraudulent.11
• One county had incorrectly removed eligible voters while updating its
voter registration records. 12
11 This county was made aware of this issue. They immediately investigated and plan to
forward the case to the appropriate prosecutor.

12 The errors we found have since been corrected.

• Voter history vote totals do not match totals from official county canvass
reports, (further explained in Chapter 3). While this problem must be
corrected, the differences did not exceed 0.4 percent of ballots cast.
• Duplicate voter registrations were not removed.
• Two voters were incorrectly merged into one record.
It is important to recognize that no election system or process is perfect, and

sufficient monitoring and oversight may not have prevented the issues we found.
However, our concern is that there is no formal monitoring process to identify
issues on a continual basis. Without sufficient monitoring, there cannot be proper
communication and training to improve the implementation of controls. Chapter
6 of this audit report provides further detail on necessary steps to provide better
oversight of the election system.
While the issues and concerns we identified within the voter registration
database must be improved, we found only isolated issues related to potential
fraud. We believe most of these issues were caused by human error, emphasizing
the need for better oversight and monitoring.
RECOMMENDATION 2.2
The Legislature should consider clarifying the oversight role of the Office of
the Lieutenant Governor with regards to voter registration. This could
include authority for regular analysis of voter records, and rulemaking
authority for minimum maintenance standards.
2.2 The LG’s Office Did Not Ensure Deceased-Voter

Records Were Removed as Required by Statute
The LG’s Office is statutorily responsible for ensuring that voter records for
deceased individuals are removed from the voter rolls.13 The office currently
provides county clerks potential matches of death records from the Department
of Health and Human Services (DHHS) through
VISTA and is working on getting death records from
We found that the
SSA. At present, the LG’s Office is not following up on
LG’s Office does not
“ensure that all this information to ensure that county clerks remove
deceased voters the names of deceased voters from the rolls. Utah
have been removed Code requires the Lieutenant Governor to compare
from the official voter rolls ninety days prior to an election to ensure
register” as required that clerks have removed the names of all deceased
by Utah Code. voters. We found that the LG’s Office has not been
providing this oversight.
13 Utah Code 20A-2-306(7)

Utah Code requires that the county remove the names of all deceased voters
from the voter database within five days of when the clerk’s office receives
notice.14 Although it appears that twenty-eight counties had removed the names
of deceased voters as required, one county had not done so for nearly four
months. As of September 1, 2022 there were over 90 records that needed to be
reviewed for removal in that county, including individuals who died prior to the
2022 primary election. Because the 90-day requirement for the 2022 general
election was August 10, 2022, the LG’s Office should have ensured that the
county removed those voters before we identified them in September.
In addition, we reviewed all DHHS death records from February 1, 2022 to
August 31, 2022 to see if all had been removed from voter rolls. We found that,
potentially, over 250 deceased voters in counties across the state were still on
voter rolls, amounting to roughly 2 percent of all deceased individuals in a seven-
month period. While there was no evidence that any of these individuals had
voting credit in VISTA, our findings highlight the importance of the LG’s Office
ensuring deceased voters are removed from voter rolls.
RECOMMENDATION 2.3
The Office of the Lieutenant Governor should monitor and ensure that the
names of deceased voters are removed from voter rolls, as required in
statute.
2.3 The LG’s Office Should Analyze

Staffing and Backup for Critical Positions
Managing the statewide voter registration system is essential to the integrity of
elections in Utah. We are concerned that the LG’s Office lacks sufficient back-up
and staffing for the management of VISTA. Currently, the LG’s Office relies on
one very knowledgeable individual to manage this database. Any entity that has
questions about VISTA turns to this person. This creates a heavy burden for this
person with no backup support. The LG’s Office reports that it has recently hired
an additional staff position to provide backup VISTA support.
In total, the LG’s Office has only five staff dedicated to elections. We are
concerned that this is not a sufficient level of staffing to provide the needed
monitoring and oversight for Utah’s election system. Compared with Utah,
surrounding states have more staff in their comparable agencies that oversee
elections. The surrounding states had staffing ranging from 21 to 50 individuals
working on elections.15 While some of these states have larger populations, we
believe the LG’s Office should review its level of staffing. Insufficient staffing may
make it difficult to fully implement the oversight recommendations in this
14 Utah Code 20A-2-305(3) “The county clerk shall remove a voter’s name from the official
register within five business days after the day on which the county clerk receives confirmation
from the Department of Health’s Bureau of Vital Records that the voter is deceased.”
15 The surrounding states included here are Colorado, California, Oregon, and Washington.

chapter to improve voter registration processes. In addition, the LG’s Office has
experienced high turnover in election staffing. This is concerning given the loss of
institutional knowledge of the election system. As addressed in the previous two
sections, the LG’s Office could provide additional statewide monitoring to ensure
that counties are performing voter roll maintenance and identify human error.
The LG’s Office may need additional dedicated staff, such as data analysts, who
can dedicate time to conducting further analysis to isolate anomalies.
If the LG’s role is expanded to provide additional monitoring, additional staff
may likely be needed. The LG’s Office should ensure sufficient staffing and
backup for critical positions managing the VISTA database.
RECOMMENDATION 2.4
The Office of the Lieutenant Governor should review its staffing plan to
determine whether critical functions have sufficient staff. If, in this review,
the Office of the Lieutenant Governor determines that they need more
resources for monitoring and maintenance, we recommend that they report
these needs to the Legislature.

CHAPTER 3 Summary
Consistent Chain of Custody Practices Are Needed to
Accurately and Transparently Account for All Ballots
BACKGROUND
Every ballot returned to a county clerk must be accounted for. The generally accepted methods to do so are
referred to as a chain of custody and are essential to conducting transparent and trustworthy elections.
Accurate chain-of-custody practices and records increase confidence that all ballots are processed correctly.
Such practices also protect against the unauthorized removal or addition of ballots.
RECOMMENDATION 3.1
The Office of the Lieutenant Governor should create rules
requiring county clerks to publicly reconcile the number of
FINDING 3.1 ballots tabulated with the number of voters given vote credit in
Canvass ballot totals from Utah’s 2022 VISTA.
primary election did not match those
recorded in the central voter database. RECOMMENDATION 3.2
The Office of the Lieutenant Governor should help clerks identify
the best data possible to reconcile canvass reports and vote credit
numbers.
RECOMMENDATION 3.3
FINDING 3.2
The Office of the Lieutenant Governor should create rules
Some counties’ chain-of-custody practices
requiring ballot envelopes to be counted as early as possible during
make it difficult to account for all ballots.
ballot processing.
FINDING 3.3 RECOMMENDATION 3.4

Counties use physical security controls The Office of the Lieutenant Governor should finalize its chain-of-
inconsistently. custody manual, including best practices for election staff in Utah’s
counties.
CONCLUSION
Our findings show that chain-of-custody practices should be improved to increase the accuracy and
transparency of ballot processing in Utah. Adopting chain-of-custody standards for tracking and reconciling
canvass reports to vote credit numbers could address the issues identified in this chapter.
23
Chapter 3
Consistent Chain-of-Custody Practices Are
Needed to Accurately and
Transparently Account for All Ballots
3.1 Canvass Ballot Totals from Utah’s 2022

Primary Election Did Not Match Those
Recorded in the Central Voter Database
Every ballot returned to a county clerk must be accounted for. Our team analyzed
ballot processing statistics and found several counties that had discrepancies
between reported ballot numbers and recorded voter totals. Such discrepancies,
when no explanation is provided, create a concern that ballots were either
counted without a link to a voter, or that a voter received vote credit but the clerk
did not count the ballot. We attempted to reconcile these discrepancies and had
some success, but the processes in some counties do not allow for a
reconciliation. Regardless of the exact cause, these
discrepancies point to ballot processing and
recordkeeping problems that should be understood Chain of custody
and explained.16 refers to the paper
trail and processes
The generally accepted methods to account for ballots that document the
are collectively referred to as a chain of custody and transfer of election
are essential to conducting transparent and materials from one
trustworthy elections. Accurate chain-of-custody person or place to
practices and records increase confidence that all the next.
ballots are processed correctly. Such practices also
protect against the unauthorized removal or addition of ballots. Policymakers
should consider options to require public reconciliation and reporting of these
key statistics as better proof of accurate ballot processing.
After the 2022 Primary Election, There Was a Discrepancy Between
Canvass Reports and Voter Registration Records in Several Counties
After the 2022 primary election, there was a mismatch between ballots counted
and voter credit assigned in 22 counties. The official count of ballots is provided
in each county’s certified canvass reports. When a clerk processes a valid ballot,
they should record vote credit for that voter in the statewide voter registration
database (VISTA).17 This vote credit in VISTA represents the authoritative record
that a given voter’s ballot was processed and counted. The vote credit is also used
16 The total number of mismatching records represented 0.4 percent of total ballots processed
in the 2022 primary election.

17 VISTA stands for Voter Information and State Tracking Application and is managed by the
Office of the Lieutenant Governor (LG’s Office) as discussed in Chapter 2.

to limit any efforts by that individual to vote again. The number of vote credits
should match the number of ballots reported in the canvass report.18
In the months following the 2022 primary election, data in many counties
showed a mismatch between canvass totals for ballots processed and voters
receiving credit in VISTA. In most counties, the deviations were not large; 9 had a
difference of five ballots or fewer, and seven counties matched perfectly.
Of Utah’s 29 counties, 13 reported more ballots processed than voters who
received vote credit in the election. This creates a concern because records show
1,031 more ballots than voters in these locations. Conversely, another 9 counties
reported fewer ballots processed than voters receiving credit in the election. This
creates the appearance of 854 votes without an associated ballot in these 9
counties. The five counties with the largest discrepancies accounted for 88
percent of all differences. Figure 3.1 highlights these numbers.
Figure 3.1 Some Counties Reported More Ballots Than Voters; Others Reported
Fewer Ballots than Voters. County canvass reports show 13 counties had 1,031 more
ballots processed than the number of voters receiving credit in VISTA. This was 0.4 percent of
all ballots processed in those 13 counties. A sample of four counties helps illustrate these ballot
discrepancies.
Source: Auditor analysis of VISTA vote data and canvass reports for the 2022 primary election
18 Because the Utah Constitution grants the right of a secret ballot, clerks separate each
ballot from its envelope after vote credit is recorded. After separation, clerks cannot trace specific
ballots back to specific voters. There are rare cases when voters are given vote credit but then
there is a problem with the ballot, making it impossible to scan. In such a scenario, a discrepancy
would exist between vote credit and the ballot scanned because the clerk cannot remove vote
credit from the affected voter. Regardless, the clerk should be able to explain such a discrepancy.

As Figure 3.1 shows, some counties reported more ballots in their canvass report,
while others recorded more vote credits in VISTA. In counties where the canvass
report shows more ballots processed than VISTA vote credit, it is unclear where
the additional ballots came from. In counties where VISTA shows more vote
credit than ballots processed in the canvass report, it is unclear why those ballots
were apparently not counted after vote credit was
given.
These findings offer
While we could not resolve county discrepancies, a clear reason for
those clerks we contacted offered multiple possible policymakers to
explanations. For example, one county discovered a consider options to
clerical error in its canvass report. The corrected create more
version shows more ballots processed and the county accountability and
transparency in
reports that it eliminates the discrepancy entirely.
ballot processing.
Officials from other counties said they may have
simply failed to record vote credit for in-person voters
on election day or for voters who cured challenged signatures. Regardless of the
underlying causes, we believe these findings offer a clear reason for policymakers
to consider policy responses such as those we explore in the next section. The
Office of the Lieutenant Governor (LG’s Office) is currently working to draft a
chain of custody rule that may address these issues.
The Director of Elections Should Establish Rules Requiring All County
Clerks to Publicly Reconcile Ballot Processing Statistics
In 2022, the Legislature empowered the director of elections within the LG’s
Office to establish administrative rules regarding the custody, documentation of
custody, handling, processing, disposition, and tabulation of ballots.19 In light of
our findings, we believe that a public reconciliation of key ballot processing
statistics should be included in the new election rules. Doing so would increase
the transparency and accountability of ballot processing throughout the state.
For example, current laws in the state of Washington20 require a public
reconciliation of ballots. At the conclusion of each election, counties must report
a one-page, high-level summary accounting for all ballots mailed, received, and
processed. Counties also reconcile total ballots processed to the vote credit given
in their central voter database. A space is provided on the report for counties to
explain any discrepancies between these key ballot processing statistics. Figure
3.2 provides an example of this process, and Appendix C contains examples of
some of these Washington reports.
19 See Utah Code 20A-3a-404, enacted by House Bill 313 in the 2022 Legislative General
Session.
20 See Washington State Statute 29A.60.235 (2).

Figure 3.2 King County, Washington Was Able to Reconcile Its Ballots in the 2022
Primary Election. The county processed more ballots than the entire state of Utah in the
2022 primary election. Even so, King County was able to reconcile ballots by identifying errors
and other potential explanations for disparities between ballots and vote credit. Each county in
Washington is required to submit a reconciliation report to the secretary of state.
Source: Auditor generated, based on ballot reconciliation data from King County, Washington, as
reported to the Washington secretary of state for the 2022 primary election
As our team observed the 2022 primary election process, we saw multiple county
clerks in Utah making significant efforts to internally track and reconcile some of
these key statistics. We believe that a requirement to publicly report and
reconcile ballot statistics—similar to the method used in Washington—would be
beneficial and would not require major process change in Utah counties. Further,
some of the potential policy changes to post-election audits discussed in Chapter
5 require strong chain-of-custody record keeping. Doing the work to track and
report key ballot statistics could put counties in a better position to employ
additional post-election audit methods and could give voters confidence that each
valid vote is counted and tracked appropriately.21
We also acknowledge that having accurate data is essential to correcting these
problems. VISTA offers multiple reports that change if, for example, voters move
or die after casting a vote. In our audit work, it took multiple attempts to identify
data sources that accurately reflect vote credit in each county. We believe that the
LG’s Office, as the steward of VISTA, should help clerks identify or create the best
data possible to track and reconcile canvass reports and vote credit numbers.
This information should then be included in the standards manual recommended
in Chapter 6.
RECOMMENDATION 3.1
The Office of the Lieutenant Governor should create rules requiring county
clerks to publicly reconcile the number of ballots tabulated with the number
of voters given vote credit in VISTA.
21 See Chapter 5 for audit findings and recommendations related to post-election audits.

