No. 04-16280-II UNITED STATES COURT OF APPEALS FOR THE ELEVENTH CIRCUIT ____________________ ) ) ) ) ) ) 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. ) _________________________________________) ROBERT WEXLER, Congressman ADDIE GREENE, Commissioner, BURT AARONSON, Commissioner, TONY FRANSETTA, PLAINTIFFS-APPELLANTS,

Appeal from the United States District Court for the Southern District of Florida Docket No. 04-80216 CIV-COHN Hon. James I. Cohn

BRIEF OF PLAINTIFFS-APPELLANTS Robert S. Peck, Esq. CENTER FOR CONSTITUTIONAL LITIGATION, PC 1050 31st Street, N.W. Washington, DC 20007 (202) 944-2874 (o) (202) 965-0920 (fax) Email: robert.peck@cclfirm.com Jeffrey M. Liggio, Esq. LIGGIO, BENRUBI & WILLIAMS, PA 1615 Forum Place Suite 3B, The Barristers Building West Palm Beach, FL 33041 (561) 616-3333 (o) (561) 616-3266 (fax) Email: jliggio@liggiolaw.com

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. 2. Commissioner Burt Aaronson Office of the Attorney General, Attorneys for Defendant-Appellee Hood 3. Center for Constitutional Litigation, P.C., Attorneys for PlaintiffsAppellants 4. 5. 6. Supervisor Kay Clem Honorable United States District Court Judge James I. Cohn Tony Fransetta, an individual
C- 1 of 2

Wexler et al. v. Lepore et al. 7. 8. 9. 10. 11. Commissioner Addie Greene Secretary of State Glenda E. Hood

Case No. 04-16280

Paul C. Huck, Jr., Attorney for Defendant-Appellee Hood Chris Kise, Attorney for Defendant-Appellee Hood Ronald Labasky, Attorney for Defendants-Appellees LePore and Clem

12. 13. 14. 15. 16. 17. 18. 19.

Landers & Parsons, P.A., Attorneys for LePore and Clem Supervisor Theresa LePore Liggio, Benrubi & Williams, P.A., Attorneys for Plaintiffs-Appellants Jeffrey M. Liggio, Attorney for Plaintiff-Appellants Robert S. Peck, Attorney for Plaintiffs-Appellants James A. Peters, Attorney for Defendant-Appellee Hood George Waas, Attorney for Defendant-Appellee Hood 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. B. C. Course of Proceedings...........................................................................1 Statement of Facts. ................................................................................3 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 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. C. D. Strict Scrutiny Applies to Evaluate Whether Equal Protection is Denied ..................................................................12 Counting Ballot Image Summaries Would Not Remedy the Equal Protection Violation ...................................22 The Emergency Rule Approach Also Contained an Unconstitutional Time Limitation ............................................27

II.

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. R4129-38. On February 12, 2004, the Division of Elections issued Opinion DE 0402, 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 machinegenerated 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

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.

2

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. R6-131539 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,

The Division of Elections reported more problems in the 2002 election with DRE equipment producing overvotes and undervotes than optical scan equipment. R4129-84 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).

3

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

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).

5

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; R5-130350 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 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. R4129-216. 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).
8

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 Q. No, I don’t believe it makes the distinction between a mistake of the voter and the voter’s intent, no. 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? 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. That’s because it doesn’t tell us; does it? Well, the statutes don’t tell you to look to that report for that. But the statutes do tell you to look to the report to try to determine the intent of the voter; do they not? That’s correct.

A.

Q. A. Q. A.

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? Yes.
23

A.

Q. Q.

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

A.

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? I believe so. 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? Yes. 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? The audit log comes from the machine. It has information on how each voter voted. It’s a ballot image, if you will. And the audit log only shows whether a vote was recorded or not? Correct. Not whether there was a reason for a vote being – failing to be recorded, other than the voter didn’t want to vote? Right.

A. Q.

A. Q.

A. Q. A. Q. A.

24

Q. A. Q.

So with this audit trail, all we can do is confirm; yes? Yes. 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? Okay. Do you agree with that? That’s what the audit log says, yes. 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? I don’t know what time they did it, but yes.

