SEPTEMBER 2020

CRITICAL
CONDITION

AMERICAN VOTER
ROLLS FILLED WITH
ERRORS, DEAD VOTERS,
AND DUPLICATE
REGISTRATIONS

Image of graveyard electioneering captured March 21,

2018 in Roma, Texas at Roma Cemetery/Queen of Peace
Memorial Park courtesy of Logan Churchwell/PILF
26.401511, -99.007149
 CRITICAL CONDITION

AMER ICA N VOTER RO LLS F ILLE D WIT H

ERR OR S , DEA D VOT E RS , AN D DU PLICAT E
R EG IST RAT IO N S

QUESTIONS?
[email protected]

I
 TABLE OF CONTENTS

LETTER FROM THE PRESIDENT 01

WHY WE DID THIS 05

SUMMARY OF FINDINGS 08

PART I — THE DATA

DECEASED REGISTRANTS & VOTING CREDITS 09

METHODOLOGY 13

DUPLICATE REGISTRANTS WITH TWO SEPARATE VOTING CREDITS IN

15
SAME ELECTION

WHAT CONSTITUTES CASTING TWO BALLOTS? 19

INTRASTATE DUPLICATES 23

WHY DO DUPLICATE REGISTRATIONS MATTER? 25

THE INTERSTATE CROSSCHECK 27

VOTES CAST FROM NON-RESIDENTIAL ADDRESSES 30

PART II — DEPLOYMENT

LITIGATION & OFFICIAL OUTREACH 34

THE FOUNDATION’S LITIGATION 35

THE COVID CASES 39

A TOOLBOX ADDITION/POTENTIAL REPLACEMENT FOR VOTER

43
REGISTRARS

SAVE PREQUEL REPORTS 45

DATA LITIGATION 47

CONCLUSION 49

END NOTES 51

II
 LETTER FROM THE
FOUNDATION PRESIDENT
In February of the Presidential Advisory
2012, the Commission on Election Integrity
PEW Research in 2018, for which I was selected
Center by President Donald J. Trump to
conducted serve. We quickly set to work
research that approaching states to collect
stunned many. public voter registration data
PEW’s report, for comparison and audit. You
Inaccurate, probably know how the rest of the
Costly, and story goes: dozens of lawsuits,
Inefficient, stonewalling and refusal by many
found millions state officials to provide the data,
of deceased registrants on the name-calling, and even some old-
voter rolls.[1] Millions more were fashioned political grandstanding,
registered multiple times in including by members of the
multiple states at the same time. Commission who opposed the
Another 10 million had incorrect President and the mission of
or outdated addresses. Few could the Commission to research the
have predicted that one in eight extent of the problem and the
voter registrations across the vulnerabilities of the election
nation had a problem. In the years process. The state of Vermont
since, I have cited that research even passed a statute prohibiting
scores of times. cooperation with the Commission.

But here’s the problem: as Millions of dollars poured into

surprising as those numbers were, lawsuits filed by advocacy
they are from eight years ago. It groups designed to shut down
was time to update them. the Commission’s work. Today,
dishonest media report that the
The first known attempt to update “Commission shut down without
these figures came in the form finding any voter fraud.” That

1
standard false line has become a federal disclosure laws and
dishonest characterization of what turn over information to which
happened. We didn’t find voter the public has access under the
fraud because we never got a requirements of the National Voter
chance to look. Registration Act.

Now that has changed.

In late 2019, the Public Interest

Legal Foundation (Foundation)
committed itself to doing what
the Commission never got the
opportunity to do: build an active
database of all the nation’s voter
registration and voter history data.
Like the Commission had planned,
this database – dubbed SAVE
(Safeguarding America’s Votes
and Elections) has comparison
capabilities against itself and
against other useful data sources
like the Social Security Death This report provides the nation
Index. SAVE is now operational. with a status check on the health
It can identify deceased and and quality of its voter rolls ahead
duplicate voters. It can identify of the 2020 Election. The timing is
registrations at ineligible addresses critical. Partisan advocates have
like post office boxes or mailbox fought for automatic vote-by-mail
stores, and even people registered systems in which all registered
in multiple states simultaneously. voters are sent ballots. These
efforts must assume the voter rolls
SAVE has all the public state are accurate enough to sustain
registration data from 42 states. this sort of radical transformation
Three states that withheld the to how we conduct elections. They
data from the Foundation were are not.
promptly sued – Illinois, Maine[2]
and Maryland. States must follow There is plenty of good news in this

2
report. Nearly a decade after those We should not confuse
notorious PEW numbers went viral, improvement with completion.
the Foundation saw how years As made evident in this report
of litigation by election integrity there is still much work to be done
groups such as the Foundation, to ensure and preserve clean,
Judicial Watch, the Election accurate voter rolls. Adding an
Integrity Project of California, the additional challenge, in most states
American Civil Rights Union, and the current system places too
others, has made a difference. much reliance on fixing bad data
Groups who fight to preserve the with communications occurring at
broken status quo – such as the the polling place. This opportunity
A. Philip Randolph Institute, the is lost with mail ballots.
League of Women Voters, and
the Brennan Center – no longer What does this report
enjoy an unopposed field. There show? It appears there are
is now advocacy, outreach, and hundreds of thousands
networking by election integrity of undetected dead
groups with local and state registrants, dead registrants
election officials to improve the casting ballots, registrants
accuracy of their voter rolls and with multiple registrations
these efforts have had measurable within the same state and
success. different states, people
voting twice across state
Instead of praising those good lines, and many registered
government efforts, liberal groups at improper commercial
have instead attacked them in the addresses like casinos, gas
public arena and in the courts, stations, and restaurants.
falsely portraying them as an
effort to disenfranchise voters.
To the average voter reading
The exact opposite is true – these
through these pages with alarm,
efforts are intended to protect the
I ask that you do just one thing
most precious right that voters
in these final months before the
have: the value of their ballot.
election: make sure that your
household’s voting records are
These groups are reactionary, and
current and free of errors – and
set in their ways.
that no one is registered at your

3
residence who does not actually When we discover that Rashawn
reside there and whose name Slade of Swissvale, Pennsylvania,
you do not recognize. Improving has seven active registrations
the nation’s voter rolls starts at because a third-party voter
home—yours to be precise. drive registered him seven times
in the weeks before the 2016
When voters have confidence Election (despite it being legal to
in the system, they are more be registered in duplicate) – do
likely to participate. Fixing something about that. When we
errors, duplications and obsolete learn that some who died in the
registrations will increase 1990s remain active on Detroit’s
confidence in the voting system voter rolls – do something about
and we hope clear the last barrier that. Stop attacking citizens and
to participation: doubt in the organizations like the Foundation
integrity of the process. who find and report these failures.
Be part of the solution, not part of
Finally, the Foundation is the first the problem.
to undertake completion of this
sort of groundbreaking study. I sincerely thank the supporters
Academics, law professors, and of the Public Interest Legal
liberal think tanks could have Foundation for helping see this
done this long ago to improve work done and you, the reader, for
the system. They did not. They taking the time to better educate
have other priorities. Instead, they yourself on the strengths and
have created a cottage industry weaknesses of our shared voter
unfairly trying to discredit those registration and election systems.
seeking to improve the system. I
would invite them to evolve from
being part of the problem to part
of the solution. It’s time to use your
vast war chest to fix things rather
than destroy state laws designed
to bring integrity and order to
our elections. Instead of trying to J. Christian Adams
impede improvements, urge states President and General Counsel
to fix the problems we find here. Public Interest Legal Foundation

4
WHY WE DID THIS
You can always count on a rolls, you must first overcome
blizzard of “reports” right before barriers erected by the keepers
an election. Though these efforts of the data. In some states, the
typically fall into one of two cost to purchase the voter roll
categories: “voter fraud” or “voter data is extraordinary. Other
suppression,” they almost always states limit access to favored
have one thing in common: groups, or worse, deny access
they are based on statistical entirely, forcing you to endure
extrapolation rather than an costly litigation to access these
audit of hard data. Whether important public records, as we
the focus is on the number of have done.
foreign nationals participating
in a given election, errors in a Even if you successfully obtain
voter registration system, the the data, you must resolve the
number of registrants supposedly challenge of making each state’s
impacted by a new voter ID law, data comparable to all others
or some other election topic, the in terms of format uniformity.
methodologies primary rely on Then you have to compare the
supposition and estimates. government data with private
data to sift out as many false
This repor t is entirely different. positives as you can.[3] Perfection
is never possible. The task is
to maximize confidence in the
Granular research concerning
results.
voter registration data is a
rarity—and for good reason.
For this effort, the Foundation
It is expensive and as the
collected every state’s list
Foundation’s litigation shows,
maintenance records where
as well as the experience of the
Presidential Commission, election ...election officials are too often
officials are too often reluctant reluctant to provide it [voter
to provide it despite laws registration data] despite laws
requiring them to do so. Before requiring them to do so.
you can audit the nation’s voter