RECOMMENDATION 3.2
The Office of the Lieutenant Governor should help clerks identify the best
data possible to reconcile the number of ballots tabulated with the number of
voters given vote credit in VISTA.
3.2 Some Counties’ Chain-of-Custody Practices Make It

Difficult to Account for All Ballots
A key part of chain-of-custody documentation is reconciling the total number of
ballots received with the final disposition of all ballots. Doing so can help ensure
that all ballots received are processed to valid outcomes and that no ballots are
removed or added along the way. Although some county clerks have developed
internal methods to do this, we observed that some counties’ practices do not
yield enough information to affirmatively demonstrate that their chain-of-
custody practices are functioning. In addition to Finding 3.1, we believe these
insights could further inform the LG’s Office as they enact and refine
administrative rules pertaining to chain of custody practices.
The Timing of Batch Formation Can
Impact Chain-of-Custody Records
While it is critical to establish an overall total of all ballots received as a
reconciliation tool, it is also important to track ballots as they move through each
step of processing. Clerks accomplish this by organizing the total number of
ballots received into more manageable portions, called batches. The size of the
batch depends on county size and preference. The most common batch size we
observed is 50 ballots. However, larger counties create batches of up to 250
ballots.
The timing of organizing ballots into batches can impact chain-of-custody
records. We observed that clerks differ in when they choose to create batches.
The two most common practices are batching envelopes prior to signature
verification, or batching after signature verification. However, a few counties wait
until after ballots are separated from the envelope before they create batches.
Figure 3.3 displays the various points at which counties created batches during
our observations of the 2022 primary election.

Figure 3.3 Counties Create Batches at Different Points in Ballot Processing. The gray
boxes indicate the different points at which Utah counties batch ballots during an election
process. The variation in timing may contribute to the difficultly in reconciling the number of
ballots processed with the number of voters receiving credit in VISTA.
Source: Auditor generated
Batching or counting later in the process may prevent accurate tracking of each
ballot through the process. Counting the total number of ballot envelopes later in
the process can also lead to a missed opportunity to establish the initial number
received, undermining later efforts to reconcile ballot processing statistics.
Some western states require that ballot counts be established in the initial phases
of ballot processing. For example, Colorado requires that envelopes be counted
and batched immediately after they arrive. Oregon requires all envelopes to be
counted as they are received.22 We recommend that the chain-of-custody rule
created by the LG’s Office also consider establishing an early timeframe for
counting ballot envelopes to better track ballots through the election process.
RECOMMENDATION 3.3
The Office of the Lieutenant Governor should create rules requiring ballot
envelopes to be counted as early as possible during ballot processing.
Some County Clerks’ Ballot Tracking and Recordkeeping

Practices Lack a Full View of Ballot Processing
Many counties use tracking sheets to account for ballot envelopes as they move
through the election process. These forms are designed to establish ballot counts
and inform election workers about ballot status at different stages in the election
process. Figure 3.4 shows examples of tracking sheets for ballot envelopes, and
the associated batch sizes, within three different counties.
22 Like Utah, both Colorado and Oregon use vote-by-mail as their primary election method.

Figure 3.4 Counties Produce Their Own Tracking Sheets for Incoming Ballot
Envelopes. The examples below, from three different counties, show the variation in batch
quantity, procedures, and whether election staff must sign off on each step.
Source: County cle s’ offices
County tracking sheets are created in-house and indicate differences in batch
size, the procedures tracked, and whether an election worker must sign off on
each step. For instance, if a hypothetical batch begins with 200 ballots before
signature verification, and 3 signatures are challenged, the ballots with
challenged signatures are removed and placed in the curing process. The batch
will then have 197 ballots, and the reason will be noted on the batch tracking
form. The next person to receive the ballots then confirms that 197 ballots are
present, thus continuing the chain-of-custody records.
However, during the 2022 primary election, we
observed some counties in which the chain-of-custody
During the 2022
practices were not adequate to show that all ballots primary election, we
had been accounted for. observed some
counties in which
County Example 1—In one county, we found that the chain-of-custody
totals in certain ballot tracking categories (e.g., practices were not
envelope not signed, signature did not match) did not adequate to show
match across different reports. After further that all ballots had
investigation, the clerk’s office was unable to explain been accounted for.
the differences or provide the correct totals. Election
staff in the same county would deliberately break the chain of custody partway
through the process by dissolving batches in order to create new, uniform batches
of 200 ballots to run through the ballot scanners. County officials explained that
they wanted to simplify the tabulation step. The county acknowledged that,
without a clear chain of custody from start to finish, any errors or deviations
would be difficult to track and correct.
County Example 2—Another county counted envelopes as they arrived at the
election office but did not batch them into specific quantities or maintain a
tracking sheet. Rather, the county placed signature-verified envelopes into a large

intake pile to be opened. Once opened, the ballots were loosely organized into
batches of 50 for tabulation. However, no tracking sheet accompanied the
envelopes or ballots throughout the election process.
County Example 3—A third county did not maintain any chain-of-custody
tracking methods during the election. It placed collected envelopes in a large
intake pile and verified signatures at random intervals.
After being separated from envelopes, various
quantities of ballots were moved at random to election Without clear and
machines for tabulation. No batching was conducted accurate processing
until after ballots were processed on tabulation statistics, it is
equipment. incredibly difficult to
demonstrate that all
Without clear and accurate processing statistics, it is ballots were
incredibly difficult to demonstrate that all ballots were processed correctly.
processed correctly without physically reviewing
archived election materials within these counties. This
falls short of the accurate and transparent processing statistics we believe can
and should be reported.
3.3 Counties Use Physical Security

Controls Inconsistently
The state’s 29 counties used various physical security measures during the 2022
primary election. For example, counties use different amounts of staff to conduct
various processes during the election. Additionally, counties use different
security measures such as camera monitoring of election rooms, locked rooms to
store envelopes and ballots, and safety locks and seals on election equipment to
avoid tampering. Any additional use of physical security measures is a policy
decision for the Legislature and state election officials.
Physical Handling of Ballots Varies by County
Our observations found that the handling of ballots also differed among counties.
For example, many counties attempted to employ two election staff members for
each step of the election process and follow other chain-of-custody best practices.
However, we also observed concerning practices in other counties, such as:
• One county rotated people in charge of the ballot tabulating machine. In
most instances, this resulted in a single person tabulating ballots at
different intervals. When a new person began tabulating ballots, they
simply continued from the same stack of ballots. No formal organization
for counting of ballots was observed.
• In another county, a single person collected ballot drop boxes at multiple

locations. This individual loaded the small ballot boxes into their vehicle
and delivered them to the county election office. The ballot drop boxes
were secured with two small luggage locks.

• The same county had a single person organizing and verifying signatures
on ballot envelopes. Once verified, the ballots are separated by a different
person. Ballots were no longer tracked or
organized once separated from the envelope.
The LG’s Office can
The Legislature established some chain-of-custody standardize
criteria with House Bill 313 (H.B. 313) in the 2022 expectations of
Legislative General Session. For example, the bill county staff to
requires that two staff members be present when enhance the
collecting ballots from a drop box or polling location— security and
a change that was to be implemented in the November transparency of
elections.
2022 general election. Additionally, ballots are to be
recorded and tracked until they are delivered to the
election office.
Even so, other practices conducted by election staff were based on individual
preferences or the resources available. The LG’s Office could standardize
expectations for county staff to improve the security and transparency of
elections. The LG’s Office is in the process of drafting a chain-of-custody rule that
could improve some of these weak controls by requiring that two staff members
be present any time ballots are handled. This includes ballot collection,
processing, adjudication, and replication.23 The LG’s Office has also reported that
it intends to provide a best practices manual for counties to reference. These
additions are timely as many new clerks take office in Utah’s election system.
RECOMMENDATION 3.4
The Office of the Lieutenant Governor should finalize its chain-of-custody

manual, including best practices for election staff in Utah’s counties.
23 In ballot replication, election workers attempt to recreate a ballot if it is damaged or not
able to be processed on tabulation machines.

Physical Security Measures Appear to Be Strong in
Most Counties but Vary According to Resources
Counties utilized a variety of physical security measures such as locked election
rooms, security cameras, multiple locks on ballot
The passage of H.B.
drop boxes, and secure election equipment. Larger
313 in the 2022 counties have more resources and may have more
Legislative General physical security measures in place. Some smaller
Session requires all counties may lack resources to maintain similar
municipalities to physical security standards at the scale of what is seen
have a ballot drop in larger counties. Individual county councils should
box. Security consider the needs of their election offices and
measures depend on
provide the appropriate funding and resources to
whether the box is
attended or ensure efficient and secure elections.
unattended. The passage of H.B. 313 in the 2022 Legislative
General Session established new standards for
physical election security. The bill broadened access to ballot drop-boxes by
requiring each municipality to have a drop box. Figure 3.5 shows an example of
the various drop boxes used in different counties.
Figure 3.5 Ballot Drop Boxes Differ from County to County. For the 2022 primary
election, external drop boxes were secured with multiple locks and camera monitoring. Smaller
counties often used smaller drop boxes that could be monitored by individuals.

H.B. 313 established requirements for cameras on drop boxes, and we observed
that clerks were working to comply. Another common security practice was the
use of cameras in the election rooms. Although not a requirement in law, some
counties used cameras to monitor the election process. Some cameras captured
the entirety of the room and processes, while others had only limited coverage.
For example, we observed a small county with a camera in the election room that
monitored every step other than ballot tabulation. We also observed counties that
did not use any cameras to monitor ballot processing.
It is a policy question for the Legislature and state election officials to determine
if camera monitoring should be required in county election offices.
Election Equipment and Software Appeared to Be Adequately
Protected and Properly Functioning During the Election Process
We examined election machines and found that they employ various physical
security features to deter tampering or manipulation. The machines have locks
and sealable access doors on data ports to prevent and detect any unauthorized
access. Election machines also require administrative passwords to ensure that
only authorized individuals have access, and the machines can produce a real-
time log of all activities performed. Election machines are not connected to the
internet. Auditors examined election equipment in multiple counties and did not
find any evidence of election machines being connected to the internet during our
observations.
Ballot scanners are designed to scan and read ballots only.24 The tally of votes
from scanners is then transferred either by specially programmed flash drives or
direct connection to election computers where election results are tabulated.
These election computers are not connected to the internet.
The external data sticks, as well as the election software and equipment, all come
from external vendors. Utah Code25 requires the LG’s Office to ensure that
election equipment is certified to federal Election
Assistance Commission (EAC) standards. The State of
Utah maintains a contract with Election Systems and s part of a pre-
Software (ES&S), a private company that specializes election test, all
in election machines and software. Twenty-seven of election equipment
the state’s twenty-nine counties utilize ES&S appeared to be
functioning properly
equipment and software. Private vendors are an
26
and accurately
essential component in Utah’s election system, identified vote
providing equipment, software, and technical support selections on test
to individual counties. ballots.
24 Twenty-seven of the twenty-nine counties in Utah use Election Systems and Software
(ES&S) brand ballot scanners. One county uses Dominion Election Software, and another uses
Unisyn Voting Solutions. These two systems use off-the-shelf office scanners that read ballots
using proprietary software installed on the counties’ election computers.
25 Utah Code 20A-5-802
26 Salt Lake County uses Dominion Election Software. Beaver County uses Unisyn Voting
Solutions. Both are private companies that use EAC-certified equipment.

As a further safeguard, counties are required to conduct a logic and accuracy
(L&A) test of election machines in the days prior to an election. An L&A test
checks the accuracy of the voting equipment before official election ballots are
processed. Election officials ensure the machines are cleared and displaying a
zero count before the test. They then run a series of test ballots on the machines
to verify the accuracy and proper functioning of the equipment. The L&A tests are
open and can be observed by the public. We observed a sample of L&A tests. The
process was conducted on machines from all three of the private companies that
provide the state’s election equipment. All election equipment appeared to be
functioning properly and accurately identified vote selections on the test ballots.
Although we did not conduct a cybersecurity penetration test on Utah’s election
systems, we did inquire about the controls that are currently in place. In addition
to the county-level controls, the Department of Public Safety’s Statewide
Information and Analysis Center (SIAC) described a mix of federal, state, and
local efforts to monitor and mitigate election threats. We were informed that
election stakeholders gather regularly to review mock scenarios involving both
physical and cybersecurity threats. SIAC also reported proactive efforts by
themselves and state entities—like the Department of Technology Services, State
Bureau of Investigation, LG’s Office, and the National Guard—to identify and
assess election threats. We observed the work done by Utah’s cybersecurity
experts on election day in November 2022 and saw efforts, including
collaboration with cybersecurity teams across the country, to proactively monitor
for election-related threats.

CHAPTER 4 Summary
Standards, Training, and Audits Could
Improve Election Signature Verification
BACKGROUND
Signature verification is one of many controls designed to ensure that eligible voters can vote, and ineligible
voters cannot. Although signature verification alone is not perfect, when combined with other election
controls, its effectiveness increases.
RECOMMENDATION 4.1
The Legislature should consider either including
FINDING 4.1
clearer standards for signature review, acceptance, and
Utah lacks clear legal standards for election signature
verification. rejection in Utah Code, or giving the Lieutenant
Governor authority to establish these standards and
instructions.
RECOMMENDATION 4.2
The Legislature should consider giving the Office of the
Lieutenant Governor explicit authority to establish
training standards and requirements for signature
FINDING 4.2 verification to better assure that practices are executed
Training standards and requirements should be consistently across the state.
consistent across the state.
RECOMMENDATION 4.3
The Legislature should consider requiring county
election staff to attend signature verification training.
RECOMMENDATION 4.4
FINDING 4.3 The Legislature should consider the options in this
Signature quality in VISTA should be improved. chapter to improve the quality of signatures
available for signature verification.
RECOMMENDATION 4.5
FINDING 4.4 The Legislature should consider giving authority to the
Mid-election performance monitoring helps identify Lieutenant Governor to establish rules requiring
and correct problems. counties to conduct mid-election audits of signatures
and performance tracking for signature reviewers.
RECOMMENDATION 4.6
FINDING 4.5
The Legislature should consider guidelines for, or the
Election workers approved challenged signatures by
necessity of, curing challenged ballot signatures by
phone without confirming voters’ identities.
phone.
CONCLUSION
Other states have clearer legal standards for signature verification, highlighting opportunities to clarify Utah
standards and to standardize training guidelines across the state. We believe the Legislature and election
officials could collaborate to consider policy options to improve the signature verification processes.
37
Chapter 4
Standards, Training, and Audits Could Improve
Election Signature Verification
4.1 Utah Lacks Clear Legal Standards

For Election Signature Verification
Of the eight states that conduct vote-by-mail elections, seven use signature
verification as a control. Our audit found that Utah’s guidelines and standards for
signature verification are not as clear as those used in other states. Adopting
clearer standards in Utah could clarify the purpose and function of signature
verification and improve the consistency of signature
verification processes for county clerks’ offices Utah does not have
statewide. clear guidance or
standards for voter
Signature verification is one of many controls signature
designed to ensure that eligible voters can vote, and verification.
ineligible voters cannot. Although signature
verification alone is not perfect, when combined with other election controls, its
effectiveness increases.
Utah Code Does Not Set Sufficient Standards
For Ballot Signature Verification
Utah Code and Administrative Rule provide limited guidance for how election
officials should evaluate and validate ballot signatures. Statutory language
regarding signatures on mail ballot envelopes is limited to a requirement that
election workers determine whether “signatures correspond,” but statute does
not provide clear guidance for how to make that determination.27
Utah Code offers comparatively more guidance for verification of petition
signatures than for ballot signatures. Statute instructs petition reviewers to
consider other evidence, such as the handwritten name and address on the
petition, in addition to whether the signature appears “substantially similar.” 28
However, similar to ballot signatures, Utah Code does not provide clear
guidance for how to determine if a signature is “substantially similar” to the
reference image. In our discussions with Utah clerks, we found that the lack of
guidance has contributed to slightly different assumptions and approaches
related to ballot signature verification.
Other states that conduct elections by mail have more specific instructions in
code and administrative rule to guide signature verification. As shown in Figure
4.1, the additional guidance addresses evaluation criteria, baseline assumptions,
27 See Utah Code 20A-3a-401(3) & (5)

28 See Utah Code 20A-7-206.3(2)

and references more instructions produced by secretaries of state (who serve as
state elections directors).
Figure 4.1 Other Vote-by-Mail States Provide More Specific Guidance for Signature
Verification. Other states have more explicit signature verification guidance in statute or rule
than what is seen in Utah Code and Administrative Rule. In addition, some authorize the
secretary of state to create additional guidelines.
Source: Law and rule from each state mentioned
In our discussions with, and extensive observation of, Utah clerks, we found that
the lack of guidance provided in law and rule has led to slightly different
assumptions and approaches as to how and when signatures should be accepted
and rejected. Observed differences include:
• Some counties believe that they are looking for reasons to reject
signatures, while other counties are looking for reasons to accept.
• Some counties train signature verifiers more rigorously than others.
• Some counties require county election staff to review signatures, while

others allow signatures to be reviewed by volunteers or other county
workers.
Utah Code and Administrative Rule are silent on the overarching guiding
principle of signature verification and lack standardized instructions.

