In RE DALLAS HERO AND CATHY CORTINA ARVIZU v. the State of Texas

Supreme Court of Texas

══════════

No. 24-0678

══════════

In re Dallas HERO and Cathy Cortina Arvizu,

Relators

═══════════════════════════════════════

On Petition for Writ of Mandamus

═══════════════════════════════════════

JUSTICE BUSBY delivered the opinion of the Court.

This dispute sits at the intersection of citizens’ power to propose

amendments to their city’s charter, the city council’s power to do

likewise, and the council’s responsibilities in preparing the ballot for an

election to approve proposed amendments. See TEX. ELEC. CODE

§ 52.072; TEX. LOC. GOV’T CODE § 9.004(a). Relators organized a citizen

petition drive and signed petitions to place three proposed charter

amendments on the upcoming election ballot, and the city council

submitted three proposed amendments of its own that relators contend

would effectively nullify their proposed amendments. The council-

initiated propositions include primacy provisions specifying that they

control in the event of a conflict.

Relators have filed a mandamus petition raising four challenges

to the council-initiated propositions. The principal question we decide

is whether the ballot language the city council selected to describe the

various propositions satisfies the standard of clarity and definiteness we

articulated in Dacus v. Parker, 466 S.W.3d 820 (Tex. 2015). We hold

that it does not: the propositions contradict each other, and the ballot

language as a whole will confuse and mislead voters because it does not

acknowledge these contradictions or address the effect of the primacy

provisions, which are chief features central to the character and purpose

of the council-initiated propositions. Because the citizen-initiated

propositions must appear on the ballot and the parties have agreed to

the ballot language for those propositions, we conclude the proper

remedy is to direct the city council not to include its duplicative

propositions on the ballot.

BACKGROUND

The qualified voters of a municipality may propose an

amendment to its charter by petition, and the governing body of the

municipality “shall submit” the amendment to the voters for approval at

an election if the petition contains sufficient signatures. TEX. LOC. GOV’T

CODE § 9.004(a). Relator Dallas HERO, a grassroots organization,

spearheaded the collection of signatures for three petitions to amend the

charter of the City of Dallas. HERO collected over 169,000 voter

signatures on all three petitions combined, and relator Cathy Arvizu

signed each of the three petitions.

On July 19, 2024, the city secretary certified that each of HERO’s

petitions included sufficient voter signatures for the measures to qualify

for places on the ballot. The Dallas City Council held multiple meetings

in mid-August to discuss a special election on proposed amendments to

the charter, including the three citizen-initiated propositions and

2

several others that an appointed commission recommended to the

council as part of the City’s decennial charter review. See id.

(authorizing municipality’s governing body to submit proposed charter

amendments to voters). Although HERO initially threatened to

challenge the City’s chosen ballot language for the citizen-initiated

propositions, HERO ultimately negotiated with the City and agreed in

writing to specific language for those propositions, which the council

designated as Propositions S, T, and U. The council passed an ordinance

ordering a special election on those propositions, among others, at its

meeting on August 14, 2024. See Dallas, Tex., Ordinance 32798 § 2

(Aug. 14, 2024).

During these meetings, several council members expressed their

disapproval of the citizen-initiated propositions, and three council

members moved to amend the ordinance to include three more proposed

charter amendments. The motion passed, and these council-initiated

propositions were designated Propositions K, M, and N. In total, the

city council voted to submit more than a dozen proposed charter

amendments to the voters in an election to be held November 5, 2024.

On August 21, relators sought emergency mandamus relief

simultaneously in the Fifth Court of Appeals 1 and this Court against the

City of Dallas and several of its elected officials in their official capacities

1 The court of appeals denied the petition for writ of mandamus on

August 23, 2024, concluding that relators failed to comply with Texas Rule of

Appellate Procedure 52 in that court and, in any event, failed to demonstrate

their entitlement to mandamus relief. In re Dallas HERO, No. 05-24-01000-

CV, ___ S.W.3d ___, 2024 WL 3912358, at *1 (Tex. App.—Dallas Aug. 23, 2024).

The court also denied a related emergency motion as moot. Id. at *2.

