Erin Elizabeth Lunceford v. Tamika Craft

Tex. App. Ct., 1st Dist. (Houston)

Erin Elizabeth Lunceford v. Tamika Craft

CourtTexas Court of Appeals, 1st District (Houston)

DecidedFebruary 3, 2026

Docket01-23-00921-CV

StatusPublished

This text of Erin Elizabeth Lunceford v. Tamika Craft (Erin Elizabeth Lunceford v. Tamika Craft) is published on Counsel Stack Legal Research, covering Texas Court of Appeals, 1st District (Houston) primary law. Counsel Stack provides free access to over 12 million legal documents including statutes, case law, regulations, and constitutions.

Bluebook

Erin Elizabeth Lunceford v. Tamika Craft, (Tex. Ct. App. 2026).

Pincite

Opinions

Opinion issued February 3, 2026.

In The

Court of Appeals For The

First District of Texas ———————————— NO. 01-23-00921-CV ——————————— ERIN ELIZABETH LUNCEFORD, Appellant V. TAMIKA CRAFT, Appellee

On Appeal from the 164th District Court Harris County, Texas Trial Court Case No. 2022-79328

OPINION

This is an appeal from a final judgment in an election contest.1 In four issues,

Appellant Erin Elizabeth Lunceford challenges the trial court’s final judgment

1 See TEX. ELEC. CODE § 221.002(f) (providing that court of appeals has appellate jurisdiction of election contests originating in district court). denying her contest to the 2022 general election for judge of the 189th District Court

in Harris County and declaring the results of that election—in which the final

canvass reflected that Tamika “Tami” Craft defeated Lunceford by a margin of 2,743

votes—to be the true outcome.

We affirm the judgment of the trial court.

2022 General Election for Judge of the 189th District Court

On November 8, 2022 (“Election Day”), Harris County conducted a general

election featuring several races for countywide positions on the ballot (“General

Election”).2 Harris County participates in the Countywide Polling Place Program,

which allows a county to establish non-precinct-based voting centers throughout the

county at which voters can cast a ballot.3 For the 2022 General Election, registered

voters were permitted to vote at any one of ninety-nine early-voting polling places

or 782 election-day polling places throughout the county. See TEX. ELEC. CODE

§ 43.007 (setting forth Countywide Polling Place Program).

One of the countywide races featured on the ballot for the 2022 General

Election was the race for District Judge of the 189th Judicial District in Harris

County between Appellant Erin Elizabeth Lunceford and Appellee Tamika “Tami”

2 Early voting was held from October 24 to November 4, 2022. 3 The Secretary of State first approved Harris County for participation in the Countywide Polling Place Program in March 2019.

2 Craft (“Contested Race”). The Harris County official canvass report reflects that a

total of 1,107,390 votes were cast in the General Election. Of that total vote count,

1,064,677 votes were cast in the Contested Race, resulting in 42,713 undervotes or

a total undervote percentage of 3.86 in the Contested Race.4 Craft received 533,710

votes while Lunceford received 530,967 votes, resulting in a 2743-vote margin of

victory for Craft.

Lunceford’s Election Contest

On December 7, 2022, Lunceford filed suit under Section 221.003 of the

Election Code alleging that the outcome of the Contested Race, “as shown by the

final canvass, [was] not the true outcome.” See TEX. ELEC. CODE § 221.003(a).

Lunceford alleged that the final canvass of the Contested Race was inaccurate

because illegal votes were cast and counted, legal votes were cast but not counted,

and election officials prevented eligible voters from voting, failed to count legal

votes, or engaged in fraud, illegal conduct, or a mistake. As a result, Lunceford

argued that the true outcome of the Contested Race was unascertainable for purposes

4 An undervote occurs when a voter casts a ballot but does not cast a vote in all races featured on the ballot. Reese v. Duncan, 80 S.W.3d 650, 653 n.2 (Tex. App.—Dallas 2002, pet. denied) (“An overvote is a vote in which the voter voted for both candidates. An undervote is a vote in which the voter did not vote for either candidate.”). The term “roll off” is sometimes used instead of “undervote.” Green v. Reyes, 836 S.W.2d 203, 211 (Tex. App.—Houston [14th Dist.] 1992, no writ) (quoting testimony stating that “roll off” means voter did not vote in particular election).

3 of Section 221.012(b) of the Election Code. Id. § 221.012(b). She requested that the

trial court declare the election void and order a new election.

At trial, Lunceford challenged the legality of votes cast in the General Election

based on several categories of alleged irregularities. Relevant to this appeal,

Lunceford alleged that (1) votes were cast by voters who were on the registrar’s

suspense list and had not presented a statement of residence, as required, before

voting, and (2) votes were cast by voters who provided an incomplete reasonable

impediment declaration. She further alleged that (1) the Harris County Elections

Administration Office (“EAO”)5 failed to provide sufficient ballot paper to polling

places in advance of the General Election and as a result, “2,535 voters were

estimated to have been turned away” from twenty-four polling places on Election

Day, and (2) “411 voters were turned away” from twenty polling places on Election

Day for other reasons, including machine malfunctions, poll workers’ inability to

reach the EAO on the phone or by other means, a lack of equipment or supplies, and

A. Scope of Inquiry

“An election contest is not an ordinary civil suit but a special statutorily

created special proceeding that provides a ‘remedy for elections tainted by fraud,

illegality or other irregularity.’” Medlin v. King, 705 S.W.3d 267, 281 (Tex. App.—

El Paso 2024, pet. denied) (quoting Blum v. Lanier, 997 S.W.2d 259, 262 (Tex.

1999)); see also Woods v. Legg, 363 S.W.3d 710, 713 (Tex. App.—Houston [1st

Dist.] 2011, no pet.) (same). “The purpose of an election contest is to determine

9 Given our disposition, we need not reach Craft’s crosspoints on appeal.

12 whether the outcome of an election is correct.” Flores v. Cuellar, 269 S.W.3d 657,

660 (Tex. App.—San Antonio 2008, no pet.).

Section 221.003 of the Election Code dictates the scope of a trial court’s

inquiry in an election contest. Under Section 221.003(a), a trial court “shall attempt

to ascertain whether the outcome of the contested election, as shown by the final

canvass, is not the true outcome” because:

(1) illegal votes were counted; or

(2) an election officer or other person officially involved in the administration of the election:

(A) prevented eligible voters from voting;

(B) failed to count legal votes; or

(C) engaged in other fraud or illegal conduct or made a mistake.

TEX. ELEC. CODE § 221.003(a). Section 221.003(b) defines “illegal vote” as “a vote

that is not legally countable.” Id. § 221.003(b). Section 221.003(b) does not define

what it means to make a “mistake.”10

During its inquiry, if the trial court can ascertain the candidate for whom an

illegal vote was cast, it “shall subtract the vote from the official total for the

candidate . . . as applicable.” Id. § 221.011(a). If, however, the trial court “finds that

10 “The language used in the election code regarding mistakes is obviously broad and could be read to include all sorts of election code violations.” Gonzalez v. Villarreal, 251 S.W.3d 763, 778 (Tex. App.—Corpus Christi–Edinburg 2008, pet. dism’d).

13 illegal votes were cast but cannot ascertain how the voters voted, the [trial court]

shall consider those votes in making its judgment.” Id. § 221.011(b). After its

inquiry, if the trial court “can ascertain the true outcome of the election, the [trial

court] shall declare the outcome.” Id. § 221.012(a). The trial court, however, “shall

declare the election void if it cannot ascertain the true outcome of the election.” Id.

§ 221.012(b). “If the number of illegal votes is equal to or greater than the number

of votes necessary to change the outcome of an election, the tribunal may declare

the election void without attempting to determine how individual voters voted.” Id.