A. Q. A. Q.

A. 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 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.
9

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 socalled 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; R5130-331. 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 flsd@flsd.uscourts.gov 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 james_peters@oag.state.fl.us George Waas, Esquire Office of the Attorney General PL-01 The Capitol Tallahassee, FL 32399-1050 george_waas@oag.state.fl.us Christopher M. Kise, Solicitor General Office of the Attorney General PL-01 The Capitol Tallahassee, FL 32399 chris_kise@oag.state.fl.us Attorneys for Defendant-Appellee Hood

ADDENDUM Page # American Civil Liberties Union et al. v. Department of State, No. 042341RX (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 PEOPLE FOR THE AMERICAN WAY FOUNDATION, Intervenor, vs. DEPARTMENT OF STATE, Respondent.

) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) )

Case No. 04-2341RX

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 Reginald J. Mitchell, Esquire 1550 Melvin Street Tallahassee, Florida 32301 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

For Intervenor:

For Respondent:

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. in this Final Order. At the final hearing, Petitioners presented the following witnesses: Howard L. Simon, Benjamin T. Wilcox, Reverend S. L. Petitioners' Exhibits A through C Intervenor presented Sharon Lettman Those facts have been incorporated

Phillips, and John Seibel. were admitted in evidence. as its witness. evidence.

Intervenor did not submit any exhibits in

Respondent called Sarah Jane Bradshaw and Paul Craft Respondent's Exhibit F was admitted in

as its witnesses. evidence.

The parties agreed to file their proposed final orders within ten days of the filing of the transcript. 2004, the one-volume Transcript was filed. On August 5,

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 The great

with approximately 22,000 members in Florida.

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 Following the 2000

litigation concerning voting issues.

presidential election, ACLU-FL filed an amicus curiae brief in Bush v. Gore. In Siegel v. LePore, ACLU-FL intervened, More recently, ACLU-FL filed

challenging the butterfly ballot.

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 The organization strives to

system and the electoral process.

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. 90,000 members in Florida. 5. FSCLC is an organization founded by the late Dr. Martin CCF has approximately

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. is preserved. The paper ballot

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, An overvote

that the DRE systems do not allow for 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. Stat. (2004). § 97.021(33), Fla.

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, MiamiDade, 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. Stat. (2004). §§ 101.015 and 101.5605, Fla.

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." Stat. (2004). § 101.5606(3), Fla.

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 If the margin of victory for a candidate or measure

situations.

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. (2004). § 102.141(6)(a), Fla. Stat.

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. (2004). § 102.166(1), Fla. Stat.

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.'" Stat. (2004). § 102.166(5)(b), Fla.

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. 16. § 102.166(6)(d), Fla. Stat. (2004).

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. calendar. 18. On April 13, 2004, Respondent amended its rule The revised proposed legislation died on special

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. 23. § 120.56(3), Fla. Stat. (2004).

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. voluntarily dismissed its petition. 28. Petitioners and Intervenor, as the parties challenging Petitioner SVREP has

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 onequarter 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. the statute by promulgating a rule. 36. It is clear that Respondent exceeded its grant of Respondent cannot amend

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 Blank and Spoiled Absentee Undervotes Absentee Overvotes Precinct Undervotes Precinct Overvotes Early Vote Undervotes Early Vote Overvotes Provisional Undervotes Provisional Overvotes Totals Turnout Ratio of No Votes 179,855

2002 Gubernatorial

2004 Presidential

4,405 1,121 33,737 689

5,441 3,119 17,516 760 4,271 168 109 69 31,453 7,641,290 0.4116%

179,855 6,137,938 2.9302%

39,952 5,144,477 0.7766%

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:
Absentee Early Voting Precinct Provisional Overall Undervotes 0.3175% 0.1840% 0.2375% 0.3949% 0.2447% Overvotes 0.0980% 0.0160% 0.0138% 0.3385% 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:
Absentee Early Voting Precinct Provisional Overall Undervotes 0.4753% 0.3176% 0.3739% 0.6849% 0.3829% Overvotes 0.1475% 0.0831% 0.1482% 2.0458% 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:
Absentee Early Voting Precinct Provisional Overall Undervotes 0.7739% 0.2997% 0.2861% 0.7513% 0.3817% Overvotes 0.5792% 0.0391% 0.0415% 0.1669% 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:
Absentee Early Voting Precinct Provisional Overall Undervotes 0.3797% 0.3420% 0.4767% 0.6282% 0.4276% Overvotes 0.1846% 0.0% 0.0% 0.1496% 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:
Absentee Early Voting Precinct Provisional Overall Undervotes 0.3698% 0.3487% 0.4056% 3.5854% 0.3950% Overvotes 0.3640% 0.0% 0.0% 2.6331% 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 UnderVotes OverVotes UnderVotes OverVotes UnderVotes OverVotes UnderVotes OverVotes InValid Write-Ins TOTAL

