No.

04-16280-II

UNITED STATES COURT OF APPEALS

FOR THE ELEVENTH CIRCUIT

____________________

ROBERT WEXLER, Congressman )

ADDIE GREENE, Commissioner, )
BURT AARONSON, Commissioner, )
TONY FRANSETTA, )
PLAINTIFFS-APPELLANTS, ) Appeal from the United
) States District Court
v. ) for the Southern
) District of Florida
THERESA LEPORE, Supervisor of Elections )
for Palm Beach County, Florida, ) Docket No. 04-80216
KAY CLEM, Supervisor of Elections ) CIV-COHN
for Indian River County, Florida and )
President of the Florida Association ) Hon. James I. Cohn
of Supervisors of Elections, )
GLENDA E. HOOD, Secretary of State )
of Florida , )
DEFENDANTS-APPELLEES. )
_________________________________________)

BRIEF OF PLAINTIFFS-APPELLANTS

Robert S. Peck, Esq. Jeffrey M. Liggio, Esq.

CENTER FOR CONSTITUTIONAL LIGGIO, BENRUBI & WILLIAMS,
LITIGATION, PC PA
1050 31st Street, N.W. 1615 Forum Place
Washington, DC 20007 Suite 3B, The Barristers Building
(202) 944-2874 (o) West Palm Beach, FL 33041
(202) 965-0920 (fax) (561) 616-3333 (o)
Email: [email protected] (561) 616-3266 (fax)
Email: [email protected]

Attorneys for Plaintiffs-Appellants

Wexler et al. v. Lepore et al. Case No. 04-16280

ROBERT WEXLER, Congressman

ADDIE GREENE, Commissioner,
BURT AARONSON, Commissioner,
TONY FRANSETTA,
PLAINTIFFS-APPELLANTS,

v.

THERESA LEPORE, Supervisor of Elections for Palm Beach County,

Florida,
KAY CLEM, Supervisor of Elections for Indian River County, Florida and
President of the Florida Association of Supervisors of Elections,
GLENDA E. HOOD, Secretary of State of Florida,
DEFENDANTS-APPELLEES.
_____________________________________________________________

CERTIFICATE OF INTERESTED PERSONS AND CORPORATE

DISCLOSURE STATEMENT

The undersigned counsel of record certifies that the following listed persons

have an interest in the outcome of this case:

1. Commissioner Burt Aaronson

2. Office of the Attorney General, Attorneys for Defendant-Appellee

Hood

3. Center for Constitutional Litigation, P.C., Attorneys for Plaintiffs-

Appellants

4. Supervisor Kay Clem

5. Honorable United States District Court Judge James I. Cohn

6. Tony Fransetta, an individual

C- 1 of 2
Wexler et al. v. Lepore et al. Case No. 04-16280

7. Commissioner Addie Greene

8. Secretary of State Glenda E. Hood

9. Paul C. Huck, Jr., Attorney for Defendant-Appellee Hood

10. Chris Kise, Attorney for Defendant-Appellee Hood

11. Ronald Labasky, Attorney for Defendants-Appellees LePore and

Clem

12. Landers & Parsons, P.A., Attorneys for LePore and Clem

13. Supervisor Theresa LePore

14. Liggio, Benrubi & Williams, P.A., Attorneys for Plaintiffs-Appellants

15. Jeffrey M. Liggio, Attorney for Plaintiff-Appellants

16. Robert S. Peck, Attorney for Plaintiffs-Appellants

17. James A. Peters, Attorney for Defendant-Appellee Hood

18. George Waas, Attorney for Defendant-Appellee Hood

19. Congressman Robert Wexler

C- 2 of 2
 REQUEST FOR ORAL ARGUMENT

Pursuant to Eleventh Circuit Rule 28-1(c), Plaintiffs-Appellants request oral

argument. This case presents an issue of great importance involving whether

Florida’s treatment of state-required manual recounts in public elections, which

differs from county-to-county on the basis of voting equipment used, violates

voters’ federal rights under the Equal Protection and Due Process Clauses of the

United States Constitution.

i
 TABLE OF CONTENTS

CERTIFICATE OF INTERESTED PERSONS AND CORPORATE

DISCLOSURE STATEMENT ..............................................................................C-1

REQUEST FOR ORAL ARGUMENT ..................................................................... i

TABLE OF CONTENTS.......................................................................................... ii

STATEMENT OF JURISDICTION....................................................................... vii

STATEMENT OF THE ISSUES...............................................................................1

STATEMENT OF THE CASE..................................................................................1

A. Course of Proceedings...........................................................................1

B. Statement of Facts. ................................................................................3

C. Statement of the Standard or Scope of Review.....................................5

SUMMARY OF THE ARGUMENT ........................................................................5

ARGUMENT .............................................................................................................6

I. FLORIDA LAW MANDATES A MANUAL RECOUNT,

WHICH CAN NEITHER BE DISPENSED WITH NOR
REPLACED WITH A MACHINE RECOUNT ...................................6

II. CURRENT FLORIDA LAW FOREGOING MANUAL

RECOUNTS FOR BALLOTING USING TOUCHSCREEN
VOTING MECHANISMS VIOLATES EQUAL
PROTECTION, AS WOULD A RULE COUNTING ONLY
BALLOT SUMMARIES FOR THAT EQUIPMENT........................12

A. Strict Scrutiny Applies to Evaluate Whether Equal

Protection is Denied ..................................................................12

C. Counting Ballot Image Summaries Would Not

Remedy the Equal Protection Violation ...................................22

D. The Emergency Rule Approach Also Contained an

Unconstitutional Time Limitation ............................................27

ii
 III. THE INABILITY TO EFFECTUATE A MEANINGFUL
STATEWIDE MANUAL RECOUNT VIOLATES
PLAINTIFFS’ DUE PROCESS RIGHTS ..........................................28

CONCLUSION........................................................................................................32

CERTIFICATE OF COMPLIANCE

CERTIFICATE OF SERVICE

ADDENDUM

iii
 TABLE OF CITATIONS

Cases
ACLU v. Department of State, Case No. 04-2341RX (Aug. 27, 2004).. 4, 10, 19, 22

American Association of People with Disabilities v. Shelley, 324

F.Supp.2d 1120 (C.D. Cal. 2004) ........................................................... 11, 25

Baker v. Carr, 369 U.S. 186, 82 S.Ct. 691 (1962) ..................................................14

Black v. McGuffrage, 209 F.Supp.2d 889 (N.D. Ill. 2002) ................................ 5, 17

Burdick v. Takushi, 504 U.S. 428, 112 S.Ct. 2059 (1992) ......................... 12, 13, 14

Bush v. Gore, 531 U.S. 98, 121 S.Ct. 525 (2000)............................................ passim

Florida Democratic Party v. Hood, 884 So.2d 1148 (Fla. 1st DCA), rev.
denied, 888 So.2d 622 (Fla. 2004).................................................................12

Georgia Manufactured Hous. Ass’n, Inc. v. Spalding County, 148 F.3d

1304 (11th Cir. 1998) ......................................................................................5

Gray v. Sanders, 372 U.S. 368, 83 S.Ct. 801 (1963)...............................................13

Hall v. Holder, 117 F.3d 1222 (11th Cir.1997) .......................................................13

Hendon v. North Carolina Board of Elections, 633 F.Supp. 454

(W.D.N.C. 1986) ...........................................................................................21

Hendon v. North Carolina State Board of Elections, 710 F.2d 177 (4th
Cir. 1983) .......................................................................................................21

Hennings v. Grafton, 523 F.2d 861 (7th Cir. 1975)..................................................29

Illinois Bd. of Elections v. Socialist Workers Party, 440 U.S. 173 (1979)..............30

Joint Anti-Fascist Refugee Comm. v. McGrath, 341 U.S. 123 (1951) ....................28

Mathews v. Eldridge, 424 U.S. 319 (1976) ................................................ 29, 30, 31

Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362 (1964)..................................... 14, 20

Roe v. Alabama, 68 F.3d 404 (11th Cir. 1995) ............................................ 16, 29, 30

iv
Roudebush v. Hartke, 405 U.S. 15, 92 S.Ct. 804 (1972).........................................15

Sable Communication of California v. FCC, 492 U.S. 115 (1989).........................22

Twining v. New Jersey, 211 U.S. 78 (1908) ............................................................28

United States v. Classic, 313 U.S. 299, 61 S.Ct. 1031 (1941) ................................14

United States v. Mosely, 238 U.S. 383, 35 S.Ct. 904 (1915)...................................20

United States v. Saylor, 322 U.S. 385, 64 S.Ct. 1101 (1944)..................................20

Wesberry v. Sanders, 376 U.S. 1, 84 S.Ct. 526 (1964)............................................13

Wexler v. Lepore, 342 F.Supp.2d 1097 (S.D. Fla. 2004)................................. passim

Wexler v. LePore, 385 F.3d 1336 (11th Cir. 2004) ...................................................2

Yick Wo v. Hopkins, 118 U.S. 356, 6 S.Ct. 1064 (1886) .................................. 13, 30

Constitutional Provisions
U.S. CONST. amend XIV .................................................................................. passim

Statutes
Fla. Admin. Code 1S-2.031 ................................................................................ 7, 10

Fla. Stat. § 101.015 ................................................................................................3, 8

Fla. Stat. § 101.294 ....................................................................................................8

Fla. Stat. § 101.5604 ..................................................................................................8

Fla. Stat. § 101.5605 ..................................................................................................8

Fla. Stat. § 102.141 ................................................................................................6, 7

Fla. Stat. § 102.166 ................................................................................................6, 7

Help America Vote Act (HAVA) of 2002, 42 U.S.C. §15301 et seq......................22

Other Authorities
AM. HERITAGE DICTIONARY OF THE ENGLISH LANGUAGE (4th ed. 2000).................27

v
Issacharoff, Samuel, Groups and the Right to Vote, 44 EMORY L.J. 869
(1995).............................................................................................................16

PIERCE, JR., RICHARD J., ADMINISTRATIVE LAW TREATISE (4th ed. 2002) ...............28

Rules
Rule 1SER04-1 ............................................................................................. 2, 11, 12

vi
TO THE HONORABLE UNITED STATES COURT OF APPEALS FOR THE
ELEVENTH CIRCUIT:

STATEMENT OF JURISDICTION

The District Court had jurisdiction over this action under 28 U.S.C. § 1331,

28 U.S.C. § 1343, and 42 U.S.C. § 1983. Venue was proper in the United States

District Court for the Southern District of Florida, Fort Lauderdale Division,

pursuant to 28 U.S.C. § 1391(b).

This is an appeal from a final judgment, issued on October 25, 2004. A

motion for reconsideration was denied as moot on November 19, 2004. Plaintiffs

filed timely Notice of Appeal on November 24, 2004. This Court has jurisdiction

over the appeal pursuant to 28 U.S.C. § 1291.

vii
 STATEMENT OF THE ISSUES

1. Consistent with the Equal Protection Clause of the Fourteenth

Amendment to the U.S. Constitution, may a State that mandates a manual recount

in certain extremely close public election contests forego that recount or rely upon

machine-generated ballot summaries for voters who use touchscreen voting

equipment, while other voters are entitled to have the paper records of their

individual ballots recounted by hand?

2. Consistent with the Due Process Clause of the Fourteenth Amendment to

the U.S. Constitution, may a State that mandates a manual recount in certain

extremely close public election contests forego utilizing some form of

individualized verification to conduct that recount for voters who use touchscreen

voting equipment and instead utilize machine-generated ballot summaries, while

other voters are entitled to have the paper records of their individual ballots

recounted by hand?

STATEMENT OF THE CASE

A. Course of Proceedings.

Plaintiffs, officeholders, candidates for electoral office, and voters, filed this

action for declaratory and injunctive relief on March 8, 2004, alleging that the

Defendants in certifying, approving, and implementing the use of touchscreen

paperless voting systems in fifteen Florida counties created a nonuniform,

1
differential standard from the systems used in fifty-two other Florida counties, in

violation of the Equal Protection and Due Process Clauses of the U.S. Constitution.

On May 24, 2004, the District Court granted Defendant Secretary of State

Glenda E. Hood’s motion to abstain on the basis that one plaintiff, Robert Wexler,

had previously filed a state constitutional challenge in state court. On September

27, 2004, this Court reversed and vacated the District Court’s order. Wexler v.

LePore, 385 F.3d 1336 (11th Cir. 2004). After denying Defendants’ motion for

reconsideration en banc, the Eleventh Circuit issued its mandate on October 7,

2004.

Trial was scheduled to begin Monday, October 18, 2004. On the eve of trial,

specifically on Friday, October 15, 2004, at 4:08 p.m., Defendant Hood issued an

emergency rule, revising its position from a “no manual recount rule” for

touchscreen machines to a recount consisting of machine-generated ballot image

summaries for that equipment. At the commencement of trial, District Court Judge

James Cohn ruled that the case would now proceed as a challenge to the

emergency rule. The case was tried over a three-day period and resulted in final

judgment for the Defendants. Because of an emergency petition filed before the

Florida Supreme Court in a different matter that could have affected interpretation

of relevant state law, Plaintiffs filed a motion for reconsideration, which was

denied as moot on November 19, 2004.

2
 Notice of appeal was timely filed on November 24, 2004.

B. Statement of Facts.

In the aftermath of the controversy surrounding the 2000 presidential

election, Florida endeavored to upgrade the voting systems utilized throughout the

state. As Secretary of State, Defendant Hood, pursuant to Fla. Stat. § 101.015, had

responsibility for adopting minimum standards for new voting machines and to

update certification standards on a continuous basis. Hood also had responsibility

for approving or disapproving each voting system. The same state statute requires

county elections supervisors, such as Defendants Teresa LePore and Kay Clem, to

establish written procedures to assure the accuracy and security of the adopted

voting systems.

Fifteen Florida counties opted to purchase touchscreen, or direct recording

electronic (DRE), voting equipment, incapable of producing paper records of each

cast ballot. Fifty-two Florida counties purchased optical-scan, or “marksense,”

equipment, that retains paper ballots for use in a manual recount. Wexler v.

Lepore, 342 F.Supp.2d 1097, 1099 (S.D. Fla. 2004). Absentee voters, and voters

who utilize provisional ballots in DRE counties vote on marksense equipment. R4-

129-38. On February 12, 2004, the Division of Elections issued Opinion DE 04-

02, declaring that no manual recounts would be conducted for the DREs because

the machines produce nothing worth counting.

3
 On March 8, 2004, four Plaintiffs instituted this action, alleging that the

election officials’ determination that they would forego manual recounts for those

voters whose ballots were cast on DREs violated the Fourteenth Amendment’s

Equal Protection and Due Process guarantees. Three of them are voters and

elected officials, who were seeking reelection at the time: U.S. Representative

Wexler and Palm Beach County Commissioners Addie Greene and Burt Aaronson.

The fourth plaintiff, Tony Fransetta, is a registered voter, who has voted in past

elections and intended to vote in the 2004 and subsequent elections.

While the case was pending, the Florida Division of Administrative

Hearings ruled on August 27, 2004, that the no-recount rule for DREs promulgated

by Defendant Hood violated state law. ACLU v. Department of State, Case No.

04-2341RX (Aug. 27, 2004)1. No appeal was taken from that ruling.

Subsequently, at 4:08 p.m. on Friday, October 15, 2004, before the trial in this case

that was to commence on Monday, October 18, 2004, Defendant Hood

promulgated an emergency rule that provided for the counting of machine-

generated ballot summary images in the event of a manual recount.

1
Attached as Addendum A.

4
C. Statement of the Standard or Scope of Review.

The District Court’s findings of fact are reviewed for clear error and its legal

conclusions are reviewed de novo. Georgia Manufactured Hous. Ass’n, Inc. v.

Spalding County, 148 F.3d 1304, 1307 (11th Cir. 1998).

SUMMARY OF THE ARGUMENT

In extremely close public election contests, Florida law mandates that a

manual recount be conducted. Although responsibility for certifying voting

equipment capable of fulfilling all the requirements of Florida law falls to

Defendant Secretary of State, her office did not make the capacity to conduct a

manual recount a requirement for approval of equipment. Fifteen of Florida’s 67

counties purchased approved voting equipment that does not allow a canvassing

board to examine individual ballots to determine whether the voter has made a

definite choice, as mandated by the state recount statute. The resulting disparate

treatment of votes subject to manual recount, which differs geographically based

on the voting equipment used, violates equal protection and due process, because

ballots do not receive equal treatment and because the requirement of fundamental

fairness is not observed. Contrary to the holding of the District Court, equal

protection does not guarantee that all voters using the same equipment be treated

the same, but that all voters participating in the same election be treated the same.