5
possible. As of this reporting series, the Foundation gathered data
from 42 states. The remainder was either blocked by local law
(for which the Foundation is pursuing litigation in federal court) or
insurmountable gaps in the poor quality of the state’s data impaired
the research. After the data were collected from the states and put
into a format that it could be studied, it was compared to commercial
data and other government data to increase confidence in the
conclusions. Also included with
the data were public records
indicating when a person ...this report seeks to give
Americans data-driven information
was credited with voting. The
of what is broken in our election
combination of state election systems and to think about ways
data, commercial data, other to fix those problems and remedy
government data such as the our vulnerabilities.
Social Security Death Index,
provides researchers with
perhaps the best platform ever constructed to analyze the health of
the voter rolls and catalog potential voter fraud vulnerabilities.

The Foundation is currently pursuing litigation against the States of

Illinois[4], Maryland[5] and Maine[6] to develop future findings.

The granular findings outlined in this and recent reports were

generated to inform the electorate and serve as the basis for voter
roll clean-up efforts and, where appropriate, criminal and civil law
enforcement investigations. More importantly, this report seeks to give
Americans data-driven information of what is broken in our election
systems and to think about ways to fix those problems and remedy
our vulnerabilities.

6
We also wish to assure the electorate that someone is pursuing
clean-up efforts in order to instill confidence in the integrity of the
system and thereby promote robust voter participation.

The problems the Foundation’s study found varied from state to state.
It is possible that a populous state might exhibit significant numbers
of deceased active registrants but show few examples of apparent
duplicate voting. A less populous state might have greater instances
of corrupted voter files – such as missing birthdate or address fields
- or potential fraud than a much more populous state. Every state
shows unique challenges and opportunities to better execute and
enforce the law.

APPARENT DUPLICATE VOTE CROSSCHECK: FLORIDA 2018

7
 SUMMARY OF FINDINGS

Deceased Registrants Across 41 States: 349,773

NY, TX, MI, FL, CA Account for Roughly 51% of National Deceased
Registrants
In 2016, 7,890 Registrants Were Credited for Voting After Death
In 2018, 6,718 Registrants Were Credited for Voting After Death
North Carolina Leads U.S. in Deceased Registrants Credited for
Voting After Death by More Than 4:1

43,760 Duplicate Registrants Appear to Have Cast Second Votes in

2016 from Same Address

37,889 Duplicate Registrants Appear to Have Cast Second Votes in

2018 from Same Address

Tens of Thousands of these Apparent Double Votes Were Exclusively

Mail Ballots

Number of Registrants Apparently Registered in 2 States and

Credited for Voting in Both States in 2018: 8,360

Number of Apparently Duplicate Registrants Credited for Voting

Twice in Same State from 2 different addresses in 2018: 5,500

Number of Registrants Credited for Voting from Apparently

Nonresidential Addresses in 2018: 34,000

Being credited for voting means that government election officials recorded a data
point whether or not a registrant voted in a particular election. A voting credit is
often referred to as “voter history” record or file.

8
 PART I – THE DATA
DECEASED REGISTRANTS & VOTING CREDITS

SAVE collected registration cast a ballot after dying were

and list maintenance data from more common in smaller states.
42 states’ registrant files, kept North Carolina and Oregon
as part of the Help America mark the clearest examples of
Vote Act's mandate regarding this phenomena where smaller
statewide voter registration lists. numbers of deceased registrants
These are public data and public exist on the rolls but a larger
information under the National percentage of them show voting
Voter Registration Act of 1993. credit.
Based upon voter roll data
collected nationwide at the end Oregon has an all vote by
of 2019, 349,773 registrants in mail system. There have been
41 states were matched against convictions in Oregon for voting
commonly utilized death records. mail ballots for those who have
These include New York (59,000 died.[7]
dead registrants) and Texas
(36,000 dead registrants), for When it comes to dead
example. From those matched registrants and voting credits,
dead registrants, 14,608 voting the Foundation’s research is
credits were highlighted as constrained by the accuracy,
occurring after respective dates of or lack of accuracy, of the
death and outside any extended government records related to
early or absentee voting periods. voting credits. Voting credits are
usually assembled well after an
Though states with larger election. Officials, for example,
populations usually had the must review poll books and
largest number of deceased sign-in sheets to see who voted,
registrants still on the voter and to modify the statewide list
rolls, those where government maintenance records to reflect
records showed the registrant this activity.

9
These modifications to the voter New York, Texas, Michigan,
credit data are not insignificant Florida, and California make up
and data entry errors can impact 51 percent of all matched dead
a registrant’s ability to remain on registrants across the United
the rolls and vote. If voting credits States. In other words, these five
are not provided after a registrant states account for a majority
votes, then that registrant is of the instances where the
shown to be inactive in the deceased are remaining on the
election, and their ability to stay rolls. Following closely behind
on the rolls if they don’t respond with at least 11,000 apiece were
to official notices from election Pennsylvania, South Carolina,
officials could be improperly North Carolina, Connecticut,
impaired. Virginia, New Jersey, and
Tennessee.
For this reason, the Foundation
has relied on voting credit data MOST DECEASED VOTING
in the official list maintenance CREDITS PER STATE - 2018
records to reach conclusions North Carolina 2,172
about duplicate votes and
deceased persons receiving Mississippi 723
credits. Because the voting credit Kentucky 652
data in the government records Minnesota 455
are significant, we attach a California 350
degree of reliability to them that
is commensurate with such an Oregon 337
important piece of information.
That being said, it is certainly MOST DECEASED VOTING
possible that election officials are CREDITS PER STATE - 2016
making mistakes regarding voting North Carolina 2,454
credit data. If that is the case, Kentucky 710
then this report serves to initiate
Mississippi 662
an examination of a previously
undetected breakdown in election New York 549
administration – false positives Minnesota 509
regarding voting credit data. California 424

10
 DECEASED REGISTRANT

TOTAL
2018 2016 TOTAL
RANK STATE MATCHED
CREDITS CREDITS CREDITS
DECEASED
TOTALS 349,773 6,718 7,890 14,608
1 New York 59,096 147 549 696
2 Texas 36,054 136 153 289
3 Michigan 34,225 97 104 201
4 Florida 25,162 100 117 217
5 California 23,414 350 424 774
6 Pennsylvania 16,685 127 177 304
7 South Carolina 14,351 50 273 323
8 North Carolina 12,940 2,172 2,454 4,626
9 Connecticut 11,948 96 103 199
10 Virginia 11,903 43 47 90
11 New Jersey 11,814 87 105 192
12 Tennessee 11,126 46 42 88
13 Missouri 8,358 27 34 61
14 Alabama 7,922 41 66 107
15 Ohio 7,368 51 51 102
16 Wisconsin 6,805 163 186 349
17 Mississippi 6,738 723 662 1,385
18 Oklahoma 5,302 27 45 72
19 Washington 5,274 26 - 26
20 Georgia 4,243 12 17 29

11
TS AND VOTING CREDITS

TOTAL
2018 2016 TOTAL
RANK STATE MATCHED
CREDITS CREDITS CREDITS
DECEASED
21 West Virginia 3,411 7 10 17
22 Nevada 3,258 4 7 11
23 Iowa 3,000 26 36 62
24 Kansas 2,338 8 10 18
25 Arizona 2,289 28 23 51
26 Utah 1,992 273 259 532
27 Arkansas 1,921 19 15 34
28 New Mexico 1,682 2 5 7
29 Rhode Island 1,637 - - -
30 Kentucky 1,280 652 710 1,362
31 Colorado 1,119 28 27 55
32 Nebraska 1,009 4 7 11
33 Minnesota 816 455 509 964
34 Idaho 708 242 246 488
35 Delaware 643 6 6 12
36 Montana 635 4 3 7
37 Oregon 469 337 342 679
38 Alaska 318 61 62 123
39 Wyoming 311 38 - 38
40 South Dakota 105 3 4 7
41 Vermont 104 - - -

12
 METHODOLOGY

The Foundation utilized a combination of three sets of primary death

record sources. The Social Security Death Index is the repository for
information about social security beneficiaries who have died. When
the Social Security Administration receives notice that someone has
died – often through an application for death benefits – the event is
incorporated into the Social Security Death Index (SSDI). The SSDI
contains false negatives. For example, not everyone who dies is in the
SSDI. Some family members never notify the SSDI or never seek death
benefits. The SSDI was designed to keep tabs on its benefits-receiving
customer base, not act as an exhaustive listing of all deceased
Americans with a Social Security number.