RECOMMENDATION 4.1
The Legislature should consider either including clearer standards for

signature review, acceptance, and rejection in Utah Code, or giving the
Lieutenant Governor authority to establish these standards and instructions.
This Report Highlights Areas for Improvement That

Could Positively Impact Signature Verification
The efficacy of signature verification improves when clerks use existing controls
in the central voter registration system (VISTA).29 This is because VISTA helps
ensure that voters who receive mail ballots are eligible, active, and living at the
address on file with the clerk. Figure 4.2 summarizes how counties use the
existing VISTA controls.
Figure 4.2 There Is Variety in How Counties in Utah Use Controls for Signature
Verification. When used properly, signature verification works in concert with other controls
to protect election integrity.

*Utah Code 20A-3a-202(2)
When used correctly, the controls preceding signature verification can help
ensure that active voters receive a unique ballot, ballots are mailed to the proper
location, and valid comparison signatures are on file. As shown in Figure 4.2, not
all counties use these controls consistently.
29 VISTA stands for Voter Information and State Tracking Application and is managed by
the Office of the Lieutenant Governor (LG’s Office) as discussed in Chapter 2.

4.2 Training Standards and Requirements Should Be
Consistent Across the State
There are no standards in Utah Code or
Administrative Rule that require training for
signature verification.30 Even so, we observed some of There are no
the same techniques for signature comparison being standards in Utah
taught at different clerks’ poll worker trainings, due to Code or
some level of collaboration between counties. There Administrative Rule
that require training
were some differences observed in the conduct and for signature
formality of training from county to county, as shown verification.
in Figure 4.3. These differences likely stem from the
lack of consistent training standards.
Figure 4.3 Signature Verification Training Is More Rigorous in Some Counties Than
Others. Prior to ta ’s 2022 primary election, auditors observed different levels of formality in
signature verification training, shown here from more to less rigorous.
In some counties, signature verification training is more established and

rigorous, while in others, training practices did not appear to be firmly
established. In one county, the training essentially consisted of an instruction
sheet posted on the wall of the office where election workers were reviewing
signatures. In another county, the clerk scheduled formal trainings that included
group practice and discussion.
The LG’s Office has provided some training resources for clerks and has made
efforts31 to promote good signature verification practices. However, the LG’s
Office has no legal authority to establish or require specific elements of signature
verification training. In contrast, three vote-by-mail states—Nevada, Washington,
and Colorado—have explicit provisions in statute or rule to provide training on
statewide standards and requirements.
Additionally, federal guidance emphasizes the importance of training. Staff
responsible for signature verification have to be trained, and training should
provide examples of different signature characteristics and allow time for study
and hands-on practice.
30 The need for better system-wide training is addressed in Chapter 6.

31The LG’s Office hosted a virtual meeting of Utah clerks in January 2022 that featured a
training presentation on signature verification by the Davis County Clerk/Auditor’s Office. A
recording of the training is available to clerks for future reference.

During a 2022 legislative committee meeting, the director of elections from the
LG’s Office spoke in support of standardizing and clarifying signature verification
standards, as well as standardizing training for election workers who are often
“left to their own devices.” To ensure that signature verification statewide
adheres to best practices, the LG’s Office should establish standards and required
training for election offices across Utah’s 29 counties.
RECOMMENDATION 4.2
The Legislature should consider giving the Office of the Lieutenant Governor
explicit authority to establish training standards and requirements for
signature verification to better assure that practices are executed
consistently across the state.
RECOMMENDATION 4.3
The Legislature should consider requiring county election staff to attend

signature verification training.
4.3 Signature Quality in VISTA Should Be Improved

Most of Utah’s voters register to vote through the Driver License Division (DLD).
This means that the signature image for many voters is taken from the digital
signature pads at DLD. The DLD signatures we observed were lower quality than
other types of signature files, primarily because of their low resolution.
Federal best practices state that the quality of signature images in registration
databases plays a major role in accepting or rejecting ballots. This is important,
because rejecting otherwise valid signatures due to a bad reference image costs
the election office time and money in the curing process.

Figure 4.4 Some Utah Counties and Hawaii Offer Options for Improving VISTA
Signature Images.32 The options below are strategies suggested by or observed in individual
Utah counties—and in Hawaii—to improve signature images available for comparison.
RECOMMENDATION 4.4
The Legislature should consider the options in this chapter to improve the
quality of signatures available for signature verification.
4.4 Mid-Election Performance Monitoring Helps

Identify and Correct Problems
Auditing a sample of signatures while conducting the election can provide
immediate information to fix inappropriate decisions during the review of voter
signatures. For example, in our review, we found one signature that the county
said should not have been approved by the human verifier. Implementing a mid-
election audit could help discover and correct inappropriately reviewed
signatures and prevent the issue from recurring in the same election. Such an
audit could play a useful role because it would introduce a second review of some
signatures that are currently reviewed only once because they are passed in the
first tier of review.
To enable such an audit, the decisions of individual signature reviewers would
need to be tracked. This would allow officials to identify reviewers who may be
accepting or rejecting signatures outside of the normal distribution, and plan
training efforts to correct observed deficiencies. Colorado established mid-
election audits in its secretary of state’s election rules, which require county
clerks to periodically audit signature verification judges. If the audit finds
irregularities, the county clerk must retrain or remove judges from the signature
verification process.
32 The options in Utah are practices that were observed in different county election offices or
suggested by county clerks. The option in Hawaii was a statewide effort.

Mid-Election Signature Audits and Performance
Tracking Can Be Done by Hand or with Software
Some election machines have software that can track the pass/rejection rates of
individual signature reviewers. While such equipment can make performance
tracking easier, performance tracking can still be done in counties where this
technology is not available. Federal best practices recommend machine audits to
assure that humans would reach the same conclusions as the machines; in the
absence of machines, a supervisor team could examine batches throughout the
day to look for outliers.
When asked about the viability of manual signature performance tracking and
subsequent mid-election auditing, county clerks in smaller counties said it could
be done with some additional effort. Many small counties already have a practice
of labeling batch tracking forms with the initials of the individual who reviewed
the batch. A mid-election audit could then be done by tracking the performance
of the specific reviewer identified on the batch tracking form.
Figure 4.5 Some Counties’ Batch Tracking Forms Include Initials That Identify
Signature Reviewers. Batch tracking forms with e ie e s’ initials could allow election
managers to track the signature review decisions of each poll worker and, thus, target training
needs.
Some Utah counties have independently developed their own signature

verification processes. For example, one county we observed reported that it
audits a minimum of 1,000 machine-verified signatures during an election.

RECOMMENDATION 4.5
The Legislature should consider giving authority to the Lieutenant Governor

to establish rules requiring counties to conduct mid-election audits of
signatures and performance tracking for signature reviewers.
4.5 The Legislature Should Consider Policies Related to

the Cure-by-Phone Process
Statute establishes a cure process that requires voters to confirm their identity
and submit a new signature when their ballot envelope signature is challenged by
the county clerk. However, it also creates an exception to this process if election
workers can “otherwise … confirm the individual’s identity.”33 Operating under
this exception, some county clerks validate challenged ballot signatures by
contacting voters by phone.
Auditors observed some county election workers approving challenged ballot
signatures over the phone without confirming the
individual’s identity, which is not consistent with Auditors observed
Utah Code. The use of phone curing is inconsistent some county
across counties, and the process can be problematic election workers
because there are no clear standards for how the approving
process can be done in a satisfactory way. challenged ballot
signatures over the
Of the eight states that conduct elections primarily by phone without
mail, only Nevada has explicit statutory instructions confirming the
and requirements for confirming voter identity and individual’s identity.
verbally curing over the phone. Unlike Utah, Nevada
34
includes specific instructions that detail a requirement for voters to verbally

provide personal data to confirm their identity, after which the voter can confirm
the validity of their ballot signature to the clerk.
Federal best practices for administering and securing elections describe a cure
process that includes the submission of a new signature before a challenged
ballot can be counted. Both Nevada’s and Utah’s laws include an alternative to
providing a new signature to cure a ballot. Nevada’s statute is written such that
either a new signature or a verbal confirmation after an identity check is
sufficient to comply with cure requirements. Utah statute initially requires the
submission of a new signature but later allows the option of providing ID
confirmation through other means (e.g., cure by phone), thus negating the legal
requirement to provide a new signature.
33 The process is described in Utah Code 20A-3a-401, with the exception in subsection (8).
34 Although “phone curing” is not explicitly mentioned, Nevada Revised Statutes 293.269927
allows a clerk to contact voters with challenged ballots over the phone and provides instructions
for orally confirming identity and confirming that a ballot signature belongs to the confirmed
voter

Other States Require the Submission of a New Signature. An official at
the office of Washington’s Secretary of State noted that voters cannot affirm the
validity of a ballot signature over the phone and, in the case of mismatched
signatures, a new signature must be provided. For the cure process in California,
voters are required to deliver a signed signature verification statement to election
officials, either by email, fax, ballot drop box or in person at a polling place. Other
states have similar requirements.
RECOMMENDATION 4.6
The Legislature should consider guidelines for, or the necessity of, curing
challenged ballot signatures by phone.

CHAPTER 5 Summary
The Legislature Should Consider New
Options for Post-Election Audits
BACKGROUND
Counties in Utah conduct both a ballot audit and a signature audit at the end of each election. The purpose of
the ballot audit is to confirm that election equipment and software correctly identify voter selections. While
this audit method fulfills its purpose, the introduction of risk-limiting audit (RLA) methods could shift the
purpose to validate election outcomes.

The LG’s Office has recently strengthened the post- The Legislature should consider options to increase the
election audit but other states have gone further. independence of any post-election audit.
RECOMMENDATION 5.2
The Legislature should consider establishing a risk-
limiting audit pilot program to enhance Utah’s post-
FINDING 5.2 election audit methods, giving the Office of the
Adopting additional post-election audit methods Lieutenant Governor rulemaking authority to establish
could increase confidence in election processing and standards.
outcomes.
RECOMMENDATION 5.3
If the Legislature authorizes a risk-limiting audit pilot
program, the Office of the Lieutenant Governor should
create rules to establish common procedures.
CONCLUSION
15 States Have Established RLA Methods in Recent Years
Policymakers and election
officials could consider
introducing an RLA option
in Utah’s election system.
RLAs are designed to

confirm that the candidate
who received the most
votes won the election
contest.
Different RLA options are

available depending on a
range of considerations,
such as the desired
confidence level, resources
required, and the
appropriate state
oversight.
49
Chapter 5
The Legislature Should Consider
New Options for Post-Election Audits
5.1 The LG’s Office Has Recently Strengthened the

Post-Election Audit but Other States Have Gone Further
Following each election in Utah, county clerks conduct
audits of both ballot machine accuracy and voter lthough Utah’s
signature verification in all 29 counties. The post- current method for
election audit method currently used in Utah is simple post-election audits
and widely used across the country, with roughly half is commonly used, it
of states conducting a similar type of ballot audit.35 could be
strengthened by
That said, the simplicity of this audit method leads to
using techniques
some disadvantages the Legislature may consider from other states.
improving.
For example, Utah’s current ballot audit examines whether machines identified
the correct vote selection but does not provide a direct confirmation that
tabulation was accurate and that the candidates with the most votes won in each
race. Additionally, current post-election audits examine only a clustered sample
of ballots rather than a broader, more random sample across all election returns.
After the 2022 primary election, we observed post-election audits in four counties
and saw that election machines correctly identified vote selection on ballots. The
Office of the Lieutenant Governor (LG’s Office) reports that no errors were found
in the other 25 counties’ ballot audits of the 2022 primary election. The LG’s
Office has also recognized the need to enhance the post-election audit process
and created more stringent requirements for the 2022 general election. In
addition, we believe that certain practices in other states offer a range of options
the Legislature could adopt to further enhance post-election audits in Utah.
The Traditional Post-Election Audit Method Fulfills Its
Objective but Lacks the Ability to Validate Election Results
The purpose of Utah’s current ballot audit method is to confirm that election
equipment and software correctly identified voter selections on sampled ballots.
A review of four counties’ traditional audits verified that election machines and
software fulfilled this purpose. As such, we do not believe that the traditional
post-election audit used in Utah is fundamentally broken, but it can be improved.
Addressing specific disadvantages of the current audit method, as listed below,
could improve public perception of election results.
35 Although ballot machine audits are common, audits of signature verification are less so.