5
 ARGUMENT

I. FLORIDA LAW MANDATES A MANUAL RECOUNT, WHICH CAN

NEITHER BE DISPENSED WITH NOR REPLACED WITH A
MACHINE RECOUNT

Florida’s election recount law is the appropriate starting point to understand

whether existing procedures amount to the type of arbitrary and disparate treatment

that gives rise to a constitutional violation. In the aftermath of the controversy

surrounding the counting and recounting of ballots during the 2000 presidential

election, the state legislature enacted new recount laws, which provide that when

the margin of difference between the candidates is one-half of one percent or less,

a “machine recount” shall take place. Fla. Stat. § 102.141(6). Understanding that

all voting mechanisms can be subject to error in recording votes, an understanding

confirmed by every expert witness who testified at trial, R5-130-482, R5-130-483;

R5-130-267, R5-130-268; R4-129-165 through R4-129-169;2 the Legislature

mandated a “manual recount of the overvotes and undervotes cast in the entire

geographic jurisdiction of such office or ballot measure” when the margin of

difference in the machine recount is one-quarter of one percent or less. Fla. Stat. §

102.166(1). The statute also permits a candidate to demand and receive a manual

2
Defendant’s Chief of the Bureau of Voting System Certification, Paul Craft,
indicated that in two instances machines were certified by the Department of State
but subsequently found to have bad source code, causing tabulation problems.
Craft, R5-130-493, R5-130-494.

6
recount when the machine recount margin falls between one-quarter and one-half

of a percent. Fla. Stat. § 102.166(2)(a). While machine recounts are conducted by

rerunning the tabulations of the existing voting equipment or examining the

counters for machines that do not use paper ballots, Fla. Stat. § 102.141(6)(b),

manual recounts require observations of individual ballots by the canvassing board

to determine “if there is a clear indication on the ballot that the voter has made a

definite choice.” Fla. Stat. § 102.141(6)(a). See also Opinion DE 04-02. Thus, the

legislative scheme evinces a distrust of a mere machine recount in the most closely

contested elections.

Recounts focus on overvotes and undervotes. Fla. Stat. § 102.166(1).

Overvotes occur when the voting system determines that the voter has cast more

votes than permitted in a particular race, Fla. Stat. § 97.021(21), while undervotes

occur when the voter improperly does not designate a choice or the voting system

finds that the voter has cast no vote in a particular race, notwithstanding the voter’s

intent to vote. Fla. Admin. Code 1S-2.031(4).

State law also tasks the Department of State with responsibility to “adopt

specific rules for each certified voting system prescribing what constitutes a ‘clear

indication on the ballot that the voter has made a definite choice,’” and “which

shall be uniform to the extent practicable.” Fla. Stat. § 102.166(5)(b) & (6)(d).

7
 In addition, the Department has responsibility for adopting minimum

standards for voting systems. Fla. Stat. § 101.015. Based on compliance with

those standards, the Department has the authority to approve or disapprove any

voting system. Id.; Fla. Stat. § 101.5605. The Department also tests and certifies

the equipment. R4-129-108; R4-129-109; R5-130-434, R5-130-492. Those tests

are designed to assure compliance with statutory requirements, as well as accuracy.

R4-129-108, R4-129-109; R5-130-434. However, the Department has chosen not

to include in its certification criteria any evaluation of a mechanism’s capacity to

comply with the manual recount statute. R4-129-108; R4-129-109.

Each board of county commissioners chooses its own voting equipment

from among those mechanisms certified by the Department. Fla. Stat. § 101.5604;

Fla. Stat. § 101.294. Fifty-two of Florida’s 67 counties opted for optical-scan, or

marksense, machines. 342 F.Supp.2d at 1099. To vote using this equipment, a

voter uses a pencil to fill in a bubble or arrow by the name of the selected

candidate on a sheet of paper. The paper is then scanned into the machine, which

tabulates the votes. The original marked paper sheets are preserved for use in any

necessary manual recount. R6-131-544 through R6-131-547. Optical scan

equipment has proven to be the voting mechanism with the least number of

8
residual votes and is less expensive than a touchscreen voting machine. R6-131-

539 through R6-131-541.3

Fifteen counties opted for touchscreen voting mechanisms, also known as

direct recording electronic voting systems (DREs), 342 F.Supp.2d at 1099, that do

not produce individual paper records for manual recounts. R4-129-84. To vote

using this equipment, a voter literally touches the screen, which then records the

vote. 342 F.Supp.2d at 1100. These touchscreen voting systems are subject to

mechanical, electronic, and programmatic errors in their functioning. Counties that

opted for DREs still use optical-scan equipment to count provisional and absentee

ballots. R6-131-624, R6-131-627; R5-130-377. Thus, DRE counties have voters

casting ballots on both types of equipment, only some of which are capable of a

manual recount.

Pursuant to her obligation to promulgate a rule for manual recounts in

counties using DREs, Defendant Hood’s elections division first issued an opinion

denying authorization “to print or review the electronic ballot images of undervotes

occurring in a recounted race.” DE 04-02, Division of Elections Opinion (Feb. 12,

3
The Division of Elections reported more problems in the 2002 election with DRE
equipment producing overvotes and undervotes than optical scan equipment. R4-
129-84 through R4-129-126. After the 2004 election, the Division issued a similar
report, indicating improvement in DREs but fewer problems still with optical
scans. Fla. Dep’t of State, Div. of Elections, Analysis and Report for the 2004
General Election (Jan. 31, 2005) (Attached as Addendum B).

9
2004). On April 13, 2004, Defendant Hood amended the rule governing recount

procedures as follows:

When a manual recount is ordered and touchscreen ballots are used,

no manual recount of undervotes and overvotes cast on a touchscreen
system shall be conducted since these machines do not allow a voter
to cast an overvote and since a review of undervotes cannot result in a
determination of voter intent as required by Section 102.166(5), F.S.
In this case, the results of the machine recount conducted pursuant to
paragraph (5)(c) shall be the official totals for the touchscreen ballots.

Fla. Admin. Code Rule 1S-2.031(7). This rule is currently in effect, even

though it was determined to be in violation of the requirements of Florida

law. ACLU, supra, at ¶ 31.

A machine recount is incapable of detecting whether an undervote is

intentional or the result of a failure of the machine to record a vote. A machine

recount is also insufficient, as a matter of law, as a substitute for a manual recount.

ACLU, supra, at ¶ 33 (“If the legislature had intended that no manual recounts be

done in counties using voting systems which did not use paper ballots, it could

have easily done so.”). Instead, Defendant Hood’s department “does not have the

authority to preclude manual recounts for a touchscreen voting system because it

does not believe that a manual recount will reveal a clear indication of the voter's

choice.” Id. at ¶ 36. 4

4
At one point, 43 percent of California voters used DREs. After experiencing
problems in the testing and certification of software, reliability, accuracy, training,
and security, California’s Secretary of State decertified the systems. American

10
 On the eve of trial, Defendant Hood promulgated an emergency rule in

response to the ACLU ruling that provided a different type of machine recount for

DREs. That rule, which expired after 90 days, provided that a canvassing board

conducting a manual recount of a DRE shall order the printing of one official copy

of a ballot image summary for each touchscreen machine that has recorded

undervotes for a race. Rule 1SER04-1(7)(a). Such reports are cumulative and

merely indicate that the machine recorded no vote for that race by use of the word

“undervote,” a numeric coding, or the absence of an “X” on the report. 342

F.Supp.2d at 1103. They do not permit the canvassing board to determine whether

the voter made an intentional choice to undervote or that the machine failed to

record the vote due to voter mistake, human error, or system error. Indeed each

and every one of the voting officials who testified conceded as much. R4-129-62

through R4-129-75; R4-129-129 through R4-129-133; R4-129-139; R4-129-140;

R5-130-401 through R5-130-405; R5-130-474 through R5-130-477; R6-131-546

through R6-131-552; R6-131-608; R6-131-628; R6-131-629; R6-131-648 through

R6-131-658. Instead, if this “manual recount” matches the machine report of the

Association of People with Disabilities v. Shelley, 324 F.Supp.2d 1120, 1124 (C.D.
Cal. 2004). A primary reason for that decertification was that DREs not outfitted
with a voter-verified paper trail “do not permit meaningful recounts.” Id. at 1128.
The same conclusion that DREs without paper trails do not permit a meaningful
recount should obtain here as well.

11
number of undervotes, the results are certified as a final official tally. Rule

1SER04-1(7)(f).5

II. CURRENT FLORIDA LAW FOREGOING MANUAL RECOUNTS

FOR BALLOTING USING TOUCHSCREEN VOTING
MECHANISMS VIOLATES EQUAL PROTECTION, AS WOULD A
RULE COUNTING ONLY BALLOT SUMMARIES FOR THAT
EQUIPMENT

A. Strict Scrutiny Applies to Evaluate Whether Equal Protection is

Denied

The Supreme Court has instructed that:

A court considering a challenge to a state election law must weigh

“the character and magnitude of the asserted injury to the rights
protected by the First and Fourteenth Amendments that the plaintiff
seeks to vindicate” against “the precise interests put forward by the
State as justifications for the burden imposed by its rule,” taking into
consideration “the extent to which those interests make it necessary to
burden the plaintiff's rights.”

Burdick v. Takushi, 504 U.S. 428, 434, 112 S.Ct. 2059, 2065 (1992) (citations

omitted).

A severe restriction on a plaintiff’s voting rights is subjected to strict

scrutiny and must be “narrowly drawn to advance a state interest of compelling

5
Among the reasons the Defendant gave for promulgating the emergency rule was
“[a]d hoc, ex-post manual recount processes of touchscreen voting systems
conducted on a county-by-county basis, reminiscent of the circumstances giving
rise to the intensely litigated 2000 General Election, may violate the equal
protection and due process clause[s] of the U.S. Constitution.” Quoted in Florida
Democratic Party v. Hood, 884 So.2d 1148, 1150 (Fla. 1st DCA), rev. denied, 888
So.2d 622 (Fla. 2004).

12
importance.” Id. (citation omitted). If the law merely imposes “reasonable,

nondiscriminatory restrictions,” the “State’s important regulatory interests are

generally sufficient” to uphold the law. Id. (citation omitted). Florida’s recount

procedures, which utilize an individualized recount by hand for those ballots cast

on optical-scan equipment but entirely forego that scrutiny of those cast on DREs,

cannot be considered nondiscriminatory. Voters in DRE counties have no chance

of having a residual vote counted in an election where it could make a difference,

while other voters do. That discriminatory treatment requires strict scrutiny.

To be sure, voting holds an exalted place in the pantheon of constitutional

values as “a fundamental political right, because [it is] preservative of all rights.”

Yick Wo v. Hopkins, 118 U.S. 356, 370, 6 S.Ct. 1064, 1071 (1886). In fact, “[n]o

right is more precious in a free country than that of having a voice in the election

of those who make the laws under which, as good citizens, we must live.”

Wesberry v. Sanders, 376 U.S. 1, 17, 84 S.Ct. 526, 535 (1964). See also Hall v.

Holder, 117 F.3d 1222, 1231 (11th Cir.1997)(voting is “a right of paramount

constitutional significance”).

Thus, “all qualified voters have a constitutionally protected right ‘to cast

their ballots and have them counted.’ . . . Every voter’s vote is entitled to be

counted once. It must be correctly counted and reported.” Gray v. Sanders, 372

U.S. 368, 380, 83 S.Ct. 801, 808 (1963)(citation omitted). See also Reynolds v.

13
Sims, 377 U.S. 533, 554, 84 S.Ct. 1362, 1378 (1964); Baker v. Carr, 369 U.S. 186,

208, 82 S.Ct. 691, 705 (1962), citing United States v. Classic, 313 U.S. 299, 61

S.Ct. 1031 (1941)(recognizing that the right to vote is infringed by a false tally).

When the voting mechanism fails to count a vote and the election hinges on

a tally of those votes, the continued failure to count that vote, as occurs under

Defendants’ system, constitutes an injury of enormous magnitude – namely,

disenfranchisement – for which there can be no justification. See Reynolds, 377

U.S. at 563, 84 S.Ct. at 1382 (“Weighting the votes of citizens differently, by any

method or means, merely because of where they happen to reside, hardly seems

justifiable.”). Equally established by virtue of the alternative technologies

available to Defendants to preserve those rights while advancing the State’s

interest in fair and honest elections,6 there is no need to burden Plaintiffs’ rights.

It is apparent, then, that weighing the Burdick factors for the appropriate

constitutional balance between the voters’ interests, the State’s legitimate interests,

and the availability of means that accommodates both interests without unduly

burdening the voters’ rights that the scale tips decisively in favor of Plaintiffs.

While every election dispute does not require strict scrutiny, the Supreme

Court has made clear that “recount mechanisms . . . [must] satisfy the minimum

6
See R4-129-170 through R4-129-200; R5-130-329 through R5 130-337; R5-130-
350 through R5-130-359.

14
requirement for nonarbitrary treatment of voters necessary to secure the

fundamental right [to vote].” Bush v. Gore, 531 U.S. 98, 105, 121 S.Ct. 525, 530

(2000)(per curiam). The Constitution’s application begins with voting as

authorized by the State:

When the state legislature vests the right to vote . . . in its people, the
right to vote as the legislature has prescribed is fundamental; and one
source of its fundamental nature lies in the equal weight accorded to
each vote and the equal dignity owed to each voter.

Id. at 104, 121 S.Ct. at 529.

Thus, although there is no constitutional right to a recount, where a state

provides for it, the recount becomes an integral part of the election process.

Roudebush v. Hartke, 405 U.S. 15, 25, 92 S.Ct. 804, 810-11 (1972). See also Gore

v. Harris, 773 So.2d 524, 531 (Fla. 2000)(Pariente, J., concurring). As such, the

Constitution safeguards votes subject to recount against arbitrary and disparate

treatment. Bush, 531 U.S. at 104-05, 121 S.Ct. at 530 (“Having once granted the

right to vote on equal terms, the state may not, by later arbitrary treatment, value

one person’s vote over that of another.”).

These holdings mean that if a state provides for a recount then every ballot

must be correctly counted in the recount and treated no differently than another

voter’s selections. As Professor Issacharoff has explained, “To be effective, a

voter’s ballot must stand a meaningful chance of effective aggregation with those

of like-minded voters to claim a just share of electoral results.” Samuel

15
Issacharoff, Groups and the Right to Vote, 44 EMORY L.J. 869, 883 (1995). Thus,

in Roe v. Alabama, 68 F.3d 404 (11th Cir. 1995), this Court recognized that unequal

criteria for counting ballots implicated fundamental fairness by diluting the votes

of some voters while disenfranchising others.

In Bush, a similar constitutional violation was identified. Two categories of

voters had their ballots denigrated: voters whose ballots were excluded from the

recount order by the Florida Supreme Court and those whose ballots were tallied

before the recount began. Both were considered harmed because their votes were

devalued by the nature of the recount process instituted. Bush, 531 U.S. at 107-08,

121 S.Ct. at 531. The Court found that recount process was “not well calculated to

sustain the confidence that all citizens must have in the outcome of elections.” Id.

at 109, 121 S.Ct. at 532.

Bush thus establishes that a state entity with the power to assure uniformity

in a statewide recount cannot allow differing standards to prevail without some

assurance that the rudimentary requirements of equal protection and fundamental

fairness are met. It further holds that rules, such as those promulgated by

Defendant Hood, must be “designed to ensure uniform treatment.” Id. at 106, 121

S.Ct. at 530. A rule that examines some voters’ ballots for a manual recount and

not others due to the voting equipment utilized – or the fortuity of geography –

does not constitute uniform treatment. See id. at 107-08, 121 S.Ct. at 531.

16
 The situation today eerily mirrors the one examined in Bush:

the citizen whose ballot was not read by a machine because he failed
to vote for a candidate in a way readable by a machine may still have
his vote counted in a manual recount; on the other hand, the citizen
who marks two candidates in a way discernible by the machine will
not have the same opportunity to have his vote count, even if a manual
examination of the ballot would reveal the requisite indicia of intent.