Relying only on the SSDI to catalog the deceased can omit younger
people who have died. To account for this issue, the Foundation
additionally compared full voter roll data against national obituary
data. Of course, not everybody who has died is in the SSDI or
has a paid-for obituary. But these missing examples of deceased
registrants, if they were known, would only push the number of
deceased registrants on the rolls higher, not lower.

Once SAVE identified matches between the voter registration lists

and SSDI and obituaries, an additional and important step was
undertaken. These matches were then screened against the three
major credit bureaus (Experian, Equifax, and TransUnion) for final
confirmation that the person on the rolls and the person who had
died were one and the same person. This process allows for high-
confidence confirmation for corresponding dates of birth, death,
and address history. There are other potential sources of data to
refine these results further but either the Foundation did not have
any access to those data or the costs to utilize those data were
exponentially higher.

13
NEW YORK FLORIDA
National Rank: 1 National Rank: 4
59,096 Dead Registrants 25,162 Dead Registrants
549 Voter Credits in 2016 117 Voter Credits in 2016
147 Voter Credits in 2018 100 Voter Credits in 2018

TEXAS CALIFORNIA
National Rank: 2 National Rank: 5
36,054 Dead Registrants 23,414 Dead Registrants
153 Voter Credits in 2016 424 Voter Credits in 2016
136 Voter Credits in 2018 350 Voter Credits in 2018

MICHIGAN NORTH CAROLINA

National Rank: 3 National Rank: 8
34,225 Dead Registrants 12,940 Dead Registrants
104 Voter Credits in 2016 2,454 Voter Credits in 2016
97 Voter Credits in 2018 2,172 Voter Credits in 2018

14
 DUPLICATE REGISTRANTS WITH
TWO SEPARATE VOTING CREDITS
IN SAME ELECTION
A duplicate registration is when has yielded some of the most
a person is registered to vote surprising results of all the in-
more than once at the same time. depth research in this report.
Sometimes duplicate registrants Simply, states are allowing
have more than two active the same persons to register
registrations. It is not necessarily multiple times without detection.
illegal to have two active In Pennsylvania, for example,
registrations, but it is a felony to the Foundation found one
vote more than once for the same registrant with seven separate
federal office under federal law. active registrations, all at the
Various state laws make it illegal same address, with the same
to vote more than once. States name and date of birth. All seven
are required under federal law registrations had separate active
to have a statewide database state issued voter identification
that detects and fixes duplicate numbers. This registrant was
registrations. The National Voter active on the rolls for four years
Registration Act (NVRA) and before the Foundation discovered
Help America Vote Act allow him, according to local officials.
for enforcement actions against
election officials for not having Why aren’t the states catching
adequate list maintenance their own duplicates? The Help
practices. America Vote Act of 2002
was specifically designed to
Since 2017, the Foundation has prevent this from happening.
studied several local jurisdictions’ The Department of Justice is
voter rolls to quantify the tasked with enforcing the laws
number of times duplicate that relate to maintaining the
registrations were created—and statewide database free from
how long they went undetected such duplicates, among other
by election officials. This work requirements.

15
The truth is that the career All of the findings of duplicates in
employees in the Voting Section this report speak directly to the
of the Civil Rights Division of ineffective enforcement of the
Department of Justice have no Help America Vote Act (HAVA)
interest in enforcing this law. If since its adoption in 2002 by
they do – as they did against the lawyers at the Department
New York and California – of Justice (DOJ). They do not
they do so only in the most even detect the problems or
extreme circumstances. New view enforcement of HAVA and
York refused to adopt modern the NVRA as outlier priorities.
voting machines, so they acted. Many formerly worked at the
California was negligent in NAACP, ACLU or similar groups
implementing a statewide voter that oppose enforcement of
database not for a year or two, HAVA and the NVRA to clean
but for over a decade. Only up and maintain accurate voter
then did the career attorneys registration rolls. An investigation
in the Voting Section bring an by the DOJ Inspector General into
action. But they have filed very the failure to enforce the NVRA
few actions to enforce these and HAVA found that DOJ staff
provisions. recruited applicants who were
likely to be opposed to enforcing

16
these laws by selecting only states are following their own
liberal lawyers to recruit and rules is reflected in the data.
alerting organizations about One thing is certain – each state
vacancies that were almost filed a HAVA plan with the U.S.
exclusively left of center. The Election Assistance Commission
Inspector General urged the on how it intended to use federal
Voting Section to reevaluate funds to comply with HAVA.
existing hiring practices. Those
[8]
The DOJ’s Voting Section has
organizations routinely appear authority to enforce HAVA’s list
as intervenors to prevent, as they maintenance mandates when
derisively call list maintenance, these HAVA plans are ignored,
“purges.” This same mindset resulting in swarms of duplicates
has infected the career ranks of as SAVE demonstrates has
the Voting Section at the Justice occurred.
Department
and every In states prone
...databases and the clerks
single instance operating them are regularly to duplications,
of duplicate creating second or third it is easy to
registrants on registrations when the registrant register to vote
statewide rolls intended for only an update to multiple times.
confirms this their existing file. Whether it
bias and their is a surname
failure to act to change,
properly enforce the law. transposed digits in a birthdate,
a typographical error, a deputy
Despite DOJ failures to enforce voter registrar’s error (phonetic
the law, this body of research confusion like “Devereaux” versus
shows that some states are “DeVero”) or even conflicting
better than others at either gender claims, voter databases
preventing registration and the clerks operating them are
duplication from occurring, or regularly creating second or third
were equipped to catch the errors registrations when the registrant
in a reasonable amount of time. intended for only an update to
Every state studied to date has a their existing file.
protocol for handling suspected
duplicate registrations. Whether It is not necessarily illegal to

17
be registered twice—whether it’s in the same house or across two
states—especially if an officer cannot prove some semblance of
intent. However, it is unlawful to cast multiple ballots in the same
election. Currently, 11 states have laws prohibiting voting in that state
and another.[9] Seven states have explicit laws barring double votes in
those particular states.[10] Thirty-one states generally prohibit double
voting in the same election.[11] Federal law also bars “voting more
than once.”[12] Within federal and most state laws, double voting is
considered a felony.

A startling number of same-address duplicate registrants apparently

took advantage of the extra votes in 2016 and 2018. In the 2016
Presidential Election, 43,760 registrants across 33 states purportedly
cast second ballots, according to voting credits issued to their records.
For the 2018 General Midterms, 37,889 registrants across 31 states
did the same. Again, these conclusions depend on the accuracy of
state voter credit data. If in fact states are bungling the recording
of voter credit data (and these registrants are not casting second
ballots) then this too must be fixed. Voter credit data is an important
list maintenance datapoint because it determines when it is legal and
appropriate to remove someone from the voter rolls. But if state voting
records are accurate, this represents potentially over 80,000 instances
of election fraud in just two elections.