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
Analysis and Report of Overvotes and Undervotes for the 2004 General Election

Attachment B

County Name Alachua Brevard Calhoun Citrus Columbia Desoto Dixie Duval Flagler Gilchrist Glades Hardee Hernando Jefferson Leon Levy Madison Manatee Monroe Okaloosa Okeechobee Osceola Polk Putnam Seminole St. Lucie Taylor Volusia Wakulla Walton

Voting System Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold Diebold

Precinct Method Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense

Canvass True Turnout Candidate Turnout Ballots Counted Votes Recorded 111,566 111,598 111,328 266,160 266,160 265,462 6,008 6,008 5,963 69,701 69,701 69,467 25,082 25,082 24,991 9,549 9,549 9,510 6,474 6,474 6,442 382,006 382,006 381,061 38,557 38,557 38,480 7,050 7,050 7,015 4,204 4,204 4,188 7,284 7,284 7,249 80,764 80,764 80,532 7,502 7,502 7,478 136,638 136,638 136,379 16,742 16,745 16,652 8,345 8,345 8,310 143,983 143,983 143,621 39,629 39,647 39,535 89,957 89,957 89,756 12,249 12,249 12,190 82,204 82,431 82,204 211,399 211,399 210,830 31,072 31,072 30,973 186,617 186,617 186,195 100,374 100,410 100,063 8,614 8,614 8,581 229,193 229,680 228,939 11,820 11,820 11,763 24,065 24,065 23,976 2,354,808 2,355,611 75,234 10,917 5,974 21,094 7,320 5,131 9,818 41,876 19,908 3,352 3,051 140,354 15,879 4,714 364,622 81,731 143,849 8,351 390,706 67,498 86,631 10,453 10,048 799,267 709,724 80,196 129,231 124,488 241,663 72,736 778,953 32,827 191,909 196,413 32,005 2,590,145 465,017 61,707 547,340 457,581 1,531,645 2,355,611 364,622 799,267 2,590,145 1,531,645 2,349,133 75,024 10,855 5,931 20,984 7,277 5,079 9,775 41,496 19,807 3,325 3,021 139,677 15,802 4,675 362,728 81,495 143,278 8,300 388,044 67,343 86,290 10,366 9,955 795,071 706,872 79,786 128,683 123,950 240,667 72,453 774,726 32,743 190,916 195,652 31,842 2,578,290 463,222 61,414 544,622 455,357 1,524,615 2,349,133 362,728 795,071 2,578,290 1,524,615

Total Undervotes 241 584 45 214 88 39 32 939 68 35 16 28 198 24 259 84 34 296 101 189 55 204 397 93 414 246 33 690 57 61

Total Total UV % Overvotes 0.2160% 29 0.2194% 114 0.7490% 0 0.3070% 20 0.3508% 3 0.4084% 0 0.4943% 0 0.2458% 6 0.1764% 9 0.4965% 0 0.3806% 0 0.3844% 7 0.2452% 34 0.3199% 0 0.1896% 0 0.5016% 9 0.4074% 1 0.2056% 66 0.2547% 11 0.2101% 12 0.4490% 4 0.2475% 23 0.1878% 172 0.2993% 6 0.2218% 8 0.2450% 101 0.3831% 0 0.3004% 51 0.4822% 0 0.2535% 28 714 22 25 14 10 7 23 10 108 28 4 6 215 18 8 498 92 169 10 671 37 114 22 30 1,145 68 36 36 0 27 10 430 13 136 17 6 779 214 30 359 377 980 714 498 1,145 779 980