Id. at 108, 121 S.Ct. at 531.

The same lack of uniformity compelled the court in Black v. McGuffrage,

209 F.Supp.2d 889, 899 (N.D. Ill. 2002), to find an equal protection violation

“when people in different counties have significantly different probabilities of

having their votes counted, solely because of the nature of the [voting] system used

in their jurisdiction.” As the court further stated:

the case at bar is not one of an accidental malfunction or unforeseen

error in counting or failing to count a particular group of ballots . . .
but rather a statutory scheme which, depending upon the choices
made by local election jurisdiction officials, will necessarily result in
the dilution of an entire group of citizens’ right to vote. . . . the votes
cast in some districts will have a significantly greater chance of being
counted than the votes cast in neighboring election districts. . . . Such
a situation does rise to the level of a constitutional violation.

Id. at 901.

The disparate treatment of recounts based on voting mechanisms adopted by

the Defendants here also violates this principle. A ballot that would be recounted

in one county will not be recounted in a neighboring county, simply because of the

equipment used. Even within a county, voters using DREs will not have their

17
votes recounted, while their neighbors who voted by absentee or provisional ballot

will. Although the voting mechanisms adopted need not be identical throughout

the state, the capacity to facilitate a manual recount must be common to all the

voting methods. See Bush, 531 U.S. at 109, 121 S.Ct. at 532. There must be an

“assurance that the recounts included in a final certification” are complete. Id. at

108, 121 S.Ct. at 532.

The need for a recount capability in elections utilizing paperless touch

screen voting machines is underscored by recent experiences in Wellington Village

and Legislative District 91, where the number of undervotes vastly outnumbered

the margin of victory. R6-131-626 through R6-131-636. In Wellington Village in

2002, the margin of victory for a council race was four votes, while 78 ballots were

deemed undervotes. In January 2004 special election District 91 race, with no

other office on the ballot, it is impossible to conceive that 134 people went to the

polling place, waited in line, signed in, stepped up to the voting machine, and

chose to vote for no candidate in an election decided by 12 votes. In both

instances, because paperless DREs were used, no statutorily mandated recount

could take place. R6-131-626 through R6-131-629 and R4-129-215; Exhibit 179.

B. The District Court Erred in Finding that Equal Protection

Applied to Voting Systems, Rather than Voters

The flaw in the District Court’s analysis is apparent from how it framed the

issue. The court said it “must decide if the rule creates a uniform, nondifferential

18
standard for conducting a manual recount in the fifteen counties using certified

touchscreen machines,” 342 F.Supp.2d at 1105 (footnote omitted), rather than

statewide. Relying on Justice Souter’s dissent in Bush, the court characterized the

equal protection problem discussed in that case as involving “disparate rules . . .

applied to determine voter intent on ‘identical types of ballots used in identical

brands of machines and exhibiting identical physical characteristics.” Id. at 1107,

quoting Bush, 531 U.S. at 134, 121 S.Ct. at 545 (Souter, J., dissenting). The court

then found conformity with the requirements of Bush because the standards

utilized in recounts are uniform for each certified voting system. Id. at 1108. One

reason the court found this sufficient is because it mystifyingly determined that the

possibility that a system might malfunction was not before the court. Id. at 1108

n.16. Yet, without a potential to malfunction, which all witnesses agreed occurs,

there would be no need for a manual recount, which the legislature has insisted

take place even with DREs.7 Moreover, because Plaintiffs’ Complaint did not

anticipate that an 11th hour emergency rule would change the lawsuit’s focus, that

allegation was deemed unnecessary, yet is still covered in Complaint ¶¶ 20, 24, 25

and 37.

7
In ACLU, the hearing officer noted that the legislature had declined to enact an
exemption from manual recounts for DREs. ACLU, supra, at ¶33.

19
 Further, the Court found that a voter who casts his or her vote on a DRE has

made a definite decision to undervote, so that the machine’s recording of an

undervote requires no further inquiry. That finding is contrary to the testimony of

every relevant witness, including Sarah Jane Bradshaw, Defendant Hood’s

assistant director of elections, the highest official defendants were willing to make

available for testimony in this case. R4-129-62 through R4-129-74; R4-129-129

through R4-129-133; R4-129-139; R4-129-140; R5-130-401 through R5-130-405;

R5-130-474 through R5-130-477; R6-131-546 through R6-131-552; R6-131-608;

R6-131-609; R6-131-628; R6-131-629; R6-131-648 through R6-131-658.

As argued earlier, the court misconceived the constitutional violation.

Officeholders “represent people, not trees or acres,” Reynolds, 377 U.S. at 562, 84

S.Ct. at 1382, and not voting equipment. Thus, the court’s focus on uniformity of

treatment of equipment was misdirected. After all, “[w]eighting the votes of

citizens differently, by any method or means, merely because of where they

happen to reside, hardly seems justifiable.” Id. at 563, 84 S.Ct. at 1382. Simply

put, the Equal Protection Clause safeguards the rights of voters to have their valid

votes counted along with the valid votes of other voters participating in that

election. United States v. Saylor, 322 U.S. 385, 388-89, 64 S.Ct. 1101, 1103

(1944); United States v. Mosely, 238 U.S. 383, 35 S.Ct. 904 (1915).

20
 The court below’s analysis was emphatically rejected by the Fourth Circuit

in Hendon v. North Carolina State Board of Elections, 710 F.2d 177 (4th Cir.

1983). There, the District Court also had rejected an equal protection challenge to

the use of different voting machines with different capacities to comply with legal

requirements “because all voters in each precinct were treated alike using the same

voting methods and counting rules.” Id. at 180 (emphasis added). That ruling was

remarkably similar to the District Court here.

At issue in Hendon were voting machines that did not permit voters to split

their tickets while largely voting for a single party, while other equipment

permitted vote splitting. The Fourth Circuit ruled that the “Constitution protects

the right of qualified citizens to vote and to have the votes counted as cast,”

necessitating review of any conditions placed on that right by strict scrutiny. Id.

The court acknowledged that a state may employ diverse methods of voting, but

stated that the state may not select methods that place a “much more onerous

burden” on some voters than others. Id. at 181. The decision found that voters

using certain equipment were unconstitutionally burdened in having their vote

counted as cast than voters using other equipment. Hendon v. North Carolina

Board of Elections, 633 F.Supp. 454 (W.D.N.C. 1986)(characterizing decision).

The inherent constitutional flaw here, like that examined in Hendon, is that

some of the voting equipment certified and used by Defendants omits an essential

21
capability that is necessary to comply with an integral part of the electoral process,

the mandatory manual recount.

The fact that Defendant Hood’s attempt to forego manual recounts in

counties utilizing paperless touchscreen voting machines was deemed ultra vires in

ACLU further demonstrates that there is no rationale that can justify the situation

that is the gravamen of this action. Moreover, even if there were some compelling

reason that Plaintiffs have failed to imagine, the requirement of narrow tailoring

that is part of the strict-scrutiny test would have advised Defendants to select a

different machine, capable of providing a paper trail, in order to avoid abridging

the rights of Plaintiffs and others similarly situated.8 See, e.g., Sable

Communication of California v. FCC, 492 U.S. 115, 131 (1989).

C. Counting Ballot Image Summaries Would Not Remedy the Equal

Protection Violation

The emergency rule, which was in effect at the time of trial, would not ameliorate

the constitutional violation if it were adopted as a permanent rule for there remains

no ability to conduct a meaningful manual recount in the 15 counties utilizing

DREs. The rule instead contemplates the counting of machine-generated ballot

8
At trial, Representative Robert Wexler testified that Congress made money
available precisely for this purpose in the Help America Vote Act (HAVA) of
2002, 42 U.S.C. §15301 et seq., which Defendant Hood had not applied for. R4-
129-216. HAVA requires that all voting systems produce a permanent paper
record for manual audits in federal elections by January 1, 2006. 42 U.S.C.
§15481(a)(2)(B).

22
image summaries. Those summaries do not reflect individual ballots, but merely

report the number of undervotes recorded by that particular DRE machine. There

remains no capability of discerning whether the voter made a definite choice.

At trial, Assistant Director of the Division of Elections Bradshaw was asked

if examining a ballot image summary would enable elections officials to tell if an

undervote was a mistake by the voter or an intentional choice. She answered:

A No, I don’t believe it makes the distinction between a mistake

of the voter and the voter’s intent, no.

Q. And, ma’am, can you and I agree that that report also is not able
to tell us to distinguish whether there was an error of by the poll
workers, a software error, a firmware error, a hardware error, as
opposed to an intentional undervote by the voter?

A. I would say that you would not ever look at that voter image
report, the ballot image report to determine a malfunction for a
hardware or software error of the machine.

Q. That’s because it doesn’t tell us; does it?

A. Well, the statutes don’t tell you to look to that report for that.

Q. But the statutes do tell you to look to the report to try to

determine the intent of the voter; do they not?

A. That’s correct.

Mr. Liggio to Curt Browning:

Q. You and I agree therefore, that these documents do not allow a

Canvassing Board to distinguish between what you and I agree
was an intent, a choice, and an undervote that’s in the system as
a result of some other cause; yes?

A. Yes.

23
 Q. And if the statute has not changed, please look at Florida
Statute 102.166 parenthesis five, close parenthesis.

Q. Does it not require the Canvassing Board, in reviewing the

undervotes, or the count, if there is clear indication that the
voter made a choice?

A. Yes.

R5-130-405. Mr. Liggio to Theresa LePore:

Q. Okay. Let me ask you a question. Last time when you and I
spoke we talked about the Senate Committee report as to the
definition of a recount; did we not?

A. I believe so.

Q. Do you agree with me that recount means to count again,

simply taking the count totals from each tabulator and adding
them together is not recounting the ballots, it’s recounting the
ballots summaries. Do you agree with that definition?

A. Yes.

Q. Do we have any way in this audit log to do a recount to

determine whether the voter made a definite choice as opposed
to some error somewhere, I’m not talking hacking or fraud, an
error?

A. The audit log comes from the machine. It has information on

how each voter voted. It’s a ballot image, if you will.

Q. And the audit log only shows whether a vote was recorded or
not?

A. Correct.

Q. Not whether there was a reason for a vote being – failing to be

recorded, other than the voter didn’t want to vote?

A. Right.

24
 Q. So with this audit trail, all we can do is confirm; yes?

A. Yes.

Q. Let me finish, I’m sorry. All we can do now is confirm, the

initial count said there were fifty undervotes, and we printed out
this piece of paper internally and, yep, it was fifty undervotes,
that’s all we can show now, right?

A. Okay.

Q. Do you agree with that?

A. That’s what the audit log says, yes.

Q. Thanks. And that’s what the audit log says that was attached to
the emergency rule, that the Secretary of State promulgated 4
o’clock this past Friday afternoon, correct?

A. I don’t know what time they did it, but yes.

R6-131-657.

Ballot image summaries reflect what the machine recorded and provides no

indication of the choice made by the voter. Defendants’ expert Dr. Shamos

testified that there is no presumption that an undervote identified by a ballot image

summary reflects the voter’s intention or choice. He noted that DREs in California

were decertified by that state’s Secretary of State because they could not provide

meaningful recounts. R4-129-201; R4-129-202; R5-130-350.9 Elections assistant

director Sarah Jane Bradshaw could not tell why ballot image summaries, available

9
The California Secretary of State decertified the DREs, after previously
defending them against a challenge, because they did not “permit meaningful
recounts.” See Shelley, 324 F. Supp.2d at 1128.

25
since May 2004, were insufficient to reflect voter choice then, but was sufficient at

the time the emergency rule was promulgated. R4-129-66 through R4-129-74.

For that reason, Ms. Bradshaw could not tell what happened in the District 91

election. Curt Browning, the Pasco County Elections Director, testified that he had

felt the original “no recount” rule for DREs complied with Florida law because it

was impossible to recount undervotes that do not exist. R5-130-398.

Recounts of ballot image summaries would constitute a sham, providing no

meaningful manual recount to determine voter choice. The Secretary’s own initial

rule agreed with that assessment, forbidding manual recounts on DREs configured

as certified by her office based on the utter uselessness of any printout that could

be conjured from the equipment. Opinion DE 04-02. It is difficult to imagine –

and testimony at trial did not establish – how those printouts suddenly became

meaningful for purposes of the 11th hour emergency rule.

In evaluating the ballot-image summary approach, the Bush decision

remains instructive:

A manual recount of all ballots identifies not only those ballots which
show no vote but also those which contain more than one, the so-
called overvotes. Neither category will be counted by the machine.
This is not a trivial concern. At oral argument, respondents estimated
there are as many as 110,000 overvotes statewide. As a result, the
citizen whose ballot was not read by a machine because he failed to
vote for a candidate in a way readable by a machine may still have his
vote counted in a manual recount; on the other hand, the citizen who
marks two candidates in a way discernible by the machine will not

26
 have the same opportunity to have his vote count, even if a manual
examination of the ballot would reveal the requisite indicia of intent.

Bush, 531 U.S. at 107-08, 121 S.Ct. at 531.

That same nontrivial concern is apparent when using ballot-image

summaries, which merely reflect what the machine recorded. Yet, a manual

recount anticipates a human examination of the ballot to determine whether the

voter made a definite choice, something that the machine report cannot substitute

for. The plain meaning of “manual” indicates that it must be done by human

hands. AM. HERITAGE DICTIONARY OF THE ENGLISH LANGUAGE, (4th ed. 2000),

available at www.bartleby.com/61/20/MOO92000.html. As a result, even when

ballot summary images are utilized, voters and candidates in DRE counties will

receive different treatment from similarly situated voters based on the

happenstance of the county or district in which those voters reside. Such different

treatment constitutes an arbitrary subversion of the electoral process that serves no

legitimate, let alone compelling, state interest and makes out a palpable violation

of the Fourteenth Amendment’s guarantee of Equal Protection of the Laws.

D. The Emergency Rule Approach Also Contained an

Unconstitutional Time Limitation

The Supreme Court identified one “further equal protection problem” that

was perpetuated by the 2000 Florida recount scheme Bush, 531 U.S. at 108, 121

S.Ct. at 531, that is repeated in Defendant Hood’s emergency rule. State law both

27
then and the emergency rule now limits the time in which a recount may be

completed. Recounts completed in timely fashion are included, while those that

are not are dispensed with. See 1SER 04-1(7)(i). However, the Court said, “the

press of time does not diminish the constitutional concern. A desire for speed is

not a general excuse for ignoring equal protection guarantees.” Id., 121 S.Ct. at

532. Just as it could not be countenanced then, it may not now, because the system

established provides “no assurance that the recounts included in a final certification

[is] complete.” Id.

III. THE INABILITY TO EFFECTUATE A MEANINGFUL STATEWIDE

MANUAL RECOUNT VIOLATES PLAINTIFFS’ DUE PROCESS
RIGHTS

The Due Process Clause “protect[s] the citizen in his private right, and

guard[s] him against the arbitrary action of government.” Twining v. New Jersey,

211 U.S. 78, 101 (1908). It guarantees appropriate process and overrides those

processes that are sufficiently “devoid of fundamental fairness.” Joint Anti-Fascist

Refugee Comm. v. McGrath, 341 U.S. 123, 161 (1951)(Frankfurter, J., concurring).

As a result, the Due Process Clause stands as a bulwark against government

decisions that are erroneous, inconsistent, and subjective by assuring protection

against arbitrary government decisionmaking. RICHARD J. PIERCE, JR.,

ADMINISTRATIVE LAW TREATISE 579 (4th ed. 2002).

28
 While federal courts do not generally involve themselves in garden variety

election disputes, enforceable due process rights attach to systemic issues that

involve patent and fundamental unfairness. Roe v. Alabama, 43 F.3d 574, 580

(11th Cir. 1995). See also Hennings v. Grafton, 523 F.2d 861, 864-65 (7th Cir.

1975)(due process is implicated when “willful conduct . . . undermines the organic

process by which candidates are elected.”). Such systemic unfairness exists in a

system chosen to fulfill statutory requirements but incapable of assuring that votes

are properly recounted as state law requires. See, e.g., Bush, 531 U.S. at 105, 121

S.Ct. at 530, (indicating that the Constitution guarantees nonarbitrary treatment of

voters in the use of recount mechanisms).