§ 221.009(b).11

B. Burden of Proof in Election Contest

Lunceford argued in the trial court that the results of the Contested Race

should be voided because alleged irregularities made it impossible to ascertain the

true outcome of the race. “An election contestant’s burden is a heavy one, and the

declared results [of an election] will be upheld in all cases except when there is clear

and convincing evidence of an erroneous result.” Medlin, 705 S.W.3d at 292; see

also Price v. Lewis, 45 S.W.3d 215, 218 (Tex. App.—Houston [1st Dist.] 2001, no

pet.) (same).

11 Section 221.009(a) states: “A voter who cast an illegal vote may be compelled, after the illegality has been established to the satisfaction of the tribunal hearing the contest, to disclose the name of the candidate for whom the voter voted or how the voter voted on a measure if the issue is relevant to the election contest.” TEX. ELEC. CODE § 221.009(a).

14 To overturn an election, an election contestant must demonstrate by clear and

convincing evidence that a voting irregularity occurred, and the irregularity

materially affected the outcome of the election. Reese v. Duncan, 80 S.W.3d 650,

655 (Tex. App.—Dallas 2002, pet. denied); see also Medlin, 705 S.W.3d at 292

(holding that “an election will not be overturned unless an election contestant proves

voting irregularities materially affected the results”); Woods, 363 S.W.3d at 713

(same). The outcome of an election is “materially affected” when a different and

correct result would have been reached in the absence of irregularities, see Woods,

363 S.W.3d at 713, or the irregularities in the election render it impossible to

determine the true will of a majority of the voters. McCurry v. Lewis, 259 S.W.3d

369, 373 (Tex. App.—Amarillo 2008, no pet.).

Clear and convincing evidence means the measure or degree of proof that will

produce in the mind of the trier of fact a firm belief or conviction as to the truth of

the allegations sought to be established. In re J.F.C., 96 S.W.3d 256, 264 (Tex.

2002). The clear and convincing standard applies when particularly important

individual interests are at risk, and it requires more proof than the preponderance of

the evidence standard applicable in most civil cases. Stary v. Ethridge, 712 S.W.3d

584, 592 (Tex. 2025). It is “an intermediate standard, falling between the

preponderance standard of ordinary civil proceedings and the reasonable doubt

standard of criminal proceedings.” In re G.M., 596 S.W.2d 846, 847 (Tex. 1980).

15 C. Standard of Review

In an election contest, we review the record to determine whether the trial

court abused its discretion. Woods, 363 S.W.3d at 713. A trial court abuses its

discretion when it acts without reference to guiding rules or principles or in an

arbitrary or unreasonable manner. In re Garza, 544 S.W.3d 836, 840 (Tex. 2018).

To determine whether the trial court abused its discretion in ruling on an election

contest, we may consider the sufficiency of the evidence supporting its findings of

fact. Rodriguez v. Rangel, 679 S.W.3d 890, 903 (Tex. App.—San Antonio 2023, pet.

denied).

When an appellant challenges the legal sufficiency of an adverse finding on

an issue on which she had the burden of proof, such as Lunceford does here, the

appellant “must demonstrate on appeal that the evidence establishes, as a matter of

law, all vital facts in support of the issue.” Dow Chem. Co. v. Francis, 46 S.W.3d

237, 241 (Tex. 2001); DeSpain v. DeSpain, 672 S.W.3d 486, 492 (Tex. App.—San

Antonio 2023, no pet.) (reviewing legal sufficiency challenge by party who bore

burden of proof on issue at trial by clear and convincing evidence). “The point of

error should be sustained only if the contrary proposition is conclusively

established.” Dow Chem. Co., 672 S.W.3d at 492. A matter is conclusively

established only if reasonable people could not differ as to the conclusions to be

drawn from the evidence. See City of Keller v. Wilson, 168 S.W.3d 802, 816 (Tex.

16 2005). If the evidence shows there was a conflict, then the elements were not

conclusively established. See Gilbreath v. Horan, 682 S.W.3d 454, 491 (Tex.

App.—Houston [1st Dist.] 2023, pet. denied) (citing Hathaway v. Gen. Mills, Inc.,

711 S.W.2d 227, 229 (Tex. 1986)).

When an appellant challenges the factual sufficiency of an adverse finding on

an issue on which she had the burden of proof, the appellant generally must establish

that the finding was against the great weight and preponderance of the evidence. See

Dow Chem. Co., 46 S.W.3d at 242. That standard, however, is not adequate where

the burden of proof at trial was clear and convincing evidence. Aguilar v. Soliz, No.

03-20-00121-CV, 2021 WL 2750456, at *3 (Tex. App.—Austin July 2, 2021, no

pet.) (mem. op.) (citing Burns v. Burns, 434 S.W.3d 223, 227–28 (Tex. App.—

Houston [1st Dist.] 2014, no pet.)). In election contests, where the standard is

heightened to clear and convincing evidence, we review the entire record to

determine whether the trial court’s failure to form a firm conviction or belief in

support of the issue is “contrary to the overwhelming weight of the evidence and

clearly wrong.” Burns, 434 S.W.3d at 227. We may not reweigh the evidence or

judge the credibility of witnesses, and we defer to the trial court’s credibility

determinations so long as they are not unreasonable. See In re A.B., 437 S.W.3d 498,

503 (Tex. 2014) (stating under clear and convincing standard “[a] court of appeals

must nevertheless still provide due deference to the decisions of the factfinder, who,

17 having full opportunity to observe witness testimony first-hand, is the sole arbiter

when assessing the credibility and demeanor of witnesses”); see also Aguilar, 2021

WL 2750456, at *3 (citing In re J.P.B., 180 S.W. 570, 573 (Tex. 2005) and

articulating factual sufficiency standard where appellant bore clear and convincing

burden of proof at trial). “We examine the record []in light of the high evidentiary

burden that [the party] bore and our required appellate deference to the trial court’s

decision that the evidence did not meet it.” Burns, 434 S.W.3d at 228.

D. Findings of Fact and Conclusions of Law

In a bench trial, we defer to the trial court’s findings of fact, and we review

conclusions of law de novo. Sw. Elec. Power Co. v. Lynch, 595 S.W.3d 678, 683

(Tex. 2020). A party may not challenge conclusions of law for factual sufficiency,

but we may review conclusions of law to determine their correctness based upon the

facts. Citizens Nat’l Bank v. City of Rhome, 201 S.W.3d 254, 256 (Tex. App.—Fort

Worth 2006, no pet.). We will uphold a conclusion of law if the judgment can be

supported on any legal theory supported by the evidence. Tex. Dep’t of Public Safety

v. Stockton, 53 S.W.3d 421, 423 (Tex. App.—San Antonio 2001, pet. denied). We

will not reverse a conclusion of law unless it is erroneous as a matter of law. Id. If a

trial court’s conclusion of law is erroneous but the court rendered the proper

judgment, the erroneous conclusion does not require reversal. BMC Software Belg.,

N.V. v. Marchand, 83 S.W.3d 789, 794 (Tex. 2002).

18 “[F]indings of fact and the conclusions of law will be construed together; and

if the findings of fact are susceptible of different constructions, they will be

construed, if possible, to be in harmony with the judgment and to support it.” Tex.

Outfitters Ltd., LLC v. Nicholson, 534 S.W.3d 65, 74 (Tex. App.—San Antonio

2017), aff’d, 572 S.W.3d 647 (Tex. 2019) (quoting Gulf Liquid Fertilizer Co. v.

Titus, 163 Tex. 260, 270, 354 S.W.2d 378, 385 (1962)).

Discussion

A. Ballot Paper Shortages at Polling Places on Election Day

In her first issue, and as part of her third and fourth issues, Lunceford

challenges the legal and factual sufficiency of the trial court’s finding that only

between 250 to 850 voters were unable to cast a vote on Election Day due to ballot

paper shortages resulting from the EAO’s failure to comply with Section 51.005 of

the Election Code. She argues that because “the evidence conclusively

demonstrated” that 2600 potential “voters tried to vote but were turned away” from

polling places on Election Day due to ballot paper shortages, the trial court erred in

factoring only 850 voters when considering whether it could ascertain whether the

reported outcome of the Contested Race was the true outcome. In other words, she

argues the trial court should have concluded that as a result of the EAO’s violation

of Section 51.005, a total of 2600 voters were unable to cast a ballot on Election Day

due to paper ballot shortages.