Total OV % 0.0260% 0.0428% 0.0000% 0.0287% 0.0120% 0.0000% 0.0000% 0.0016% 0.0233% 0.0000% 0.0000% 0.0961% 0.0421% 0.0000% 0.0000% 0.0537% 0.0120% 0.0458% 0.0277% 0.0133% 0.0327% 0.0279% 0.0814% 0.0193% 0.0043% 0.1006% 0.0000% 0.0222% 0.0000% 0.1164% 0.0303% 0.0292% 0.2290% 0.2343% 0.0474% 0.0956% 0.4483% 0.1019% 0.2579% 0.1406% 0.1193% 0.1967% 0.1532% 0.1134% 0.1697% 0.1366% 0.1126% 0.1175% 0.1197% 0.1717% 0.0548% 0.1316% 0.2105% 0.2986% 0.1433% 0.0096% 0.0449% 0.0279% 0.0000% 0.0112% 0.0137% 0.0552% 0.0396% 0.0709% 0.0087% 0.0187% 0.0301% 0.0460% 0.0486% 0.0656% 0.0824% 0.0640% 0.0303% 0.1366% 0.1433% 0.0301% 0.0620%

Absentee Ballots 18,691 50,070 842 15,400 3,238 1,830 919 63,352 5,972 919 746 1,052 15,865 1,037 22,900 2,917 993 26,721 11,018 19,928 2,173 10,451 57,148 4,513 30,428 18,616 1,518 44,218 1,671 4,571 439,717 18,639 1,613 1,277 3,602 1,475 1,039 1,520 7,136 2,567 543 363 25,829 2,732 1,509 69,844 17,664 35,334 946 67,414 11,472 15,299 2,122 1,326 151,577 98,971 19,744 27,723 23,971 47,479 16,597 98,466 6,991 34,794 40,412 4,104 419,252 64,031 12,152 91,802 88,360 256,345 439,717 69,844 151,577 419,252 256,345

Absentee Undervotes 56 121 8 106 17 11 12 201 11 4 4 11 34 6 93 21 9 70 31 31 10 37 137 22 79 53 9 156 20 16 1,396 61 7 13 40 10 9 6 44 12 4 4 105 17 0 332 51 86 3 941 15 42 13 22 1,173 350 73 114 142 159 75 338 15 161 140 25 1,592 218 47 434 249 948 1,396 332 1,173 1,592 948