In Mathews v. Eldridge, 424 U.S. 319, 335 (1976), the Supreme Court

articulated “three distinct factors” that ought to be considered in evaluating an

alleged due process violation: 1) whether the private interest affected by the

official action is within “life, liberty and property”; 2) whether there is a risk of an

erroneous deprivation of that interest through the procedures used, and the

probable value, if any, of additional or substitute procedural safeguards; and 3)

whether the government maintains an overriding interest that includes

consideration of the function involved and the fiscal and administrative burdens

that the additional or substitute procedural requisites would entail.

29
 All three Mathews considerations militate in favor of the Plaintiffs. The first

– whether “life, liberty or property” is affected – is undeniable. The right to vote is

“of the most fundamental significance under our constitutional structure.” Illinois

Bd. of Elections v. Socialist Workers Party, 440 U.S. 173, 184 (1979). It is a

fundamental liberty. Yick Wo, supra. As such, when the election process reaches a

point of “patent and fundamental unfairness, a violation of the due process clause

may be indicated and relief under § 1983 therefore in order.” Roe, 43 F.3d at 580.

The second Mathews consideration – whether alternative procedures could

avoid the risk of an erroneous deprivation – also favors the Plaintiffs’ position. In

a close election, the Florida Legislature has determined that the proper way to

assure a fair and honest election is by engaging in a manual recount. The paperless

DRE machines adopted in 15 counties are incapable of allowing such a recount.

As a result, voters whose ballot is wrongfully read not to record a vote will

erroneously be deprived of their right to vote because of the procedures authorized

by Defendants. Experience demonstrates that this occurrence is not merely

theoretical but real. R5-130-377. The testimony was overwhelming that additional

safeguards are imperative. Defendant’s expert, Dr. Shamos, testified to a system in

shambles, both at trial and before a congressional committee. R5-130-330; R5-

130-331. Dr. Shamos went on to agree that adding paper back-up to the machines

would be useful. R5-130-337. See also R4-129-159 through R4-129-200.

30
 Finally, the third Mathews consideration, the government’s interest, also

favors the Plaintiffs. The government shares with Plaintiffs an interest in honest

and fair elections with candidates and voters. There is no divergence of interest

between the two. Moreover, as evidenced by the decision to employ optical scan

equipment in the 52 other Florida counties, it is entirely possible to meet that

common interest without creating insuperable fiscal or administrative burdens by

adopting voting machines capable of complying with Florida’s manual recount

requirement. In fact, optical scan equipment costs less than DREs. R6-131-540;

R6-131-541.

While Defendants may argue that there is no budget to purchase new

machines after having spent the allocated funds for the machines at issue here, that

mistake cannot justify the violation of Plaintiffs’ rights. If it did, then government

actors would always be immune from a challenge like this when they expend

limited funds on equipment inadequate to meet constitutional requirements. Such

a result would render Plaintiffs’ rights entirely voidable as a result of wrongful

choices by Defendants, which can just as easily be made by design as by

inadvertence.

Moreover, testimony established that Florida had not applied for available

federal funds to finance the purpose of better voting equipment, which will be

31
 CERTIFICATE OF SERVICE

I certify that on this date I caused a true and correct copy of the foregoing

BRIEF OF PLAINTIFFS-APPELLANTS, in both paper and electronic format, to

be served on the following individuals by depositing same with Federal Express,

using Standard Overnight Delivery and by email:

Honorable United States District Court Judge James I. Cohn

299 East Broward Boulevard, Room 206
Fort Lauderdale, Florida 33301
[email protected]

Paul C. Huck, Jr., Esquire*

Regional Deputy Attorney General
110 Southeast 6th Street, 10th Floor
Fort Lauderdale, FL 33301

James A. Peters, Esquire

Office of the Attorney General
400 S. Monroe Street
Tallahassee, FL 32399
[email protected]

George Waas, Esquire

Office of the Attorney General
PL-01 The Capitol
Tallahassee, FL 32399-1050
[email protected]

Christopher M. Kise, Solicitor General

Office of the Attorney General
PL-01 The Capitol
Tallahassee, FL 32399
[email protected]

Attorneys for Defendant-Appellee Hood

 ADDENDUM

Page #

American Civil Liberties Union et al. v. Department of State, No. 04-

2341RX (Aug. 27, 2004) ...........................................................................A

Florida Department of State, Division of Elections, Analysis and Report of

Overvotes and Undervotes for the 2004 General Election .......................B

 Addendum A

STATE OF FLORIDA
DIVISION OF ADMINISTRATIVE HEARINGS

AMERICAN CIVIL LIBERTIES UNION )

OF FLORIDA, INC. (ACLU-FL); )
SOUTHWEST VOTER REGISTRATION )
EDUCATION PROJECT (SVREP); )
COMMON CAUSE FLORIDA (CCF); )
FLORIDA SOUTHERN CHRISTIAN )
LEADERSHIP CONFERENCE CHAPTER )
(FSCLC); AND THE FLORIDA VOTERS )
LEAGUE, INC. (FVL), )
)
Petitioners, )
)
and ) Case No. 04-2341RX
)
PEOPLE FOR THE AMERICAN WAY )
FOUNDATION, )
)
Intervenor, )
)
vs. )
)
DEPARTMENT OF STATE, )
)
Respondent. )
)

FINAL ORDER

Pursuant to notice, a final hearing was held in this case

on July 27, 2004, in Tallahassee, Florida, before Susan B.

Kirkland, a designated Administrative Law Judge of the Division

of Administrative Hearings.
 APPEARANCES

For Petitioners: Benjamin R. Patterson, Esquire

Jerry G. Traynham, Esquire
Patterson & Traynham
315 Beard Street
Post Office Box 4289
Tallahassee, Florida 32315-4289

For Intervenor: Reginald J. Mitchell, Esquire

1550 Melvin Street
Tallahassee, Florida 32301

For Respondent: George L. Waas, Esquire

Office of the Attorney General
The Capitol, Plaza Level 01
Tallahassee, Florida 32399-1050

STATEMENT OF THE ISSUE

Whether Florida Administrative Code Rule 1S-2.031(7) is an

invalid exercise of delegated legislative authority under

Subsection 120.56(3), Florida Statutes (2004).

PRELIMINARY STATEMENT

On July 6, 2004, Petitioners, the American Civil Liberties

Union of Florida, Inc. (ACLU-FL); Southwest Voter Registration

Education Project (SVREP); Common Cause Florida (CCF); Florida

Southern Christian Leadership Conference Chapter (FSCLC); and

The Florida Voters League, Inc. (FVL), filed, pursuant to

Subsection 120.56(3), Florida Statutes (2003), a Petition to

Determine the Invalidity of Rules challenging the validity of

Florida Administrative Code Rule 1S-2.031(7).

On July 9, 2004, Respondent, Department of State, filed an

unopposed Request for Official Recognition, requesting

2
that official recognition be taken of the following:

(1) Subsections 97.021(3), 101.5603(4), 101.5603(5),

and 101.5612, Florida Statutes (2003) (Exhibit A); (2) Order

Granting Defendants' Motions to Dismiss in Wexler v. LePore,

319 F. Supp. 2d 1354 (S.D. Fla. 2004) (Exhibit B); (3) Order

Granting Defendants [sic] Motion to Dismiss and Determining

Motion for Change of Venue Moot in Wexler v. LePore,

No. 502004CA 000491XXXXMB AA (Fla. 15th Cir. Ct. Feb. 11, 2004)

(Exhibit C); and (4) DE 04-02, Division of Elections Opinion

issued February 12, 2004 (Exhibit D). On July 12, 2004,

Respondent filed Respondent's Second Request for Official

Recognition, requesting that official recognition be taken of

Review of Voting Irregularities of the 2000 Presidential

Election, Report Number 2001-201, Prepared for the Florida

Senate by the Commission on Ethics and Elections, March 2001

(Exhibit E). An Order Granting Official Recognition was entered

on July 26, 2004, granting Respondent's requests for official

recognition.

On July 21, 2004, People for the American Way Foundation

(PFAWF) filed a Petition for Leave to Intervene. The petition

was heard by telephonic conference call on July 26, 2004, and an

Order Granting Petition to Intervene was entered on the same

date.

3
 The parties filed a Prehearing Stipulation on July 27,

2004, and stipulated to certain facts contained in Section E of

the Prehearing Stipulation. Those facts have been incorporated

in this Final Order.

At the final hearing, Petitioners presented the following

witnesses: Howard L. Simon, Benjamin T. Wilcox, Reverend S. L.

Phillips, and John Seibel. Petitioners' Exhibits A through C

were admitted in evidence. Intervenor presented Sharon Lettman

as its witness. Intervenor did not submit any exhibits in

evidence. Respondent called Sarah Jane Bradshaw and Paul Craft

as its witnesses. Respondent's Exhibit F was admitted in

evidence.

The parties agreed to file their proposed final orders

within ten days of the filing of the transcript. On August 5,

2004, the one-volume Transcript was filed. On August 12, 2004,

Petitioners filed an Unopposed Request for Enlargement of Filing

Period requesting additional time to file the proposed final

orders. On August 13, 2004, an order was entered extending the

time for filing the proposed final orders to August 18, 2004.

The parties timely filed their proposed final orders, which have

been considered in rendering this Final Order.

On August 25, 2004, Petitioner SVREP filed a Notice of

Voluntary Dismissal of SVREP, stating that SVREP had been unable

to appear at the hearing to provide evidence of its standing.

4
 FINDINGS OF FACT

1. Petitioner ACLU-FL is a public interest organization

with approximately 22,000 members in Florida. The great

majority of the ACLU-FL members are registered Florida voters.

ACLU-FL has an interest in advancing civil liberty principles,

including fair, accurate, and reliable electoral processes.

2. As part of its activities, ACLU-FL is involved in

litigation concerning voting issues. Following the 2000

presidential election, ACLU-FL filed an amicus curiae brief in

Bush v. Gore. In Siegel v. LePore, ACLU-FL intervened,

challenging the butterfly ballot. More recently, ACLU-FL filed

an amicus curiae brief in litigation filed by Congressman Wexler

concerning the manual recount of votes cast on touchscreen

voting systems. In Florida Caucus of Black State Legislators,

Inc. v. Crosby, __ So. 2d __, 29 Fla. L. Weekly D1629 (Fla. 1st

DCA July 14, 2004), ACLU-FL was successful in getting a writ of

mandamus issued against the Department of Corrections to require

the Department to provide to offenders, prior to their

discharge, an application form and other forms necessary to

obtain the Governor's review for restoration of their civil

rights and to advise the offenders that the Department will

assist in the preparation of the forms at the offenders'

request.

5
 3. ACLU-FL attempts to advance confidence in the voting

system and the electoral process. The organization strives to

ensure that the voting systems used in Florida are accurate,

reliable, and capable of providing for recounts and audits.

4. CCF is a government reform advocacy group with a basic

interest in good government, including issues such as campaign

finance reform, ethics, reapportionment, and election reform.

Its members believe that each voter's ballot should be counted

as cast and it is important to be able to verify one's vote and

have a record of each vote as cast. CCF has approximately

90,000 members in Florida.

5. FSCLC is an organization founded by the late Dr. Martin

Luther King, Jr.; Ralph Abernathy; and Reverend C. K. Steele.

This organization has an interest in election issues and was

instrumental in the passage of the Voting Rights Act in the

1960's. It is concerned with having the vote of each voter

counted as cast.

6. PFAWF is a civil rights and civil liberties

organization, which promotes constitutional rights and the value

of the vote. It has numerous voter registration projects,

including the Election Protection Project, which came about as a

result of the 2000 presidential election. The Election

Protection Project in Florida involves educating citizens on

6
their rights, the process of voting, and their legal recourse in

the event of election irregularities.

7. No evidence was presented concerning the standing of

the FVL to bring this rule challenge.

8. After the 2000 presidential election, the Committee on

Ethics and Elections prepared a report for the Florida Senate

entitled Review of the Voting Irregularities of the 2000

Presidential Election (the Report). Among other things, the

Report generally discussed the types of voting systems,

including the Direct Recording Electronic (DRE) systems and the

optical scan systems. At the time of the Report, the DRE

systems, sometimes called touchscreen systems, were not used in

Florida.

9. In the optical scan systems, the voter marks a paper

ballot, which is fed into an optical scanning device which

"reads" the ballot and tabulates the results. The paper ballot

is preserved. In a DRE system, the voter indicates his or her

choice by the use of a touchscreen, pushbuttons, or similar

devices, rather than a paper ballot. The Report indicated that

most of the DRE systems do no offer a paper trail as backup for

the vote cast. The evidence presented at the final hearing

established that there are DRE systems which can provide a paper

trail by printing a picture of the screen as it appears to the

voter when the voter indicates that his voting is completed or

7
by storing the screen image on microfilm. However, the DRE

systems currently certified by Respondent for use in Florida do

not provide a paper trail.

10. It is clear from the evidence at the final hearing,

that the DRE systems do not allow for an overvote. An overvote

occurs when a voter designates more names than there are persons

to be elected to an office or designates more than one answer to

a ballot question. § 97.021(20), Fla. Stat. (2004). DRE

systems either require the voter to unselect a previously

designated candidate before voting for another candidate or

automatically default to the last candidate selected.

11. There can be an undervote using the DRE systems. An

undervote occurs when a voter does not properly designate any

choice for an office or a ballot question. § 97.021(33), Fla.

Stat. (2004). Respondent requires that the DRE systems it

certifies must provide a notification to the voter of an

undervote before the voter casts his or her ballot.

12. After the difficult experiences with the 2000

presidential election, the following 15 counties in Florida

switched to a touchscreen voting system: Broward, Charlotte,

Collier, Hillsborough, Indian River, Lake, Lee, Martin, Miami-

Dade, Nassau, Palm Beach, Pasco, Pinellas, Sarasota, and Sumter.

The aggregate voting population of the above-listed counties

comprises slightly more than 50 percent of Florida's registered

8
voters. Petitioners, ACLU-FL, CCF, and FSCLC, and Intervenor

PFAWF have members who are registered voters in all 15 counties

with touchscreen voting systems.

13. Respondent has the statutory responsibility to set the

minimum standards for the hardware and software for electronic

and electromechanical voting systems, which include touchscreen

voting systems. § 101.015(1), Fla. Stat. (2004). Respondent

must approve electronic and electromechanical voting systems

prior to their use in Florida. §§ 101.015 and 101.5605, Fla.

Stat. (2004). One of the requirements for approval of an

electronic or electromechanical voting system is that the system

"must immediately reject a ballot where the number of votes for

an office or measure exceeds the number of votes which the voter

is entitled to cast or where the tabulating equipment reads the

ballot as a ballot with no votes cast." § 101.5606(3), Fla.

Stat. (2004). If the system uses a paper ballot, the system

must be capable of accepting a ballot rejected pursuant to

Subsection 101.5606(3), Florida Statutes (2004), and recording a

no vote for any office that has been overvoted or undervoted.

14. Florida law provides for recounts of votes in certain

situations. If the margin of victory for a candidate or measure

in the first set of unofficial returns is one-half of a percent

or less, a recount of the votes cast is to be ordered by the

board certifying the results on that race or measure.

9
§ 102.141(6), Fla. Stat. (2004). If paper ballots are used, the

paper ballots are put through the automatic tabulating

equipment, and a determination is made whether the returns

correctly reflect the votes cast. § 102.141(6)(a), Fla. Stat.

(2004). If no paper ballots are used, the canvassing board

examines the counters on the precinct tabulators to determine

whether the total on the precinct tabulators are equal to the

overall election return. If there is a discrepancy, the

counters on the tabulators are presumed to be correct.

§ 102.141(6)(b), Fla. Stat. (2004).

15. If the margin of victory in the second set of

unofficial returns is one-quarter of one percent or less, the

board responsible for certifying the election results shall

order a manual recount of the overvotes and undervotes cast for

that particular office or measure. § 102.166(1), Fla. Stat.

(2004). A vote is to be counted "if there is a clear indication

on the ballot that the voter has made a definite choice."

§ 102.166(5)(a), Fla. Stat. (2004). Respondent has the

responsibility to adopt rules for each certified voting system

stating what constitutes "'a clear indication on the ballot that

the voter has made a definite choice.'" § 102.166(5)(b), Fla.

Stat. (2004). Respondent also has the authority to promulgate

rules which set forth recount procedures for each certified

10
voting system in addition to the statutorily mandated recount

procedures. § 102.166(6)(d), Fla. Stat. (2004).