19 Craft responds that Lunceford is seeking to lower her burden of proof by

requesting that this Court, in sustain her first issue, “assume,” without any evidence,

“that all 2,600 [persons who left their polling place on Election Day] were unable to

vote elsewhere because proving otherwise” would be “both impossible and

impractical.” In her crosspoint on this issue, Craft also argues that the trial court

abused its discretion in finding that 2600 lawful voters left a polling place without

voting because of ballot paper shortages and estimating that between 250 to 850 of

those potential voters did not vote elsewhere on Election Day. She argues the trial

court abused its discretion because there is no clear and convincing evidence that

any of the 2600 potential voters who were eligible to vote “left because of ballot

paper issues as opposed to any other reasons,” were “unable to vote elsewhere or did

not vote at a different polling location after leaving,” and would have voted in the

Contested Race. There is thus no competent evidence from which the trial court

could have reasonably “estimated” that 250 to 850 of 2600 potential voters did not

vote on Election Day. And she further disputes that the EAO violated Section 51.005

of the Election Code.

1. Section 51.005 of the Election Code

Section 51.005 of the Election Code, titled “Number of Ballots,” states:

(a) The authority responsible for procuring the election supplies for an election shall provide for each election precinct a number of ballots equal to at least the percentage of voters

20 who voted in that precinct in the most recent corresponding election plus 25 percent of that number.

(a-1) The number of ballots provided may not exceed the total number of registered voters in the precinct unless the county participates in the countywide polling place program under Section 43.007.

TEX. ELEC. CODE § 51.005(a), (a-1).

Lunceford challenged the results of the Contested Race, arguing, among other

things, that the EAO did not comply with Section 51.005 of the Election Code,

because, contrary to its provisions, the EAO supplied polling places with the same

amount of ballot paper on Election Day—1200 pages, enough for 600 voters—rather

than allocating ballot paper by accounting for “125% of the voters from the last-like

election.” She argued that due to the EAO’s violation of Section 51.005, several

polling places experienced ballot paper shortages on Election Day and, as direct

result, 2535 persons who were in line to vote on Election Day were “turned away.”12

Craft argued that Section 51.005 is not applicable because it applies to

precinct-based elections, and the 2022 General Election involved a countywide

polling place election. She further argued that the EAO could not have used the

information from the most recent corresponding election to allocate ballot paper on

12 Although the trial court found that 2600 potential voters left a polling place on Election Day due to ballot paper shortages, Lunceford acknowledges that she presented evidence of only 2535 such potential voters. It is unclear how the trial court arrived at the 2600 figure.

21 Election Day because the most recent corresponding election—the 2018 general

election—had been a precinct-based election and not a countywide polling place

election, and further, Harris County underwent redistricting and “re-precincting” in

early 2022.

In its final judgment and findings of fact and conclusions of law, the trial court

held that Section 51.005 applied to the 2022 General Election and that the EAO had

not complied with it because it had allocated an identical number of ballot paper to

766 of the 782 polling places on Election Day rather than “calculat[ing] the 2022

need for ballot paper scientifically by looking at known numbers from” the 2018

general election. As a result, the court concluded that “several polling locations ran

out of paper” on Election Day and 2600 voters left the polling places where such

shortages occurred without voting. “Given the state of the evidence,” the trial court

“estimated” that “between 250 and 850 voters left [the first polling place] and did

not vote elsewhere” on Election Day. The trial court found that the “EAO’s ballot

paper decision to ignore section 51.005 was both ‘illegal conduct’ and a mistake.” It

thus factored a total of 850 “affected” votes when making its final determination.

We need not decide whether Section 51.005 applied to the 2022 General

Election or whether the EAO’s purported failure to comply with Section 51.005

constitutes illegal conduct or a mistake because even assuming Section 51.005

applied and that the EAO failed to comply with it—issues we do not reach—

22 Lunceford failed to establish conclusively that 2600 potential eligible voters were

denied the privilege of voting in the Contested Race on Election Day due to ballot

paper shortages.

2. Analysis

To satisfy her burden at trial with respect to the 2600 voters, Lunceford had

to establish not only that the voters left polling places on Election Day due to ballot

paper shortages, but also that those voters (1) were eligible to vote,13 (2) would have

cast a vote in the Contested Race, and (3) were unable to cast a vote on Election Day.

See Medlin, 705 S.W.3d at 309 (“The relevant inquiry on appeal is whether []

election officials prevented eligible voters from voting in the contested elections.”);

Reese, 80 S.W3d at 656 (holding that in multi-race election, contestant “must []show

the illegal votes were cast in the race being contested”).

The evidence at trial established that on Election Day, 24 of the 782 polling

places throughout the county experienced ballot paper shortages at some point

during election hours. Three election officials testified at trial (one live and two by

deposition), and twenty-one others submitted testimony via depositions by DWQs

13 To qualify as an eligible voter, a voter must meet the requirements of “a qualified voter as defined by Section 11.002 [of the Election Code] on the day the person offers to vote,” be “a resident of the territory covered by the election for the office or measure on which the person desires to vote,” and “satisfy all other requirements for voting prescribed by law for the particular election.” TEX. ELEC. CODE § 11.001(a).

23 on this issue.14 While they all testified that at some point during Election Day they

experienced ballot paper shortages at their polling places and that some voters left

due to the shortages, with the exception of one election official, no other election

official testified that they knew that specific number of those voters were unable to

cast a vote elsewhere. Indeed, some election officials conceded it was possible that

the voters they “turned away” from their polling place could have voted at one of

the other 782 election-day polling places, and some testified that they informed the

“turned away” voters of nearby polling places at which they could vote.

Victoria Williams, a presiding judge on Election Day, was the only witness

who testified that she knew that some of the voters at her polling place who left due

to ballot paper shortages were unable to cast a ballot elsewhere on Election Day. She

testified that she started running low on ballot paper, and when she had about 100

pages left—enough for fifty voters to cast their vote—she counted the first fifty

people waiting in line. She and her clerks informed the rest of the people in line that

they were running low on ballot paper, and they directed them to a posted list of

nearby polling locations. See TEX. ELEC. CODE § 43.007(O) (“Each countywide

polling place must post a notice of the four nearest countywide polling place

locations by driving distance.”). According to Williams, some of the people looked

14 The trial court considered a total of thirty-eight DWQs from election officials: (1) twenty-one addressed ballot paper shortages, and (2) the remainder addressed alleged irregularities unrelated to ballot paper shortages.

24 at the list of nearby polling places, some waited, and others left. Asked whether she

had any evidence that any of the “turned away” voters had been unable to vote

elsewhere, she said she could only “guarantee without a doubt” that seventy-one

people, who were “in [her] line at 7:00 pm” had not been able to cast a vote, because

there were seventy-seven people inside her polling place and “[b]y the time the paper

arrived [at 10:15 p.m.] there were only six left.” She testified that with respect to the

people who left her polling place before 8 p.m., those voters could have cast a ballot

elsewhere if that location had paper.

The trial court therefore was not presented with conclusive evidence that 2600

eligible voters were unable to cast a vote on Election Day due to paper ballot

shortages. See City of Keller, 168 S.W.3d at 816 (stating matter conclusively

established only if reasonable people could not differ as to conclusions to be drawn

from evidence). At most, the court was presented with evidence that seventy-one

people were unable to cast a ballot on Election Day due to ballot paper shortages.