18
 WHAT CONSTITUTES CASTING TWO BALLOTS?
A registrant who has multiple active registrations at the same address, for
the purposes of this report, must match across the following parameters:
• First name
• Last name characters (allows matching for a maiden name with a
subsequent hyphenated name)
• Date of birth (with matching for transposed or incomplete
numbers)
• Full address and applicable apartment/unit number
• General election voting credit

SAME-ADDRESS DUPLICATES SHOWING

SECOND VOTE CREDIT IN 2016
RANK STATE COUNT RANK STATE COUNT
1 North Carolina 9,734 18 Texas 13
2 Georgia 9,619 19 New Jersey 11
3 Michigan 7,140 20 Connecticut 8
4 Colorado 3,445 21 South Carolina 8
5 Mississippi 3,368 22 Florida 7
6 Arizona 3,277 23 Alabama 4
7 Wisconsin 2,066 24 Minnesota 4
8 Delaware 1,816 25 Arkansas 3
9 California 903 26 Nebraska 3
10 New Mexico 863 27 Tennessee 3
11 Utah 585 28 Ohio 2
12 Idaho 301 29 Kansas 1
13 Alaska 209 29 Kentucky 1
14 Vermont 134 29 Oklahoma 1
15 South Dakota 122 29 Oregon 1
16 New York 94 29 West Virginia 1
17 Pennsylvania 13 43,760

19
 SAME-ADDRESS DUPLICATES SHOWING
SECOND VOTE CREDIT IN 2018
RANK STATE TOTAL RANK STATE TOTAL
1 Georgia 9,899 17 New York 36
2 North Carolina 7,026 18 Pennsylvania 14
3 Michigan 6,457 19 New Jersey 10
4 Colorado 3,175 20 Texas 8
5 Arizona 3,077 21 Connecticut 7
6 Mississippi 2,428 22 Florida 6
7 Wisconsin 1,828 23 Alabama 3
8 Delaware 1,304 24 Arkansas 3
9 New Mexico 721 25 South Carolina 3
10 California 624 26 Kentucky 2
11 Utah 549 27 Kansas 2
12 Idaho 235 28 Ohio 1
13 Alaska 190 28 Oklahoma 1
14 South Dakota 100 28 Oregon 1
15 Vermont 96 28 West Virginia 1
16 Wyoming 82 37,889

An alarming trend that emerged from both the 2018 and 2016
datasets was the prominence of duplicate registrants being credited
for casting two absentee/mail-in ballots at the same time in the same
election. The Foundation was able to determine, when detailed data
was provided by the various states, how often these apparently
duplicate registrants either participated by voting twice by mail or by
voting both in-person and again by mail in a single election.

Mail ballots appear to be the problem. Remember, someone casting a

mail ballot need not present themselves in front of any election official.
Moreover, groups like the League of Women Voters are even seeking
to invalidate witness requirements for mail ballots in the 2020 Election.

20
 2016 GENERAL ELECTION While in conversations with
RANK STATE MAIL-MAIL % election officials about these
1 North Carolina 51% data, one question is uniformly
2 Georgia 43% asked: what does it take to get
3 Michigan 68% assigned a mail ballot credit
4 Colorado 90% as opposed to a polling place
6 Arizona 9% one? Generally speaking, a
7 Wisconsin 14% polling place credit means that
9 California 34% a poll check-in was successfully
10 New Mexico 4% completed by the voter. One way
11 Utah 59% a counted ballot is not behind
that credit is if the voter chose
22% of Arizona’s 2016 double not to complete the balloting
vote credits involved mail and process at the site after he or
subsequent in-person vote she checked in. Another way
combinations. a counted ballot is not behind
the credit is if the election
official – either at the polling
34% of New Mexico’s 2016
double vote credits involved mail
site or processing the poll books
and subsequent in-person vote after the election – mistakenly
combinations. assigned an in-person credit to
the wrong registrant. As for mail
ballot credits, only returned and
2018 GENERAL ELECTION
counted ballots result in such
RANK STATE MAIL-MAIL %
credits. Rejected, surrendered,
1 Georgia 42%
spoiled, or otherwise missing
2 North Carolina 42%
ballots do not result in a credit.
3 Michigan 69%
4 Colorado 95%
5 Arizona 57%
7 Wisconsin 10%
9 New Mexico 5%
10 California 37%
11 Utah 84%

21
22
 INTRASTATE DUPLICATES
Even though HAVA required that states build a single database to
house all registered voters to prevent—in part—duplication across
county lines, the same can apparently still happen in 2020. The
premise is simple: an 18-year-old registers in her hometown in county
A, goes off to college in county B, starts
a career and family in county C. All along
the way, she is updating her registration
and canceling the previous one with each
transaction—or so she thought.

In 24 states, apparently duplicate registrants

were given double voting credits in 2018
across each state’s county lines or even
within the same county. As an example, a
registrant in Wake County, North Carolina,
with matching full name, date of birth, and Social Security number
appearing on the voter rolls twice with two different registration
addresses was credited for voting at each location with two absentee
ballots. This type of scenario—regardless of voting method—occurred
more than 5,500 times during the 2018 Midterms nationwide.

In theory, personally identifying information (PII) like a Social Security

number would serve as a common denominator when a person
changes addresses and wishes for her voter registration record to
reflect those life changes. But in the vast majority of states registrants
are not required to provide a full Social Security number (SSN). As
described above, some states’ database systems are designed in a
way to confuse Michael Smith and Michael M. Smith for two different
people even though they were both born on July 4, 1976 and live at
123 Main Street Apartment A. It is, therefore, no wonder a duplicate
registration occurs when Michael and Michael M. appear at two
different addresses and one is without a SSN.

23
The table outlines the states with the highest numbers of apparently
duplicate voters with differing addresses and overlapping voter
credits for the 2018 General Election. Reviews of Social Security
numbers and address histories from credit reporting databases
served to validate identity. The review process allowed the Foundation
to access personal address histories maintained by credit reporting
bureaus going back a decade.

SAME-ADDRESS DUPLICATES WITH

DOUBLE VOTE CREDITS

RANK STATE 2018 2016 TOTALS RANK STATE 2018 2016 TOTALS

1 Georgia 9,899 9,619 19,518 18 Pennsylvania 14 13 27

2 North Carolina 7,026 9,734 16,760 19 Texas 8 13 21

3 Michigan 6,457 7,140 13,597 20 New Jersey 10 11 21

4 Colorado 3,175 3,445 6,620 21 Connecticut 7 8 15

5 Arizona 3,077 3,277 6,354 22 Florida 6 7 13

6 Mississippi 2,428 3,368 5,796 23 South Carolina 3 8 11

7 Wisconsin 1,828 2,066 3,894 24 Alabama 3 4 7

8 Delaware 1,304 1,816 3,120 25 Arkansas 3 3 6

9 New Mexico 721 863 1,584 26 Minnesota - 4 4

10 California 624 903 1,527 27 Nebraska - 3 3

11 Utah 549 585 1,134 28 Tennessee - 3 3

12 Idaho 235 301 536 29 Ohio 1 2 3

13 Alaska 190 209 399 30 Kansas 2 1 3

14 Vermont 96 134 230 31 Kentucky 2 1 3

15 South Dakota 100 122 222 32 Oklahoma 1 1 2

16 New York 36 94 130 33 Oregon 1 1 2

17 Wyoming 82 0 82 34 West Virginia 1 1 2

24
 WHY DO DUPLICATE
REGISTRATIONS MATTER?
On the surface, seeing that someone accidentally became registered
twice may seem like a minor matter. After all, there are a number of
ways in which a person can become registered to vote beyond simply
picking up an application and mailing it to the elections office. Even
community college applications can double as voter registration forms
in some states. People are likely to fill out these forms differently over
time, but officials should be able to merge the disparate information
into one file, right?

Not in every state.

Duplicate voter registrations are still very much an underappreciated

concern—but the Foundation isn’t the only outfit sounding the alarm
about duplicates anymore.

“Duplicates are an important indicator in voter file quality,” an

academic paper from CalTech noted in September 2019.[13] Georgia
was the only state during the 2018 election cycle to systematically
report the number of duplicate registrations it cleaned up before
votes were cast. Other states should follow Georgia’s lead for
transparency. Georgia voluntarily revealed it corrected almost 63,000
such files.[14] In late 2019 and early 2020, the Foundation filed federal
lawsuits in Detroit, Michigan, and Pittsburgh, Pennsylvania, where
long-undetected duplicate registrations by the same individual were
specifically cited as voter list maintenance failures under the National
Voter Registration Act of 1993 and state laws.

We believe data such as these detailed here makes enforcement of

the National Voter Registration Act – both by private parties and by
the Justice Department – more targeted, efficient and precise.

25
Duplicates are the easiest voter roll errors to detect and fix. States
possess the data already. Reliance on third party data isn’t needed.
All it requires is sorting the statewide database file to detect
duplicates. In litigation, the Foundation once asked an election official
in a deposition if they had ever sorted the registrant file to detect
duplicates. Not only had they never done it, they never even thought
of doing it. That’s the sad reality of why there are so many duplicate
registrations in America. The existence of so many duplicates has
direct relevance to any proposed movement to vote by mail. Those
who most strongly advocate for vote-by-mail probably never
considered the possibility of so many duplicates on the voter rolls
either.