16. On February 12, 2004, Respondent issued DE 04-02,

Division of Elections Opinion on manual recount procedures for

the touchscreen voting systems. Respondent opined:

Because it is impossible for a voter to

overvote or make stray marks on an
electronic ballot, the manual recount
provisions of section 102.166, Florida
Statutes, do not apply and therefore,
counties utilizing touchscreen voting
systems are not required or authorized to
print or review the electronic ballot images
of undervotes occurring in a recounted race.

17. During the 2004 legislative session, legislation was

filed, SB 3004, which would have renamed paper ballots as

marksense ballots and would have required manual recounts only

on voting systems using marksense ballots and would have

provided that manual recounts may not be conducted of undervotes

on touchscreen machines. The CS for CS SB 3004 deleted the

provisions that recounts would be done only on marksense

ballots. The revised proposed legislation died on special

calendar.

18. On April 13, 2004, Respondent amended its rule

governing recount procedures, and Petitioners have challenged

the validity of the portion of the rule dealing with recounts of

touchscreen votes, Florida Administrative Code Rule 1S-2.031(7),

which provides:

11
 When a manual recount is ordered and
touchscreen ballots are used, no manual
recount of undervotes and overvotes cast on
a touchscreen system shall be conducted
since these machines do not allow a voter to
cast an overvote and since a review of
undervotes cannot result in a determination
of voter intent as required by Section
102.166(5), F.S. In this case, the results
of the machine recount conducted pursuant to
paragraph (5)(c) shall be the official
totals for the touchscreen ballots.

19. Petitioners, ACLU-FL, CCF, and FSCLC, and Intervenor

PFAWF, contend that Florida Administrative Code

Rule 1S-2.031(7), prohibiting manual recounts of undervotes on

the touchscreen voting systems, violates the statutory

requirement for a manual recount and denies voters a method to

insure the accuracy of the votes cast.

20. Respondent argues that if a manual recount were to be

conducted of undervotes using the touchscreen voting systems it

would be useless in determining that there was "a clear

indication on the ballot that the voter has made a definite

choice" because the screen would be blank. This assertion may

be true if the software or hardware on the voting machine does

not malfunction. It does not deal with a malfunction situation

in which what actually appeared on the screen when the vote was

cast was not what was actually recorded by the machine. It is

not known whether such a malfunction has occurred in Florida

because a copy of the screen image is not maintained.

12
 21. Respondent's expert, Paul Craft, acknowledged that the

touchscreen voting systems have had problems with system

software, but pointed out that the touchscreen voting systems

certified by Respondent currently have no known system software

problems.

CONCLUSIONS OF LAW

22. The Division of Administrative Hearings has

jurisdiction over the parties to and the subject matter of this

proceeding. § 120.56(3), Fla. Stat. (2004).

23. Petitioners, ACLU-FL, CCF, and FSCLC, and Intervenor

PFAWF, have standing to bring this rule challenge.

Subsection 120.56(3)(a), Florida Statutes (2004), provides that

"[a] substantially affected person may seek an administrative

determination of the invalidity of an existing rule at any time

during the existence of the rule."

24. In Florida Homebuilders v. Department of Labor and

Employment Security, 412 So. 2d 351, (Fla. 1982), the Florida

Supreme Court set forth the requirements for associational

standing in a rule challenge.

To meet the requirements of section

120.56(1), an association must demonstrate
that a substantial number of its members,
although not necessarily a majority, are
"substantially affected" by the challenged
rule. Further the subject matter of the
rule must be within the association's
general scope of interest and activity, and
the relief requested must be of the type

13
 appropriate for a trade association to
receive on behalf of its members.

Id. at 353-354.

25. Petitioners ACLU-FL, CCF, and FSCLC, and Intervenor

PFAWF, have a substantial number of members who are registered

voters in Florida, and have members who are registered voters in

the 15 counties using the touchscreen voting systems who will

not have their votes manually recounted in the event of a close

election because of the challenged rule. These associations'

interests and activities include ensuring that it members' votes

are recorded accurately and maintaining the integrity of the

voting process. Having a rule declared invalid is within the

types of relief that would be appropriate for an association to

ask on behalf of its members.

26. Respondent argues that Petitioners and Intervenor do

not have standing because they have not demonstrated a

particularized injury that is different in kind from that

sustained by the public generally and that they have not

demonstrated an "injury in fact." Petitioners, ACLU-FL, CCF,

FSCLC, and Intervenor PFAWF, will sustain an injury different

from the public in general. Members voting in the 15 counties

using a touchscreen voting system, will not be entitled to have

their undervotes counted because of the challenged rule, while

voters using other types of voting systems will be entitled to

14
have their undervotes counted. In NAACP, Inc. v. Florida Board

of Regents, 863 So. 2d 294, 300 (Fla. 2003), the Florida Supreme

Court rejected the type of argument that would require the

members having an injury in fact, such as cast unintended

undervotes, in order to have standing.

It also appears that the First District

was adopting a rule of standing that would
require a challenge to demonstrate immediate
and actual harm, i.e., rejection of
admission to a state university by a member
before standing would be granted. We
required no such showing in Florida Home
Builders. Indeed, such a holding would
constitute a narrowing of the concept of
standing as defined in Florida Home
Builders. Under our holding there the
required showing is that there would be a
substantial effect of the rule change on a
substantial number of its members.

27. Petitioner FVL presented no evidence on the issue of

its standing, and, thus, has failed to establish that it has

standing to bring this rule challenge. Petitioner SVREP has

voluntarily dismissed its petition.

28. Petitioners and Intervenor, as the parties challenging

an existing rule, have the burden to establish by a

preponderance of the evidence that Florida Administrative Code

Rule 1S-2.031(7) is an invalid exercise of delegated legislative

authority. § 120.56(3), Florida Statutes (2004). Petitioners

and Intervenor contend that the rule at issue exceeds

Respondent's grant of rule making authority, modifies and

15
contravenes the specific provisions of law it implements, and is

arbitrary and capricious.

29. Section 102.166, Florida Statutes (2004), deals with

manual recounts and provides:

(1) If the second set of unofficial

returns pursuant to s. 102.141 indicates
that a candidate for any office was defeated
or eliminated by one-quarter of a percent or
less of the votes cast for such office, that
a candidate for retention to a judicial
office was retained or not retained by one-
quarter of a percent or less of the votes
cast on the question of retention, or that a
measure appearing on the ballot was approved
or rejected by one-quarter of a percent or
less of the votes cast on such measure, the
board responsible for certifying the results
of the vote on such race or measure shall
order a manual recount of the overvotes and
undervotes cast in the entire geographic
jurisdiction of such office or ballot
measure.
(2)(a) If the second set of unofficial
returns pursuant to s. 102.141 indicates
that a candidate for any office was defeated
or eliminated by between one-quarter and
one-half of a percent of the votes cast for
such office, that a candidate for retention
to judicial office was retained or not
retained by between one-quarter and one-half
of a percent of the votes cast on the
question of retention, or that a measure
appearing on a ballot was approved or
rejected by between one-quarter and one-half
of a percent of the votes cast on such
measure, any such candidate, the political
party of such candidate, or any political
committee that supports or opposes such
ballot measure is entitled to a manual
recount of the overvotes and undervotes cast
in the entire geographic jurisdiction of
such office or ballot measure, provided that

16
 a request for a manual recount is made by
5 p.m. on the third day after the election.
(b) For federal, statewide, state, and
multicounty races and ballot issues,
requests for a manual recount shall be made
in writing to the state Elections Canvassing
Commission. For all other races and ballot
issues, requests for a manual recount shall
be made in writing to the county canvassing
board.
(c) Upon receipt of a proper and timely
request, the Elections Canvassing Commission
or county canvassing board shall immediately
order a manual recount of overvotes and
undervotes in all affected jurisdictions.

30. Florida Administrative Code Rule 1S-2.031(7) provides

that "[w]hen a manual recount is ordered, and touchscreen

ballots are used, no manual recount of undervotes and overvotes

cast on a touchscreen system shall be conducted." The rule is

contrary to the plain language of Section 102.166, Florida

Statutes (2004), which requires manual recounts of overvotes and

undervotes when the margin of victory is one-quarter of a

percent or less or when there is a proper and timely request for

a manual recount.

31. In State v. Burris, 875 So. 2d 408, 410 (Fla. 2004),

the Florida Supreme Court stated:

Our purpose in construing a statute is to

give effect to the Legislature's intent.
State v. J.M., 824 So. 2d 105, 109 (Fla.
2002). When a statute is clear, courts will
not look behind the statute's plain language
for legislative intent or resort to rules of
statutory construction to ascertain intent.
Lee County Elec. Coop., Inc. v. Jacobs, 820
So. 2d 297, 303 (Fla. 2002). Instead the

17
 statute's plain and ordinary meaning must
control, unless this leads to an
unreasonable result or a result clearly
contrary to legislative intent.

32. In Subsection 102.141(6), Florida Statutes (2004), the

Florida Legislature made a distinction between recounts of the

first set of unofficial returns for voting systems using paper

ballots and voting systems that do not use paper ballots, such

as the touchscreen voting systems. For a recount pursuant to

Subsection 102.141(6), Florida Statutes (2004), with systems

using paper ballots, each ballot is to be put through automatic

tabulating equipment and a determination is to be made whether

the returns correctly reflect the votes cast. The tabulating

equipment is to be tested immediately before and after the

recount, and, if no error is detected in the equipment, the

recount tabulation shall be presumed correct. For a recount

pursuant to Subsection 102.141(6), Florida Statutes (2004), for

voting systems that do not use paper ballots, the counters on

the precinct tabulators are to be examined to ensure that the

total of the returns on the tabulators equals the overall

election return. The counters on the tabulators are presumed

correct if there is a discrepancy between the overall election

returns and the tabulator counters.

33. The Florida Legislature made no distinction between

voting systems using paper ballots and those not using paper

18
ballots when requiring manual recounts. If the legislature had

intended that no manual recounts be done in counties using

voting systems which did not use paper ballots, it could have

easily done so; it did not. In the 2004 legislative session, SB

3004 was introduced containing language which would essentially

prohibit manual recounts for the touchscreen voting systems.

The conference committee substitute legislation rejected this

language, leaving the provisions requiring manual recounts

undisturbed.

34. Other language in Section 102.166, Florida Statutes

(2004), supports the conclusion that the legislature intended

that manual recounts be conducted for all types of voting

systems. Subsection 102.166(5)(b), Florida Statutes (2004),

states that Respondent "shall adopt specific rules for each

certified voting system prescribing what constitutes a 'clear

indication on the ballot that the voter has made a definite

choice.'" (emphasis supplied) Subsection 102.144(6)(d),

Florida Statutes (2004), provides that Respondent "shall adopt

detailed rules prescribing additional recount procedures for

each certified voting system which shall be uniform to the

extent practicable" and lists specific areas which the rules

must address. (emphasis supplied) The statutes clearly

contemplate that manual recounts will be done on each certified

voting system, including the touchscreen voting systems.

19
 35. The Florida Supreme Court emphasized the plain meaning

of a statute in determining legislative intent in Overstreet v.

State, 629 So. 2d 125, 126 (Fla. 1993):

Legislative intent must be determined

primarily from the language of the statute.
S.R.G. Corp. v. Department of Revenue, 365
So. 2d 687 (Fla. 1978). The legislature is
assumed to know the meaning of the words in
the statute and to have expressed its intent
by the use of those words. . . . We decline
to add words to a statute where, as in this
case, the language is clear and unambiguous.
"It is a settled rule of statutory
construction that unambiguous language is
not subject to judicial construction,
however wise it may seem to alter the plain
language." State v. Jett, 626 So. 2d 691
(Fla. 1993). If the legislature did not
intend the results mandated by the statute's
plain language, then the appropriate remedy
is for it to amend the statute.

If the Legislature does not intend for manual recounts to be

done for the touchscreen voting systems, it is the Legislature's

responsibility to amend the statute. Respondent cannot amend

the statute by promulgating a rule.

36. It is clear that Respondent exceeded its grant of

rulemaking authority in promulgating Florida Administrative Code

Rule 1S-2.031(7). It has the authority to promulgate procedures

for manual recounts in addition to those set forth in

Section 102.166, Florida Statutes (2004), and is required to

address minimum areas in those rules, but it does not have the

authority to abolish manual recounts for certain types of voting

20
equipment. Respondent is required to adopt rules prescribing

what constitutes a "clear indication on the ballot that the

voter has made a definite choice," but Respondent does not have

the authority to preclude manual recounts for a touchscreen

voting system because it does not believe that a manual recount

will reveal a clear indication of the voter's choice.

37. Petitioners and Intervenor have not presented evidence

to establish that the rule is arbitrary and capricious as those

terms are defined in Subsection 120.52(8)(e), Florida Statutes

(2004).

38. Subsections 120.52(8)(b) and (c), Florida Statutes

(2004), provide that if an agency has exceeded its grant of

rulemaking authority or if the rule enlarges, modifies, or

contravenes the provisions of law implemented, the rule is

deemed to be an invalid exercise of delegated legislative

authority. Thus, Petitioners ACLU-FL, CCF, and FSCLC, and

Intervenor PFAWF, have established that Florida Administrative

Code Rule 1S-2.031(7) is an invalid exercise of delegated

legislative authority.

39. In the Petition to Determine the Invalidity of Rules,

Petitioners requested attorney's fees and costs pursuant to

Subsection 120.595(3), Florida Statutes (2003). Jurisdiction is

retained to determine if Petitioners and Intervenor are entitled

to attorney's fees and costs, and, if so, the amount.

21
 ORDER

Based on the foregoing Findings of Fact and Conclusions of

Law, it is

ORDERED that:

1. Petitioners, Southwest Voter Registration Education

Project and Florida Voters League, Inc., are dismissed.

2. Florida Administrative Code Rule 1S-2.031(7) is an

invalid exercise of delegated legislative authority.

3. The issue of attorney's fees and costs will be

determined in a separate hearing.

DONE AND ORDERED this 27th day of August, 2004, in

Tallahassee, Leon County, Florida.

S
SUSAN B. KIRKLAND
Administrative Law Judge
Division of Administrative Hearings
The DeSoto Building
1230 Apalachee Parkway
Tallahassee, Florida 32399-3060
(850) 488-9675 SUNCOM 278-9675
Fax Filing (850) 921-6847
www.doah.state.fl.us

Filed with the Clerk of the

Division of Administrative Hearings
this 27th day of August, 2004.

22
COPIES FURNISHED:

Benjamin R. Patterson, Esquire

Jerry G. Traynham, Esquire
Patterson & Traynham
315 Beard Street
Post Office Box 4289
Tallahassee, Florida 32315-4289

George L. Waas, Esquire

Office of the Attorney General
The Capitol, Plaza Level 01
Tallahassee, Florida 32399-1050

Reginald J. Mitchell, Esquire

1550 Melvin Street
Tallahassee, Florida 32301

Alma R. Gonzalez, Esquire

AFSCME Florida Council 79
3064 Highland Oaks Terrace
Tallahassee, Florida 32301

Honorable Glenda E. Hood

Secretary of State
R.A. Gray Building
500 South Bronough Street
Tallahassee, Florida 32399-0250

Gerry York, General Counsel

Secretary of State
The Collins Building
107 West Gaines Street
Tallahassee, Florida 32399-0250

Scott Boyd, Executive Director and General Counsel

Joint Administrative Procedures Committee
120 Holland Building
Tallahassee, Florida 32399-1300

23
 NOTICE OF RIGHT TO JUDICIAL REVIEW

A party who is adversely affected by this Final Order is

entitled to judicial review pursuant to Section 120.68, Florida
Statutes. Review proceedings are governed by the Florida Rules
of Appellate Procedure. Such proceedings are commenced by
filing the original Notice of Appeal with the agency clerk of
the Division of Administrative Hearings and a copy, accompanied
by filing fees prescribed by law, with the District Court of
Appeal, First District, or with the District Court of Appeal in
the Appellate District where the party resides. The notice of
appeal must be filed within 30 days of rendition of the order to
be reviewed.