Absentee Absentee Absentee Early Vote Early Vote Early Vote Early Vote Early Vote UV % Overvotes OV % Ballots Undervotes UV % Overvotes OV % 0.2996% 22 0.1177% 16,389 26 0.1586% 1 0.0061% 0.2417% 58 0.1158% 30,564 60 0.1963% 8 0.0262% 0.9501% 0 0.0000% 1,447 9 0.6220% 0 0.0000% 0.6883% 15 0.0974% 20,117 0 0.0000% 0 0.0000% 0.5250% 0 0.0000% 7,339 30 0.4088% 0 0.0000% 0.6011% 0 0.0000% 2,310 11 0.4762% 0 0.0000% 1.3058% 0 0.0000% 1,091 3 0.2750% 0 0.0000% 0.3173% 0 0.0000% 58,693 113 0.1925% 0 0.0000% 0.1842% 6 0.1005% 8,873 17 0.1916% 1 0.0113% 0.4353% 0 0.0000% 1,616 11 0.6807% 0 0.0000% 0.5362% 0 0.0000% 345 3 0.8696% 0 0.0000% 1.0456% 4 0.3802% 2,831 10 0.3532% 0 0.0000% 0.2143% 23 0.1450% 13,942 31 0.2223% 3 0.0215% 0.5786% 0 0.0000% 2,486 10 0.4023% 0 0.0000% 0.4061% 0 0.0000% 17,974 17 0.0946% 0 0.0000% 0.7199% 0 0.0000% 2,078 10 0.4812% 2 0.0962% 0.9063% 1 0.1007% 2,446 8 0.3271% 0 0.0000% 0.2620% 29 0.1085% 10,274 7 0.0681% 4 0.0389% 0.2814% 6 0.0545% 8,917 18 0.2019% 3 0.0336% 0.1556% 9 0.0452% 10,383 17 0.1637% 0 0.0000% 0.4602% 3 0.1381% 3,507 17 0.4847% 0 0.0000% 0.3540% 13 0.1244% 19,161 40 0.2088% 1 0.0052% 0.2397% 123 0.2152% 6,023 6 0.0996% 0 0.0000% 0.4875% 0 0.0000% 3,775 7 0.1854% 0 0.0000% 0.2596% 0 0.0000% 20,671 18 0.0871% 5 0.0242% 0.2847% 87 0.4673% 22,882 32 0.1398% 2 0.0087% 0.5929% 0 0.0000% 2,421 8 0.3304% 0 0.0000% 0.3528% 28 0.0633% 40,255 86 0.2136% 3 0.0075% 1.1969% 0 0.0000% 4,776 19 0.3978% 0 0.0000% 0.3500% 4 0.0875% 6,462 0 0.0000% 23 0.3559% 0.3175% 0.3273% 0.4340% 1.0180% 1.1105% 0.6780% 0.8662% 0.3947% 0.6166% 0.4675% 0.7366% 1.1019% 0.4065% 0.6223% 0.0000% 0.4753% 0.2887% 0.2434% 0.3171% 1.3959% 0.1308% 0.2745% 0.6126% 1.6591% 0.7739% 0.3536% 0.3697% 0.4112% 0.5924% 0.3349% 0.4519% 0.3433% 0.2146% 0.4627% 0.3464% 0.6092% 0.3797% 0.3405% 0.3868% 0.4728% 0.2818% 0.3698% 0.3175% 0.4753% 0.7739% 0.3797% 1.1325% 431 7 3 5 0 1 0 8 23 9 1 0 42 4 0 103 65 151 7 499 25 101 14 16 878 63 36 36 0 27 10 430 13 136 17 6 774 200 27 329 377 933 431 103 878 774 933 0.0980% 0.0376% 0.1860% 0.3915% 0.0000% 0.0678% 0.0000% 0.5263% 0.3223% 0.3506% 0.1842% 0.0000% 0.1626% 0.1464% 0.0000% 0.1475% 0.3680% 0.4274% 0.7400% 0.7402% 0.2179% 0.6602% 0.6598% 1.2066% 0.5792% 0.0637% 0.1823% 0.1299% 0.0000% 0.0569% 0.0603% 0.4367% 0.1860% 0.3909% 0.0421% 0.1462% 0.1846% 0.3123% 0.2222% 0.3584% 0.4267% 0.3640% 0.0980% 0.1475% 0.5792% 0.1846% 0.3640% 350,048 18,572 2,429 1,586 5,467 1,725 937 2,481 11,385 4,677 873 864 14,799 2,466 369 68,630 6,770 12,807 1,987 79,171 15,810 18,758 2,653 2,843 140,799 176,743 13,659 44,155 32,331 32,182 12,243 244,022 9,989 29,584 33,872 16,517 645,297 86,642 17,328 49,831 69,902 223,703 350,048 68,630 140,799 645,297 223,703 644 35 9 2 20 6 3 8 53 15 9 5 38 15 0 218 5 39 7 249 24 61 21 16 422 498 57 164 116 109 53 901 19 114 100 76 2,207 277 67 196 240 780 644 218 422 2,207 780 0.1840% 0.1885% 0.3705% 0.1261% 0.3658% 0.3478% 0.3202% 0.3225% 0.4655% 0.3207% 1.0309% 0.5787% 0.2568% 0.6083% 0.0000% 0.3176% 0.0739% 0.3045% 0.3523% 0.3145% 0.1518% 0.3252% 0.7916% 0.5628% 0.2997% 0.2818% 0.4173% 0.3714% 0.3588% 0.3387% 0.4329% 0.3692% 0.1902% 0.3853% 0.2952% 0.4601% 0.3420% 0.3197% 0.3867% 0.3933% 0.3433% 0.3487% 0.1840% 0.3176% 0.2997% 0.3420% 0.2611% 56 2 0 1 1 0 1 0 22 5 0 0 22 3 0 57 5 4 0 31 2 4 2 7 55 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 56 57 55 0 0 0.0160% 0.0108% 0.0000% 0.0631% 0.0183% 0.0000% 0.1067% 0.0000% 0.1932% 0.1069% 0.0000% 0.0000% 0.1487% 0.1217% 0.0000% 0.0831% 0.0739% 0.0312% 0.0000% 0.0392% 0.0127% 0.0213% 0.0754% 0.2462% 0.0391% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0160% 0.0831% 0.0391% 0.0000% 0.0000%