% OF DUPLICATES CREDITED
RANK STATE
FOR CASTING 2 MAIL BALLOTS

1 Georgia 42.53%

2 North Carolina 47.34%

3 Michigan 68.29%

4 Colorado 92.19%

5 Arizona 31.95%

6 Mississippi 0.00%

7 Wisconsin 12.02%

8 Delaware 0.00%

9 New Mexico 4.23%

10 California 35.17%

11 Utah 71.08%

26
 THE INTERSTATE CROSSCHECK

Checking a state’s voter roll against another state can be a harder

task, but the SAVE Database did just that.

More than 8,360 single registrants were identified in the SAVE

Database as apparently being registered and credited for voting twice
in two different states during the 2018 General Election, a potential
violation of state and federal election laws. That represents over 8,000
potential cases of election fraud.

Again, state election officials are better suited to make final

determinations if indeed these apparently credited double voters are
indeed the same people. State officials will have access to additional
identifying information to confirm SAVE’s findings, but here’s why we
can have confidence in them.

One of the challenges to detecting cross state duplicate registrations

is that there are plenty of people with the same name and date of
birth. When two separate voter registrations of apparently the same
person are detected across state lines, there must be some validation
with PII that these are in fact the same person. The SAVE database
supplemented the voter registration information received from states
with commercial database and credit bureau information to find PII
unique to each registered voter to avoid false positives. It resulted in
150,000 potential double voters being reduced to 8,360 when unique
PII was included in the analysis. This was not an inexpensive process
in building the SAVE Database, but it was important to increase the
confidence of the conclusions.

The SAVE Database focused on a single state and then compared

registration lists from that state against all the others. It then repeated
the process with all of the states in the database.

27
Americans have long heard about the “snowbird” effect on Florida’s
elections over the years featuring New Yorkers and residents of other
northeastern states seeking warmer weather come late fall. There is
considerable truth to the meme—but not all apparent duplicates were
generated by owners of second homes. A relatively even distribution
of apparent duplications existed between coupled states like New
York and Florida (snowbird second home types) and domestic
migrants (nearly any state and Texas). Wherever former Californians
fled to, you could see duplications.

To better demonstrate the findings, this table lays out the states with
the most duplicate sets of registrants who voted in more than one
state.

APPARENT INTERSTATE DUPLICATES

STATE WITH VOTING CREDITS IN MORE
THAN ONE STATE (2018)

California 2,005

Florida 1,836

Texas 1,125

North Carolina 1,029

Pennsylvania 917

Georgia 844

New Jersey 830

New York 793

Michigan 791

Virginia 592

28
As the Foundation has stressed many times in the past, being
registered to vote twice in two (or potentially more) states is not a
criminal act. Registrants do not always notify their former state that
they have moved and some states do not always send information
about a new registrant voter to that registrant's state of former
residence.

Without mandatory cross-state reporting protocols, duplicate

registrations can be hard to avoid. If a person relocates from a
state with substandard voter list maintenance practices, it can take
sometimes up to eight years to fall off their former roll due to inactivity
if they do not successfully alert the jurisdiction of their move. This
highlights the need for states to participate in cooperative ventures to
identify these duplicate registrants. States would be well advised to
take the leads of private organizations – such as the Foundation – that
have spent private dollars to do what the states ought to be doing
more effectively--detect obsolete registrations. Instead of spurning
private efforts, states should embrace them as partners. After all, a
confirmation notice can be sent to the new address to complete the
process, or a subsequent registration can be used as a written notice
from the registrant to cancel the obsolete registration.

29
 VOTES CAST FROM
NON-RESIDENTIAL ADDRESSES
SAVE compared the nation’s voter rolls with commercial addresses. The
data shows that instead of being registered at actual residences as
required by state laws, 35,000 registrations are at commercial locations
such as casinos, gas stations, and restaurants.

In July 2020, Congressman Steve Watkins (R-Kansas) was charged

on three felony counts of alleged voter fraud for claiming a UPS
Store address as
his residence and RANK STATE TOTAL RANK STATE TOTAL

then voting in 2019 1 California 7,244 22 Nebraska 484

municipal elections.[15] 2 Texas 1,952 23 Mississippi 456

Voter registration 3 Virginia 1,772 24 Iowa 451

applications printed 4 Wisconsin 1,653 25 Nevada 429
by the State of Kansas 5 Florida 1,623 26 Tennessee 395
ask the user directly to 6 North Carolina 1,597 27 Missouri 387
give their “residential 7 Arizona 1,435 28 Utah 265
address.”[16] Kansas 8 Alabama 1,336 29 West Virginia 242
law is also clear
9 New York 1,312 30 Connecticut 235
that a permanent
10 Pennsylvania 1,115 31 Wyoming 206
address used for
11 Washington 1,083 32 Minnesota 199
voter registration
12 Ohio 1,047 33 New Mexico 188
purposes must be a
“place of residence.”[17] 13 Georgia 912 34 Idaho 178

Before this case 14 Kentucky 824 35 Michigan 135

arose, the Foundation 15 Colorado 805 36 Rhode Island 115

was ascertaining if 16 Montana 765 37 South Dakota 70

zoned nonresidential 17 Oklahoma 636 38 New Jersey 64
addresses were ever 18 Oregon 617 39 Massachusetts 41
claimed as homes 19 Kansas 599 40 Delaware 16
for the purpose of 20 Alaska 565 41 Vermont 15
voting—and whether 21 South Carolina 528 TOTAL 33,991

30
any voting credits were assigned STATE APPLICATION LANGUAGE

for such voters in the 2018 federal Alabama Address Where You Live
midterm elections. You MUST provide the Alaska
residence address where you claim
Alaska
residency. Do not use PO, PSC, HC
Voting credits were assigned or RR.
to 34,000 registrants across Arizona
Residential Address (where you
41 states at addresses zoned live–no P.O.Box/business address)

for non-residential buildings in Arkansas Address Where You Live

federal midterm elections in 2018. Not a P.O. Box or business address

California (Number, Street, Ave., Drive, etc.
The study focused on addresses Include N, S, E, W)
zoned for light/heavy industrial, Colorado Address (no P.O. Boxes)
office parks, single-use retail, and Connecticut Address Where You Live
service stations. Delaware Street address

Address Where You Live (legal

Although most states are relatively Florida
residence-no P.O. Box)
generous on what they consider RESIDENCE ADDRESS: House No.
Georgia
a residence—a group shelter or and street name

even an overpass is fine as long Residence Address (Do not use

PO Box or business address. If no
as you’re specific on the location— street address, describe location of
Idaho
they tend to draw the line at residence by cross streets, section,
township, range, or other physical
rented mailboxes measuring not description.)
much bigger than a shoebox. A Iowa Address Where You Live
very small minority of states allow Kansas Residential Address
you to claim a commercial address Kentucky Residential Address
as your home if you do indeed live
Massachusetts Address where you live now
there.
address where you live – house
Michigan
number & street name

address where you live (residence)

Minnesota if mail cannot be delivered to the
address above, provide P.O. Box

Physical Home Address(Where you

Mississippi
live)

ADDRESS WHERE YOU LIVE

(HOUSE NO., STREET, APT. NO, OR
Missouri
RURAL ROUTE AND BOX - NO PO
BOXES)
1450 E. Compton Blvd., Compton, CA
17 votes in 2018

31
 STATE APPLICATION LANGUAGE STATE APPLICATION LANGUAGE

Montana Montana Residence Address Residence Address: Street Address

and Apartment Number. If none,
Nebraska Current Residential Address Texas describe where you live. Do not
Nevada Residential Address –See include P.O. Box, Rural Rt. or
Nevada Instructions on Back (No P.O.Box / Business Address
Business Address) Physical Address (required, principal
Utah
Home Address (DO NOT use PO place of residence, no P.O. Box)
New Jersey
Box) My principal dwelling place is
Vermont
Physical Address Where You Live located at
New Mexico
Now Residence address (May not be a
Virginia
New York The address where you live P.O. Box)

Provide your residential address - Washington residential address in Washington

where you physically live. Do not Provide your residence address
North Carolina
enter a P.O. Box or a mail drop (the address where you live). Check
location. West Virginia the box if you live within city limits.
Your voting residence is the Include the name of the county
location that you consider to be where you live.
a permanent, not a temporary, The Address Where You Live your
residence. Your voting residence is Wisconsin residential voting address, which
the place in which your habitation cannot be a P.O. Box
is fixed and to which, whenever you
are absent, you intend to return. Wyoming Wyoming address where you live
Ohio
If you do not have a fixed place of
habitation, but you are a consistent
or regular inhabitant of a shelter or
other location to which you intend
to return, you may use that shelter
or other location as your residence
for purposes of registering to vote.