• Current audit methods verify that election equipment and software
identify the correct voter selections on a small sample of ballots. However,
the audit does not confirm that the candidates with the most votes won the
election.
• Utah’s method calls for a sample of 1 percent or 1,000 ballots (whichever is

smaller). This can lead to very small samples in counties where only a
small number of ballots are received. For example, three of Utah’s smallest
counties audited two to nine ballots each after the 2022 primary election.
• With the current method, ballots to be audited are selected using a

random sample of batches. This type of clustered sample does not give a
broad representation of ballots across many batches.
Figure 5.1 displays how counties organize ballots into many batches. The
traditional post-election audit only examines a few batches in each county.
Figure 5.1 Only a Few Batches of Ballots Are Reviewed in the Traditional Post-
Election Audit. ta ’s t aditional audit et od equi es t at counties select one percent of all
ballots by batch. This results in reviewing only a small portion of t e county’s total nu e of
batches.
Source: Auditor generated 36
Because this audit method only examines a select few batches, it is not intended
to recount votes or verify election results. The purpose of Utah’s current audit
method is simply to verify that the election machines were operating correctly
during the processing of the audited batches. The purpose of the risk-limiting
36 For illustrative purposes, Figure 5.1 utilizes the most common batch quantity of 50 ballots.
Some counties batch in numbers as high as 250 ballots, while others may have batches with
varying ballot quantities. However, the method for the post-election audit is the same regardless
of batch size.

audit methods, discussed later in this chapter, is to validate that the candidate
with the most votes won an election.
The Current Post-Election Ballot Audit Tests Election Machines
To Determine If Identified Vote Selections are Correct
In Utah, specific practices for the post-election ballot
audit were established by a 2006 directive from the
The purpose of
LG’s Office, prior to the now common method of mail- Utah’s current ballot
in ballots.37 The audit directive requires that counties audit is to ensure
review 1 percent, or 1,000 (whichever is less), of ballots that election
to verify that tabulation machines correctly read the tabulation machines
marks voters made on their ballots. This audit is and software
conducted soon after an election and is open to the correctly identified
public. individual voters’
selections.
The purpose of Utah’s current ballot audit is to ensure
that election tabulation machines and election software correctly identified the
voter’s selection. For example, if a voter marks a vote for Candidate A on a ballot,
the audit is meant to verify that the machine correctly identified the selection for
Candidate A. In our observation of four counties’ post-election audits, we found
that election machines and software identified the correct voter selections in each
of the 1 percent or 1,000 ballots examined.38
Utah’s method of auditing a set percentage of ballots is the most common method
used by states to conduct post-election audits. This method is known as a
traditional post-election audit and represents a simple, valid way to verify the
accuracy of tabulation machines in identifying voter selections on ballots.
The Legislature Added an Audit of Voter Signatures
In 2017 as the State Moved Toward Mail-in Voting
In 2017, the Legislature added a requirement that the post-election audit begin
reviewing voter’s signatures on mail-in ballots.39 Similar to the ballot audit, the
signature audit reviews one percent of the signatures approved by election
workers or verification machines to ensure that the signatures match the voters’
reference signatures on file.
In our review, we observed that the signature audit appeared to verify that
signatures on the ballot envelopes matched signatures in the VISTA voter
database. Although the post-election signature audits we observed appeared to be
satisfactory, we had other questions about the overall signature verification
process that we discuss in greater detail in Chapter 4.
37 Statute is unclear regarding the LG’s Office having oversight for election audits. Even if no
changes are made to post-election audit practices, statute could be clarified to better support
current practices.
38 Three counties used ES&S election machines and software. One county used Dominion
election software.
39 Utah Code 20A-3a-202 (9).

The LG’s Office Recently Took Steps to
Enhance the Post-Election Audit Procedure
To address some of the areas highlighted above, the
The LG’s Office
LG’s Office recently implemented more detailed established more
election audit procedures for the November 2022 detailed election
general election. The LG’s Office noted that requiring audit procedures for
more rigorous standards for the audit process can the November 2022
enhance public perception of election results. Some of General Election.
the new guidelines include:
• A minimum of 50 ballots and signatures must be examined, regardless of
whether this number exceeds the 1 percent requirement in some counties.
• The audit must be conducted by dedicated staff who are free from other
responsibilities and distractions. At least two staff must be present for the
audit.
• The signature audit will also include a verification of the voter’s

registration information in the statewide VISTA database.
We believe these new requirements build on the existing process for post-election
audits and strengthen audit expectations for election officials. However, there are
additional audit methods used in other states that policymakers should consider
to further enhance post-election audits in Utah.
The Legislature could bolster confidence in election outcomes by considering
requiring that an independent third party perform any post-election audits. Audit
standards require that the entity that performed the work cannot audit the same
work. One option for the Legislature to consider is to have counties audit each
other. Another option to increase audit independence would be to have an
independent third party, such as the State Auditor’s Office, conduct the audit.
RECOMMENDATION 5.1
The Legislature should consider options to increase the independence of any

post-election audit.
The second half of this chapter focuses on various options policymakers could
adopt to improve post-election audits and increase public confidence in election
processes and results.
5.2 Adopting Additional Post-Election Audit

Methods Can Enhance Audit Outcomes
Adopting RLA methods could add to Utah’s election system by shifting the
purpose of the post-election audit to validating that election outcomes are
correct. RLAs are conducted after an election to verify, statistically, that the

candidates with the most votes won. Many states are moving to RLAs as the new
standard for post-election audits and allow election officials to adjust the number
of examined ballots in an audit to provide statistical confidence in election
results. Many states have recently adopted versions of RLAs in their election
systems to boost confidence in elections. Because there are various ways to
conduct RLAs, policymakers should deliberate on whether RLA methods could be
beneficial in Utah.
The Purpose of Risk-Limiting Audits Is to Confirm
That the Candidate with the Most Votes Won
Risk-limiting audits are relatively new and focus on verifying that the candidate
who received the most votes won an election contest.
An RLA can be defined as “a method to ensure that at the end of
the canvass, the hardware, software, and procedures used to tally
votes found the real winners.” 40
One study noted that RLAs are an innovative method with a specified chance of
confirming a correct outcome. This audit method could shift the purpose toward
election outcomes rather than Utah’s current traditional audit which focuses on
whether election machines identified the correct voter selection.
RLAs may improve the sample of audited ballots
because they do not rely on a set percentage of ballots n RL is an
being reviewed—like the one percent requirement in examination of an
Utah’s current post-election audit. Rather, an RLA adjustable number
of ballots that stops
allows an adjustable number of ballots to be reviewed
when there is strong
based on the margin of victory. For example, if there is evidence that a full
a large margin of victory, fewer ballots need to be recount would
reviewed. However, if the margin of victory is close, confirm the same
election officials can review more ballots to ensure outcome.
statistical confidence that the correct candidate won.
In short, an RLA is an examination of an adjustable
number of ballots that stops when there is strong evidence that a full recount
would confirm the same outcome.
RLAs require that a confidence level be established to determine how many
ballots need to be examined to meet the desired confidence level. For example, if
the confidence level is set at 90 percent, it would mean that there is a 10 percent
chance that election results are incorrect. A higher confidence level (e.g., 95
percent) would suggest a lower risk (e.g., 5 percent) that the election results are
incorrect. Thus, more ballots would need to be examined if the confidence level is
higher. The desired confidence level for an RLA would need to be determined by
policymakers and election officials.
40M. Lindeman M. and P. Stark, “A Gentle Introduction to Risk-limiting Audits,” IEEE Security and Privacy
10, no. 05 (March 2012)

Risk-Limiting Audits Select Many Ballots
From Multiple Batches in an Election
RLAs also differ from the traditional audit because they examine individual
ballots from multiple batches, rather than selecting a few batches for
examination. The traditional audit method of examining a few batches as a
cluster does not pull from the larger universe of ballots and does not account for
final election results. For example, an examined batch may come from a specific
precinct that leans heavily toward a particular candidate; however, even though
that batch favors one candidate, it does not confirm that the candidate won the
election.
Conversely, an RLA method allows a random sample across the broader universe
of all ballots received by a county, not just batches. A review of ballots from many
batches allows for a more diversified examination of the entire election. As
mentioned, election officials may adjust the size of the sample of ballots to meet
the statistical confidence level desired to verify the correct outcome of an
election. Figure 5.2 shows a simplified example of how individual ballots may be
pulled from many batches in an RLA.
Figure 5.2 Risk-Limiting Audits Review Ballots from Many Batches. RLAs pull random
ballots from many batches to gain statistical confidence that the correct candidates won the
election. The number of ballots examined can be adjusted until the confidence level is met.
Examining Physical Paper Ballots Is

Necessary in Risk-Limiting Audits
Utah’s election system is conducted entirely on paper. However, many counties in
Utah rely on election software to provide a scanned image of ballots when they
are processed in tabulation machines. During post-election audits, most election
officials review only the scanned image of the ballot to determine if the machine
correctly identified the voter’s intended selection.
For an RLA, best practices state that election auditors examine physical paper
ballots to reduce the risk of potential software programming errors that may not
correctly identify the voter’s intent. Some election experts question the reliance
on election software or scanned images—which cannot be guaranteed to be

perfect—and suggest that election audits be conducted using the official paper
record.
In our post-election audit observations in four counties in the days following the
2022 primary election, only one of the four counties examined physical paper
ballots as part of its audit process. The other three counties conducted the audit
by examining ballot images on screens, aided by election software to indicate
voter intent.
Although rare, it is possible for election software to identify nonexistent vote
selections. For example, in 2020, one county noted in its audit report that an
election machine had incorrectly identified a shadow on a ballot as a vote
selection. The ballot image was initially viewed on a computer monitor rather
than the actual paper ballot.
If policymakers determine to incorporate RLAs into Utah’s election system,
counties should be required to examine physical paper ballots during the audit.
Additionally, a strong chain-of-custody standard would be necessary in each
county to locate paper ballots needed for the RLA. Chain-of-custody methods are
discussed in Chapter 3.
Risk-Limiting Audits Can Provide Greater
Assurance that the Candidate with the Most
Votes Won Because the
purpose of an RL is
Because the purpose of an RLA is to validate the to validate the
winners of the election, using this method as the post- winners of the
election audit method may help to improve public election, using this
perception of election results. For example, one method may help
county council addressed a letter to the Legislature public perception of
and LG’s Office, asking for a better audit process due election results.
to some confusion about audit results. The confusion
centered around a lack of clarity regarding current procedures for tabulation and
ballot sampling. While the county council stood by the election results, council
members argued for different audit methods to encourage faith and integrity in
elections.
While one batch of ballots is not indicative of the county’s total election results,
the current audit process resulted in a lack of confidence about the election
outcome in that county. Using an RLA option may clarify election results by
providing statistical confidence regarding who won an election contest.
Several States Have Moved toward
Risk-Limiting Audits in Recent Years
Several states have introduced various RLA methods in recent years to boost
public confidence in election systems. For example, Colorado’s RLA is meant to
“ensure effective election administration and public confidence in the election
process” and “provide a more effective manner of conducting audits than
traditional audit methods in that risk limiting audit methods typically require

only limited resources for election races with wide margins of victory while
investing greater resources in close races.” Some states have established election
RLAs in statute. Others have introduced a pilot program in select counties, while
some states give counties the option of whether to conduct an RLA.
Figure 5.3 shows which states currently use an RLA as a method for post-election
audits, along with the year RLAs were introduced.
Figure 5.3 Fifteen States Introduced Risk Limiting Audits for Elections within the
Last Five Years. Eight of these states established an RLA pilot program for select counties.
Source: Auditor generated from National Conference of State Legislatures (NCSL) data
Requirements for an RLA differ by state. Some states, such as Colorado, require
each county to produce a cast vote record (CVR), which follows each ballot
through the election process. Other RLA methods may not require a CVR and
simply pull a random sample of ballots from multiple batches. Additionally, the
statistical confidence level varies between states. For example, Colorado initially
required a 91 percent confidence level (i.e., 9 percent risk that the outcome is
incorrect) but then increased the confidence level to 97 percent (3 percent risk
that the outcome is incorrect) as counties became more comfortable performing
the RLA. Conversely, Oregon has a lower confidence level of 90 percent (10
percent risk that the outcome is incorrect). Many states also examine the physical
paper ballots as part of the RLA.
As shown in Figure 5.3, eight states are currently testing an RLA pilot program.
This allows select counties within these states to perform an RLA and become

comfortable with the guidelines and procedures. For example, Indiana initially
started its RLA pilot program with five counties in the 2020 general election. The
state has expanded the number of participating counties to ten in the 2022
general election. Utah policymakers and election officials may wish to establish
an RLA method that balances the state’s unique needs and goals for the election
system.
Policymakers and Election Officials Should Deliberate on Which
Risk Limiting udit Options Could be Used for Utah’s Election System
There are various ways to conduct RLAs to
n RL option that accommodate the needs of policymakers and election
balances the officials. There may be limits to what is feasible and
preferred confidence efficient for Utah elections; balancing the preferred
level with the confidence level with the resources needed to execute
resources needed to audit procedures is a policy decision for the
execute audit Legislature. Election officials likely would need to
procedures is a policy
study options for RLAs and consult with statisticians
decision for the
Legislature. and election audit experts to produce an RLA model
appropriate for Utah’s elections. The LG’s Office could
then create standards to support counties in conducting RLAs.
The resources necessary to conduct RLAs differ based on the method chosen.
Some types of RLAs require that election officials maintain a cast vote record, or
CVR. The CVR is a record of the order and location of all ballots scanned during
an election. An RLA that uses a CVR must have a record of the order in which
ballots were scanned on tabulation machines during an election. However, RLAs
can also be conducted without the use of a CVR, instead relying on a random
sample of ballots. The two most common methods for an RLA are:
• Ballot Polling RLA—Ballots are randomly selected (polled) from
multiple batches. These RLAs can be conducted without a voter list and
are often used with margins of victory that exceed 10 percent.
• Ballot Comparison RLA—Individual, specific ballots are selected and

compared against a CVR. This option requires a strong chain of custody
because the exact location of every ballot is needed. It is used for margins
of any size, including closer margins of victory.
Figure 5.4 shows that RLA options can be determined on a range that balances
different methods, confidence levels, and the desired level of state control.

Figure 5.4 Policymakers Could Deliberate on an Appropriate RLA Option for Utah
Elections. RLA options can be determined on a scale, depending on the resource
requirements and confidence level policymakers desire to balance. Any option will require the
Legislature to codify RLA requirements and the LG’s Office to establish RLA procedures.
Some states do not require an RLA in every election contest. Rather, they give
counties the option to perform an RLA if the election contest is close. For
example, some states set the basic RLA requirement in statute, with detailed
procedures found in administrative rule. These options are provided to assist
policymakers and election officials to determine the specific practices and
methods for an RLA method in Utah. The appropriate practices for a potential
RLA are a decision for the Legislature.
Although Utah’s current traditional audit practice confirms that election
equipment correctly identified vote selections on ballots, the inclusion of an RLA
method could give Utah counties additional confirmation that election results are
correct. Additionally, an RLA pilot program could gradually introduce this audit
method to balance local clerks’ limited resources and to help inform future audit
considerations.
We recommend that the Legislature consider the inclusion of an RLA pilot
program in statute. The Legislature could determine the amount of detail to
include in code, such as specific procedures, confidence level, county-level
options, and other areas. If, however, code merely establishes an RLA
requirement, the LG’s Office could further study RLA options and produce
specific procedures in Administrative Rule, in line with statute.
RECOMMENDATION 5.2
The Legislature should consider establishing a risk-limiting audit pilot

program to enhance Utah’s post-election audit methods, giving the Office of
the Lieutenant Governor rulemaking authority to establish standards.

RECOMMENDATION 5.3
If the Legislature authorizes a risk-limiting audit pilot program, the Office of

the Lieutenant Governor should create rules to establish common
procedures.