Even giving due deference to the trial court’s finding that as many as 850 voters

were unable to cast a ballot on Election Day due to paper shortages attributable to

EAOs’ alleged illegal conduct or mistake, the result would be the same because there

25 is no conclusive evidence that the remaining 1236 voters were unable to cast a vote

on Election Day as a result of the EAO’s conduct.15

As part of her fourth issue on this point, Lunceford argues that because the

“evidence conclusively demonstrated” how many voters tried to vote but were

turned away”—2600—the trial court’s findings were in error. She argues the trial

court was “obligated” to find that it was not necessary for her to prove the identity

of the 2600 voters who left without voting, whether they ultimately voted elsewhere

on Election Day, whether they would have cast a ballot in the Contested Race, and

for whom they would have voted, because it was “impossible and impractical” for

her to prove these “unprovable facts.”

According to Lunceford, the trial court should have factored not just 850

voters, but a total of 2600 voters in its final determination. She reasons the trial court

was required to do so because she proved that as a result of the EAO’s purported

illegal conduct and mistakes, at least twenty-four polling places ran out of ballot

paper on Election Day and 2600 potential voters left those polling places without

voting due to the ballot paper shortages. She argues that “all that is required [of her]

is to show that these things occurred, so that the Court may take them into account

15 Even assuming the 850 number is supported by the evidence, as we conclude later in our opinion, the trial court did not abuse its discretion in concluding that the total number of affected votes was “not large enough to put the true outcome [of the Contested Race] in doubt.”

26 when determining whether the true outcome of the election may be ascertained.” We

disagree.

As the election contestant in this case, Lunceford bore the heavy burden of

proving by clear and convincing evidence that voting irregularities occurred in the

Contested Race and that those voting irregularities materially affected the outcome

of the race. See Reese, 80 S.W.3d at 655 (stating contestant must prove by clear and

convincing evidence that “voting irregularities materially affected the outcome of

the election”); Medlin, 705 S.W.3d at 292 (“An election contestant’s burden is a

heavy one, and the declared results will be upheld in all cases except when there is

clear and convincing evidence of an erroneous result.”). This heavy burden required

her to establish that eligible voters were unable to cast a vote in the Contested Race

due to ballot paper shortages.

Citing to Section 221.012 of the Election Code, Lunceford argues she was not

required to prove whether any of 2600 potential voters ultimately voted elsewhere

and whether they would have cast a ballot in the Contested Race, because it was

impossible and logistically impractical for her to do so, and thus the trial court should

have taken all 2600 potential votes into account “when determining whether the true

outcome of the election may be ascertained.” Lunceford’s argument does not survive

scrutiny.

27 Under Section 221.003, titled “Scope of Inquiry,” the tribunal must first

“attempt to ascertain whether the outcome of the contested election, as shown by the

final canvass, is not the true outcome” because illegal votes were counted or election

officials prevented eligible voters from voting, failed to count legal votes, or engaged

in other fraud or illegal conduct or made a mistake. TEX. ELEC. CODE § 221.003(a).

This inquiry is based on the evidence presented to the trial court. After engaging in

this analysis, Section 221.012 dictates the action of the tribunal. Section 221.012,

titled “Tribunal’s Action on Contest,” provides that if a “tribunal hearing an election

contest can ascertain the true outcome of the election, the tribunal shall declare the

outcome.” Id. § 221.012(a). If, however, the tribunal “cannot ascertain the true

outcome of the election,” the tribunal “shall declare the election void.” Id.

§ 221.012(b).

Read together, these two provisions clarify that an election contestant must

first “allege and prove (1) particularized material irregularities in the conduct of an

election,” and after, the trial court will determine whether (2) the irregularities render

it impossible to determine the true outcome of the election. See Reese, 80 S.W.3d at

656 (explaining that “[a]fter the contestant has proved that illegal votes were cast in

the contested race,” pursuant to Section 221.012(b), “if the trial court can ascertain

the true outcome, it must declare the outcome,” but if the “trial court cannot ascertain

the true outcome of the election, it must declare the election void”) (emphasis

28 added); Guerra v. Garza, 865 S.W.2d 573, 576 (Tex. App.—Corpus Christi–

Edinburg 1993, writ dism’d w.o.j.) (stating contestants “must allege and prove

particularized material irregularities in the conduct of the election” and, separately,

“that the irregularities rendered impossible a determination of the majority of the

voters’ true will”).

Section 221.012(b) thus did not absolve Lunceford of her burden to establish

that 2600 eligible voters were unable to cast a vote in the Contested Race due to

ballot paper shortages—the alleged irregularity. See Guerra v. Ramirez, 364 S.W.2d

720, 724 (Tex. App.—San Antonio 1963, writ dism’d) (“The fact that a number of

witnesses are thereby necessary does not relieve the contestant of the heavy burden

he must carry to overthrow the presumption that the election officials discharged

their duty properly in rejecting the ballots.”).16

Lunceford cites to Gonzalez v. Villarreal, 251 S.W.3d 763 (Tex. App.—

Corpus Christi–Edinburg 2008, pet. dism’d) and Green v. Reyes, 836 S.W.2d 203

(Tex. App.—Houston [14th Dist.] 1992, no writ) in support of her argument.

Lunceford’s reliance on Gonzalez and Green is misplaced because the issue

16 Lunceford argues in her brief that “[n]owhere in the documentary or testimonial evidence is there any proof whatsoever whether, and to what extent, any of the 2600 turned away voters ended up voting somewhere else.” But it was not Craft’s burden to establish that these voters were able to cast a vote on Election Day; it was Lunceford’s burden to establish they were unable to do so because of EAO’s actions.

29 presented in those cases was whether the outcome of the contested elections could

be determined based on evidence that purported illegal votes had been cast in the

contested races.17 See TEX. ELEC. CODE § 221.003(a)(1) (stating court “shall attempt

to ascertain whether the outcome of the contested election, as shown by the final

canvass, is not the true outcome because. . . illegal votes were counted”).

Gonzalez and Green are also distinguishable because the contestants in those

cases challenged the legality of votes that had been cast in the contested races,

whereas Lunceford argues that the trial court should have concluded, based on

evidence that 2600 voters left twenty-four polling locations on Election Day, and

that all 2600 voters were unable to cast a vote on Election Day. If the 2022 General

Election had been a precinct-based election requiring voters to cast a vote at a

17 In Green, the contestant presented evidence that over 400 voters had cast a vote in both the Republican and Democratic primaries in a congressional race rendering those “crossover” ballots void. Green v. Reyes, 836 S.W.2d 203, 204 (Tex. App.— Houston [14th Dist.] 1992, no writ). The trial court heard evidence “concerning the voting conduct of 313 alleged crossover voters,” including testimony from 246 voters who appeared in person or through telephone depositions, and it concluded, based on the evidence, that there had been 305 illegal crossover votes. Id. at 206. For the remaining votes, the trial court found “there was no reliable way of determining how the 126 unascertained votes should be considered to further adjust the votes for each candidate.” Id. at 207. Similarly, in Gonzalez, the contestant presented evidence that “voters who were not on the registration rolls and did not have voter registration cards were allowed to vote with only their identification cards.” Gonzalez v. Villarreal, 251 S.W.3d 763, 768 (Tex. App.—Corpus Christi– Edinburg 2008, pet. dism’d). In determining whether the court could ascertain the true outcome of the contested race based on the challenged votes, the court concluded that “disheveled election records” made “it impossible to determine whether those votes were legally cast and countable.” Id. at 768, 781. Thus, in both Green and Gonzalez, the contestants presented evidence of the alleged irregularity.

30 designated precinct, Lunceford’s argument might carry more weight. But the

General Election was a countywide polling place election, allowing voters to vote at

any one of 782 election-day polling places. The evidence established that only

twenty-four of 782 polling places were affected by ballot paper shortages on Election

Day, that they were affected only during some portion of the day, that additional

ballot paper was delivered to some of these polling places, and that at some of these

affected polling places, voters were informed about nearby polling places at which

they could vote. Evidence that a person left a polling place on Election Day without

voting is thus not conclusive evidence that that voter was unable to cast a vote

elsewhere at one of the remaining 758 unaffected polling places on Election Day,

nor does it establish that the trial court’s finding on this issue was contrary to the

overwhelming weight of the evidence and clearly wrong.