3

(collectively, the City). 2 The City has filed a response to the mandamus

petition in this Court, which we consider along with the parties’ court of

appeals filings. 3

ANALYSIS

The petition for writ of mandamus raises four issues: whether

Propositions K, M, and N (1) fail to satisfy the Dacus standard,

(2) violate the single-issue rule, and (3) violate the home-rule provisions

of the Texas Constitution; and (4) whether the agreed-upon ballot

language for Propositions S, T, and U should be amended. We conclude

that relators are entitled to relief in part.

2 The real parties in interest are: (1) Chad West, in his official capacity

as City of Dallas Council Member, District 1; (2) Jesse Moreno, in his official

capacity as City of Dallas Council Member, District 2; (3) Zarin D. Gracey, in

his official capacity as City of Dallas Council Member, District 3; (4) Carolyn

King Arnold, in her official capacity as City of Dallas Council Member,

District 4; (5) Jaime Resendez, in his official capacity as City of Dallas Council

Member, District 5; (6) Omar Narvaez, in his official capacity as City of Dallas

Council Member, District 6; (7) Adam Bazaldua, in his official capacity as City

of Dallas Council Member, District 7; (8) Tennell Atkins, in his official capacity

as City of Dallas Council Member, District 8; (9) Paula Blackmon, in her

official capacity as City of Dallas Council Member, District 9; (10) Kathy

Stewart, in her official capacity as City of Dallas Council Member, District 10;

(11) Jaynie Schultz, in her official capacity as City of Dallas Council Member,

District 11; (12) Cara Mendelsohn, in her official capacity as City of Dallas

Council Member, District 12; (13) Gay Donnell Willis, in her official capacity

as City of Dallas Council Member, District 13; (14) Paul E. Ridley, in his official

capacity as City of Dallas Council Member, District 14; (15) Billierae Johnson,

in her official capacity as the City of Dallas City Secretary; and (16) Eric L.

Johnson, in his official capacity as Mayor of the City of Dallas.

3 See In re Williams, 470 S.W.3d 819, 821 (Tex. 2015) (collecting cases

in which “we have granted relief without requesting additional briefing—

especially in election cases—when time is critical, the issues are clear, and all

parties have had a chance to respond”).

4

As we explain below, Propositions K, M, and N “mislead the voters

by omitting certain chief features that reflect [their] character and

purpose” in violation of Dacus, 466 S.W.3d at 826, and removing those

propositions from the ballot would remedy the redundancy and

confusion they create without delaying the election. We therefore hold

that Propositions K, M, and N may not be included on the ballot as

presently written. Because our holdings on the first issue support the

relief requested, we do not reach relators’ second or third issues.

Regarding Propositions S, T, and U, we hold relators are estopped

from challenging the ballot language they agreed the City would use.

See Robinson v. Plano Bd. of Educ., 514 S.W.2d 135, 137-38 (Tex. Civ.

App.—Dallas 1974, no writ). We thus deny relief on their fourth issue.

I. Governing law

This Court has jurisdiction to issue a writ of mandamus

“compel[ling] the performance of any duty imposed by law in connection

with the holding of an election.” TEX. ELEC. CODE § 273.061.

“[M]andamus relief is appropriate if the relator establishes a clear abuse

of discretion for which there is no adequate appellate remedy.” In re

Durnin, 619 S.W.3d 250, 252 (Tex. 2021) (citing In re AutoNation, Inc.,

228 S.W.3d 663, 667 (Tex. 2007)). A clear abuse of discretion is shown

where the law has imposed a ministerial duty on a city “regarding the

handling of a referendum petition and any resulting election,” but the

city does not comply with that law. In re Williams, 470 S.W.3d 819, 820

(Tex. 2015). “If a ballot can be corrected prior to the election, a post-

election contest is an inadequate remedy for mandamus purposes.”

Durnin, 619 S.W.3d at 252.

5

“Home-rule cities possess the full power of local self-government,

which includes the power of qualified voters in the city to adopt or

amend a city charter by majority vote.” Hotze v. Turner, 672 S.W.3d

380, 385 (Tex. 2023) (internal quotation marks and footnote omitted).