CHAPTER 6 Summary
In Election Code Could Be Clarified
BACKGROUND
The Lieutenant Governor is the state’s chief election officer, responsible for overseeing election
administration in the state. Like other states’ chief election officers there are several duties spelled out in
Utah Code, but unlike other states, in Utah, there is currently no entity responsible for ensuring that
election controls are used.
RECOMMENDATION 6.1
The Legislature should consider adding election
standard oversight and enforcement responsibilities
and mechanisms to Utah Code.
FINDING 6.1
Utah election code does not specify oversight and RECOMMENDATION 6.2
enforcement roles. The Office of the Lieutenant Governor should
implement any oversight and enforcement
responsibilities as required by the Legislature in
response to this audit.
RECOMMENDATION 6.3
The Legislature should consider whether to require
county election staff to participate in election
trainings.
FINDING 6.2
County election workers are not required to RECOMMENDATION 6.4
participate in election training. The Office of the Lieutenant Governor should
determine the best way to support and provide
training for clerks on the controls and procedures of
Utah elections.

Although the LG’s Office is working on a standards The Office of the Lieutenant Governor should finalize
manual, there is currently no such resource for Utah its manual of standards to help guide clerks’ election
clerks. administration.
CONCLUSION
Given (1) clarity over their oversight and enforcement roles, (2) a requirement that training for clerks be
mandatory, and (3) a standardized reference manual for all clerks, the LG’s Office can help tighten controls
and ensure elections are run consistently by counties.
63
Chapter 6
In Election Code Could Be Clarified
6.1 Utah Election Code Does Not Specify

To ensure elections are more uniformly administered, we believe Utah Code
could clarify and define oversight and enforcement roles for elections as seen in
other states. Utah Code designates the Lieutenant Governor (LG) as the state’s
chief election officer.
This report uses the description of oversight and oversight bodies offered by the
Government Accountability Office. They state that “an oversight body oversees
the entity’s operations; provides constructive criticism…; and where appropriate,
makes oversight decisions so that the entity achieves its objectives.” Specifically,
it requires that the oversight body “provide oversight…in the development and
performance of control activities.” In relation to Utah’s election systems, the
Office of the Lieutenant Governor (LG’s Office) is the oversight body.
Greater Clarity Could be Given to the
Lieutenant Governor’s Oversight Role
Utah Code tasks the LG to “exercise general supervisory authority over all elections.”
This allows ambiguity as to what activities should be conducted as part of that “general
supervisory authority.”
The same section of code further requires, in part, that the LG:
• “Exercise direct authority” for federal, state-wide, or multi-county races

• “Assist county clerks in unifying the election ballot”
• Publish election information
• Answer election questions
• “Maintain election returns and statistics”
• Certify the winners of elections
• “Coordinate with local, state, and federal officials to ensure compliance
with state and federal election laws.”
Aside from these specific duties, the only other oversight requirement is that “the
Lieutenant Governor may not assume the responsibilities assigned to the county
clerks… or other local election officials.” In part because of non-specific oversight
roles, counties’ use of controls is inconsistent as shown in the bulk of this report.
Examples include:
• Chapter 2 – The VISTA tool to check for voters who have died is used
inconsistently.
Audit Title
• Chapter 3 – Ballots counted are not reconciled with the vote in VISTA.
• Chapter 4 – Counties have varying standards for signature verification

training.
Clarifying the extent of the LG’s oversight responsibilities could help to ensure
that these and other control weaknesses do not continue.
Other States’ Election Officers Have Greater
Responsibility for Election Oversight
The election officers in neighboring states (usually a secretary of state) have more
clear oversight authority than Utah’s LG. Specifically, Colorado and Oregon both
require their election officers to ensure that their elections follow the law.
CRS
the secretary of state has the following duties
To superv ise the conduct of elections in this state; to enforce the
prov isions of this code; to make uniform interpretations of this code
CRS
O To promulgate, publish and distribute such rules as the secretary of
state finds necessary for the proper administration and enforcement
of the election laws
To inspect and rev iew the practices and procedures of county clerk
and recorders
To enforce the prov ision of this code by injunctiv e action brought by
the attorney general
RS
It is the secretary s responsibility to obtain and maintain uniformity in
the application, operation and interpretation of the election laws.
RS O
The Secretary of State and each county clerk shall diligently seek out
any ev idence of v iolation of any election law.
In contrast, the oversight authority granted to the LG is more limited in scope

and specificity. By specifying the extent of the LG’s oversight responsibilities, the
Legislature can empower them to not only train counties, but to determine
whether they are meeting standards and notify counties of the need for
improvement.
Some Common Enforcement Responsibilities of the
Chief Elections Officer Are Absent in Utah Code
While statute mentions oversight in a vague way, it is silent on enforcement
mechanisms or responsibilities. This means that if a county clerk, intentionally or
not, does not follow Utah Code or Administrative Rule, the default mechanism
to enforce compliance is through the courts. Attorneys at the Office of Legislative
Audit Title

Research and General Counsel (OLRGC) agree with that assessment as shown on
the next page.41
OLAG Does
Does the
the Lieutenant Governorhave
lieutenant governor haveany
any
OLAG enforcement authority if a county election
Question enforcement authority if a county election
uestion official fails to
official fails to comply
comply with
with law
law or
or rule
rule?
No.
No. The
The Lieutenant Governordoes
lieutenant governor doesnot nothave
have
OLRGC
OLRGC specific
specific enforcement authority if
enforcement authority if aa county
county
Response
Response election official fails
election official fails to
to comply
comply with
with lawlaw or
or
rule.
rule.
Some other entities within Utah show potential options for enforcement
mechanisms. Both the Attorney General (AG) and the State Auditor have
enforcement authority in their respective fields.
State Attorney General Authority

Utah Code 67-5-1(2)(a)
Require a district attorney or county attorney to…report on the status of public
business.
Utah Code 67-5-1(2)(b)(i)(B)
Review first degree felonies when a county or district attorney fails to screen a case or
file charges.
State Auditor Authority

Utah Code 67-3-1(7)(g)
Withhold state allocated funds or the disbursement of property taxes…to ensure that
officials…comply with state laws and procedures in the budgeting, expenditures, and
financial reporting of public funds.
Utah Code 67-3-3(1)
Suspend any disbursement of public funds whenever, in the state auditor’s opinion,
the disbursement is contrary to law.
In addition to these options, the Legislature could also give the LG’s Office the
authority to fine counties in consistent need of improvement.
41 For the complete OLRGC opinion, see Appendix D.
Audit Title
ther States’ Election fficers Have Greater
Responsibility for Election Enforcement. Some Without an
other states have granted their chief election officer enforcement
increased enforcement responsibilities. For example, mechanism, a
Oregon statute reads: clerk’s failure to
meet existing
Whenever it appears to the Secretary of State standards could
that a county clerk, city elections officer or a consistently result
local elections official has failed to comply with in more errors and
weaker security in
an interpretation of any election law…or has their jurisdiction,
failed to comply with a rule, directive or which in turn could
instruction made by the Secretary of State …the undermine
Secretary of State may apply to the appropriate confidence in
elections statewide.
circuit court for an order to compel the county
clerk, city elections officer or local elections
official to comply.42
This and other examples show that other states have adopted a mechanism to
ensure that the actual execution of state election law meets the legislature’s and
chief election officer’s expectations.43
RECOMMENDATION 6.1
The Legislature should consider adding election standard oversight, and

enforcement responsibilities and mechanisms to Utah Code.
RECOMMENDATION 6.2
The Office of the Lieutenant Governor should implement any oversight and
enforcement responsibilities as required by the Legislature in response to
this audit.
6.2 County Election Workers Are Not
Required to Participate in Election Training
While each county has different election needs in terms of geographic and
demographic size, there is a need for universal security standards. There are
universal principles and available controls that have generally been adopted by
all counties; however, not all controls have been adopted with the same strength,
consistency, and reliability. We believe it is important that each clerk is
consistently trained and apprised on industry best practices and aware of the
42Oregon Revised Code 246.820

43Washington code includes a cover all penalties statute for election officials who fail to uphold
election law or rule.
Audit Title

interpretation of state statute. As previously mentioned, oversight of the
application of these standards should then occur.
Current Efforts to Educate Clerks Are Decentralized and Voluntary
Each separate clerk in each county must determine how best to learn their duties,
as well as train their staff in carrying out their duties. There are some existing
options, including participating in voluntary events offered by the LG’s office, or
meetings at the clerk’s professional association, the Utah Association of Counties
(UAC). In fact, clerks cite UAC as their primary forum for improving the
administration of elections and understanding election laws. The LG’s Office
offers bi-weekly training and discussion online, but these meetings are voluntary.
Whether a clerk or their staff attend these events is
There appears to be entirely determined by the preference of the clerk.
no evaluation to This leaves the state open to the risk of a spotty
ensure that the system where some counties may be unprepared to
principles are being
handle election irregularities or security breaches.
learned or applied in
each jurisdiction. The need for system-wide training has been shown
throughout this report. Namely:
• Chapter 2 – Clerks may not be aware of some of the tools in VISTA
intended to simplify election roll maintenance. There is currently no
official VISTA training.
• Chapter 3 – Because there are no standards or training on ballot

reconciliation, the number of votes counted does not match the number of
voters who got credit for voting.
• Chapter 4 – Finding 4.2 explains in detail the need for more thorough,
consistent training on what can and cannot be accepted for a legitimate
signature.
• Chapter 5 – Any new auditing practices would need a thorough

explanation and training to orient clerks to the new process.
Training Can Help Ensure That Clerks

Understand and Implement Best Practices. In Systematic,
addition to improving election controls and processes, mandatory training
of clerks on current
required in-person trainings would give the LG a good
practices and
opportunity to gauge the knowledge and experience of standards could (1)
clerks around the state. increase the
consistency of use
Washington, Oregon, and Colorado all have statutory
of the practices and
requirements that the election officer train or provide standards, and (2)
contracted training in clerks’ election duties. In reveal areas for
Oregon, for example, required bi-yearly trainings are additional training.
used both to inform and to evaluate clerks and see
whether state minimum standards are being met.
Audit Title
Auditor Observations Revealed Inconsistent Knowledge of
Election Law and Best Practices Among Clerks
There was at least one auditor from our office in every county reviewing and
observing elections during the 2022 primary election. The auditors were trained
to review for the existence and application of controls and best practices. We
found county clerks and their staff to be committed to election security and
protocols. However, we also found several items that need to be addressed from
training to oversight. Counties should be supported by the LG’s office in these
areas moving forward.
During our observations, we found that clerks often had contradicting views on
the implementation of best practices and whether certain practices were listed in
Utah Code. For example:
• In one county ballots were kept in a pile on an office desk with no labels.
Piles with different purposes such as envelopes that are ready for
separation, signatures that require further review, and ballots that were
returned via USPS were often stored next to each other without clear
distinction.
• Some counties require election workers to conduct in-process audits as

ballots are being counted to ensure the accuracy of the signature
verification process, while others only review signatures a second time if
there is a discrepancy or during the post-election audit.
• In one county, same-day voters were allowed to same-day register; they

were not then issued provisional ballots as required, but instead received
standard ballots.
• Some counties have a policy of using two people to collect ballots from
county drop boxes. However, in clerks’ offices with limited staff, there are
instances where only one employee collected sealed ballot drop boxes.
These issues, among others, are ripe for standards and training. Finding 6.3 addresses
this further.
Fifty-Nine Percent of Utah Clerks Will Begin

2023 with Less than Two Years’ Experience
Like much of the country, the last five years have resulted in high turnover
among county clerks. Figure 6.1 shows the amount of experience among Utah
clerks.
Audit Title

Figure 6.1 In January, 8 of 29 Utah Clerks Will Begin Their First Term. An additional
nine a e less t an t o yea s’ ex e ience
i st Te
i st Te
i st Te io to
Source: Office of the Lieutenant Governor
While some counties use professional, unelected election directors, the relative
inexperience of clerks still poses a concern. Clerks have the ultimate say in
determining a jurisdiction’s approach to an election within statutory limitations.
They decide how to use controls, which controls to use, and how to comply with
election code. The lack of uniformity in the conduct of elections is likely to
continue and institutional knowledge of election administration is being lost as
experienced clerks retire. Because of this, there needs to be mandatory
standardized training on these important controls.
RECOMMENDATION 6.3
The Legislature should consider whether to require county election staff to

participate in election trainings.
RECOMMENDATION 6.4
The Office of the Lieutenant Governor should determine the best way to
support and provide training for clerks on the controls and procedures of
Utah elections.
6.3 The LG’s Office Is Working on a Manual of
Standards and Guidance for Clerks to Prepare for and
Administer Elections
Communication regarding changes to statute and election deadlines is currently
done primarily on an ad hoc basis with no central resource for clerks to turn to
for assistance, aside from reaching out to the LG’s Office directly. Currently, the
Audit Title
LG’s staff reaches out to county clerks via email to inform them of upcoming
deadlines, changes to election law, and opportunities for grant funding and
occasionally performs in-person visits. This differs from eight western states that
provide a comprehensive election administration manual in addition to fielding
questions from clerks.44
In Utah, issues and questions that clerks face often do not get brought to the LG’s
attention until the administration of an election. However, the LG’s Office is
working on a centralized election handbook that they plan to release the first part
of 2023. This manual or set of standards could preempt many issues, thus saving
time for both the clerk and the LG’s Office. Distributing such a manual also sets a
uniform standard for every officer in every election that, if followed, could
increase security and accountability in every county. This set of standards also
creates a training foundation as discussed in Finding 6.2.
Figure 6.2 Eight of Eleven Western States Have a Form of Election Administration
Manual. Utah is working on their own manual, but it has not yet been completed.
WA
MT
OR ID
W
as anual
o anual
NV UT CO
CA
A NM
Source: Auditor analysis
44 For example, Oregon’s vote-by-mail manual.
Audit Title

There Is No Statutory Requirement for a Manual
Statute currently requires that the LG distribute copies
of Utah Code, Chapter 20A, Election Code. However, Utah’s lack of an
this practice is ineffective in ensuring compliance with easy-to-understand
standards manual
law or best practices. Each county’s specific operations leave clerks to
are different. Some counties use signature verification decide for
machinery, while it is done all by election workers in themselves complex
other counties. While the LG cannot prescribe step-by- aspect of the
step instructions for every case in every county, a set of process.
overall guidelines for the fundamentals would still be
useful.
The importance of an easily accessible manual is heightened by the relative
inexperience of many clerks in Utah. If the LG’s Office were to collect
institutional knowledge of best practices into a single place, it creates a
clearinghouse of election knowledge that everyone is working from. New clerks
could be better prepared to transition into their new roles. One clerk we visited
with less than two years of experience noted that there is a steep learning curve to
running an election. This clerk’s response to many of our questions was “that is a
good question. I’ll have to call [another clerk] and ask.” While we applaud the
clerks for being a resource to each other, a set of standards and guidance, with
subsequent training, could enable new clerks to find the information themselves.
Other States Have Published
Comprehensive and Easy-to-Read Manuals
Other western states such as Oregon, Washington, Colorado, and California all
publish election administration manuals which allow clerks to easily understand
the law and how to effectively apply it. The manuals published in other states still
offer a wide degree of latitude in making decisions within preestablished
minimum standards. Because counties are so different in population makeup,
this level of latitude is essential. Some of the topics the manuals address include
creating a physical and cyber security plan, a chain-of-custody process, and
dealing with outlying issues such as homeless individuals and
reissued/supplemental ballots. Figure 6.3 shows a section of Oregon’s manual on
security needs.
Audit Title
Figure 6.3 This Section of the Oregon Vote-By-Mail Manual Lays Out Best Practices
for Creating a Mandatory Security Plan. There are other sections that expound on best
practices, most of which have statutory or administrative rule references for support.
Source: Vote by Mail Procedures Manual, Oregon Secretary of State, Elections

Division
In our discussions with representatives of other states, they expressed that they
get fewer questions and clerks are able to deal with situations because the manual
exists to guide them more independently.
The director of elections in the LG’s Office reports that its planned new election
manual contains sections on:
• Audits • Voter registration
• Canvassing • Public service requests
• Equipment maintenance logs • Ballot design
• Undeliverable ballots and • Ballot delivery
mail • Election results
• Ballot collection • Accessibility and inclusivity
• Ballot processing • Recounts
• Chain of custody • Petition processing
• Logic and accuracy tests • Physical and cyber security
• Adjudication of voter intent
Many of the specific recommendations throughout this report fit within one of
these areas and should provide specific focus. We support the LG’s efforts and
encourage them to complete this manual, distribute it, train election staff on its
principles, and update it when required.
Audit Title

RECOMMENDATION 6.5
The Office of the Lieutenant Governor should finalize its manual of standards
to help guide clerks’ election administration.