24
 Addendum B

Analysis and Report of

Overvotes and Undervotes
for the 2004 General Election
Pursuant to Section 101.595, Florida Statutes

January 31, 2005

Florida Department of State

Division of Elections
Room 316, R.A. Gray Building
500 South Bronough Street
Tallahassee, Florida 32399-0250
(850) 245-6200
Introduction

The 2004 election cycle was a great success. Our citizens engaged in the election as
never before and were energized to participate in their democracy at unprecedented
levels. Over 7.6 million Floridians voted in the presidential election either by voting
early, casting an absentee ballot, voting at the polls on election day, or by using a
provisional ballot. The reforms put in place since 2000 have served Florida's voters well.
Through continued commitment and hard work we will build on these reforms and
continue to improve the state’s elections process for all Florida voters.

Section 101.595, Florida Statutes, requires that the Department of State gather statistics
regarding the number of overvotes and undervotes cast in the first race of each General
Election and report its findings to the Legislature and the Governor no later than
January 31 of the year following that General Election. The purpose of this report is to
look at factors relating to no valid votes being cast. The two circumstances where this
occurs are ballots containing overvotes and undervotes. It is important to emphasize that
an undervote is not an error or “lost vote” but rather the prerogative of the voter. Today,
Florida's "undervote" rate is at the lowest level in our history. "Undervotes" occur when
voters exercise their right to withhold their vote, and often do for a number of reasons,
that include maintaining a perfect voting record, as a protest vote, or for reasons known
only to the voter. An overvote occurs when a voter casts more votes than allowable in
any race.

Although the election reforms enacted in 2001 resulted in a dramatic reduction in the
number of overvotes and undervotes cast in the 2002 election, this is the first time the
results of a Presidential Election have been available for study since those reforms were
put in place. In addition, it is important to note this was the first election cycle in which
early voting was mandated statewide and the first presidential election in which
provisional ballots were utilized. This report can thus serve as a new baseline with which
to evaluate future election results.

Executive Summary

Since 2000, there has been a continuing trend in the reduction of the number of ballots
cast without valid votes. In 2000, undervotes and overvotes accounted for 2.9302% of
the votes cast in the presidential race. In 2002, that number dropped to 0.7766% due, in
large part, to a change in the voting systems that were certified for use in the State. The
most recent election cycle saw yet another reduction in that number, dropping it to a
historically low 0.4116%. An unprecedented statewide, non-partisan voter education
campaign combined with the fact that voters continue to become more accustomed to
using new technologies, have contributed greatly to the continuation of this trend. This
conclusion is well documented in the Report on Voter Education Programs During the
2004 Election Cycle.

The difference in the rates of undervotes for touchscreen ballots and marksense (optical
scan) ballots is marginal, 0.4214 of one percent for touchscreen and 0.3056 of one
percent for marksense. The fact that a voter is able to cast an overvote on a marksense
ballot, but not on a touchscreen, resulted in the ratio of ballots with no valid vote cast on
the marksense systems being nearly identical to the ratio for touchscreens. The overvote
rate for marksense ballots was 0.0980 of one percent. When these additional 4,116
overvotes cast in the election are factored in with the marksense undervote rate, the ratio
for no valid vote cast in the presidential race becomes 0.4036 of one percent for
marksense ballots and 0.4214 of one percent for touchscreen ballots. When reviewing
this report, one should consider that the difference in voter performance between the two
systems is a statistically insignificant; 0.0178 of one percent. Applied against the
7,641,290 ballots in this election this difference is 1,360 ballots.

The extreme high and low, as well as the average rate of overvotes and undervotes for
each voting system and each style of ballot used by these systems has been examined.
Although there are some minor statistical differences, the numbers generated by this
study are small enough that they may well be the result of intangible factors, such as
voter intent, are not related to system design, and cannot be clearly documented. The
Department of State should continue to monitor overvotes and undervotes from each
General Election in order to detect any future fluctuations in these rates.

It should be noted that the Technical Development Guidelines Development Committee

of the United States Elections Assistance Commission has initiated a project to study the
human/machine interaction in voting systems with the intention to develop standards and
“best practices” for electronic voting machines and for ballot design.

Due to the recent increase in the use of provisional and early voting and the need to
capture the associated statistics for use in this, as well as other reports, the data necessary
to complete this report have become more complex and difficult to validate in the short
timeframe available. It is recommended that the Legislature expand the data required as
part of the official canvass report to include the data required for this report, for any
reports required by the federal government, and give the Department of State rule making
authority to specify the content and format of the data.

Data Sources and Data Validation

Data for this report was gathered from three sources. First, a survey regarding overvotes
and undervotes was sent to each Supervisor of Elections (Attachment A). This survey
became the basis for developing statistics for the number of overvotes and undervotes
cast for each type of ballot: precinct, absentee, early, and provisional.

Next, each county’s official Canvass Report was used to derive the number of candidate
votes used in this study. These votes were totaled to calculate the “Candidate Votes
Recorded” statistic in the spreadsheet titled “Sorted_Over_Under.xls” (Attachment B).
Although the canvass reports were an adequate source to obtain information concerning
the candidate votes, variations existed from county to county in the determination of the
total number of votes cast in the election. Many counties only included the total number
of votes cast for each candidate in their total number of votes cast, while other counties

2
included a count of all ballots cast in the race. The difference between these two
approaches is whether or not overvotes and undervotes are included as part of the statistic
for the total number of votes cast. Florida law is silent as to which approach is correct.
Since candidate votes cast plus overvotes and undervotes must always equal “total ballots
cast”, adjustments to the number of “Total Votes Cast” in the Canvass Reports were
required for several counties. These adjustments are detailed in the “Data Validation
Notes” included as Attachment C to this report.

The third source of data was a survey taken earlier in the month for the United States
Election Assistance Commission (EAC). The EAC survey became the basis for total
ballots cast for each type of ballot: precinct, absentee, early, and provisional. During the
validation process of this data, it became apparent that, as in the case of the Canvass
Reports, some counties reported only actual candidate votes for each type of ballot, while
others reported the actual number of votes (including overvotes and undervotes) cast for
each type. In addition, some counties included military overseas ballots in their absentee
count and others did not.

The use of two page optical scan ballots by some of the counties is yet another reason for
the differing totals between the Canvass Report and the column entitled “True Turnout
Ballots Counted” on the spreadsheet. Several counties reported that there were
differences in the numbers of ballots cast for pages one and two of the ballot. It appears
that some voters may have cast only one of the two ballot pages. In this situation, some
of the counties used the largest number of ballots cast for each page in calculating their
votes cast for each precinct while others used an average of the number of pages read.

There were nineteen counties that reported no absentee overvotes. When these counties
were contacted by the Division of Elections, six of them reported that, pursuant to section
101.5614(5), Florida Statutes, duplicate ballots were made of absentee ballots containing
an overvoted race. If the canvassing board found that the overvoted race showed a clear
indication of the voter’s choice, then the ballot was duplicated to indicate that choice. If
they could not determine a clear indication of voter’s choice in that race, they duplicated
the ballot to include all valid votes. This action had the net effect of converting overvotes
to undervotes. The counties are as follows: Columbia, two ballots; Duval, twenty
ballots; Leon, thirty-two ballots; Seminole, twenty ballots; Taylor, one ballot; Lake,
seven ballots. This resulted in approximately eighty-two votes statewide.

It should be noted that the Department has not audited the data provided by the counties.
Our calculation for “True Turnout Ballots Counted” is based on the premise that the
candidate votes reported by the county, plus the number of overvotes and undervotes
reported by the county, equal “True Turnout Ballots Counted” which should then equal
the “Total Ballots Counted” in the EAC report. After all the inconsistencies were
addressed, there remained a difference of fifty-nine votes between the corrected EAC
numbers for “Total Ballots Counted” and our numbers for “True Turnout Ballots
Counted”, spread over eleven counties, that could not be reconciled. These unresolved
differences were addressed by adding the difference to both the corrected “Canvass
Turnout” and to “Precinct Ballots.”

3
Discussion of the analysis

An analysis of the reduction in undervotes and overvotes is displayed below:

2000 Presidential 2002 Gubernatorial 2004 Presidential

Blank and Spoiled 179,855

Absentee Undervotes 4,405 5,441
Absentee Overvotes 1,121 3,119
Precinct Undervotes 33,737 17,516
Precinct Overvotes 689 760
Early Vote Undervotes 4,271
Early Vote Overvotes 168
Provisional Undervotes 109
Provisional Overvotes 69

Totals 179,855 39,952 31,453

Turnout 6,137,938 5,144,477 7,641,290
Ratio of No Votes 2.9302% 0.7766% 0.4116%

The numbers for the 2000 and 2002 elections are derived from the 2002 Report on
Overvotes and Undervotes, published by the Division of Elections. The analysis of the
2004 data is displayed in the spreadsheet titled “Sorted_Over_Under.xls”. A hard copy
of the spreadsheet is included as “Attachment B” and a CD containing the spreadsheet
has also been provided with the published version of this report. If the report is
downloaded from the Division of Elections Website,
http://election.dos.state.fl.us/reports/04OverUnderVotes.shtml, the index provides a link
to download the spreadsheet.

The county statistics provided in the spreadsheet have been sorted according to the voting
system and by the type of ballot used in the precincts. In understanding this data it is
important to remember that touchscreen machines are used exclusively for precinct and
early voting. All counties use marksense ballots for their absentee and provisional
voting.

The first section of the spreadsheet examines the Diebold system, using marksense oval
targeted ballots in the precincts. There was some use of Diebold touchscreens in early
voting in Duval County and for accommodating voters with disabilities in other counties.
We have not considered those counties as a separate category.

4
In the Diebold Counties:

The averages in the Diebold Counties were:

Undervotes Overvotes
Absentee 0.3175% 0.0980%
Early Voting 0.1840% 0.0160%
Precinct 0.2375% 0.0138%
Provisional 0.3949% 0.3385%
Overall 0.2447% 0.0303%

Calhoun County had the greatest ratio of ballots cast without a valid vote in the
Presidential race. Undervotes were 0.7490 of one percent and overvotes were
zero. The undervotes were fairly evenly distributed across the four types of
ballots.

Leon County had the smallest ratio of ballots cast without a valid vote in the
Presidential race. Undervotes were reported at 0.1896 of one percent and
overvotes were reported at zero. The distribution of undervoted ballots was
heaviest among the absentee ballots.

The second section of the spreadsheet discusses the ES&S system, using the ES&S model
100 precinct scanner and marksense oval targeted ballots. Many of the counties using
this system also use ES&S touchscreen terminals to accommodate voters with
disabilities. Highlands County has an ES&S touchscreen in each precinct to
accommodate voters with disabilities. We have not separated those counties into a
separate category.

In the ES&S 100 Counties:

The averages in the ES&S 100 Counties were:

Undervotes Overvotes
Absentee 0.4753% 0.1475%
Early Voting 0.3176% 0.0831%
Precinct 0.3739% 0.1482%
Provisional 0.6849% 2.0458%
Overall 0.3829% 0.1366%

Although the ratio of overvotes on the provisional ballots appears extremely high,
it represents only three out of one hundred and forty six provisional ballots.

Hamilton County had the greatest ratio of ballots without a valid vote cast in the
Presidential race. Undervotes were 0.5652 of one percent and overvotes were

5
 0.4483 of one percent. The ratio of undervotes among absentee ballots is
reasonably high at 0.8662 of one percent.

Bay County had the smallest ratio of ballots without a valid vote cast in the
Presidential race. Undervotes were reported at 0.2499 of one percent and
overvotes were reported at 0.0292 of one percent.

The third section of the spreadsheet discusses the ES&S system, using the ES&S Optech
and the Sequoia Optech precinct scanners and marksense arrow targeted ballots.

In the ES&S and Sequoia Optech Counties:

The averages in the ES&S and Sequoia Optech Counties were:

Undervotes Overvotes
Absentee 0.7739% 0.5792%
Early Voting 0.2997% 0.0391%
Precinct 0.2861% 0.0415%
Provisional 0.7513% 0.1669%
Overall 0.3817% 0.1433%

Baker County had the greatest ratio of ballots without a valid vote cast in the
Presidential race. Undervotes were 0.6270 of one percent and overvotes were
0.2986 of one percent. The undervotes and overvotes among absentee ballots
were fairly high at 1.6591 percent and 1.2066 percent.

Santa Rosa County had the smallest ratio of ballots without a valid vote cast in the
Presidential race. Undervotes were reported at 0.1748 of one percent and
overvotes were reported at 0.0548 of one percent.

The fourth section of the spreadsheet discusses the ES&S system, using the ES&S
touchscreen system and marksense oval targeted ballots for absentee and provisional
ballots.

In the ES&S Touchscreen Counties:

The averages in the ES&S Touchscreen Counties were:

Undervotes Overvotes
Absentee 0.3797% 0.1846%
Early Voting 0.3420% 0.0%
Precinct 0.4767% 0.0%
Provisional 0.6282% 0.1496%
Overall 0.4276% 0.0301%

Note: Early and precinct ballots are on touchscreen systems, which do not allow
overvotes. Absentee and provisional ballots are on oval targeted marksense
ballots which can be overvoted.

6
 Miami-Dade County had the greatest ratio of ballots without a valid vote cast in
the Presidential race. Undervotes were 0.4874 of one percent and overvotes were
0.0552 of one percent. The undervotes among precinct ballots were fairly high at
0.5868 of one percent.

Nassau County had the smallest ratio of ballots without a valid vote cast in the
Presidential race. Undervotes were reported at 0.2163 of one percent and
overvotes were reported at 0.0396 of one percent.

The fifth section of the spreadsheet discusses the Sequoia system, using the Sequoia
touchscreen system and marksense arrow targeted ballots for absentee and provisional
ballots.

In the Sequoia Touchscreen Counties:

The averages in the Sequoia Touchscreen Counties were:

Undervotes Overvotes
Absentee 0.3698% 0.3640%
Early Voting 0.3487% 0.0%
Precinct 0.4056% 0.0%
Provisional 3.5854% 2.6331%
Overall 0.3950% 0.0640%

Note: Early and precinct ballots are on touchscreen systems, which do not allow
overvotes. Absentee and provisional ballots are on arrow targeted marksense
ballots which can be overvoted.

Palm Beach County had the greatest ratio of ballots without a valid vote cast in
the Presidential race. Undervotes were 0.4310 of one percent and overvotes were
0.0656 of one percent

Hillsborough County had the smallest ratio of ballots without a valid vote cast in
the Presidential race. Undervotes were reported at 0.3400 of one percent and
overvotes were reported at 0.0460 of one percent. However, the provisional ballot
ratios were high with undervotes reported at 8.2126 percent and overvotes
reported at 3.3816 percent

The sixth section of the spreadsheet contains a summary of the totals from the preceding
five sections.

The seventh section of the spreadsheet provides a breakdown of the total numbers for
oval marksense ballots, arrow marksense ballots and touchscreens. In this analysis,
counties using arrow marksense ballots had a higher ratio of ballots without a valid vote
cast in the Presidential race. Total undervotes were at a ratio of 0.3842 of one percent for
arrows and 0.2791 of one percent for ovals. Total overvotes were at a ratio of 0.2010 of
one percent for arrows and 0.0634 for ovals. Totals for ballots cast on touchscreen voting
systems were at a ratio of 0.4214 for undervotes and had zero overvotes.

7
The eighth section of the spreadsheet compares undervotes cast on touchscreen ballots to
undervotes and overvotes cast on marksense ballots and to the overall undervotes and
overvotes for all ballots. The ratio of undervotes on touchscreen ballots is 0.4214 of one
percent and the ratio of combined undervotes and overvotes on marksense ballots is
0.4036 of one percent. The ratio of combined undervotes and overvotes for all ballots is
0.4116 of one percent. The result is a statistically insignificant .0178.

The two columns on the far right side of the spreadsheet sheet contain estimated
percentages of true undervotes cast on absentee ballots verses mismarked ballots on a
county by county basis. Supervisors of Elections were asked to make an estimate based
upon their experiences during the canvass of absentee ballots. Miami-Dade had the
highest estimate at fifteen percent. Three counties reported ten percent. Twenty-two of
the remaining counties reported five percent or less, and forty-one reported zero. We
have attempted no analysis on these estimates but have recorded them and presented
them as reported by the counties.

8
 Florida Department Of State - Division Of Elections

Attachment A

2004 GENERAL ELECTION

REPORT ON OVERVOTES AND UNDERVOTES
(s. 101.595, F.S.)