Precinct Precinct Ballots Undervotes 76,381 159 185,325 403 3,718 28 34,182 108 14,497 41 5,408 17 4,464 17 258,746 625 23,694 40 4,514 20 3,101 9 3,395 7 50,942 133 3,979 8 95,397 149 11,748 53 4,905 17 106,917 219 19,703 52 59,566 141 6,565 28 52,660 127 148,150 254 22,780 64 135,161 313 58,795 159 4,672 16 144,568 441 5,372 18 12,996 44 1,562,301 37,981 6,875 3,111 12,024 4,120 3,147 5,817 23,351 12,655 1,935 1,823 99,661 10,666 2,836 226,002 57,209 95,647 5,418 243,536 40,177 52,151 5,676 5,879 505,693 431,488 46,741 57,323 68,121 161,953 43,892 435,887 15,839 127,526 122,101 11,382 1,522,253 313,930 32,015 404,666 299,201 1,049,812 1,562,301 226,002 505,693 1,522,253 1,049,812 3,710 92 21 14 40 20 16 19 175 46 10 15 319 27 31 845 88 277 31 801 78 117 30 25 1,447 1,916 244 234 280 700 145 2,558 37 582 504 56 7,256 1,052 149 1,701 1,356 4,258 3,710 845 1,447 7,256 4,258

Precinct UV % 0.2082% 0.2175% 0.7531% 0.3160% 0.2828% 0.3143% 0.3808% 0.2415% 0.1688% 0.4431% 0.2902% 0.2062% 0.2611% 0.2011% 0.1562% 0.4511% 0.3466% 0.2048% 0.2639% 0.2367% 0.4265% 0.2412% 0.1714% 0.2809% 0.2316% 0.2704% 0.3425% 0.3050% 0.3351% 0.3386% 0.2375% 0.2422% 0.3055% 0.4500% 0.3327% 0.4854% 0.5084% 0.3266% 0.7494% 0.3635% 0.5168% 0.8228% 0.3201% 0.2531% 1.0931% 0.3739% 0.1538% 0.2896% 0.5722% 0.3289% 0.1941% 0.2243% 0.5285% 0.4252% 0.2861% 0.4440% 0.5220% 0.4082% 0.4110% 0.4322% 0.3304% 0.5868% 0.2336% 0.4564% 0.4128% 0.4920% 0.4767% 0.3351% 0.4654% 0.4203% 0.4532% 0.4056% 0.2375% 0.3739% 0.2861% 0.4767% 0.2436%

Precinct Overvotes 6 48 0 5 3 0 0 0 2 0 0 3 8 0 0 7 0 32 2 3 1 8 48 6 3 11 0 18 0 1 215 13 22 8 9 6 22 2 63 14 3 6 149 10 8 335 22 14 3 141 10 7 6 7 210 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 215 335 210 0 0