Complete List of Audit
Recommendations
A Performance Auditor of Utah’ Election System and Controls 77

Audit Title

Complete List of Audit Recommendations
This report made the following seven recommendations. The numbering convention assigned to
each recommendation consists of its chapter followed by a period and recommendation number
within that chapter.
Recommendation 2.1
We recommend that the Office of the Lieutenant Governor finalize standards for the frequency
and use of key VISTA maintenance tools, and then monitor their implementation.
Recommendation 2.2
We recommend that the Legislature consider clarifying the oversight role of the Office of the
Lieutenant Governor with regards to voter registration. This could include authority for regular
analysis of voter records, rulemaking authority for minimum maintenance standards.
Recommendation 2.3
We recommend that the Office of the Lieutenant Governor monitor and ensure that the names
of deceased voters are removed from voter rolls, as required in statute.
Recommendation 2.4
We recommend that the Office of the Lieutenant Governor review its staffing plan to determine
whether critical functions have sufficient staff. If, in this review, the Office of the Lieutenant
Governor determines that they need more resources for monitoring and maintenance, we
recommend that they report these needs to the Legislature.
Recommendation 3.1
We recommend that the Office of the Lieutenant Governor create rules requiring county clerks
to publicly reconcile the number of ballots tabulated with the number of voters given vote credit
in VISTA.
Recommendation 3.2
We recommend that the Office of the Lieutenant Governor help clerks identify the best data
possible to reconcile the number of ballots tabulated with the number of voters given vote credit
in VISTA.
Recommendation 3.3
We recommend that the Office of the Lieutenant Governor create rules requiring ballot
envelopes to be counted as early as possible during ballot processing.
Recommendation 3.4
We recommend that the Office of the Lieutenant Governor finalize its chain-of-custody
manual, including best practices for election staff in Utah’s counties.
Recommendation 4.1
We recommend that the Legislature consider either including clearer standards for signature
review, acceptance, and rejection in Utah Code, or giving the Lieutenant Governor authority to
establish these standards and instructions.

Recommendation 4.2
We recommend that the Legislature consider giving the Office of the Lieutenant Governor
explicit authority to establish training standards and requirements for signature verification to
better assure that practices are executed consistently across the state.
Recommendation 4.3
We recommend that the Legislature consider requiring county election staff to attend signature
verification training.
Recommendation 4.4
We recommend that the Legislature consider the options in this chapter to improve the quality
of signatures available for signature verification.
Recommendation 4.5
We recommend that the Legislature consider giving authority to the Lieutenant Governor to
establish rules requiring counties to conduct mid-election audits of signatures and performance
tracking for signature reviewers.
Recommendation 4.6
We recommend that the Legislature consider guidelines for, or the necessity of, curing
Recommendation 5.1
We recommend that the Legislature consider options to increase the independence of any post-
election audit.
Recommendation 5.2
We recommend that the Legislature consider establishing a risk-limiting audit pilot program to
enhance Utah’s post-election audit methods, giving the Office of the Lieutenant Governor
rulemaking authority to establish standards.
Recommendation 5.3
We recommend that, if the Legislature authorizes a risk-limiting audit pilot program, the Office
of the Lieutenant Governor should create rules to establish common procedures.
Recommendation 6.1
We recommend that the Legislature consider adding election standard oversight, and
enforcement responsibilities and mechanisms to Utah Code.
Recommendation 6.2
We recommend that the Office of the Lieutenant Governor implement any oversight and
enforcement responsibilities as required by the Legislature in response to this audit.
Recommendation 6.3
We recommend that the Legislature consider whether to require county election staff to
Audit Title

Recommendation 6.4
We recommend that the Office of the Lieutenant Governor determine the best way to support
and provide training for clerks on the controls and procedures of Utah elections.
Recommendation 6.5
We recommend that the Office of the Lieutenant Governor finalize its manual of standards to
help guide clerks’ elections administration.

Audit Title

Appendix

Audit Title

A. Full Detail of Election Processes and Controls
Highlighted in Figure 1.1

Audit Title

Figure 1.1 in Chapter 1 was designed to give a high-level view of the key election
processes we feel are critical to protecting election integrity. While the chapter
provides additional detail and context to support our conclusion that Utah’s
election system is functioning well overall, there are many more detailed controls
within each element of Figure 1.1. We want
interested readers to be able to see and
Link to interactive
understand those details. version of Figure 1.1:
To that end, our team built an online version of
the figure where users can click on each process
and explore the overview of the process. Clicking
on each item there will show the full list of
controls, risks, and audit recommendations
related to each. We have also embedded relevant
charts and figures from the audit report in the
online version. Please use the QR code shown to
the right to access that version.
We also wanted readers of the full report to have access to this process and
control information and prepared this appendix for that purpose. All ten
elements shown in Figure 1.1 are represented here, along with the detailed
process and control elements contained in the interactive, online version of the
figure. Relevant recommendations from the report are also shown here with
chapter references so readers can explore areas where our team performed a full
analysis (e.g., recommendation 2.3 is the third recommendation from Chapter 2).
It should be noted that some areas found in this election process do not have
specific audit recommendations and not all recommendations from the audit are
found here. Recommendations are focused on the areas with the most risk or
areas that could be improved upon.
1. County-Level Election Administration

Overview
• With some oversight from the Office of the Lieutenant Governor (LG’s Office),
elected county clerks administer elections at the county level.
• Within the limits of state law, decisions about election equipment, ballot and
envelope printing, ballot boxes, polling places, etc. are made and largely
funded at the county level.
Controls
• Subdividing election administration makes it extremely difficult to
compromise statewide and/or national races because of the complexity of
coordinating and executing such widescale fraud against so many independent
offices.
Audit Title
• Local control allows for experimentation and innovation to suit unique needs
and circumstances. Vote-by-mail in Utah grew from such grassroots
beginnings.
• Local-level elected clerks are in direct control of essential election processes,

giving voters more direct power to hold election officials accountable.
Risks
• Variation in quality of election administration could open vulnerabilities in
areas with inferior processes.
• Variation in interpretation of election laws could threaten eligible voters’ right

to vote.
Recommendations
• None
2. State-Level Voter Registration System

Overview
• County clerks use a centralized, state-level voter registration system called
VISTA. The system is managed by the Lieutenant Governor.
• To participate in elections, all eligible voters must register to vote with the
clerk for the county in which the voter resides. The information is then saved in
VISTA.
• Registration seeks to verify both identity and place of residency.
Controls
• VISTA enables voter management across county lines and provides central
controls over ballot security.
• With support from the LG’s Office, the system enables clerks to validate voter
information against data sources like the Driver License Division, Office of
Vital Records and Statistics, Department of Corrections, Social Security
Administration, the Electronic Registration Information Center (ERIC), etc.
• Using a central database can help control voting activity in real time, making it
very difficult to vote more than once.
Risks
• Because of the decentralized use of VISTA, oversight can be lacking.
• As with all systems, there is a possibility of human error when entering data.
• There can be a lack of ongoing voter record maintenance and use of VISTA
tools checking for duplicate voter registrations.
Audit Title

Recommendations
• None
3. Routine Voter List Maintenance

Overview
• As required by state law, the LG’s Office uploads data regarding voters who are
incarcerated or deceased and those who have moved or changed personal
details with the Driver License Division (e.g., marriage name change).
• Clerks then use the information provided by the LG’s Office, in addition to
several duplicate voter reports and methods, to review and modify voter
records in an attempt to keep the voter roll as current and accurate as
reasonably possible.
• Clerks will also enter new voter information when a voter registers somewhere
other than the Driver License Division.
Controls
• Constant maintenance of voter records helps ensure that registered voters are
eligible and voting in the appropriate location/elections.
• This maintenance also helps ensure that ballots are sent to the voter’s correct
address.
• Because mail ballots are only sent to active voters, accurate tracking of voter
participation reduces the risk and cost of sending ballots to voters who have
not participated in multiple elections.
Risks
• Human error can occur, especially considering the large volume of records to
be reviewed and modified and the critical, specific details like address, date of
birth, and driver license number that must be entered.
• Clerks’ offices can neglect voter record maintenance, increasing the risk that
voters will not receive a ballot or receive one other than their own.
• Oversight of voter maintenance can be neglected.
• There is a lack of standards and guidance for key processes within voter
registration database.
Audit Title
Recommendations
• 2.1 We recommend that the Office of the Lieutenant Governor finalize
standards for the frequency and use of key VISTA maintenance tools, and then
monitor their implementation.
• 2.2 We recommend that the Legislature consider clarifying the oversight role of
the Office of the Lieutenant Governor with regards to voter registration. This
could include authority for regular analysis of voter records, and rulemaking
authority for minimum maintenance standards.
• 2.3 We recommend that the Office of the Lieutenant Governor monitor and
ensure that the names of deceased voters are removed from voter rolls, as
required in statute.
• 2.4 We recommend that the Office of the Lieutenant Governor review its
staffing plan to determine whether critical functions have sufficient staff. If, in
this review, the Office of the Lieutenant Governor determines that they need
more resources for monitoring and maintenance, we recommend that they
report these needs to the Legislature.
4. Pre-Election Logic and Accuracy Testing

Overview
• Tabulation machines are tested before each election to ensure to the clerk's
satisfaction, that they are programmed to accurately read each race and ballot
type.
• This process culminates in a public meeting where the machine programming

(i.e., the machines' logic and accuracy) is demonstrated with test ballots.
Controls
• With a limited exception for ranked-choice voting equipment, the Lieutenant
Governor ensures that all voting equipment used in Utah is certified to the
standards of the US Election Assistance Commission (EAC).
• Tabulation machine maintenance and software updates are performed only by

authorized individuals.
• Tabulation machines are physically secured and locked to prevent

unauthorized access.
Audit Title

Risks
• Election machines may not correctly identify test ballots during the pre-
election L&A test. No reported instances of this risk were reported. If
complications arise, they can be remedied prior to the election.
• Auditors observed an L&A demonstration in a large county that was not as

clearly explained as it could have been, undermining the opportunity to
increase the confidence of those in attendance.
Recommendations
• None
5. Mail Ballot Preparation

Overview
• Clerks' offices produce a list of eligible, active voters in their counties. These
lists are sent to specialized election printing vendors, selected by each county,
for ballot and envelope printing.
Controls
• Each envelope is assigned a unique control number by the state’s central voter
database (VISTA). This, in combination with the voter’s ID number, is used by
the voter registration database (VISTA) to validate ballots upon return.
• If a voter needs a new mail-in ballot, the unique control number will be
'spoiled' and will no longer be accepted by VISTA. VISTA will automatically
assign a new control number for the new envelope/ballot.
Risks
• Printing errors can create logistical and processing problems that clerks must
overcome.
• Some counties have on-demand ballot printers and some counties order
surplus ballots from printers to ensure that all eligible voters can get a ballot
if/when needed. Inadequate controls over these surplus ballots could provide
opportunities for election workers to cast unlawful surplus votes.
• Failure to complete the voter maintenance from element #3 can result in

ballots prepared for and sent to voters who no longer reside in their address of
record.
Audit Title
Recommendations
• None
6. Ballot Intake and Chain of Custody

verview
• Clerks offices receive ballots from ballot drop boxes and the US Postal Service
and must ensure that all ballots returned are accurately tracked and accounted
for.
• Clerks use ballot chain of custody practices to create a verifiable document trail
to account for all ballots received.
Controls
• Best practices say accurate, transparent chain of custody practices and records
can increase voter confidence that ballots were handled correctly and protect
against the unauthorized removal or addition of ballots.
• Two individuals should retrieve ballots to reduce the risk of someone adding or
removing ballots as required by code.
• Various tracking forms provide an audit trail of ballot retrieval and processing.
• Ballots are typically combined for processing in batches of 25-250 (depending

on the size of the county). Batching provides smaller units of work that can
each be tracked and accounted for.
• If ballots received are reconciled to the total number of ballots processed, it

ensures that there are neither more nor fewer than there were at the beginning
of the process.
Risks
• We observed significant variations in chain of custody practices across
counties. Some counties lacked custody documentation sufficient to fully or
clearly account for all ballots received.
• Failure to reconcile ballot intake to the final total of ballots processed leaves
questions about full ballot accountability.
• Lack of proper segregation of duties between ballot intake and other processing
steps can create an opportunity for an election worker to compromise the
count (i.e., by adding or removing ballots) without detection.
Audit Title

Recommendations
• 3.1 We recommend that the Office of the Lieutenant Governor create rules
requiring county clerks to publicly reconcile the number of ballots tabulated
with the number of voters given vote credit in VISTA.
• 3.2 We recommend that the Office of the Lieutenant Governor help clerks
identify the best data possible to reconcile the number of ballots tabulated with
the number of voters given vote credit in VISTA.
• 3.3 We recommend the Office of the Lieutenant Governor create rules

requiring ballot envelopes to be counted as early as possible during ballot
processing.
• 3.4 We recommend that the Office of the Lieutenant Governor finalize its
chain-of-custody manual, including best practices for election staff in Utah’s
counties.
7. Signature Verification and Curing