“Given the state of the evidence,” the trial court found that “2600 voters who

tried to vote at their polling place of choice left without voting” due to “paper

shortages,” “estimated that 250 to 850 lawful voters did not cast votes because of

the EAO’s ballot-paper decision,” and took 850 of those voters into account when

considering whether it could “ascertain the true outcome of the election.” Because

Lunceford did not conclusively establish that 2600 eligible voters were unable to

cast a ballot on Election Day due to ballot paper shortages, the trial court did not

abuse its discretion by failing to factor all 2600 votes into account when considering

31 whether it could ascertain the outcome of the Contested Race. “The law is well

settled that irregularities in the conduct of an election which cannot be shown to have

materially affected the results of the election are immaterial.” Woods, 363 S.W.3d at

716 (quoting Kennelly v. Gates, 406 S.W.2d 351, 358 (Tex. Civ. App.—Houston

1966, no writ)).

After examining the record in light of Lunceford’s high evidentiary burden,

and “defer[ring] to the trial court’s decision that the evidence did not meet it,” we

conclude that Lunceford did not conclusively establish that 2600 voters were unable

to cast a vote on Election Day due to ballot paper shortages resulting from the EAO’s

failure to comply with Section 51.005 of the Election Code. Nor can we say that the

trial court’s failure to form a firm conviction or belief that 2600 voters were unable

to cast a vote on Election Day due to ballot paper shortages resulting from the EAO’s

failure to comply with Section 51.005 of the Election Code was contrary to the

overwhelming weight of the evidence and clearly wrong. See DeSpain, 672 S.W.3d

at 492 (legal sufficiency); Burns, 434 S.W.3d at 227–28 (factual sufficiency); see

generally City of Keller, 168 S.W.3d at 816 (stating matter conclusively established

only if reasonable people could not differ as to conclusions to be drawn from

evidence).

We overrule Lunceford’s first issue and the portion of Lunceford’s third and

fourth issues on this point.

32 B. Non-Ballot Paper Related Problems at Polling Places

In her second issue, Lunceford argues that in addition to the voters who left

polling places due to ballot paper shortages, she established through the DWQ

testimony of several election officials that an additional 411 potential voters left their

first polling place for reasons unrelated to ballot paper shortages, such as machine

malfunctions, poll workers’ inability to reach the EAO on the phone or by other

means, a lack of equipment or supplies, and other problems. Notwithstanding, she

argues the trial court “completely ignored” and failed to take these 411 voters into

account when it considered whether it could ascertain the true outcome of the

election.

In its final judgment and calculation of “affected votes,” the trial court did not

include the 411 voters who purportedly left their polling places for reasons unrelated

to ballot paper shortages. In its finding of fact number 17, the trial court stated:

From the evidence, the court finds that because of paper shortages 2600 voters who tried to vote at their polling place of choice left without voting. These numbers do not include voters discouraged by long lines who voted elsewhere due to machine malfunctions or paper jams, which were not caused by EAO decisions.

(Emphasis added).

Contrary to Lunceford’s position, Craft argues that the trial court addressed

the 411 voters in this finding of fact when it found that the 2600 voters who left

polling places without voting due to ballot paper shortages did “not include voters

33 discouraged by long lines who voted elsewhere due to machine malfunctions or

paper jams, which were not caused by EAO decisions.” Craft argues the trial court

properly refused to consider these 411 voters who left a polling place without voting

because “there was no nexus between the cause of long lines and the conduct or

decisions of an election official,” and furthermore, “the evidence of those 411

potential voters suffers from all of the same flaws as the 2600 potential voters that

supposedly left for ballot paper reasons.”

We agree with Craft that in its findings of fact, the trial court implicitly

addressed the approximately 411 potential voters who left polling places for reasons

unrelated to ballot paper shortages, concluding they left the polling place without

voting for reasons “not caused by EAO decisions.” We also conclude that the trial

court did not abuse its discretion in rejecting Lunceford’s contention that the trial

court should have taken these 411 voters into account when considering whether the

outcome of the Contested Race, as shown by the final canvass, was not the true

outcome.

Under Section 221.003(a)(2), the “tribunal hearing an election contest shall

attempt to ascertain whether the outcome of the contested election, as shown by the

final canvass, is not the true outcome” because an election officer or other person

officially involved in the administration of the election “prevented eligible voters

from voting” or “engaged in other fraud or illegal conduct or made a mistake.” TEX.

34 ELEC. CODE § 221.003(a)(2)(A), (C). Lunceford did not argue that election officials

“prevented eligible voters from voting” or that election officials “engaged in fraud

or illegal conduct or made a mistake” with respect to the alleged problems at polling

places unrelated to ballot paper shortages. The trial court thus did not abuse its

discretion by failing to include in its calculation of “affected votes” the number of

“voters discouraged by long lines who voted elsewhere due to machine malfunctions

or paper jams, which were not caused by EAO decisions.” (Emphasis added).

Moreover, even assuming there was sufficient evidence that 411 potential

voters left a polling place without voting on Election Day due to non-ballot paper

shortages attributable to the EAO’s conduct, like the 2600 potential voters who left

a polling place without voting due to ballot paper shortages, Lunceford did not

conclusively establish that these 411 voters were unable to cast a vote elsewhere on

Election Day.18 Nor did she establish that the trial court’s finding on this issue was

contrary to the overwhelming weight of the evidence and clearly wrong.

We overrule Lunceford’s second issue.

18 Indeed, the only voter to testify at trial testified through a DWQ that he left three polling locations without voting on Election Day for non-ballot paper related reasons, but that he ultimately was able to cast a vote at a fourth polling place.

35 C. Reasonable Impediment Declarations (“RIDs”)

In her third issue, Lunceford argues there is legally and factually insufficient

evidence supporting the trial court’s finding that 350 illegal votes were cast by voters

who presented incomplete RIDs. She argues the correct number should have been

380 illegal votes based on incomplete RIDs.19

In its finding of fact number 44, the trial court stated:

The court concludes that 380 of the 532 challenged RIDs are so lacking in the statutory information that they are improper, and votes cast by these 350 voters should not have been counted.

(Emphasis added). Similarly, in its final judgment, the trial court held:

A RID is the voter’s chance to comply with the code’s effort to make sure that voters can demonstrate who they are with documents. The court concludes that 380 of the 532 challenged RIDs are so lacking in the statutory information that they are improper, and votes cast by these 350 voters should not have been counted.

(Second and third emphasis added).20

19 Under Section 63.001(b) of the Election Code, a voter must present to an election officer at the polling place “(1) one form of identification listed in Section 63.0101(a); or (2) one form of identification listed in Section 63.0101(b) accompanied by the declaration described by Subsection (i).” TEX. ELEC. CODE § 63.001(b). Under Section 63.001(i), if the requirement for identification prescribed in Section (b)(1) is not met, “an election official shall notify the voter that the voter may be accepted for voting if the voter meets the requirement for identification prescribed by Subsection (b)(2) and executes a declaration declaring the voter has a reasonable impediment to meeting the requirement for identification prescribed by Subsection (b)(1).” Id. § 63.001(i).

36 Lunceford “assumes the Trial Court made a mistake and intended to say 380

instead of 350” and that her “view makes sense when viewed in concert with Finding

of Fact number 73, which adds up all the votes accepted by the Trial Court as

illegal.”21

It is unclear from the trial court’s finding of fact number 44 and from the

above-quoted final judgment language whether the trial court concluded that 380 or

350 “illegal votes” were cast based on its findings that the voters’ “RIDs [were]

lacking statutory information.” We agree with Lunceford, however, that given the

trial court’s conclusion that “a total of 2041 illegal votes” were cast, for that

“number[s] to add up correctly, 380 is required, not 350.” We thus agree with

Lunceford that the trial court’s finding that “350 voters should not have been

counted” is the result of a typographical error.