This direct legislative power “is the exercise by the people of a power

reserved to them, and not the exercise of a right granted.” 4 Because we

presume that voters are familiar with every measure on the ballot, 5

proposed amendments to a city charter need not be reprinted verbatim

on the ballot. Dacus, 466 S.W.3d at 826. “Except as otherwise provided

by law, the authority ordering the election shall prescribe the wording

of a proposition that is to appear on the ballot.” TEX. ELEC. CODE

§ 52.072(a). 6 Cities thus have broad discretion in crafting the ballot

language describing propositions. Dacus, 466 S.W.3d at 826.

But this discretion has important statutory and common-law

limits. For example, a proposition “may appear on the ballot only once.”

TEX. ELEC. CODE § 52.072(b). And “[t]he common law protects the

4 Taxpayers Ass’n of Harris County v. City of Houston, 105 S.W.2d 655,

657 (Tex. 1937); see TEX. CONST. art. I, § 2 (“All political power is inherent in

the people[.]”). The ballot-initiative process has deep historical roots in Texas’s

legislative tradition, stemming from “the people’s dissatisfaction with

officialdom’s refusal to enact laws.” Coalson v. City Council of Victoria, 610

S.W.2d 744, 747 (Tex. 1980).

5 Election notices, including “a substantial copy of the proposed

amendment” to the city charter, must be published in the newspaper before

the election. TEX. LOC. GOV’T CODE § 9.004(c)(1).

6 Under the Election Code, a “proposition” is “the wording appearing on

a ballot to identify a measure,” TEX. ELEC. CODE § 1.005(15), which is “a

question or proposal submitted in an election for an expression of the voters’

will,” id. § 1.005(12).

6

integrity of the election with a minimum standard for the ballot

language.” Dacus, 466 S.W.3d at 823. In addition to allowing voters “to

identify and distinguish the different propositions from each other,” id.

at 824, “[t]he language appearing on the ballot, regardless of its source,

must ‘substantially submit[] the question . . . with such definiteness and

certainty that the voters are not misled.’” In re Petricek, 629 S.W.3d 913,

919 (Tex. 2021) (quoting Dacus, 466 S.W.3d at 823); see also Reynolds

Land & Cattle Co. v. McCabe, 12 S.W. 165, 165 (Tex. 1888).

In Dacus, we recognized two ways in which a proposition’s ballot

language may mislead voters. “First, it may affirmatively misrepresent

the measure’s character and purpose or its chief features. Second, it

may mislead the voters by omitting certain chief features that reflect its

character and purpose.” Dacus, 466 S.W.3d at 826. Relators argue that

the ballot language for the council-initiated propositions is misleading

in this second way.

II. At least one relator has standing.

Before reaching the dispositive issue, we must confront the City’s

argument that relators lack standing to challenge the language of the

council-initiated propositions. In Blum v. Lanier, we recognized that

“[c]itizens who exercise their rights under initiative provisions act as

and ‘become in fact the legislative branch of the municipal government.’”

997 S.W.2d 259, 262 (Tex. 1999) (quoting Glass v. Smith, 244 S.W.2d

645, 649 (Tex. 1951)). “[A]s sponsors of the initiative, [these signers]

have a justiciable interest in seeing that their legislation is submitted to

the people for a vote” without misleading language. Id. Thus, “a

qualified voter . . . has standing to seek . . . an injunction forbidding the

7

City’s use of a misleading ballot proposition so long as the injunction

does not operate to delay or cancel the called election.” Id. at 264. This

“reasoning also applies to [pre-election] mandamus proceedings.”

Williams, 470 S.W.3d at 821; see also Petricek, 629 S.W.3d at 917

(“Voters who sign an initiative petition have standing to seek mandamus

relief against the city council if they can establish the elements for such

relief.”).

As we have explained, the signers’ “interest in the valid execution

of the charter amendment election” constitutes “an interest in that

election distinct from that of the general public.” Blum, 997 S.W.2d at

262 (emphasis added). Some courts have also recognized that a

particularity of interest in the electoral process may stem from activities

like “organiz[ing] and financ[ing] a petition drive to get [a proposition]

on the ballot.” In re Robinson, 175 S.W.3d 824, 828 (Tex. App.—Houston

[1st Dist.] 2005, no pet.).