Street or 911 address or directions

Oklahoma to your home Do not use a rural
route or P.O. box

Oregon Oregon residence address 1695 N. Nellis Blvd. Las Vegas, NV

2 votes in 2018
Pennsylvania Address (not P.O. Box)

Home Address (Do not enter a post

Rhode Island
office box)

South Carolina Address Where You Live

If Residence Address is a PO Box,

rural box, or general delivery, you
South Dakota
must give the location of your
residence

Address Where You Live (legal

Tennessee NPR West Headquarters
residence-no P.O. Box)
9909 Jefferson Blvd. Culver City, CA
4 mail ballots cast from here in 2018

32
33
 PART II – DEPLOYMENT
LITIGATION & OFFICIAL OUTREACH
We have all heard the refrain that voter fraud is rare, and not
widespread. The same people and groups who say this the loudest
never bother to look. They have not undertaken a project such as
this SAVE Database. Indeed, rather than support an effort to fix the
problems SAVE Database has detected, Americans can count on them
to interfere with any proposed solutions, and even attack the existence
of this report.

Shame on them. Had we learned 30 years ago that voter rolls were full
of thousands of dead but active registrants, or thousands of duplicates
potentially voting twice inside states and across state lines, Americans
– both Republican and Democrat - would have been united in calls to
find solutions to fix the problem. Liberal interest groups would not have
dared pick a fight over these findings. Federal employees in the Voting
Section at the Department of Justice would have spent more energy
considering ways to act to fix the problems rather essentially ignoring
this problem.

The SAVE data findings are also relevant in 2020 to the rush to move
to automatic vote-by-mail. It has been used by the Foundation in, as of
this writing, 11 briefs in COVID-19-related litigation brought by liberal
groups seeking to cancel or modify state election laws that protect the
integrity and security of the election and voting process.[18]

SAVE could not have come at a better time.

34
THE FOUNDATION’S LITIGATION

ALLEGHENY COUNTY, PENNSYLVANIA

Over the past two years, Pennsylvania has received national attention
for admitting to registration glitches over decades that resulted
in thousands of foreign nationals being registered to vote. The
Foundation is currently litigating against the Commonwealth to collect
all documents explaining the full scale of the failure.[19] Pennsylvania
has fought tooth and nail to
conceal the extent of the problem,
including how many aliens illegally
cast ballots in elections.

The Foundation released a report

about foreign national registration
collected from Allegheny County,
home to Pittsburgh.[20] In 2019,
careful reviews of that county’s
voter roll revealed thousands
of duplicates, triplicates,
quadruplicates, and even one man
simultaneously registered to vote
seven times. On January 15, the
County was formally notified of the
nearly 7,500 registrants flagged
for duplication concerns; 1,500
matched against the Social Security Death Index; another 1,500 who
were older than 100 years of age (49 born in the 1800s); nearly 1,200
registrants showing no date of birth; and more.

35
Shortly before the onset of the pandemic, the ACLU tried to intervene
in the case, essentially arguing that Allegheny County and the court
could not be trusted to handle the Foundation’s data. That’s a shame,
because the errors on the rolls were objective, clear and easy to fix.
The parties reached an agreement to fix the problems and improve
the process. This is how the SAVE Database can be used to improve
the administration of American elections and protect voters from
dilution of their franchise.

“Allegheny County deserves credit for agreeing to fix a serious

problem with elections there. People have been getting registered
two, three, four, even seven times over to vote in Pittsburgh and
the suburbs. We found those problems, and the County agreed to
fix them. We also outlined plans to address registration files which
may be outdated, incomplete, or belong to deceased persons. This
settlement demonstrates what can be accomplished when good
government groups work with election officials in good faith without
the inter ference of ideologically driven activists who oppose such
measures. Those same activists push radical changes to vote by mail,
which shows how important this settlement was for a clean election
in Pennsylvania."
— J. Christian Adams

36
THE FOUNDATION’S LITIGATION

CITY OF DETROIT

In May 2019, the City was notified about a variety issues that, by all
appearances and communications, were being ignored. Detroit was
warned about the thousands of “active” registrants aged 85-plus
that were matched against the Social Security Death Index. Officials
were alerted to registrants being shown as born in “1823.” A child
appeared to be registered. Yet again, thousands were registered in
duplicate and triplicate. A federal lawsuit to correct these matters was
filed on December 10, 2019.

Once discovery was fully underway, documents and statements

by the defendants began to illuminate how nearly every one of the
thousands of duplicates had been cleaned up. Local officials had
started hunting for death records across the state. By June, the
Foundation was satisfied by the amount of action evident coming
from the City and dismissed the case.

37
“This case wasn’t complicated. The City of Detroit could have started
to fix these problems before litigation, but didn’t. Other jurisdictions
should take note--if you don’t act on solid data that your voter rolls
are corrupted with dead and duplicate registrations, you will be
sued. It is also a message to left wing groups who sought to stop the
cleanups. They need to stop standing in the way of clean elections
and stop wasting court time with their anti-integrity agenda. Election
officials can get rolls clean without removing valid registrants. It’s
time groups like the League of Woman Voters realized that data-
driven list maintenance is something most Americans want.”
— J. Christian Adams

38
THE COVID CASES
Liberals have sought to cancel state election laws and re-write the
rules of the 2020 election that protect the integrity and security of
voting and the election process because of the pandemic. A hyper
funded litigation effort was deployed, asking courts to cancel state
laws enacted by democratically elected legislatures and signed
by governors. This antidemocratic effort has a common theme:
demanding mail ballots are made available with fewer rules; loosen
or get rid of procedures to verify the identity of the voter and prevent
the forgery and alteration of ballots; force states to legalize vote
harvesting and allow campaign staffers, party activists, and political
consultants access to voted absentee ballots; and change deadlines.
As lawsuits are filed, the Foundation – using the important data from
SAVE – files amicus briefs to demonstrate to courts what risks would
be incurred by implementing automatic circulation of mail-in ballots
and loosening or voiding security protocols.

39
ISSA V. NEWSOM
RNC V. NEWSOM
U.S. Eastern District of California
Plaintiff(s): Darrell Issa, James B. Oerding, Jerry Griffin, Michelle
Bolotin, Michael Sienkiewicz, Republican National Committee
Core Issues: Lawsuit to block California from performing an all vote-
by-mail election for the 2020 General Election
The Foundation’s Brief: 23,000 deceased registrants, 1,800 apparently
duplicated voters, 2,000 apparent interstate duplicate voters

FAY V. MERRILL
Connecticut Supreme Court
Plaintiff(s): Mary Fay
Core Issues: Enjoining the automatic circulation of mail ballots during
the 2020 General Election
The Foundation’s Brief: 12,000 deceased registrants, duplicate
registrations, risks of sending mail to outdated addresses

THE NEW GEORGIA PROJECT V. RAFFENSPERGER

U.S. Northern District of Georgia
Plaintiff(s): The New Georgia Project (Stacey Abrams)
Core Issues: Error notification, absentee age restriction, postage
requirement, receipt deadline, voiding ballot harvest ban
The Foundation’s Brief: 4,200 deceased registrants, nearly 20,000
duplicates were credited for second votes in 2016 and 2018, nearly
850 Georgians credited for voting in 2 states

40
DEMOCRACY NC V. NC STATE BOARD OF ELECTIONS
U.S. Middle District of North Carolina
Plaintiff(s): Democracy North Carolina, The League of Women Voters
North Carolina, Donna Permar, John P. Clark, Margaret B. Cates, Lelia
Bentley, Regina Whitney Edwards, Robert K. Priddy II, Susan Shaffer,
and Walter Hutchins
Core Issues: Waive 25-day application deadline, remove witness
requirements, other-than-mail ballot returns
The Foundation’s Brief: # Apparently duplicated registrants credited
for 2nd votes in 2016 General Election: 9,700, # apparently duplicated
registrants credited for 2nd votes in 2018 General Election: 7,000