Because the trial court’s final calculation of “illegal votes” included the 380

“illegal votes” tied to incomplete RIDs, however, the trial court’s statement that “350

voters should not have been counted” did not cause the rendition of an improper

judgment or prevent Lunceford from presenting her case on appeal. Any such error

21 In its finding of fact number 73, the trial court stated: “Mistake and illegal-vote findings. The court has estimated that 250 to 850 lawful voters did not cast votes because of the EAO’s ballot-paper decision, which was ‘illegal conduct’ and also a mistake under section 221.003. There were 2041 illegal votes as discussed above. Assuming and using the largest estimated number (850), this yields a total of 2891 affected votes.” (Emphasis in original).

37 was thus harmless. See TEX. R. APP. P. 44.1(a) (court cannot reverse judgment in

civil case based on error unless error “probably caused the rendition of an improper

judgment” or “probably prevented the appellant from properly presenting the case

to the court of appeals”).

We overrule this portion of Lunceford’s third issue.

In her crosspoint on this issue, Craft argues that the evidence is legally and

factually insufficient to support the trial court’s finding that “380 RIDs constitute[d]

illegal votes.” She first argues that Lunceford waived the issue because she “never

contended that the votes tied to improper RIDs were ‘illegal’” but only complained

that the faulty RIDs “constituted ‘mistakes.’” Separately, Craft argues that “an

incomplete RID, in and of itself, [cannot] serve as the basis to disqualify [a] vote

attached to that RID” and the trial court ignored the “presumption in place that [an]

election judge working with the voter acted properly.” We do not reach these issues,

because even assuming, without deciding, that sufficient evidence supported the trial

court’s finding that 380 votes tied to incomplete RIDs were cast—an issue we do

reach—as we explain below, the trial court did not abuse its discretion in declaring

that the outcome of the election was the true outcome.

Voters on the Suspense List

In finding of fact number 67, the trial court stated:

Lunceford contended that 1995 voters whose names were on the suspense list were permitted to vote without showing that they still

38 resided in the county. This contention was not proved to the court’s satisfaction and it is respectfully denied.

In her third issue, Lunceford argues the trial court erred by failing to take into

consideration the votes of these 1995 voters because “the evidence conclusively

demonstrated that these 1,995 voters who cast a ballot without a Statement of

Residence (“SOR”) cast a vote that was illegal.” In other words, she complains about

the trial court’s failure to find that she proved that 1995 voters on the suspense list

cast illegal votes. Relying in part on the testimony from her expert, Steve Carlin,

Lunceford argues that she presented undisputed evidence that 2039 voters on the

official Roster had a suspense notation (“S”) next to their name and that 1995 of

those voters cast a vote in the General Election without submitting a SOR.22

Craft responds that the trial court properly found that Lunceford failed to bring

forth clear and convincing evidence that the 1995 voters on the suspense list cast

illegal votes because Lunceford did not present evidence establishing the date the

22 Section 63.0011 of the Election Code, titled “Statement of Residence,” provides that “before a voter may be accepted for voting, an election officer shall ask the voter if the voter’s residence address on the precinct list of registered voters is current and whether the voter has changed residence within the county.” TEX. ELEC. CODE § 63.0011(a). If a “voter’s residence address is not current because the voter has changed residence within the county, the voter may vote, if otherwise eligible” but “[b]efore being accepted for voting, the voter must execute and submit to an election officer a statement” that includes “a statement that the voter satisfies the applicable residence requirements,” “all of the information that a person must include in an application to register to vote under Section 13.002,” and the “date the statement is submitted to the election officer.” Id. § 16.0011(b)–(c). The statement must also “include a field for the voter to enter the voter’s current county of residence.” Id. § 16.0011(c-1).

39 voters were placed on the suspense list, ruling out that the voters voted using a

provisional ballot, accounting for any voters who were mistakenly placed on the

suspense list, determining whether any voter had cured their suspense status, or

proving how many of the challenged voters cast a ballot in the Contested Race.

According to Craft, Carlin’s methodology was critically flawed because, among

other things, he never considered whether any of the voters on the suspense list had

cast a vote in the General Election using a provisional ballot and he failed to verify

that the voters’ registrations were suspended when they voted on Election Day.

A. Applicable Law

Section 15.081 of the Election Code, titled “Suspense List,” provides that the

“registrar shall maintain a suspense list containing the name of each voter: (1) who

fails to submit a response to the registrar in accordance with Section 15.053;23 (2)

whose renewal certificate is returned to the registrar . . . or (3) who appears on the

list of nonresidents of the county provided to the registrar . . . .” TEX. ELEC. CODE

§ 15.081(a). The registrar is required to “enter the notation ‘S’, or a similar notation

approved by the secretary of state, on the list of registered voters beside each voter’s

name that also appears on the suspense list.” Id. § 15.111.

23 Section 15.053 provides that a “voter shall submit to the registrar a written, signed response to the notice that confirms the voter’s current residence.” TEX. ELEC. CODE § 15.053(a).

40 A “voter whose name appears on a precinct list of registered voters with the

notation ‘S’” may cast a vote in “an election held on or after the date the voter’s

name is entered on the suspense list” if the “voter satisfies the residence

requirements prescribed by Section 63.0011 and submits a statement of residence

in accordance with that section.” Id. § 15.112.

B. Analysis

On December 8, 2022, Harris County released the official Roster for the 2022

General Election.24 Relying exclusively on the Roster, Carlin testified that 2039

voters on the Roster were on the suspense list. Carlin acknowledged that the Roster

contained a lot of inaccurate and missing information. Despite these shortcomings,

Carlin relied exclusively on the Roster to compile the list of 2039 voters on the

suspense list. Carlin, who did not verify when each voter’s registration was

suspended, assumed that when the voters cast their vote on Election Day—

November 8, 2022—the voters’ registration was in suspense status. When he

compared the suspense list with the SOR data transcribed by his team of volunteers,

which he admitted at trial was not “100 percent accurate,” Carlin determined that of

the 2039 voters on the suspense list, eighty-eight had submitted a SOR before casting

24 Lunceford states in her brief that the Harris County Voter Roster is a “list of everyone who voted in the November 8, 2022 General Election,” while Craft states that the Roster “indicates the number of people who checked in at a polling location, but that does not necessarily reflect votes because it does not include who left after checking in without voting (‘fleeing votes).’”

41 a vote in the General Election. Because Carlin was unable to match the remaining

1995 voters on the suspense list with a SOR, Lunceford argues those voters had cast

an illegal vote.

Carlin admitted that he did not use the most reliable method to match a voter

on the suspense list with the SOR data because he only had a Voter Unique Identifier

(VUID”) for 100 or so of the approximately 40,000 SORs in the database. Carlin

testified that he was not aware that a voter on the suspense list may cast a vote with

a provisional ballot, and he did not compare the list of suspense voters with the

provisional ballots cast on Election Day.

Beth Stevens, Craft’s expert, testified that a person on the suspense list can

cast a vote using a provisional ballot without submitting a SOR. She also testified

that the Roster reflects only the number of people who checked in at a polling place

on Election Day, not the number of people who actually cast a ballot.25 She testified

that Carlin’s comparison of the suspense list with the SOR data was also flawed

because Carlin and his team did not have VUIDs for every voter who submitted a

SOR, allowing them to match a voter on the suspense list with a submitted SOR.

25 One election official testified through DWQs that his polling place was unable to process voters for forty-five minutes because their scanner was broken and “[s]everal people waited in line to cast their ballots but others that were discouraged by the long wait time, left with their ballots in hand.” He knew of one voter who left with her ballot but returned to the polling place later after the scanner was working to cast her ballot.