Here, relator Arvizu signed the petitions for the citizen-initiated

propositions, so she has an interest in the electoral process sufficient to

confer standing. Blum, 997 S.W.2d at 262-64. The parties do not

extensively address relator HERO’s standing either as an entity that

organized the petition drive or through the doctrine of associational

standing. See generally Abbott v. Mexican Am. Legis. Caucus, 647

S.W.3d 681, 690-95 (Tex. 2022). But we need not determine HERO’s

standing, as Arvizu’s is sufficient to pursue the full extent of relators’

requested relief. See Heckman v. Williamson County, 369 S.W.3d 137,

152 n.64 (Tex. 2012).

8

The City argues, however, that relators’ standing extends only to

challenges regarding the electoral process for the citizen-initiated

propositions and does not extend to the council-initiated propositions.

Under the particular circumstances of this case, we disagree. The injury

relators assert is that the ballot language for the council-initiated

propositions misleads voters because it omits the effect those

propositions would have on the propositions for which Arvizu signed

petitions. Consistent with our cases described above, we conclude the

same particularity of interest arises when there is a colorable basis for

arguing that another proposition on the same ballot would have the

effect of negating a proposition the voter signed. The invasion of that

interest is no less distinct or particularized when allegedly misleading

or confusing ballot language is located in a separate proposition that

otherwise duplicates the same substantive measure.

III. The ballot language for Propositions K, M, and N is

misleading.

Turning to the merits of relators’ challenge, we conclude that the

ballot language for council-initiated Propositions K, M, and N violates

the Dacus standard’s second prong by “omitting certain chief features

that reflect [their] character and purpose.” 466 S.W.3d at 826. 7 When

various propositions on the ballot interact with each other, the ballot

should be considered as a whole in applying this standard. 8

7 In conducting this analysis, we have no occasion to consider the

validity or effect of any proposition if approved by the voters.

8 See City of McAllen v. McAllen Police Officers Union, 221 S.W.3d 885,

895 (Tex. App.—Corpus Christi–Edinburg 2007, pet. denied) (upholding trial

9

Although Section 9.005 of the Texas Local Government Code

“requires a City Council to adopt and give effect to a citizen-initiated

amendment that the voters approve by a majority vote, a dilemma arises

when two approved amendments conflict.” Hotze, 356 S.W.3d at 389-

390. Here, each of the three council-initiated propositions will, if

approved by voters, create precisely that dilemma for one or more of the

citizen-initiated propositions. Yet the ballot language does not

acknowledge the conflicting character of these propositions so that

voters can attempt to avoid the dilemma by casting consistent votes.

Nor does it inform voters of conflict provisions the council included in its

propositions for the purpose of resolving this dilemma in favor of the

council-initiated propositions.

Specifically, the agreed-upon ballot language describing citizen-

initiated Proposition S reads:

Shall the Dallas City Charter be amended by adding a new

chapter that grants standing to any resident of Dallas to

bring a lawsuit against the city to require the city to comply

with provisions of the city charter, city ordinances, and

state law; entitles claimants to seek declaratory and

injunctive relief against the city and recover costs and

reasonable attorney’s fees; and waives the city’s

governmental immunity from suit and liability in claims

brought under this amendment?

(emphases added). Compare the ballot language describing council-

initiated Proposition M, which reads:

court’s conclusion that “the propositions [as described on the ballot] were

misleading, both in and of themselves, and in juxtaposition to one another”),

disapproved of in part on other grounds by Dacus, 466 S.W.3d at 826 & n.12.

10

Shall Chapter II of the Dallas City Charter be amended by

adding a section that states that Nothing in the City

charter is intended to waive the City’s governmental

immunity from suit liability or damages, or to grant

standing to residents to sue the city, its employees or

officials?

(emphases added).