ANY COUNTY

President of the United States

(Turnout - Votes: 0 - 0 = 0)
Absentee Absentee Early Voting Early Voting Precinct Precinct Provisional Provisional InValid
TOTAL
UnderVotes OverVotes UnderVotes OverVotes UnderVotes OverVotes UnderVotes OverVotes Write-Ins
0 0 0 0 0 0 0 0 0 0

Reasons for Absentee UnderVotes

(An Estimated Percentage)
True UnderVotes: 0%
Mismarked Ballots: 0%

Problems Encountered with the Ballot Design or Instructions

which may have contributed to voter confusion
None

Recommendations for Correcting Ballot Confusion

None

Problems Encountered with the Voting System Design

None

Recommendations for Correcting Design Problems

None

_______________________________________ ______________
Supervisor of Elections Date

http://election.dos.state.fl.us/reports/pdf/Attachment_A.htm5/4/2005 7:00:13 AM
Sorted Over Under.xls Florida Department of State Attachment B
Analysis and Report of Overvotes and Undervotes
for the 2004 General Election

County Voting Precinct Canvass True Turnout Candidate Total Total Total Total Absentee Absentee Absentee Absentee Absentee Early Vote Early Vote Early Vote Early Vote Early Vote Precinct Precinct Precinct Precinct Precinct Provisional Provisional Provisional Provisional Provisional True Undervote % on Mismarked Undervote %
Name System Method Turnout Ballots Counted Votes Recorded Undervotes UV % Overvotes OV % Ballots Undervotes UV % Overvotes OV % Ballots Undervotes UV % Overvotes OV % Ballots Undervotes UV % Overvotes OV % Ballots Undervotes UV % Overvotes OV % Absentees (Estimate) on Absentees (Estimate
Alachua Diebold Marksense 111,566 111,598 111,328 241 0.2160% 29 0.0260% 18,691 56 0.2996% 22 0.1177% 16,389 26 0.1586% 1 0.0061% 76,381 159 0.2082% 6 0.0079% 137 0 0.0000% 0 0.0000% 95 5
Brevard Diebold Marksense 266,160 266,160 265,462 584 0.2194% 114 0.0428% 50,070 121 0.2417% 58 0.1158% 30,564 60 0.1963% 8 0.0262% 185,325 403 0.2175% 48 0.0259% 201 0 0.0000% 0 0.0000% 99 1
Calhoun Diebold Marksense 6,008 6,008 5,963 45 0.7490% 0 0.0000% 842 8 0.9501% 0 0.0000% 1,447 9 0.6220% 0 0.0000% 3,718 28 0.7531% 0 0.0000% 1 0 0.0000% 0 0.0000% 100 0
Citrus Diebold Marksense 69,701 69,701 69,467 214 0.3070% 20 0.0287% 15,400 106 0.6883% 15 0.0974% 20,117 0 0.0000% 0 0.0000% 34,182 108 0.3160% 5 0.0146% 2 0 0.0000% 0 0.0000% 100 0
Columbia Diebold Marksense 25,082 25,082 24,991 88 0.3508% 3 0.0120% 3,238 17 0.5250% 0 0.0000% 7,339 30 0.4088% 0 0.0000% 14,497 41 0.2828% 3 0.0207% 8 0 0.0000% 0 0.0000% 98 2
Desoto Diebold Marksense 9,549 9,549 9,510 39 0.4084% 0 0.0000% 1,830 11 0.6011% 0 0.0000% 2,310 11 0.4762% 0 0.0000% 5,408 17 0.3143% 0 0.0000% 1 0 0.0000% 0 0.0000% 100 0
Dixie Diebold Marksense 6,474 6,474 6,442 32 0.4943% 0 0.0000% 919 12 1.3058% 0 0.0000% 1,091 3 0.2750% 0 0.0000% 4,464 17 0.3808% 0 0.0000% 0 0 0.0000% 0 0.0000% 100 0
Duval Diebold Marksense 382,006 382,006 381,061 939 0.2458% 6 0.0016% 63,352 201 0.3173% 0 0.0000% 58,693 113 0.1925% 0 0.0000% 258,746 625 0.2415% 0 0.0000% 1,215 0 0.0000% 6 0.4938% 100 0
Flagler Diebold Marksense 38,557 38,557 38,480 68 0.1764% 9 0.0233% 5,972 11 0.1842% 6 0.1005% 8,873 17 0.1916% 1 0.0113% 23,694 40 0.1688% 2 0.0084% 18 0 0.0000% 0 0.0000% 99 1
Gilchrist Diebold Marksense 7,050 7,050 7,015 35 0.4965% 0 0.0000% 919 4 0.4353% 0 0.0000% 1,616 11 0.6807% 0 0.0000% 4,514 20 0.4431% 0 0.0000% 1 0 0.0000% 0 0.0000% 100 0
Glades Diebold Marksense 4,204 4,204 4,188 16 0.3806% 0 0.0000% 746 4 0.5362% 0 0.0000% 345 3 0.8696% 0 0.0000% 3,101 9 0.2902% 0 0.0000% 12 0 0.0000% 0 0.0000% 100 0
Hardee Diebold Marksense 7,284 7,284 7,249 28 0.3844% 7 0.0961% 1,052 11 1.0456% 4 0.3802% 2,831 10 0.3532% 0 0.0000% 3,395 7 0.2062% 3 0.0884% 6 0 0.0000% 0 0.0000% 100 0
Hernando Diebold Marksense 80,764 80,764 80,532 198 0.2452% 34 0.0421% 15,865 34 0.2143% 23 0.1450% 13,942 31 0.2223% 3 0.0215% 50,942 133 0.2611% 8 0.0157% 15 0 0.0000% 0 0.0000% 100 0
Jefferson Diebold Marksense 7,502 7,502 7,478 24 0.3199% 0 0.0000% 1,037 6 0.5786% 0 0.0000% 2,486 10 0.4023% 0 0.0000% 3,979 8 0.2011% 0 0.0000% 0 0 0.0000% 0 0.0000% 100 0
Leon Diebold Marksense 136,638 136,638 136,379 259 0.1896% 0 0.0000% 22,900 93 0.4061% 0 0.0000% 17,974 17 0.0946% 0 0.0000% 95,397 149 0.1562% 0 0.0000% 367 0 0.0000% 0 0.0000% 98 2
Levy Diebold Marksense 16,742 16,745 16,652 84 0.5016% 9 0.0537% 2,917 21 0.7199% 0 0.0000% 2,078 10 0.4812% 2 0.0962% 11,748 53 0.4511% 7 0.0596% 2 0 0.0000% 0 0.0000% 100 0
Madison Diebold Marksense 8,345 8,345 8,310 34 0.4074% 1 0.0120% 993 9 0.9063% 1 0.1007% 2,446 8 0.3271% 0 0.0000% 4,905 17 0.3466% 0 0.0000% 1 0 0.0000% 0 0.0000% 100 0
Manatee Diebold Marksense 143,983 143,983 143,621 296 0.2056% 66 0.0458% 26,721 70 0.2620% 29 0.1085% 10,274 7 0.0681% 4 0.0389% 106,917 219 0.2048% 32 0.0299% 71 0 0.0000% 1 1.4085% 100 0
Monroe Diebold Marksense 39,629 39,647 39,535 101 0.2547% 11 0.0277% 11,018 31 0.2814% 6 0.0545% 8,917 18 0.2019% 3 0.0336% 19,703 52 0.2639% 2 0.0102% 9 0 0.0000% 0 0.0000% 100 0
Okaloosa Diebold Marksense 89,957 89,957 89,756 189 0.2101% 12 0.0133% 19,928 31 0.1556% 9 0.0452% 10,383 17 0.1637% 0 0.0000% 59,566 141 0.2367% 3 0.0050% 80 0 0.0000% 0 0.0000% 100 0
Okeechobee Diebold Marksense 12,249 12,249 12,190 55 0.4490% 4 0.0327% 2,173 10 0.4602% 3 0.1381% 3,507 17 0.4847% 0 0.0000% 6,565 28 0.4265% 1 0.0152% 4 0 0.0000% 0 0.0000% 0 0
Osceola Diebold Marksense 82,204 82,431 82,204 204 0.2475% 23 0.0279% 10,451 37 0.3540% 13 0.1244% 19,161 40 0.2088% 1 0.0052% 52,660 127 0.2412% 8 0.0152% 159 0 0.0000% 1 0.6289% 100 0
Polk Diebold Marksense 211,399 211,399 210,830 397 0.1878% 172 0.0814% 57,148 137 0.2397% 123 0.2152% 6,023 6 0.0996% 0 0.0000% 148,150 254 0.1714% 48 0.0324% 78 0 0.0000% 1 1.2821% 99 1
Putnam Diebold Marksense 31,072 31,072 30,973 93 0.2993% 6 0.0193% 4,513 22 0.4875% 0 0.0000% 3,775 7 0.1854% 0 0.0000% 22,780 64 0.2809% 6 0.0263% 4 0 0.0000% 0 0.0000% 90 10
Seminole Diebold Marksense 186,617 186,617 186,195 414 0.2218% 8 0.0043% 30,428 79 0.2596% 0 0.0000% 20,671 18 0.0871% 5 0.0242% 135,161 313 0.2316% 3 0.0022% 357 4 1.1204% 0 0.0000% 95 5
St. Lucie Diebold Marksense 100,374 100,410 100,063 246 0.2450% 101 0.1006% 18,616 53 0.2847% 87 0.4673% 22,882 32 0.1398% 2 0.0087% 58,795 159 0.2704% 11 0.0187% 117 2 1.7094% 1 0.8547% 100 0
Taylor Diebold Marksense 8,614 8,614 8,581 33 0.3831% 0 0.0000% 1,518 9 0.5929% 0 0.0000% 2,421 8 0.3304% 0 0.0000% 4,672 16 0.3425% 0 0.0000% 3 0 0.0000% 0 0.0000% 100 0
Volusia Diebold Marksense 229,193 229,680 228,939 690 0.3004% 51 0.0222% 44,218 156 0.3528% 28 0.0633% 40,255 86 0.2136% 3 0.0075% 144,568 441 0.3050% 18 0.0125% 639 7 1.0955% 2 0.3130% 100 0
Wakulla Diebold Marksense 11,820 11,820 11,763 57 0.4822% 0 0.0000% 1,671 20 1.1969% 0 0.0000% 4,776 19 0.3978% 0 0.0000% 5,372 18 0.3351% 0 0.0000% 1 0 0.0000% 0 0.0000% 100 0
Walton Diebold Marksense 24,065 24,065 23,976 61 0.2535% 28 0.1164% 4,571 16 0.3500% 4 0.0875% 6,462 0 0.0000% 23 0.3559% 12,996 44 0.3386% 1 0.0077% 36 1 2.7778% 0 0.0000% 100 0

Totals Diebold Marksense 2,354,808 2,355,611 2,349,133 5,764 0.2447% 714 0.0303% 439,717 1,396 0.3175% 431 0.0980% 350,048 644 0.1840% 56 0.0160% 1,562,301 3,710 0.2375% 215 0.0138% 3,545 14 0.3949% 12 0.3385%

Bay ES&S 100 Marksense 75,234 75,234 75,024 188 0.2499% 22 0.0292% 18,639 61 0.3273% 7 0.0376% 18,572 35 0.1885% 2 0.0108% 37,981 92 0.2422% 13 0.0342% 42 0 0.0000% 0 0.0000% 99 1
Bradford ES&S 100 Marksense 10,917 10,917 10,855 37 0.3389% 25 0.2290% 1,613 7 0.4340% 3 0.1860% 2,429 9 0.3705% 0 0.0000% 6,875 21 0.3055% 22 0.3200% 0 0 0.0000% 0 0.0000% 90 10
Franklin ES&S 100 Marksense 5,974 5,974 5,931 29 0.4854% 14 0.2343% 1,277 13 1.0180% 5 0.3915% 1,586 2 0.1261% 1 0.0631% 3,111 14 0.4500% 8 0.2572% 0 0 0.0000% 0 0.0000% 100 0
Gadsden ES&S 100 Marksense 21,094 21,094 20,984 100 0.4741% 10 0.0474% 3,602 40 1.1105% 0 0.0000% 5,467 20 0.3658% 1 0.0183% 12,024 40 0.3327% 9 0.0749% 1 0 0.0000% 0 0.0000% 99 1
Gulf ES&S 100 Marksense 7,320 7,320 7,277 36 0.4918% 7 0.0956% 1,475 10 0.6780% 1 0.0678% 1,725 6 0.3478% 0 0.0000% 4,120 20 0.4854% 6 0.1456% 0 0 0.0000% 0 0.0000% 100 0
Hamilton ES&S 100 Marksense 5,131 5,131 5,079 29 0.5652% 23 0.4483% 1,039 9 0.8662% 0 0.0000% 937 3 0.3202% 1 0.1067% 3,147 16 0.5084% 22 0.6991% 8 1 12.5000% 0 0.0000% 100 0
Hendry ES&S 100 Marksense 9,818 9,818 9,775 33 0.3361% 10 0.1019% 1,520 6 0.3947% 8 0.5263% 2,481 8 0.3225% 0 0.0000% 5,817 19 0.3266% 2 0.0344% 0 0 0.0000% 0 0.0000% 100 0
Highlands ES&S 100 Marksense 41,876 41,876 41,496 272 0.6495% 108 0.2579% 7,136 44 0.6166% 23 0.3223% 11,385 53 0.4655% 22 0.1932% 23,351 175 0.7494% 63 0.2698% 4 0 0.0000% 0 0.0000% 100 0
Jackson ES&S 100 Marksense 19,908 19,908 19,807 73 0.3667% 28 0.1406% 2,567 12 0.4675% 9 0.3506% 4,677 15 0.3207% 5 0.1069% 12,655 46 0.3635% 14 0.1106% 9 0 0.0000% 0 0.0000% 99 1
Lafayette ES&S 100 Marksense 3,352 3,352 3,325 23 0.6862% 4 0.1193% 543 4 0.7366% 1 0.1842% 873 9 1.0309% 0 0.0000% 1,935 10 0.5168% 3 0.1550% 1 0 0.0000% 0 0.0000% 100 0
Liberty ES&S 100 Marksense 3,051 3,051 3,021 24 0.7866% 6 0.1967% 363 4 1.1019% 0 0.0000% 864 5 0.5787% 0 0.0000% 1,823 15 0.8228% 6 0.3291% 1 0 0.0000% 0 0.0000% 99 1
Marion ES&S 100 Marksense 140,354 140,354 139,677 462 0.3292% 215 0.1532% 25,829 105 0.4065% 42 0.1626% 14,799 38 0.2568% 22 0.1487% 99,661 319 0.3201% 149 0.1495% 65 0 0.0000% 2 3.0769% 100 0
Suwannee ES&S 100 Marksense 15,879 15,879 15,802 59 0.3716% 18 0.1134% 2,732 17 0.6223% 4 0.1464% 2,466 15 0.6083% 3 0.1217% 10,666 27 0.2531% 10 0.0938% 15 0 0.0000% 1 6.6667% 99 1
Union ES&S 100 Marksense 4,714 4,714 4,675 31 0.6576% 8 0.1697% 1,509 0 0.0000% 0 0.0000% 369 0 0.0000% 0 0.0000% 2,836 31 1.0931% 8 0.2821% 0 0 0.0000% 0 0.0000% 100 0

Totals ES&S Marksense 364,622 364,622 362,728 1,396 0.3829% 498 0.1366% 69,844 332 0.4753% 103 0.1475% 68,630 218 0.3176% 57 0.0831% 226,002 845 0.3739% 335 0.1482% 146 1 0.6849% 3 2.0548%