42 In addition to creating a suspense list based on the Roster, Carlin testified that

he relied exclusively on the Roster to compile a list of voters whose registration had

been cancelled as of Election Day. Like the suspense list, Carlin assumed that the

voters’ registration was in cancelled status when they voted on Election Day because

there was a “C” in the status column next to their name. The record reflects that

Carlin’s cancellation list was inaccurate because only five of the 2970 voters on his

list had their registration cancelled prior to Election Day. The remaining 2965 voters

were placed on the cancellation registration list after the General Election.26

Although Lunceford argues on appeal that the Roster is “a snapshot of what

occurred on the day the voter voted” and “[t]hus, by definition, each of these voters

were on the suspense list at the time they presented themselves to the Qualifying

Table for voting,” Lunceford did not submit evidence that the Roster was an accurate

reflection of a voter’s status on Election Day. Moreover, the record reflects that the

Roster, which was released a month after the General Election, does necessarily

reflect a voter’s status on Election Day, as demonstrated by the numbers of cancelled

voters who appear on the list but whose voter registration was cancelled after

Election Day.27

26 As a result, Lunceford withdrew those 2695 voters from her election contest. 27 Lunceford’s counsel argued during oral argument that although a voter may be placed on the cancellation list after the election but before the official roster is released one month later, the suspense list is “determined in the fall” and “the books

43 Lunceford also responds that she was not required to rule out all possible

scenarios in which a voter on the suspense list could have voted legally, such as

voting with a provisional ballot affidavit, because “it was not necessary to disprove

a negative.” In an election contest, however, the contestant “has the burden of

proving a negative proposition—that a challenged vote is not legal.” Rivera v. Lopez,

No. 13-14-00581-CV, 2014 WL 8843788, at *3 n.2 (Tex. App.—Corpus Christi–

Edinburg May 14, 2014, no pet.) (mem. op.). It is the appellate court’s “duty [to

then] determine on what potential basis an uncounted vote may have been legal, then

close in about October” in advance of early voting. According to Lunceford’s counsel, “So 100% of everybody who had an S on the voter roster had the S on the date of the election and in fact had it probably 45 days earlier than the date if the election because of the way the S notation works.” This argument was not made to the trial court and no evidence supporting it was before the trial court when it rendered its final judgment. We have not found, and Lunceford has not directed us to, any evidence in the record establishing this fact or explaining that a person’s suspense status does not change between election day and the date the official roster is released. Rather, the only evidence before the trial court, and before us on appeal, is that Carlin compiled his suspense list and cancellation list from notations in the Roster and that some of the voters on Carlin’s cancelation list had their registration cancelled after Election Day, thus indicating that the Roster and Carlin’s corresponding lists were not a reliable and an accurate reflection of a voter’s status on Election Day. We note that the Election Code provides that people are added to the suspense list on a monthly basis. See TEX. ELEC. CODE § 15.081(a)(3) (stating “registrar shall maintain a suspense list containing the name of each voter. . . who appears on the list of nonresidents of the county provided to the registrar under Section 62.114, Government Code”); TEX. GOV’T CODE § 62.114 (a)–(b) (requiring clerk of court to “maintain a list containing the name and address of each person who is disqualified under this subchapter from jury service because the person is not a resident of the county” and send on “the third business day of each month. . . a copy of the list of persons disqualified in the previous month because the persons do not reside in the county” to county’s voter registrar and secretary of state).

44 review whether [the contestant] presented evidence that the vote was illegal” before

applying “the traditional rule of legal sufficiency review.” Id.

As the person contesting the election, it was thus Lunceford’s burden to prove

that the 1995 voters on the suspense list cast illegal votes. She did not meet her

burden. Lunceford’s expert did not consider how many voters on the suspense list

could have voted with a provisional ballot, whether any of the challenged voters had

cured their suspense status, or the date on which the voters were placed on the

suspense list. As the sole factfinder, the trial court could have reasonably determined

that Carlin’s suspense list and cross-reference efforts, like the cancellation list, were

inaccurate and that Carlin’s testimony that 1995 voters had cast an illegal vote

because they were on the suspense list was similarly flawed and disregarded Carlin’s

testimony on this issue. Gilbreath, 682 S.W.3d at 507 (stating when clear and

convincing evidence standard applies, courts “assume the fact finder resolved any

disputed facts in favor of its finding if a reasonable fact finder could have done so

and we disregard all evidence that a reasonable fact finder could have disbelieved”).

After examining the record in light of Lunceford’s high evidentiary burden,

and “defer[ring] to the trial court’s decision that the evidence did not meet it,” we

conclude that Lunceford did not conclusively establish that 1995 illegal votes were

cast in the General Election by voters on the suspense list, nor can we say that the

trial court’s failure to form a firm conviction or belief that1995 illegal votes were

45 cast in the General Election by voters on the suspense list is contrary to the

overwhelming weight of the evidence and clearly wrong. See DeSpain, 672 S.W.3d

at 492 (legal sufficiency); Burns, 434 S.W.3d at 227–28 (factual sufficiency); see

also City of Keller, 168 S.W.3d at 816 (stating matter conclusively established only

if reasonable people could not differ as to conclusions to be drawn from evidence).

Whether Outcome of Final Canvas is True Outcome of Contested Race

The ultimate question before the trial court was whether the court was able to

“ascertain the true outcome of the election” based on the evidence presented. TEX.

ELEC. CODE § 221.012(b) (stating if trial court “can ascertain the true outcome of the

election, the [trial court] shall declare the outcome”); id. § 221.012(a) (stating trial

court “shall declare the election void if it cannot ascertain the true outcome of the

election”).

According to the final canvass, 1,064,677 voters cast a ballot in the Contested

Race. Craft received 533,710 while Lunceford received 530,967, resulting in a

margin of victory of 2743 votes. In its final judgment, the trial court concluded that

“a total of 2041 illegal votes” had been cast in the General Election based on several

46 alleged irregularities.28 29 The trial court also concluded that between “250 to 850

votes were not cast due to the EAO’s ballot paper decision, which was illegal

conduct and also a mistake.”

28 In footnote 22 to Finding of Fact No. 71, the trial court accounted for the following “illegal votes:” “[v]oting by out-of county residents (1236), provisional ballots (43), mail ballots (45), photo identification (380), erroneous instructions to the SVC (7), instructions for unscannable ballots (zero), mistakes regarding TRO (325), and voting after registration was cancelled (5). These findings . . . equal 2041 illegal votes.” 29 Lunceford argues that the trial court “correctly concluded that” the 325 net votes cast for Craft during the extended hour of voting on Election Day “were illegal votes” and thus Section 221.011 of the Election Code required the trial court to subtract the votes from the margin of victory. On the contrary, in Conclusion of Law Nos. 32 and 34, the trial court concluded that agreeing “to the extension was not illegal’ and that the “325 net votes for Craft resulted from the EAO’s mistaken approval of the extra hour.” (Emphasis in original). Section 221.011(a) only applies to illegal votes. See TEX. ELEC. CODE § 221.011(a) (stating if trial court can ascertain candidate for whom illegal vote was cast, court “shall subtract the vote from the official total for the candidate . . . as applicable”). Thus, assuming, without deciding, that the trial court correctly found that the EAO made a “mistake” by agreeing to the extended hour of voting on Election Day, the trial court did not err by not subtracting the 325 votes from Craft’s margin of victory. To the extent that footnote 22 to Finding of Fact No. 71 conflicts with Conclusions of Law Nos. 32 and 34, we must construe them together and if possible, in harmony with the judgment. A finding of mistake is consistent with the trial court’s inclusion of the 325 votes with the “affected votes,” as opposed to subtracting them from the margin of victory. Construing footnote 22, Finding of Fact No. 71, and Conclusion of Law Nos. 32 and 34 together, we conclude that the trial court found that the 325 votes were the result of a mistake, not “illegal votes.” See Tex. Outfitters Ltd., LLC v. Nicholson, 534 S.W.3d 65, 74 (Tex. App.—San Antonio 2017), aff’d, 572 S.W.3d 647 (Tex. 2019) (“[F]indings of fact and the conclusions of law will be construed together; and if the findings of fact are susceptible of different constructions, they will be construed, if possible, to be in harmony with the judgment and to support it.”).