Precinct Provisional Provisional Provisional Provisional Provisional OV % Ballots Undervotes UV % Overvotes OV % 0.0079% 137 0 0.0000% 0 0.0000% 0.0259% 201 0 0.0000% 0 0.0000% 0.0000% 1 0 0.0000% 0 0.0000% 0.0146% 2 0 0.0000% 0 0.0000% 0.0207% 8 0 0.0000% 0 0.0000% 0.0000% 1 0 0.0000% 0 0.0000% 0.0000% 0 0 0.0000% 0 0.0000% 0.0000% 1,215 0 0.0000% 6 0.4938% 0.0084% 18 0 0.0000% 0 0.0000% 0.0000% 1 0 0.0000% 0 0.0000% 0.0000% 12 0 0.0000% 0 0.0000% 0.0884% 6 0 0.0000% 0 0.0000% 0.0157% 15 0 0.0000% 0 0.0000% 0.0000% 0 0 0.0000% 0 0.0000% 0.0000% 367 0 0.0000% 0 0.0000% 0.0596% 2 0 0.0000% 0 0.0000% 0.0000% 1 0 0.0000% 0 0.0000% 0.0299% 71 0 0.0000% 1 1.4085% 0.0102% 9 0 0.0000% 0 0.0000% 0.0050% 80 0 0.0000% 0 0.0000% 0.0152% 4 0 0.0000% 0 0.0000% 0.0152% 159 0 0.0000% 1 0.6289% 0.0324% 78 0 0.0000% 1 1.2821% 0.0263% 4 0 0.0000% 0 0.0000% 0.0022% 357 4 1.1204% 0 0.0000% 0.0187% 117 2 1.7094% 1 0.8547% 0.0000% 3 0 0.0000% 0 0.0000% 0.0125% 639 7 1.0955% 2 0.3130% 0.0000% 1 0 0.0000% 0 0.0000% 0.0077% 36 1 2.7778% 0 0.0000% 0.0138% 0.0342% 0.3200% 0.2572% 0.0749% 0.1456% 0.6991% 0.0344% 0.2698% 0.1106% 0.1550% 0.3291% 0.1495% 0.0938% 0.2821% 0.1482% 0.0385% 0.0146% 0.0554% 0.0579% 0.0249% 0.0134% 0.1057% 0.1191% 0.0415% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0138% 0.1482% 0.0415% 0.0000% 0.0000% 3,545 42 0 0 1 0 8 0 4 9 1 1 65 15 0 146 88 61 0 585 39 423 2 0 1,198 2,522 52 30 65 49 4 578 8 5 28 2 3,343 414 212 1,041 118 1,785 3,545 146 1,198 3,343 1,785 14 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 1 7 1 0 9 20 0 0 0 1 0 0 0 0 0 0 21 34 0 28 2 64 14 1 9 21 64 0.3949% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 12.5000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.6849% 0.0000% 0.0000% 0.0000% 0.0000% 2.5641% 1.6548% 50.0000% 0.0000% 0.7513% 0.7930% 0.0000% 0.0000% 0.0000% 2.0408% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.6282% 8.2126% 0.0000% 2.6897% 1.6949% 3.5854% 0.3949% 0.6849% 0.7513% 0.6282% 2.0168% 12 0 0 0 0 0 0 0 0 0 0 0 2 1 0 3 0 0 0 0 0 2 0 0 2 5 0 0 0 0 0 0 0 0 0 0 5 14 3 30 0 47 12 3 2 5 47 0.3385% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 3.0769% 6.6667% 0.0000% 2.0548% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.4728% 0.0000% 0.0000% 0.1669% 0.1983% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.0000% 0.1496% 3.3816% 1.4151% 2.8818% 0.0000% 2.6331% 0.3385% 2.0548% 0.1669% 0.1496% 0.9524%

True Undervote % on Mismarked Undervote % Absentees (Estimate) on Absentees (Estimate 95 5 99 1 100 0 100 0 98 2 100 0 100 0 100 0 99 1 100 0 100 0 100 0 100 0 100 0 98 2 100 0 100 0 100 0 100 0 100 0 0 0 100 0 99 1 90 10 95 5 100 0 100 0 100 0 100 0 100 0

Totals Diebold Marksense Bay Bradford Franklin Gadsden Gulf Hamilton Hendry Highlands Jackson Lafayette Liberty Marion Suwannee Union ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 ES&S 100 Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense Marksense

5,764 0.2447% 188 37 29 100 36 29 33 272 73 23 24 462 59 31 0.2499% 0.3389% 0.4854% 0.4741% 0.4918% 0.5652% 0.3361% 0.6495% 0.3667% 0.6862% 0.7866% 0.3292% 0.3716% 0.6576%

75,234 10,917 5,974 21,094 7,320 5,131 9,818 41,876 19,908 3,352 3,051 140,354 15,879 4,714 364,622 81,731 143,849 8,351 390,706 67,498 86,631 10,453 10,048 799,267

99 90 100 99 100 100 100 100 99 100 99 100 99 100

1 10 0 1 0 0 0 0 1 0 1 0 1 0

Totals ES&S Marksense Clay Escambia Holmes Orange Santa Rosa St. Johns Washington Baker ES&S Optech Marksense ES&S Optech Marksense ES&S Optech Marksense ES&S Optech Marksense ES&S Optech Marksense ES&S Optech Marksense ES&S Optech Marksense Sequoia OptechMarksense

1,396 0.3829% 144 402 41 1,991 118 227 65 63 0.1762% 0.2795% 0.4910% 0.5096% 0.1748% 0.2620% 0.6218% 0.6270%

100 98 100 99 100 98 99 99

0 2 0 1 0 2 1 1

Totals Optech Marksense Broward Charlotte Collier Lake Lee Martin Miami-Dade Nassau Pasco Sarasota Sumter ES&S ES&S ES&S ES&S ES&S ES&S ES&S ES&S ES&S ES&S ES&S Touchscreen Touchscreen Touchscreen Touchscreen Touchscreen Touchscreen Touchscreen Touchscreen Touchscreen Touchscreen Touchscreen