Clay ES&S Optech Marksense 81,731 81,731 81,495 144 0.1762% 92 0.1126% 17,664 51 0.2887% 65 0.3680% 6,770 5 0.0739% 5 0.0739% 57,209 88 0.1538% 22 0.0385% 88 0 0.0000% 0 0.0000% 100 0
Escambia ES&S Optech Marksense 143,849 143,849 143,278 402 0.2795% 169 0.1175% 35,334 86 0.2434% 151 0.4274% 12,807 39 0.3045% 4 0.0312% 95,647 277 0.2896% 14 0.0146% 61 0 0.0000% 0 0.0000% 98 2
Holmes ES&S Optech Marksense 8,351 8,351 8,300 41 0.4910% 10 0.1197% 946 3 0.3171% 7 0.7400% 1,987 7 0.3523% 0 0.0000% 5,418 31 0.5722% 3 0.0554% 0 0 0.0000% 0 0.0000% 100 0
Orange ES&S Optech Marksense 390,706 390,706 388,044 1,991 0.5096% 671 0.1717% 67,414 941 1.3959% 499 0.7402% 79,171 249 0.3145% 31 0.0392% 243,536 801 0.3289% 141 0.0579% 585 0 0.0000% 0 0.0000% 99 1
Santa Rosa ES&S Optech Marksense 67,498 67,498 67,343 118 0.1748% 37 0.0548% 11,472 15 0.1308% 25 0.2179% 15,810 24 0.1518% 2 0.0127% 40,177 78 0.1941% 10 0.0249% 39 1 2.5641% 0 0.0000% 100 0
St. Johns ES&S Optech Marksense 86,631 86,631 86,290 227 0.2620% 114 0.1316% 15,299 42 0.2745% 101 0.6602% 18,758 61 0.3252% 4 0.0213% 52,151 117 0.2243% 7 0.0134% 423 7 1.6548% 2 0.4728% 98 2
Washington ES&S Optech Marksense 10,453 10,453 10,366 65 0.6218% 22 0.2105% 2,122 13 0.6126% 14 0.6598% 2,653 21 0.7916% 2 0.0754% 5,676 30 0.5285% 6 0.1057% 2 1 50.0000% 0 0.0000% 99 1
Baker Sequoia OptechMarksense 10,048 10,048 9,955 63 0.6270% 30 0.2986% 1,326 22 1.6591% 16 1.2066% 2,843 16 0.5628% 7 0.2462% 5,879 25 0.4252% 7 0.1191% 0 0 0.0000% 0 0.0000% 99 1

Totals Optech Marksense 799,267 799,267 795,071 3,051 0.3817% 1,145 0.1433% 151,577 1,173 0.7739% 878 0.5792% 140,799 422 0.2997% 55 0.0391% 505,693 1,447 0.2861% 210 0.0415% 1,198 9 0.7513% 2 0.1669%

Broward ES&S Touchscreen 709,724 709,724 706,872 2,784 0.3923% 68 0.0096% 98,971 350 0.3536% 63 0.0637% 176,743 498 0.2818% 0 0.0000% 431,488 1,916 0.4440% 0 0.0000% 2,522 20 0.7930% 5 0.1983% 95 5
Charlotte ES&S Touchscreen 80,196 80,196 79,786 374 0.4664% 36 0.0449% 19,744 73 0.3697% 36 0.1823% 13,659 57 0.4173% 0 0.0000% 46,741 244 0.5220% 0 0.0000% 52 0 0.0000% 0 0.0000% 99 1
Collier ES&S Touchscreen 129,231 129,231 128,683 512 0.3962% 36 0.0279% 27,723 114 0.4112% 36 0.1299% 44,155 164 0.3714% 0 0.0000% 57,323 234 0.4082% 0 0.0000% 30 0 0.0000% 0 0.0000% 100 0
Lake ES&S Touchscreen 124,488 124,488 123,950 538 0.4322% 0 0.0000% 23,971 142 0.5924% 0 0.0000% 32,331 116 0.3588% 0 0.0000% 68,121 280 0.4110% 0 0.0000% 65 0 0.0000% 0 0.0000% 100 0
Lee ES&S Touchscreen 241,663 241,663 240,667 969 0.4010% 27 0.0112% 47,479 159 0.3349% 27 0.0569% 32,182 109 0.3387% 0 0.0000% 161,953 700 0.4322% 0 0.0000% 49 1 2.0408% 0 0.0000% 99 1
Martin ES&S Touchscreen 72,736 72,736 72,453 273 0.3753% 10 0.0137% 16,597 75 0.4519% 10 0.0603% 12,243 53 0.4329% 0 0.0000% 43,892 145 0.3304% 0 0.0000% 4 0 0.0000% 0 0.0000% 100 0
Miami-Dade ES&S Touchscreen 778,953 778,953 774,726 3,797 0.4874% 430 0.0552% 98,466 338 0.3433% 430 0.4367% 244,022 901 0.3692% 0 0.0000% 435,887 2,558 0.5868% 0 0.0000% 578 0 0.0000% 0 0.0000% 85 15
Nassau ES&S Touchscreen 32,827 32,827 32,743 71 0.2163% 13 0.0396% 6,991 15 0.2146% 13 0.1860% 9,989 19 0.1902% 0 0.0000% 15,839 37 0.2336% 0 0.0000% 8 0 0.0000% 0 0.0000% 99 1
Pasco ES&S Touchscreen 191,909 191,909 190,916 857 0.4466% 136 0.0709% 34,794 161 0.4627% 136 0.3909% 29,584 114 0.3853% 0 0.0000% 127,526 582 0.4564% 0 0.0000% 5 0 0.0000% 0 0.0000% 100 0
Sarasota ES&S Touchscreen 196,413 196,413 195,652 744 0.3788% 17 0.0087% 40,412 140 0.3464% 17 0.0421% 33,872 100 0.2952% 0 0.0000% 122,101 504 0.4128% 0 0.0000% 28 0 0.0000% 0 0.0000% 100 0
Sumter ES&S Touchscreen 32,005 32,005 31,842 157 0.4905% 6 0.0187% 4,104 25 0.6092% 6 0.1462% 16,517 76 0.4601% 0 0.0000% 11,382 56 0.4920% 0 0.0000% 2 0 0.0000% 0 0.0000% 100 0

Totals ES&S Touchscreen 2,590,145 2,590,145 2,578,290 11,076 0.4276% 779 0.0301% 419,252 1,592 0.3797% 774 0.1846% 645,297 2,207 0.3420% 0 0.0000% 1,522,253 7,256 0.4767% 0 0.0000% 3,343 21 0.6282% 5 0.1496%

Hillsborough Sequoia Touchscreen 464,849 465,017 463,222 1,581 0.3400% 214 0.0460% 64,031 218 0.3405% 200 0.3123% 86,642 277 0.3197% 0 0.0000% 313,930 1,052 0.3351% 0 0.0000% 414 34 8.2126% 14 3.3816% 100 0
Indian River Sequoia Touchscreen 61,707 61,707 61,414 263 0.4262% 30 0.0486% 12,152 47 0.3868% 27 0.2222% 17,328 67 0.3867% 0 0.0000% 32,015 149 0.4654% 0 0.0000% 212 0 0.0000% 3 1.4151% 90 10
Palm Beach Sequoia Touchscreen 547,340 547,340 544,622 2,359 0.4310% 359 0.0656% 91,802 434 0.4728% 329 0.3584% 49,831 196 0.3933% 0 0.0000% 404,666 1,701 0.4203% 0 0.0000% 1,041 28 2.6897% 30 2.8818% 99 1
Pinellas Sequoia Touchscreen 457,581 457,581 455,357 1,847 0.4036% 377 0.0824% 88,360 249 0.2818% 377 0.4267% 69,902 240 0.3433% 0 0.0000% 299,201 1,356 0.4532% 0 0.0000% 118 2 1.6949% 0 0.0000% 100 0

Totals Sequoia Touchscreen 1,531,477 1,531,645 1,524,615 6,050 0.3950% 980 0.0640% 256,345 948 0.3698% 933 0.3640% 223,703 780 0.3487% 0 0.0000% 1,049,812 4,258 0.4056% 0 0.0000% 1,785 64 3.5854% 47 2.6331%

Totals Diebold Marksense 2,354,808 2,355,611 2,349,133 5,764 0.2447% 714 0.0303% 439,717 1,396 0.3175% 431 0.0980% 350,048 644 0.1840% 56 0.0160% 1,562,301 3,710 0.2375% 215 0.0138% 3,545 14 0.3949% 12 0.3385%
Totals ES&S Marksense 364,622 364,622 362,728 1,396 0.3829% 498 0.1366% 69,844 332 0.4753% 103 0.1475% 68,630 218 0.3176% 57 0.0831% 226,002 845 0.3739% 335 0.1482% 146 1 0.6849% 3 2.0548%
Totals Optech Marksense 799,267 799,267 795,071 3,051 0.3817% 1,145 0.1433% 151,577 1,173 0.7739% 878 0.5792% 140,799 422 0.2997% 55 0.0391% 505,693 1,447 0.2861% 210 0.0415% 1,198 9 0.7513% 2 0.1669%
Totals ES&S Touchscreen 2,590,145 2,590,145 2,578,290 11,076 0.4276% 779 0.0301% 419,252 1,592 0.3797% 774 0.1846% 645,297 2,207 0.3420% 0 0.0000% 1,522,253 7,256 0.4767% 0 0.0000% 3,343 21 0.6282% 5 0.1496%
Totals Sequoia Touchscreen 1,531,477 1,531,645 1,524,615 6,050 0.3970% 980 0.0620% 256,345 948 1.1325% 933 0.3640% 223,703 780 0.2611% 0 0.0000% 1,049,812 4,258 0.2436% 0 0.0000% 1,785 64 2.0168% 47 0.9524%

Marksense Ballots - Precinct, Absentee, Early, and Provisional

Oval Targeted Marksense 3,142,828 8,773 0.2791% 1,991 0.0634% 928,813 3,320 0.3574% 1,308 0.1408% 418,678 862 0.2059% 113 0.0270% 1,788,303 4,555 0.2547% 550 0.0308% 7,034 36 0.5118% 20 0.2843%
Arrow Targeted Marksense 1,057,397 4,063 0.3842% 2,125 0.2010% 407,922 2,121 0.5200% 1,811 0.4440% 140,799 422 0.2997% 55 0.0391% 505,693 1,447 0.2861% 210 0.0415% 2,983 73 2.4472% 49 1.6426%
All Marksense Ballots - Precinct, Absentee, Early, and Provisional 4,200,225 12,836 0.3056% 4,116 0.0980% 1,336,735 5,441 0.4070% 3,119 0.2333% 559,477 1,284 0.2295% 168 0.0300% 2,293,996 6,002 0.2616% 760 0.0331% 10,017 109 1.0882% 69 0.6888%

All Touchscreen Ballots - Precinct and Early 3,441,065 14,501 0.4214% 0 0.0000% 869,000 2,987 0.3437% 0 0.0000% 2,572,065 11,514 0.4477% 0 0.0000%

Total All Systems 7,641,290 27,337 0.3578% 4,116 0.0539% 1,336,735 5,441 0.4070% 3,119 0.2333% 1,428,477 4,271 0.2990% 168 0.0118% 4,866,061 17,516 0.3600% 760 0.0156% 10,017 109 1.0882% 69 0.6888%

No Valid Vote on Touchscreens (Unders) 14,501 0.4214%

No Valid Vote on Marksense (Overs and Unders) 16,952 0.4036%
No Valid Vote Overall (Overs and Unders) 31,453 0.4116%
 Attachment C

Data Validation Notes

Analysis and Report of Overvotes and Undervotes
For the 2004 General Election

Alachua County – There was a reported turnout of 111,566 on the official Canvass of
Votes Cast. The county advised us by letter that there were actually 111,598 ballots cast
which is the number used in the “True Turnout Ballots Counted” column on the
spreadsheet.

Brevard County – The total number of votes cast on the official Canvass of Votes Cast
and the total number of votes cast reported on the Report on Overvotes and Undervotes
completed by the county balanced to each other. Candidate votes of 265,462, plus
undervotes of 584, plus overvotes of 118 equals a total number of ballots cast of 266,160.
The EAC report did not match these numbers. The county reviewed their records and
determined that 175 overseas absentees had not been included in the EAC report.
Another 20 ballots had been left off the EAC report but it could not be determined
whether those were absentee, precinct, early or provisional votes. Accordingly, the EAC
report number of 49,895 for absentee ballots was increased by 175 ballots to 50,070. The
remaining 20 ballots were added to the EAC number for precinct ballots of 185,305 for a
corrected number of precinct ballots of 185,325.

Hillsborough County – There was a reported turnout of 464,849 on the official Canvass
of Votes Cast. There were 465,017 Reported Total Ballots Counted on the EAC report.
On the Report on Overvotes and Undervotes completed by the county, there was a
reported Turnout of 465,017. Candidate votes were reported as 463,222 and overvotes
and undervotes totaled 1,765. Accordingly, 465,017 is the number used in the “True
Turnout Ballots Counted” column.

Levy County – There was a reported turnout of 16,742 and 16,652 candidate votes
reported on the official Canvass of Votes Cast. The Report on Overvotes and Undervotes
completed by the county had a total of 93 total overvotes and undervotes. Overvotes plus
undervotes plus candidate votes totaled 16,745. The County confirmed the error by e-
mail. Accordingly, 16,745 is the number used in the “True Turnout Ballots Counted”
column.

Monroe County – There was a reported turnout of 39,629 on the official Canvass of
Votes Cast. This figure was from an unofficial tabulation. The correct number was
39,647. The County both explained this by telephone and provided a page from one of
the constitutional amendments which showed the correct turnout. Accordingly, 39,647 is
the number used in the “True Turnout Ballots Counted” column. The county then
provided recalculations of the numbers reported to the EAC as follows.

1
 Attachment C

Monroe County Reported to EAC County Corrected

and Official Canvass Corrections Totals
Precinct Ballots 19,706 -3 19,703
Absentee Ballots 10,949 69 11,018
Early Voting
Ballots 8,918 -1 8,917
Provisional Ballots 9 0 9
Total Ballots 39,582 65 39,647

Osceola County – There was a reported turnout of 82,204 and an equal number of
candidate votes on the official Canvass of Votes Cast. The county reported a total of 227
undervotes and overvotes. Accordingly, actual turnout was 82,204 candidate votes, plus
227 overvotes and undervotes to equal 82,431 votes cast. That is the number used in the
“True Turnout Ballots Counted” column. This same error was made in the EAC Report.
Overvotes and undervotes were not included in each type of ballot. To correct it, the
overvotes and undervotes were added to the totals reflected on the EAC report. Osceola
County will be asked to provide the Division with a corrected Report on Overvotes and
Undervotes.

Osceola County Reported to EAC Overvotes and Corrected

and Official Canvass Undervotes Totals
Precinct Ballots 52525 135 52660
Absentee Ballots 10401 50 10451
Early Voting Ballots 19120 41 19161
Provisional Ballots 158 1 159
Total Ballots 82204 227 82431

St. Lucie County – There was a reported turnout of 100,374 on the official Canvass of
Votes Cast. The county advised us by letter that there were actually 100,410 ballots cast
which is the number used in the “True Turnout Ballots Counted” column.

Santa Rosa – There were numerous discrepancies reported by this county. All numbers
are based upon a recalculation of statistics by the Supervisor of Elections. The total
number of candidate votes was not restated.

Volusia County – There was a reported turnout of 229,193 and 228,939 candidate votes
reported on the official Canvass of Votes Cast. There were 228,939 reported Total
Ballots Counted on the EAC report. The county reported a total of 741 of overvotes and
undervotes. Candidate votes plus overvotes and undervotes must equal the total ballots
counted (228,939 + 741 = 229,680). Accordingly, this number is used in the “True
Turnout Ballots Counted” column. Volusia will be asked to provide the Division with a
corrected Report on Overvotes and Undervotes.

After the above errors were corrected, there were several counties where the numbers
reported on their EAC reports did not match the numbers on their Official Canvass and
Reports of Overvotes and Undervotes by a matter of a few votes. To bring all of the

2
 Attachment C

denominators for ratios into balance and minimize distortion of the statistics, the
discrepancies were added and subtracted to the EAC report’s “Precinct Ballots Cast”
totals as follows:

County Reported on Reported on EAC Precinct Adjusted Precinct

Official Canvass EAC Report Difference Ballot Count Ballot Count
Bradford 10,917 10,913 4 6,871 6,875
Citrus 69,701 69,696 5 34,177 34,182
Columbia 25,082 25,075 7 14,490 14,497
Desoto 9,549 9,534 15 5,393 5,408
Dixie 6,474 6,472 2 4,462 4,464
Gulf 7,320 7,299 21 4,099 4,120
Hardee 7,284 7,289 -5 3,400 3,395
Marion 140,354 140,321 33 99,628 99,661
Putnam 31,072 31,067 5 22,775 22,780
Taylor 8,814 8,813 1 4,671 4,672