47 Taking the 2041 illegal votes and “using the largest estimated number [of

votes not cast] (850),” the trial court concluded that there had been a total of “2,891

affected votes” in the General Election and that of those total votes, 2779 votes had

been cast in the Contested Race

As part of her fourth issue, Lunceford argues that the “trial court was required

to find that [she] was not and is not required to demonstrate whether an illegal vote

was cast and counted in the Contested [Race] to be afforded a new election.” She

argues that pursuant to Section 221.009(b) of the Election Code, a tribunal may reach

its decision without attempting to determine how individual voters voted so long as

the number of illegal votes is equal to or greater than the number of votes necessary

to change the outcome of an election. We disagree that Section 221.009(b) dispenses

with Lunceford’s burden to establish that illegal votes were cast in the Contested

Race.

The trial court in an election contest “shall attempt to ascertain whether the

outcome of the contested election, as shown by the final canvass, is not the true

outcome.” See TEX. ELEC. CODE § 221.003(a). In a multi-county race, the contestant

has the burden of proving that illegal votes were cast in the specific contested race.

See Reese, 80 S.W.3d at 656 (stating in multi-race election, contestant “must first

show that (1) illegal votes were counted or (2) an election official prevented eligible

voters from voting, failed to count legal votes, or engaged in other fraud, illegal

48 conduct or mistake,” and “must next show the illegal votes were cast in the race

being contested”); Medrano v. Gleinser, 769 S.W.2d 687, 688 (Tex. App.—Corpus

Christi–Edinburg 1989, no writ) (“The burden of proving illegality in an election

contest is on the contestant, who must prove that illegal votes were cast in the

election being contested and that a different and correct result would have been

reached by not counting the illegal vote.”); Miller v. Hill, 698 S.W.2d 372, 375 (Tex.

App.—Houston [14th Dist.] 1985, writ dism’d w.o.j.) (reversing trial court’s

judgment voiding county sheriff’s election because no evidence established that

illegal votes were cast in contested race).

Section 221.009(a) does not provide otherwise. Section 221.009(a) permits a

trial court to compel a voter who has cast an illegal vote “to disclose the name of the

candidate for whom the voter voted or how the voter voted on a measure if the issue

is relevant to the election contest.” TEX. ELEC. CODE. § 221.009(a) (emphasis added).

Section 221.009(b) states that, “[i]f the number of illegal votes is equal to or greater

than the number of votes necessary to change the outcome of an election, the tribunal

may declare the election void without attempting to determine how individual voters

voted.” Id. Under Section 221.011(a), if the trial court “can ascertain the candidate

or side of a measure for which an illegal vote was cast,” it must “subtract the vote

from the official total for the candidate or side of the measure, as applicable.” Id.

§ 221.011(a) (emphasis added). If, however, the trial court “finds that illegal votes

49 were cast but cannot ascertain how the voters voted, the tribunal shall consider those

votes in making its judgment.” Id. § 221.011(b). In the context of Sections

221.009(a) and 221.011(a), the references to “how voters voted” are thus references

to the name of the candidate or side of a measure for which a voter voted—not

whether the voter voted in the contested race.30

Recognizing that in an election contest, a tribunal must consider only those

votes cast in a contested race, the trial court found that while 2891 votes in the

General Election had been “affected,” “[n]ot all of the 2891 would have been cast in

the Lunceford vs. Craft contest.” Explaining there had been “a 3.86% undervote in

the race for the189th District Court,” the court held that “roughly the same undervote

percentage in the [Contested Race] would have occurred with the affected votes—

[that] 96.14% of the 2041 illegal votes (plus the estimated 850 that were deterred

from voting by the ballot paper decision) would have been cast in the [Contested

30 Lunceford also relies on findings of fact and conclusions of law entered by the trial court in Leal v. Peña, No. 2020-DCL-06433 in the 107th District Court of Cameron County, Texas. Although the court of appeals affirmed the trial court’s judgment in that case, the appellate court’s opinion did not discuss the findings of fact and conclusions of law upon which Lunceford relies. Nor did the court affirm the trial court’s judgment based on those findings and conclusions. See Peña v. Leal, No. 13-22-00204-CV, 2023 WL 3116752 (Tex. App.—Corpus Christi–Edinburg Apr. 27, 2023, pet. denied) (mem. op.).

50 Race].” Thus, the trial court concluded that a total of “2779 votes in the 189th

[Contested Race] (96.14% of 2891) [had been] affected.”31 32

The court found that while the “2779 affected votes slightly exceed[ed] Craft’s

margin of victory, 2743” the number of affected votes was “not large enough to put

the true outcome [of the Contested Race] in doubt.” Contrasting this case to Green

v. Reyes, 836 S.W.2d 203 (Tex. App.—Houston [14th Dist.] 1992, no pet.), where

the trial court held it could not ascertain the true outcome of the election because the

31 In her third issue, Lunceford argues the trial court correctly analyzed the undervote percentage and she does not challenge that portion of the trial court’s findings. She argues, however, that the trial court erred by applying the undervote percentage to the 325 net votes cast for Craft during the extended hour of voting on Election Day. We agree. The purpose of the undervote calculation was to calculate the number of “affected” votes cast in the Contested Race. The trial court had before it the total number of votes cast during the extra hour of voting in the Contested Race and it knew for whom they had been cast because the Supreme Court of Texas ordered that those provisional ballots be segregated and separately counted. Excluding the 325 votes, the trial court found that a total of 1716 “illegal votes” and up to 850 votes were not counted due to alleged mistake or illegal conduct, resulting in a total count of 2566 “affected votes.” Applying the undervote percentage to that number would yield 2467 “affected votes” in the Contested Race, and adding the 325 votes to that total number would result in “2792” total “affected votes.” That is only a thirteen-vote difference compared to the trial court’s calculation of “2779 affected votes.” We thus conclude that the trial court’s error was not material because it only results in a thirteen-vote difference. 32 Craft argues in a crosspoint that the trial court abused its discretion in applying the undervote percentage to calculate the number of affected votes in the Contested Race because there was no expert testimony or other evidence supporting the trial court’s application of the undervote percentage across all categories of challenged votes to determine the number of “affected votes” in the Contested Race. Because it is not necessary to our disposition, we do not address this issue, nor do we express an opinion on the sufficiency or correctness of the trial court’s decision to apply the undervote percentage to calculate the number of affected votes in the Contested Race.

51 number of illegal votes was roughly three times as large as the margin of victory, the

trial court concluded that “2779 illegal votes [was] not enough to make the true

outcome unknowable in an election with a 2743-vote margin in the canvassed final

result.” The trial court denied Lunceford’s contest and declared “Craft’s victory in

the contest for Judge of the 189th District Court [] to be the true outcome.”

Given the number of votes cast in the Contested Race, and the number of votes

the trial court concluded were “affected” in the Contested Race, we cannot say that

the trial court abused its discretion by finding that it could ascertain the true outcome

of the Contested Race because the number of affected votes was not large enough to

put the true outcome in doubt. See Woods, 363 S.W.3d at 716 (“The statute, however,

expressly leaves the discretion to make such a declaration to the trial court.”).

We hold the trial court did not abuse its discretion by finding that Craft’s

victory over Lunceford in the Contested Race, as set forth in the final canvas, was

the true outcome of the Contested Race and denying Lunceford’s election challenge.

52 Conclusion

We affirm the trial court’s judgment.33 34

Veronica Rivas-Molloy Justice

Panel consists of Justices Rivas-Molloy, Gunn, and Morgan.

Morgan, J., concurring.

33 Because we affirm the trial court’s judgment, we do not consider Craft’s remaining crosspoints. 34 Our opinion should not be construed as a finding of sufficiency or correctness on any of the trial court’s findings of fact and conclusions of law not expressly resolved by our opinion or the merits of this appeal.

Cite This Page — Counsel Stack

Bluebook (online)

Erin Elizabeth Lunceford v. Tamika Craft, Counsel Stack Legal Research, https://law.counselstack.com/opinion/erin-elizabeth-lunceford-v-tamika-craft-txctapp1-2026.