NEW MEXICO EX REL. RIDDLE V. OLIVER

New Mexico Supreme Court
Plaintiff(s): New Mexico County Clerks
Core Issues: Automatic ballot mailings to all registrants in time for June
2020 Primary Election
The Foundation’s Brief: 3,100 registrants flagged for duplicate voting
concerns, 1,680 apparently deceased registrants, 1,500 aged 100+

FISHER V. HARGETT
Tennessee Supreme Court
Plaintiff(s): Earle Fisher, Benjamin Lay
Core Issues: TN Legislature’s rights to rule over conduct of elections
procedures
The Foundation’s Brief: No constitutional right to a mail ballot

41
TEXAS DEMOCRATIC PARTY V. ABBOTT
U.S. Western District of Texas, Fifth Circuit Court of Appeals
Plaintiff(s): TX Democratic Party, Gilberto Hinojosa, Joseph Daniel
Cascino, Shanda Marie Sansing, Brenda Li Garcia
Core Issues: Voter intimidation, mail voting security, alleged 26th
Amendment violations
The Foundation’s Brief: Plaintiffs fail to make intimidation claim, mail
voting is fraught with fraud, ballot harvesting risks, Texas does not
violate 26th Amendment

LEAGUE OF WOMEN VOTERS VIRGINIA V. VIRGINIA STATE

BOARD OF ELECTIONS
U.S. Western District of Virginia
Plaintiff(s): League of Women Voters Virginia
Core Issues: Remove witness signature requirements from mail ballot
materials
The Foundation’s Brief: 11,600 potentially deceased registrants, 1,700
apparent commercial address votes, nearly 600 Virginians credited for
casting a second ballot in another state

ZIGNEGO V. WISCONSIN ELECTIONS COMMISSION

Wisconsin Supreme Court
Plaintiff(s): Wisconsin Ex Rel, Timothy Zignego, David W. Opitz,
Frederick Luehrs III
Core Issues: Ensure clean voter rolls prior to 2020 Election
The Foundation’s Brief: 6,000 apparently deceased registrants, nearly
4,000 duplicates were credited for second votes in 2016 and 2018

42
 A TOOLBOX ADDITION/POTENTIAL
REPLACEMENT FOR VOTER REGISTRARS
Voter registration offices across the nation need all the help they can
get to keep records accurate and up-to-date. America is a highly mobile
society and good help can sometimes be hard to find in keeping pace with
the electorate. Furthermore, locales can sometimes be limited by statute
or budget in affirmatively engaging outside help. More than a decade
ago, several states joined forces to pool data and expertise to keep tabs
on more mobile registrants. The Interstate Voter Registration Crosscheck
program grew initially under the management of the Kansas and Missouri
Secretaries of State to a majority of states participating prior to the 2016
Elections. But alongside the government cooperative, Pew Research’s
ERIC (Electronic Registration Information Center) co-op continued to grow
without the affixed political lightning rods.

The “Kansas Crosscheck” (Interstate Voter Registration Crosscheck)

sustained heavily funded left-wing attacks for years as ERIC managed
to pick up clients in their wake. Unlike the Kansas/Missouri model,
that nonprofit organization charged member states expensive dues,
required more data from states than just voter rolls, and required that list
maintenance leads be shielded from public record laws. ERIC now has a
majority of participating states with recent recruits like Texas and Florida.
Despite the pedigree of ERIC being squarely in the left-of-center camp,
foes of election integrity have even turned on ERIC’s crosscheck program.

“ERIC should be called ERROR because it’s that erroneous and that full of
flaws. ERIC claims to find people who’ve moved. Now, apparently almost
half the people on the list are young people.” – Barbara Arnwine, Lawyers
Committee for Civil Rights Under Law

One nonprofit publication promoting Arnwine’s critiques went so far as

to label the outfit “ERIC Crow—Jim Crow’s Liberal Cousin.” [21] The group
featured content suggesting that PEW President Rebecca W. Rimel “looks

43
like the kind old lady down the street with too many cats who will look in
on yours when you’re on vacation… Just don’t leave your voter rolls with
her. When you come home, you’ll find them bleached white and oddly
shrunken.”

The Foundation’s SAVE Database offers a well-timed substitute or addition

to the voter registrar’s toolbox.
• SAVE does not cost states a dime to participate. The Foundation’s
mission is in part to cooperatively improve voter rolls across the
country. Taxpayers should not have to pay extra for that.
• SAVE does not require states to provide confidential, non-public
information on its registered voters and unregistered residents. ERIC
requires voter roll and confidential driver’s license information for
membership.
• SAVE does not require members to mass mail voter registration
offers to licensed drivers who are not registered to vote.
• In fact, there is no membership structure in SAVE whatsoever.

States need as many tools and lead-generating systems as they can

handle. The Barack Obama-impaneled Presidential Commission on
Election Administration was on target when it recommended that
“interstate exchanges of voter registration should be expanded.” [22]

44
SAVE PREQUEL REPORTS
The methodologies and audit capacities underpinning the SAVE Database
have been explored and reported since October 2018. This section
summarizes each edition. All are available online at the Foundation’s
website for public education uses.

IS FLORIDA READY FOR THE 2020 ELECTIONS?

July 2020

Florida’s role in determining the outcome in national

elections, combined with its mobile population of
snowbirds, makes it vulnerable to election fraud,
irregularities, and errors. We know that foreign
nationals register and vote but proof is not typically
available until they are forced to self-report their records
in the face of an immigration proceeding. We know
that deceased registrants received mail-in ballots at
their former addresses and that those ballots were later
counted in federal elections. We also know that some
Floridians voted more than once in the same election by virtue of having a
second residence in another state.

DOWN THE HATCH: HOW LEFT-WING ELECTION REFORMS CAN

BE FORCED ON UNSUSPECTING COMMUNITIES
May 2020

In late 2019, the Public Interest Legal Foundation

encountered alarm about a rather novel situation, where
election officials were creating voting process errors
and also acting as sovereign partners with outside
ideological interest groups. With the help of New
Mexico’s open records statutes, these theories were
investigated.

45
CALM BEFORE THE STORM: ARE PALM BEACH COUNTY’S
ELECTIONS PROTECTED AGAINST EMERGING THREATS?
November 2019

The 2018 Midterm was marred by recount delays and

other alarming events, which went on to gain national
attention in the days and weeks following election day.
Shortly before Governor Ron DeSantis ended Palm
Beach County Supervisor of Elections Susan Bucher’s
tenure in office, the Public Interest Legal Foundation
began audits and record collection efforts to identify
specific flaws within voter registration files and any
systemic glitches creating them.

MOTOR VOTER MAYHEM: MICHIGAN’S VOTER ROLLS IN

DISREPAIR
October 2018

Michigan demonstrates how noncitizens become

registered to vote through DMV transactions
and others like it required by the National Voter
Registration Act of 1993. The State does not have a
verification system keeping false claims of citizenship
(intentionally given or otherwise) from being accepted
during voter registration. Immigrants and citizens alike
continue to suffer the consequences. Several Michigan
jurisdictions also exhibited alarming problems with
other voter roll maintenance obligations like duplications and potentially
deceased registrants remaining on the rolls for years on end.

46
DATA LITIGATION
A minority of states refuse to disclose or sell voter registration data to
non-partisan or private entities. One must be a state resident and/or
be a representative of a recognized political party organization. Any
moral high ground that these states tried to stand on when the Trump
Presidential Advisory Commission on Election Integrity came asking for
data was dubious at best.[23]

During data collection, several states similarly refused the Foundation’s

requests on the basis that it was not a qualifying partisan entity, or it
was not local.

PUBLIC INTEREST LEGAL FOUNDATION V. MATTHEW DUNLAP

U.S. District of Maine

February 19, 2020, the Foundation sued Maine Secretary of State

Matthew Dunlap for refusing to disclose voter registration records
under the National Voter Registration Act’s public inspection provision.
A copy of Maine’s statewide roll was originally requested on October
17, 2019, and rejected the same day—explaining the data could not be
shared since the Foundation had zero interest in leveraging the data for
partisan ends.

“Secretary Dunlap purpor ts to be a champion of

transparency, until it comes to his own office. Maine law
conflicts with federal statute. A person or organization’s
lack of partisan interests should not disqualify them from
reviewing list maintenance records.”
— J. Christian Adams

As of the release of this report, litigation is pending and proceeding

toward a potential trial.