Overview
• The primary means of validation and control of incoming ballot envelopes is
the comparison of voters' signatures against known signature examples.
• Signature verification and its surrounding controls are supported by the

central voter registration database (VISTA).
• Many counties manually review all signatures against voter registration

records. Some counties have signature verification software that automatically
reviews and accepts a certain number of signatures considered to be acceptably
accurate. Those not accepted by the software are manually reviewed by election
staff.
Controls
• County clerks can use statewide driver’s license data to obtain valid signature
images for most Utah voters.
• County practices include additional levels of more intensive review for any
signatures that are not obvious matches with reference signatures.
• Many clerks provide some training to election staff on the techniques and
standards for proper signature validation.
• If a signature is challenged, clerks must contact voters to give them the

opportunity to correct (or “cure”) the ballot so it can still be counted.
Audit Title
Risks
• Utah lacks clear legal standards for election signature verification.
• Ballot signature verification training has not been consistently deployed

statewide. Therefore, standards and quality may vary somewhat.
• If counties do not work to maintain and improve voters’ signature files, valid
signatures may be rejected or accepted inappropriately.
• Signatures that are passed on the first review are analyzed only once.
• Auditors observed that curing ballot signature challenges over the phone
resulted in election staff not appropriately verifying the identity of voters in
some cases.
Recommendations
• 4.1 We recommend that the Legislature consider either including clearer
standards for signature review, acceptance, and rejection in Utah Code, or
giving the Lieutenant Governor authority to establish these standards and
instructions.
• 4.2 We recommend that the Legislature consider giving the Office of the
Lieutenant Governor explicit authority to establish training standards and
requirements for signature verification to better assure that practices are
executed consistently across the state.
• 4.3 We recommend that the Legislature consider requiring county election staff
to attend signature verification training.
• 4.4 We recommend that the Legislature consider the options in this chapter to
improve the quality of signatures available for signature verification.
• 4.5 We recommend that the Legislature consider giving authority to the

Lieutenant Governor to establish rules requiring counties to conduct mid-
election audits of signatures and performance tracking for signature reviewers.
• 4.6 We recommend that the Legislature consider guidelines for, or the

necessity of, curing challenged ballot signatures by phone.
Audit Title

8. Paper Ballot Tabulation
Overview
• Ballots are designed and printed to facilitate machine tabulation in all Utah
counties.
• All ballots are cast on paper. Because nothing is fully digital, a full audit trail is
available for all ballots cast.
• Machines generate a 'Cast Vote Record' for each individual ballot showing how
the machine read the voter's ballot markings. This can be compared to the
digital ballot image (scanned by the machine) or the physical paper ballot for
auditing purposes.
Controls
• Voted paper ballots are processed and stored in batches to preserve a full
paper-based audit trail that can be used to validate the machine count if
needed.
• Tabulation machine totals of ballots scanned are matched to batch control

amounts (recorded as part of the chain of custody) to verify piece count
accuracy.
• A post-election audit in each county reviews machines' Cast Vote Records

compared to the voters' markings on the actual ballots to ensure accurate
machine programming and performance.
Risks
• Ballots could be misplaced in county elections, especially if the county uses
poor chain of custody practices (element #6).
Recommendations
• None
9. Post-Election Audit
Overview
• Clerks submit aggregated ballot/batch tracking data to the LG’s Office. The
LG’s Office selects batches for audit comprising 1% or 1,000 ballots, whichever
is fewer.
• An audit of ballots is done to ensure ballot tabulation machines are

programmed to accurately read each race and ballot type.
• An audit of signatures is done to confirm that signatures were appropriately

validated.
Controls
• The LG’s Office selects batches for the audit independent of clerks and sends
that information immediately before the audit begins.
Audit Title
• The process is open to observation by candidates, political parties, and
members of the public at all times.
Risks
• Disadvantages of the traditional audit include: small sample sizes, clustered
ballot samples, no direct machine tally verification.
• Currently, the individual clerk’s offices audit themselves. Audits standards

require an independent party conduct the audit.
Recommendations
• 5.1 We recommend that the Legislature consider options to increase the
independence of any post-election audit.
• 5.2 We recommend that the Legislature consider establishing a risk-limiting

audit pilot program to enhance Utah’s post-election audit methods, giving the
Office of the Lieutenant Governor rulemaking authority to establish standards.
• 5.3 We recommend that, if the Legislature authorizes a risk-limiting audit pilot

program, the Office of the Lieutenant Governor should create rules to establish
common procedures.
10. Voter Feedback and Public Observation

Overview
• Voters play an important role in maintaining election integrity. Voters can and
do provide valuable information to election officials and workers.
• Voters have a responsibility to update their voter records, which often occurs
when updating the same information on their driver’s licenses.
• Citizens, candidates, party leaders, etc. can observe most of the key election
processes listed here to independently check exactly how elections are
administered by their local clerk’s office.
Controls
• If ballots don’t arrive when expected or other anomalies occur, voters regularly
contact clerks to find answers. This can lead to the detection of situations in
need of investigation and correction.
• Public observation of election administration can increase the accountability of

clerks and election workers.
• Showing voters that clerks have nothing to hide can increase the overall level
voter confidence in Utah’s election integrity.
Audit Title

Risks
• Some voters do not update their voter information as diligently as they should.
• We found that very few people actually observe election administration or

public demonstrations.
• Public observers can interfere with proper ballot processing or act in

belligerent or disruptive ways.
Recommendations
• None
Audit Title
Audit Title

B. Voter Status and Record Type

Audit Title

Voter Status:
Voter Utah Code
Definition
Status Reference
Registered voter not classified as inactive by
Active 20A-1-102(1)
county clerk. Eligible to receive mail ballots.
Placed inactive by county clerk. Individuals may
still vote and participate in elections but will not
be mailed a ballot. Voters become inactive by
not voting in two consecutive federal elections 20A-1-102(28);
Inactive
without updating their registration, or when 20A-2-306
there is evidence the voter has moved (NCOA
list, undeliverable ballots, undeliverable mail,
etc.)
Typically used when a voter has submitted an Not defined in
In Suspense
incomplete registration form. code
Used for underage voters who are preregistered Not defined in
Not Eligible
to vote. code
Determined to be no longer eligible to vote i.e.,
20A-2-305
Removable due to death, written request, or moving out of
20A-2-306(4)(b)
state.
Voter Record Type (Record Protection):

Voter
Utah Code
Record Definition
Reference
Type
Unless requested by the voter, voter records are
Public 20A-2-104
considered public.
Voter registration records are withheld from
20A-2-104(1)(b);
persons other than government entities, political
Private 20A-2-104(4)(d);
parties, candidates for public office, and their
20A-2-104(4)(h)
contractors, employees, and volunteers.
Must request to be a withheld record and be a
victim of domestic violence, a law enforcement
officer, a member of the armed forces, a public
figure, protected by a protective order, or whose
20A-2-104(1)(b);
Withheld registration was classified as a private record
20A-2-104(7)
prior to May 12, 2020. This protects registration
records from all political parties, candidates for
public office, and their contractors, employees,
and volunteers.

Audit Title

C. Examples of Washington
State Reconciliation Reports

Audit Title

King County Auditor
Reconciliation Form
General information Registered voters eligible to participate
County name King Active registered voters 1,420,898
Election date 11/3/2020 Inactive registered voters 104,014
Category Reconciliation (detailed accounting of ballots)

Ballots not counted
Issued
Credited voters in VoteWA Ballots Discrepancy
(number of Ballots Accepted
(envelopes containing forwarded to (If zero, category
voters issued (Counted / Tabulated) Ballots Rejected
ballots) other balances)
ballots)
counties
UOCAVA 37,376 23,043 22,774 N/A 269 0
Federal write-in N/A 370 370 N/A 0 0
Provisional 74 74 16 0 58 0
DREs 0 0 0 N/A 0 0
All voters not reported in
1,479,986 1,207,893 1,196,902 0 10,991 0
above categories
Total for all voter categories 1,517,436 1,231,380 1,220,062 0 11,318 0
Summary
Reconciliation Additional Information
Overall Ballot Reconciliation Replacement Ballots
Ballots Received 1,231,380 Requested 72,842
Ballots Accepted 1,220,062 Issued 72,842
Ballots Not Counted 11,318 Returned 61,690
Discrepancy 0 Counted 60,691
Rejected 999
Voters credited to ballots counted Return method

Credited voters in VoteWA 1,219,824 Email 13,306
Credited envelopes without ballots (124) Fax 274
Deposited at staffed, unstaffed deposit sites and voting
Voters not credited in VoteWA (examples: FWAB or ACP) 362 909,987
centers
Total valid ballots 1,220,062 Non-UOCAVA returned by Fax or Email 0
Discrepancy 0
Generation method
VoteWA 19
Other online programs 37,764
PDF originating from county 0
Non-UOCAVA ballots issued electronicaly 15,548
AVU
Estimated number of paper ballots printed by AVU 46
Category discrepancy explanation VoteWA discrepancy explanation
Report prepared by: Jerelyn Hampton Date: November 23, 2020

Contact number: 206-477-4107
Rev 2.4 (08/13/2019)

Audit Title

Cowlitz County Auditor
Reconciliation Form
County name Cowlitz Active registered voters 66,452
Election date 11/5/2019 Inactive registered voters 6,067

Ballots not counted
Issued
Credited voters in Ballots Discrepancy
VoteWA (envelopes forwarded to (If zero, category
containing ballots) other balances)
ballots)
counties
UOCAVA 688 60 58 N/A 2 0
DREs 0 0 0 N/A 0 0
66,593 29,836 29,592 0 244 0
above categories
Total for all voter categories 67,281 29,896 29,650 0 246 0
Summary
Ballots Received 29,896 Requested 573
Ballots Accepted 29,650 Issued 573
Ballots Not Counted 246 Returned 376
Discrepancy 0 Counted 375
Rejected 1

Credited voters in VoteWA 29,638 Email 26
centers
Total valid ballots 29,648 Non-UOCAVA returned by Fax or Email 0
Discrepancy (2)
Generation method
VoteWA 111
Other online programs 0
Non-UOCAVA ballots issued electronicaly 320
AVU

Voter returned General Election envelope with Primary ballot inside. Discovered during ballot 1. Two ballots counted for one registered voter. Two household members signed each others'
inspection but unable to tie ballot back to voter. This is noted in "Credited envelopes without envelopes. One envelope was given the correct label to match the signature. The other envelope
ballots." was not. The same voter received credit twice in the same batch. Later, uncredited voter returned
a second ballot, which was accepted. Although caught, ballots had already gone to final
processing. 2.
Two ballots counted for one registered voter. One household member signed their own envelope
and also their housemate's envelope. Household member's envelope was given a new label to
match the signature of the voter who signed. The same voter received credit twice in the same
batch. Although caught, ballots had already gone to final processing.
Report prepared by: Hayley Johnson Date: November 26, 2019

Contact number: 360-577-3005
Rev 2.4 (08/13/2019)

Audit Title

Yakima County Auditor
Reconciliation Form
County name Yakima Active registered voters 18,437
Election date 4/27/2021 Inactive registered voters 601

Ballots not counted
Issued
Credited voters in Ballots Discrepancy
VoteWA (envelopes forwarded to (If zero, category
containing ballots) other balances)
ballots)
counties
UOCAVA 258 26 25 N/A 1 0
DREs 0 0 0 N/A 0 0
18,318 6,283 6,216 0 67 0
above categories
Total for all voter categories 18,576 6,309 6,241 0 68 0
Summary
Ballots Received 6,309 Requested 66
Ballots Accepted 6,241 Issued 66
Ballots Not Counted 68 Returned 25
Discrepancy 0 Counted 25
Rejected 0

Credited voters in VoteWA 6,242 Email 2
centers
Total valid ballots 6,241 Non-UOCAVA returned by Fax or Email 0
Discrepancy 0
Generation method
VoteWA 115
Other online programs 0
Non-UOCAVA ballots issued electronicaly 9
AVU
Report prepared by: K Fisher Date: 5.7.2021

Contact number: 509.574.1340
Rev 2.4 (08/13/2019)

Audit Title

D. OLRGC Opinion

Audit Title

OFFICE OF LEGISLATIVE RESEARCH AND GENERAL COUNSEL
John Q. Cannon, Director
John L. Fellows, General Counsel
Jake Dinsdale
Senior Audit Supervisor
W315 State Capitol Complex
Salt Lake City, Utah 84114
November 23, 2022
Re: Lieutenant Governor’s Election Oversight Authority
Jake,
You have asked me to respond to three questions relating to the lieutenant governor’s election oversight
authority. I address each below.
Question 1: Does statute give the lieutenant governor clearly-defined oversight authority over elections?
Short Answer: No, not in relation to local election officials. While the lieutenant governor is designated
as the state’s chief election officer and is granted general supervisory authority over elections,
this authority is limited and does not grant specific enforcement power.
Federal election laws require that each state have a “single” or “chief” officer designated to fulfill certain
duties in relation to those laws. These include:
• The Uniformed and Overseas Citizens Absentee Voting Act, 52 U.S.C. Sec. 20301, et, seq.
(UOCAVA)
• The National Voter Registration Act, 52 U.S.C. Sec. 20501, et. Seq. (NVRA)
• The Help America Vote Act, 52 U.S.C. 20901, et seq. (HAVA)
To ensure Utah’s compliance with federal law, Section 20A-2-300.6 of the Utah Code designates the
lieutenant governor as Utah’s “chief elections officer” and makes her responsible to oversee all of Utah’s
voter registration activities (this includes the UOCAVA voter registration requirements), oversee all other
responsibilities established by the NVRA and HAVA, and “coordinate with local, state, and federal
officials to ensure compliance with state and federal election laws.”
In light of the specific requirements of the federal laws cited above, and the language of Subsection 20A-
2-300.6(2)(a), the grant of oversight authority in that subsection is, except in relation to voter registration,
likely limited to ensuring compliance with the NVRA and HAVA. The grant of oversight authority in
relation to voter registration is broader and encompasses compliance with UOCAVA and the entire voter
registration process within Utah. This voter registration oversight authority, and the division of
responsibilities between the lieutenant governor and the county clerks, is delineated in Title 20A, Chapter,
2, Voter Registration.
In addition to the fact that the oversight authority described in Subsection 20A-2-300.6(2)(a) is limited in
scope, the designation of the lieutenant governor as the “chief elections officer” and the requirement that
she oversee certain activities are vague and do not grant specific enforcement authority. Further, the
language of Subsection 20A-2-300.6(2)(b) clarifies that, with respect to all other provisions of state and
federal law not specified in Subsection 20A-2-300.6(2)(a), the lieutenant governor’s role is limited to
coordinating with federal, state, and local officials to ensure compliance. This subsection does not give
any enforcement authority with respect to state and local officials (and cannot with respect to federal
officials). Finally, while the designation of the lieutenant governor as the election officer “for all