3,051 0.3817% 2,784 374 512 538 969 273 3,797 71 857 744 157 0.3923% 0.4664% 0.3962% 0.4322% 0.4010% 0.3753% 0.4874% 0.2163% 0.4466% 0.3788% 0.4905%

709,724 80,196 129,231 124,488 241,663 72,736 778,953 32,827 191,909 196,413 32,005 2,590,145

95 99 100 100 99 100 85 99 100 100 100

5 1 0 0 1 0 15 1 0 0 0

Totals ES&S Touchscreen Hillsborough Indian River Palm Beach Pinellas Sequoia Sequoia Sequoia Sequoia Touchscreen Touchscreen Touchscreen Touchscreen

11,076 0.4276% 1,581 263 2,359 1,847 0.3400% 0.4262% 0.4310% 0.4036%

464,849 61,707 547,340 457,581 1,531,477 2,354,808 364,622 799,267 2,590,145 1,531,477

100 90 99 100

0 10 1 0

Totals Sequoia Touchscreen Totals Diebold Marksense Totals ES&S Marksense Totals Optech Marksense Totals ES&S Touchscreen Totals Sequoia Touchscreen

6,050 0.3950% 5,764 1,396 3,051 11,076 6,050 0.2447% 0.3829% 0.3817% 0.4276% 0.3970%

Marksense Ballots - Precinct, Absentee, Early, and Provisional Oval Targeted Marksense Arrow Targeted Marksense All Marksense Ballots - Precinct, Absentee, Early, and Provisional All Touchscreen Ballots - Precinct and Early Total All Systems No Valid Vote on Touchscreens (Unders) No Valid Vote on Marksense (Overs and Unders) No Valid Vote Overall (Overs and Unders)

3,142,828 1,057,397 4,200,225 3,441,065 7,641,290

8,773 0.2791% 4,063 0.3842% 12,836 0.3056% 14,501 0.4214% 27,337 0.3578% 14,501 0.4214% 16,952 0.4036% 31,453 0.4116%

1,991 2,125 4,116 0 4,116

0.0634% 0.2010% 0.0980% 0.0000% 0.0539%

928,813 407,922 1,336,735

3,320 2,121 5,441

0.3574% 0.5200% 0.4070%

1,308 1,811 3,119

0.1408% 0.4440% 0.2333%

418,678 140,799 559,477 869,000

862 422 1,284 2,987 4,271

0.2059% 0.2997% 0.2295% 0.3437% 0.2990%

113 55 168 0 168

0.0270% 0.0391% 0.0300% 0.0000% 0.0118%

1,788,303 505,693 2,293,996 2,572,065 4,866,061

4,555 1,447 6,002 11,514 17,516

0.2547% 0.2861% 0.2616% 0.4477% 0.3600%

550 210 760 0 760

0.0308% 0.0415% 0.0331% 0.0000% 0.0156%

7,034 2,983 10,017

36 73 109

0.5118% 2.4472% 1.0882%

20 49 69

0.2843% 1.6426% 0.6888%

1,336,735

5,441

0.4070%

3,119

0.2333%

1,428,477

10,017

109

1.0882%

69

0.6888%

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 email. 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 Precinct Ballots Absentee Ballots Early Voting Ballots Provisional Ballots Total Ballots

Reported to EAC and Official Canvass 19,706 10,949 8,918 9 39,582

County Corrections -3 69 -1 0 65

Corrected Totals 19,703 11,018 8,917 9 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 Precinct Ballots Absentee Ballots Early Voting Ballots Provisional Ballots Total Ballots Reported to EAC and Official Canvass 52525 10401 19120 158 82204 Overvotes and Undervotes 135 50 41 1 227 Corrected Totals 52660 10451 19161 159 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 Bradford Citrus Columbia Desoto Dixie Gulf Hardee Marion Putnam Taylor

Reported on Official Canvass 10,917 69,701 25,082 9,549 6,474 7,320 7,284 140,354 31,072 8,814

Reported on EAC Report 10,913 69,696 25,075 9,534 6,472 7,299 7,289 140,321 31,067 8,813

Difference 4 5 7 15 2 21 -5 33 5 1

EAC Precinct Ballot Count 6,871 34,177 14,490 5,393 4,462 4,099 3,400 99,628 22,775 4,671

Adjusted Precinct Ballot Count 6,875 34,182 14,497 5,408 4,464 4,120 3,395 99,661 22,780 4,672

3