47
PUBLIC INTEREST LEGAL FOUNDATION V. LINDA LAMONE
U.S. District of Maryland

On December 16, 2019, the Foundation sued the Maryland State Board
of Elections (MSBOE) for refusing to disclose voter registration records
under the National Voter Registration Act’s public inspection provision.
The State informed the Foundation the November prior that because
it was not based locally, the application to copy the data would be
rejected. Shortly before this report published, the MSBOE agreed to
settle the case and released all requested documents.

PUBLIC INTEREST LEGAL FOUNDATION V.

ILLINOIS STATE BOARD OF ELECTIONS
U.S. Central District of Illinois

On July 27, 2020, the Foundation sued the Illinois State Board of
Elections for refusing to disclose voter registration records under the
National Voter Registration Act’s public inspection provision. The State
prevents private, non-partisan groups from accessing full extracts of
the registered voter file.

48
CONCLUSION
In the aftermath of the 2016 Election, voter registration list maintenance
took on a new meaning and sense of urgency. If states’ voter databases
were fair game for outside attackers then simply relying on strong
passwords and fully funded IT departments was an incomplete strategy.
The Foundation went to work arguing how effective voter list maintenance
practices served as natural defenses against hostile parties who would
sabotage whole systems. If voter data was a soft target, then officials
and the general public needed to
know where the weaknesses were.
But you cannot hope to improve
the contents of American voter
registration systems unless you
are first willing to digest them in
their entirety. Building the SAVE
Database became an absolute
necessity.

During the first half of the 2020

Election cycle, the Foundation and
some state governments built tools
for the next time that Russian
hackers or others came calling.
When the Chinese Coronavirus
came first, SAVE took on a whole
new use.

As America grappled with the demands and attendant logistics for

more mail voting, the Foundation sounded alarms to the fact that the
most important piece of the process—voter registration data—was not
being maintained for the task. In an all-mail voting scenario, hundreds of
thousands of dead registrants would get ballots; many thousands more
would again have opportunities to vote twice; and many thousands more
would have chances to claim mailbox rental stores, warehouses, or gas

49
stations as homes. Accounting for only the past two federal elections,
SAVE generated more than 500,000 leads for voter registrars and law
enforcement to act upon.

Large scale, cooperative voter registration best practices are now under
full assault by increasingly well-funded leftist activist groups. When
established and respected organizations too fall under racialist attacks,
alternatives and supplements must come forward, like SAVE has done. If
early trends show a sustaining effect, America could become even more
mobile as citizens relocate to avoid virus hotspots.[24]

As noted at the outset of the report, the nation’s voter rolls have shown
improvement in some states over the past decade. Year-round pressure
and a consistent drumbeat for improved collaboration among officials
is the key to long-term success. But even in this period of improvement,
problems are still appearing on a widespread scale. There is substantial
evidence demonstrating how clumsy or even negligent acts by voter
registrars can result in bad actors taking advantage in increasingly tight
election contests. Once thought to only impact local elections, voter fraud
(by mail in particular) has now marred at least one federal contest. SAVE
is the only national tool built to account for this sobering challenge and
confront it.

These data show that the nation is not yet ready for an all-mail election.
Put aside the wholesale ineptitude of the United States Postal Service to
run a national election, the voter rolls remain a mess. If voter rolls are a
mess, mail elections will be a mess.

50
ENDNOTES
1 PEW Research Foundation; One in Eight Voter Registrations Inaccurate (February
14, 2012), https://www.pewtrusts.org/en/about/news-room/press-releases-and-
statements/2012/02/14/pew-one-in-eight-voter-registrations-inaccurate-51-million-
citizens-unregistered .

2 The Secretary of State of Maine, Matt Dunlap, was one of the members of the
presidential commission.

3 A false positive, generally speaking, is an incorrect finding that appears to be

otherwise accurate. For example, “John James Smith” born July 4, 1976 appearing in two
states may appear to be the same person, but without performing any identity validation
to ensure they are indeed the same person, risk becoming a false positive.

4 PILF v Illinois SBE (C.D. Ill.), https://publicinterestlegal.org/blog/pilf-sues-illinois-for-

failing-to-disclose-registered-voter-data/.

5 PILF v. Linda Lamone (D. Md.), https://publicinterestlegal.org/cases/pilf-v-linda-

lamone/.

6 PILF v. Matthew Dunlap (D. Me.), https://publicinterestlegal.org/cases/pilf-v-

matthew-dunlap/.

7 https://www.oregonlive.com/portland/2010/09/north_portland_man_80_probably.
html.

8 Inspector General report, pages 210-221, https://oig.justice.gov/reports/review-

operations-voting-section-civil-rights-division.

9 Arizona, Colorado, Kansas, Kentucky, Maine, Missouri, New Hampshire, Oregon,

South Dakota, Virginia, and Washington, https://www.ncsl.org/research/elections-and-
campaigns/double-voting.aspx.

10 Alabama, Delaware, Hawaii, Illinois, Maryland, Mississippi and West Virginia (see

prior link).

11 Alaska, Arkansas, California, Connecticut, Florida, Georgia, Idaho, Iowa, Louisiana,

Massachusetts, Michigan, Minnesota, Montana, Nebraska, Nevada, New Jersey, New
Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, Pennsylvania, Rhode
Island, South Carolina, Tennessee, Texas, Utah, Vermont, Wisconsin and Wyoming (see
prior link).

51
12 Federal law prohibits voting more than once for the same federal office. 52 U.S.C. §
10307(e).

13 Seo-young Silvia Kim, California Institute of Technology, HSS, Caltech, 1200

East California Blvd., Pasadena, CA 91125-0002, USA, https://journals.sagepub.com/
doi/10.1177/1532673X19870512.

14 EAC.gov.

15 POLITICO; Kansas Republican Steve Watkins charged with voter fraud (July 14, 2020),
https://www.politico.com/news/2020/07/14/kansas-gop-rep-watkins-indicted-for-voter-
fraud-362072.

16 Kansas voter registration application ENGLISH, https://www.kssos.org/forms/elections/

voterregistration.pdf.

17 K.S. 25-2309.

18 PILF; PILF Leads in Defending Mail Voting Protections amid Pandemic (July 20, 2020),
https://publicinterestlegal.org/blog/pilf-leads-in-defending-mail-voting-protections-amid-
pandemic/.

19 PILF v. Pennsylvania (M.D. Pa.), https://publicinterestlegal.org/cases/pilf-v-torres-et-al/.

20 PILF; Steeling the Vote (July 12, 2018), https://publicinterestlegal.org/blog/steeling-

the-vote-allegheny-county-reveals-how-citizenship-verification-protects-citizens-and-
immigrants-alike/.

21 Nation of Change; ERIC Crow, Jim Crow’s Liberal Twin (July 15, 2020), https://www.
nationofchange.org/2020/07/15/eric-crow-jim-crows-liberal-twin/.

22 The American Voting Experience: Report and Recommendations of the Presidential

Commission on Election Administration (January 2014), https://www.eac.gov/sites/default/
files/eac_assets/1/6/Amer-Voting-Exper-final-draft-01-09-14-508.pdf.

23 NBC News; 45 States Refuse to Give Voter Data to Trump Panel (July 6, 2017), https://
www.nbcnews.com/politics/white-house/forty-four-states-refuse-give-voter-data-trump-
panel-n779841.

24 Breitbart News; Flight from the Cities: New Home Sales Jump 13.8% in June – Highest
Since 2007 (July 24, 2020), https://www.breitbart.com/economy/2020/07/24/new-home-sales-
jump-13-8-in-june/.

52
 The Public Interest Legal Foundation, a 501(c)(3) non-profit
organization, relies on contributions to conduct the research
and develop findings like those contained in this report. PILF
is the only organization performing this level of work with
respect to voter registration system integrity in America.
Time, travel, and technology help deliver new insights into our
election systems to better educate citizens and policymakers
alike. We also bring litigation to pry this public information
from government officials when necessary. None of this would
be possible without your support. Please help us expand our
efforts by visiting
publicinterestlegal.org/donate
to offer your fully tax-deductible gift today.

PUBLIC INTEREST LEGAL FOUNDATION

PUBLICINTERESTLEGAL .ORG
32 EAST WASHINGTON STREET
SUITE 1675 3 1 7. 2 0 3 . 5 5 9 9
INDIANAPOLIS, INDIANA 46204