statewide ballots and elections” (Subsection 20A-1-102(23)(a)) with the authority to render “all
interpretations and make all initial decisions about controversies or other matters arising under this
chapter” (Section 20A-1-402), in relation to statewide ballots and elections, is significant, it does not
grant specific enforcement authority if a local elected official fails to comply.
Utah Code Subsection 67-1a-2(2)(a)(i) provides that the lieutenant governor shall “exercise general
supervisory authority over all elections.” While this provision arguably grants additional authority to the
lieutenant governor, it lacks the specificity necessary to empower her to exercise direct authority over
local election officials who are, themselves, elected officials and who perform a large portion of the work
relating to elections. In contrast, Subsection 67-1a-2(2)(a)(ii) does grant direct authority to the lieutenant
governor, but that authority relates only to elections that extend beyond the jurisdictions of local election
officials. In those situations, the law specifically states that she is responsible to “exercise direct authority
over the conduct of elections for federal, state, and multicounty officers and statewide or multicounty
ballot propositions and any recounts involving those races.” Id. This provision goes well beyond the
vague grant of “general supervisory authority over all elections.”
In conclusion, while the lieutenant governor is designated as the state’s chief election officer and is
granted certain oversight authority under Utah Code Subsection 20A-2-300.6(2)(a), this authority is
limited in scope and does not grant specific enforcement power. Similarly, the grant of general
supervisory authority over elections in Utah Code Subsection 67-1a-2(2) is vague and lacks enforcement
power. This is further evidenced by Subsection 67-1a-2(2)(b), stating that, “[a]s chief election officer, the
lieutenant governor may not assume the responsibilities assigned to the county clerks, city recorders, town
clerks, or other local election officials by Title 20A, Election Code.”
The Legislature, of course, may expand the authority of, or grant additional authority to, the lieutenant
governor by enacting legislation that clearly expresses that authority.
Question 2: Does the lieutenant governor have power to set election performance standards?
Short Answer: Yes, but only to a limited extent some extent.
The lieutenant governor may set election performance standards, as follows:
• As part of her oversight authority over voter registration requirements and the state’s
responsibilities under the NVRA and HAVA (Subsection 20A-2-300.6(2)(a)).
• To assist in her coordination with local, state, and federal officials to ensure compliance with
state and federal election laws (Subsection 20A-2-300.6(2)(b)).
• To make rules establishing requirements for verifying certain statuses in relation to withholding
registration records (Subsection 20A-2-104(8)(c)).
• To make rules “establishing requirements for election officials regarding ballot security,
including the custody, documentation of custody, handling, processing, disposition, and
tabulation of ballots” (Subsection 20A-3a-404(1)).
• To make rules “describing the procedures that a counting judge is required to follow for counting
ballots in an instant runoff voting race” (Subsection 20A-4-101(2)(f)(i)).
• To make rules “establishing software validation procedures that an election officer is required to
comply with to verify that voting system files have not been tampered with” (Subsection 20A-5-
905(1)).
• To make rules “establishing minimum requirements that a vendor must meet to be eligible to
print ballots to be used in an election” (Subsection 20A-3a-404(1)).
• To make rules relating to the submission and verification of nomination petition signatures
(Subsection 20A-9-403(3)(f)).
• To make rules establishing “procedures for complying with, and verifying compliance with, the
candidate nominating process” (Section 20A-9-410).
OFFICE OF LEGISLATIVE RESEARCH AND GENERAL COUNSEL | Page 2 of 3

Question 3: Does the lieutenant governor have any enforcement authority if a county election official
fails to comply with law or rule and, if not, are there templates for enforcement mechanisms in state
statute?
Short Answer: No. The lieutenant governor does not have specific enforcement authority if a county
election official fails to comply with law or rule.
The lieutenant governor is not granted specific enforcement authority over county election officials who
fail to comply with law or rule. Moreover, Subsection 67-1a-2(2)(b), provides that, “[a]s chief election
officer, the lieutenant governor may not assume the responsibilities assigned to the county clerks, city
recorders, town clerks, or other local election officials by Title 20A, Election Code.”
Traditionally, the lieutenant governor has exercised its oversight authority by consulting with county
election officials to advise them regarding legal requirements and assist them in fulfilling those
requirements. If a circumstance were to arise where a county election official fails to comply with law or
rule after consultation, the lieutenant governor could consider the following:
• Exerting political pressure to enforce compliance.
• Enlisting federal authority to enforce compliance with federal law.
• Exploring judicial enforcement options.
While there is not a template in state statute that would be an exact fit for legislation granting specific
enforcement authority to the lieutenant governor in relation to county election officials who fail to comply
with law or rule, the attorney general is granted some power with respect to local prosecution officials
that may be instructive. The attorney general has:
• Authority to “require a district attorney or county attorney of the state to, upon request, report on
the status of public business entrusted to the district or county attorney's charge” (Subsection 67-
5-1(2)(a)).
• Authority, under certain circumstances, to review investigations and file charges in relation to
first degree felonies when a county or district attorney fails to screen a case or file charges
(Subsection 67-5-1(2)(b)).
• Authority to seek a court order to compel compliance by a district attorney or county attorney
with requirements to provide certain records to the attorney general (Subsection 67-5-1(2)(e)).
If the Legislature is interested in granting specific enforcement authority to the lieutenant governor to
enforce compliance with the law by local election officials, the Office of Legislative Research and
General Counsel can conduct research and present them with options to consider.
Thomas R. Vaughn
Managing Associate General Counsel
OFFICE OF LEGISLATIVE RESEARCH AND GENERAL COUNSEL | Page 3 of 3

Audit Title

Agency Response
Audit Title

Audit Title

STATE OF UTAH
SPENCER J. COX OFFICE OF THE LIEUTENANT GOVERNOR DEIDRE M. HENDERSON
GOVERNOR SALT LAKE CITY, UTAH LIEUTENANT GOVERNOR
84114-2220
December 5, 2022
Auditor General Kade R. Minchey
Office of the Auditor
Utah State Capitol Complex
East Office Building, STE 310
Salt Lake City, UT 84114
Dear Mr. Minchey,

It has been a pleasure to work with you and your team. First, I appreciate the professional
manner with which the legislative auditors conducted themselves and their efforts to listen, learn,
and gain a complete understanding of the complexity of our election systems.
We are pleased the legislative audit found no evidence of systematic problems, widespread
errors, or significant fraud. Further, the audit found that any bad actors would need to defeat
multiple layers of defensive control measures to undermine election integrity in Utah. The
systems in place are the result of careful and deliberate efforts by the Legislature over many
years to ensure that Utah voters have access to the ballot, confidence in the security of our
elections, and the ability to hold those who run them accountable for their job performance.
We will always have more work when it comes to election systems in Utah. We are never
satisfied with the status quo. Instead, we are continually looking for ways to improve. To that
end, we appreciate the findings and recommendations of this audit—many of which we have
already implemented. We look forward to working with the Legislature to continue our shared
commitment to strengthening Utah’s elections.
Sincerely,
Deidre M. Henderson
Lieutenant Governor

Recommendation 2.1
We recommend that the Office of the Lieutenant Governor finalize standards for the frequency
and use of key VISTA maintenance tools, and then monitor their implementation.
Response: We agree.
Recommendation 2.2
We recommend that the Legislature consider clarifying the oversight role of the Office of the
Lieutenant Governor with regards to voter registration. This could include authority for regular
analysis of voter records and rulemaking authority for minimum maintenance standards.
Response: We agree. During the 2022 legislative session, our office worked with county clerks
and legislators on a bill (H.B. 313) to improve election law and procedures. A critical part of this
legislation was the implementation of a voter registration audit. The initial voter registration
audit we conducted in August and September identified findings like those in this legislative
audit. We are confident that the recently implemented controls will allow us to refine our process.
Recommendation 2.3
We recommend that the Office of the Lieutenant Governor monitor and ensure that the names of
deceased voters are removed from voter rolls, as required in statute.
Response: We agree.
Recommendation 2.4
We recommend that the Office of the Lieutenant Governor review its staffing plan to determine
whether critical functions have sufficient staff. If, in this review, the Office of the Lieutenant
Governor determines that they need more resources for monitoring and maintenance, we
recommend that they report these needs to the Legislature.
Response: We agree. Our office currently oversees elections with five employees, 16 to 45 fewer
FTEs than the audit’s comparison states. We will continue doing as much as possible with our
current resources; however, any additional training, oversight, and critical function redundancy
recommendations in this audit would require more staff to implement.
Recommendation 3.1
We recommend that the Office of the Lieutenant Governor create rules requiring county clerks to
publicly reconcile the number of ballots tabulated with the number of voters given vote credit in
VISTA.
Response: We agree. We implemented a similar policy for the 2022 General Election. Pursuant
to authorization from the Legislature, we will adopt rules regarding this recommendation.

Recommendation 3.2
We recommend that the Office of the Lieutenant Governor help clerks identify the best data
possible to reconcile canvass reports and vote credit numbers.
Response: We agree. We implemented this recommendation during the 2022 General Election.
Recommendation 3.3
We recommend that the Office of the Lieutenant Governor create rules requiring ballot envelopes
be counted as early as possible during ballot processing.
Response: We agree. Our office has submitted a rule currently in the rulemaking review process.
This new rule (R623-8) requires county clerks to batch ballots and account for them from the
time ballots arrive in their offices until they are tabulated and stored.
Recommendation 3.4
We recommend that the Office of the Lieutenant Governor finalize its chain-of-custody manual,
including best practices for election staff in Utah’s counties.
Response: We agree. Our office is in the process of finalizing a chain of custody manual.
Recommendation 4.1
We recommend that the Legislature consider including either clearer standards for signature
review, acceptance, and rejection in Utah Code, or giving the lieutenant governor authority to
establish these standards and instructions.
Response: We agree. During 2021 and 2022, our office worked with a legislator on a bill that
failed to pass in either legislative session. The bill would have granted rulemaking authority to
develop signature standards and training. It would have also created a certification program for
election workers and provided the funding necessary for implementation.
Recommendation 4.2
We recommend that the Legislature consider giving the Office of the Lieutenant Governor
explicit authority to establish training standards and requirements for signature verification to
better assure that practices are executed consistently across the state.
Response: We agree.
Recommendation 4.3
We recommend that the Legislature consider requiring county election staff to attend signature
verification training.
Response: We agree.

Recommendation 4.4
We recommend that the Legislature consider the options in this chapter to improve the quality of
signatures available for signature verification.
Response: We agree.
Recommendation 4.5
We recommend that the Legislature consider giving authority to the lieutenant governor to
establish rules requiring counties to conduct mid-election audits of signatures and performance
tracking for signature reviewers.
Response: We agree.
Recommendation 4.6
We recommend that the Legislature consider guidelines for, or the necessity of, curing
Response: We agree.
Recommendation 5.1
We recommend that the Legislature consider options to increase the independence of any
post-election audit.
Response: We agree.
Recommendation 5.2
We recommend that the Legislature consider establishing a risk-limiting audit pilot program to
enhance Utah’s post-election audit methods, giving the Office of the Lieutenant Governor
rulemaking authority to establish standards.
Response: We agree. In addition to the audit policy that we enacted for the 2022 General
Election, we will continue to seek ways to strengthen our post-election audit methods.
Recommendation 5.3
We recommend that, if the Legislature authorizes a risk-limiting audit pilot program, the Office
of the Lieutenant Governor should create rules to establish common procedures.
Response: We agree.

Recommendation 6.1
We recommend that the Legislature consider adding election standard oversight and enforcement
responsibilities and mechanisms to Utah Code.
Response: We agree.
Recommendation 6.2
We recommend that the Office of the Lieutenant Governor implement any oversight and
enforcement responsibilities as required by the Legislature in response to this audit.
Response: We agree.
Recommendation 6.3
We recommend that the Legislature consider whether to require county election staff to
Response: We agree.
Recommendation 6.4
We recommend that the Office of the Lieutenant Governor determine the best way to support and
provide training for clerks on the controls and procedures of Utah elections.
Response: We agree. Our office will develop additional training plans and materials to the
extent the Legislature gives us more resources and authority to supplement the training we
currently provide.
Recommendation 6.5
We recommend that the Office of the Lieutenant Governor finalize its manual of standards to
help guide clerks’ elections administration.
Response: We agree. Our election handbook is in the final drafting stages before its scheduled
release in the first quarter of 2023.

Office of the Legislative Auditor General Audit Title
120
Olag.utah.gov

2022 Utah Election Audit

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

2022 Utah Election Audit

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2022 Utah Election Audit

Uploaded by

Copyright:

Available Formats

Report No.

A Performance Audit of Utah’s

Office of the Legislative

Senator Evan J. Vickers Representative Mike Schultz

Senator Karen Mayne Representative Brian S. King

Office of the Legislative Auditor General

Audit Subcommittee of the Legislative Management Committee

December 13, 2022

Transmitted herewith is our report:

Kade R. Minchey, CIA, CFE

PERFORMANCE ELECTION SYSTEMS AND CONTROLS

AUDIT REQUEST KEY FINDINGS

3.1 Canvass Ballot Totals from 2022

1.1 Despite Opportunities to Improve, Utah’s

A Performance Audit of Utah’s Election System and Controls 3

4 Office of the Legislative Auditor General

Figure 1.1 provides a high-level view of election

An Example Highlights Election Controls in

• If the fraudster is unable to steal authorized ballots, creating envelopes

A Performance Audit of Utah’s Election System and Controls 5

• It is also important to note that each instance of election forgery is a felony

6 Office of the Legislative Auditor General

• Chapter 4: A lack of clarity in standards and training for signature

• Chapter 5: Opportunities to refine the purpose and results of post-election

1.2 Differences in Other States’ Election Laws Can

• Instead of Wisconsin’s rigid legal procedure to send absentee/mail ballots

A Performance Audit of Utah’s Election System and Controls 7

1.3 Prosecution Is Rare Because Reported

3 The other case was related to gathering petition signatures.

8 Office of the Legislative Auditor General

2.1 Mistakes Within the Voter Registration Database

4 Utah Code 67-1a-2(2)

A Performance Audit of Utah’s Election System and Controls 11

Standards for Internal Control in the Federal Government

Management should remediate identified internal control

The LG’s Office has recently taken steps toward a

12 Office of the Legislative Auditor General

Maintenance of voter data is an ongoing task that requires constant vigilance as

A Performance Audit of Utah’s Election System and Controls 13

VIST Data Sources

VIST Maintenance Tools

Annual aintenance Vote

• Evaluated compliance with laws that require timely removal of deceased

• Compared voter history statistics against official canvass reports to

14 Office of the Legislative Auditor General

voter statuses, see Appendix B.

A Performance Audit of Utah’s Election System and Controls 15

Twenty-seven out of twenty-nine counties have used the annual maintenance

16 Office of the Legislative Auditor General

Name Mismatch 1.4%

SSN Mismatch 1.0%

DOB Mismatch 0.2%

A Performance Audit of Utah’s Election System and Controls 17

ounties Sufficiently ec ounties Sufficiently ec

Source: Auditor analysis of VISTA

10 Utah is a member of Electronic Registration Information Center, Inc. (ERIC). ERIC is a

18 Office of the Legislative Auditor General

Additional Statewide Monitoring Could

forward the case to the appropriate prosecutor.