Our review of the proposed charter amendments’ text confirms

that Proposition M would prevent anything in the charter from granting

residents standing to sue the City or waiving the City’s governmental

immunity from suit and liability. Conversely, Proposition S would

(among other things) amend the charter to grant residents standing to

sue the City and waive the City’s governmental immunity from suit and

liability. In other words, a chief feature of Proposition M is that it

conflicts entirely with Proposition S—or, as the City puts it, the

propositions use “identical terms” to “advance different outcomes.” Yet

the ballot language for Proposition M—which appears first on the

ballot—does not acknowledge this characteristic. 9

Furthermore, the Proposition M charter amendment includes a

primacy provision that reads: “If there is a conflict between this

provision and another provision of this charter, this provision controls.”

As this language reveals, a purpose of the Proposition M amendment is

9 We do not hold that the ballot description for any charter amendment

that clarifies or contradicts other existing or proposed parts of the city charter

to any degree must flag that inconsistency to comply with Dacus. Instead, we

hold that when the conflict between a proposed charter amendment and

another proposition on the same ballot is substantial enough to be considered

a chief feature, as it unquestionably is here, failing to identify that conflict in

the ballot language is misleading.

11

to negate the Proposition S amendment, which appears on the same

ballot. 10 But the ballot language for Proposition M omits this chief

feature that reflects its purpose. For these reasons, we hold the

Proposition M ballot language is misleading.

Similar conflicts arise from the juxtaposition of the other two

pairs of propositions. The agreed-upon ballot language describing

citizen-initiated Proposition T reads:

Shall Chapter VI of the Dallas City Charter be amended by

adding a new section compelling the city to conduct the

city-commissioned Community Survey on an annual basis,

to be completed by a minimum of 1,400 Dallas residents on

their satisfaction on quality of life issues, the results of

which will result in the city manager earning additional

performance compensation (between 0 percent and 100

percent of the city manager’s annual base salary) or the

termination of the city manager?

(emphases added). The ballot language describing council-initiated

Proposition N reads:

The appointment-removal and compensation-language in

chapter six, section one of the city charter controls over any

other conflicting provision of the city charter.

10 We express no view regarding how a court would reconcile the

conflicting propositions discussed in this section of our opinion—including

what effect the primacy provisions of the council-initiated propositions would

have—if all were approved as currently drafted. See Hotze, 672 S.W.3d at 389-

390. Under the Dacus standard, it is sufficient to conclude that the ballot

description is also misleading because it does not identify that a purpose of the

council-initiated propositions—as shown by their primacy provisions—is to

negate the citizen-initiated propositions. We also observe that if it is an open

question how a court would resolve the conflicts, it is unlikely that voters will

understand how to cast their votes to do so without providing any guidance in

the ballot language.

12

(emphasis added). 11

In addition to establishing an annual survey of city residents,

Proposition T ties the city manager’s compensation and removal from

office to the results of that survey. But Proposition N adds a primacy

provision that seeks to nullify any alternative means for determining

compensation and removal, which would include Proposition T’s new

survey. Because the ballot language for Proposition N does not state the

chief feature that it proposes to nullify Proposition T, which is part of

the same ballot, we hold that Proposition N’s ballot language is

misleading.

Finally, both Propositions K and U seek to affect appropriations

for public-safety officers in direct conflict with one another. The agreed-

upon ballot language describing citizen-initiated Proposition U reads:

Shall Chapter XI of the Dallas City Charter be amended by

adding a new section compelling city council to appropriate

no less than 50 percent of annual revenue that exceeds the

total annual revenue of the previous year to fund the

Dallas Police and Fire Pension, with any monies remaining

of that 50% to be appropriated to increasing the starting

compensation of officers of the Dallas Police Department

and to increase the number of police officers to a minimum

of 4,000, and to maintain that ratio of officers to the City of

Dallas population as of the date of passage of this

amendment?

(emphases added). The language describing council-initiated

Proposition K reads:

11 Chapter Six, Section One of the charter gives the city council

authority over the appointment, compensation, and removal of the city

manager. See Dallas, Tex., City Charter ch. VI, § 1.

13

Shall Chapter XI, Section 3 and Chapter XXIV, Section 18

of the Dallas City Charter be amended to state that city

council shall make the final determinations regarding

appropriations of City funds. And that any instructions in

the charter regarding the appropriation of City funds for

City employee wages serve only as recommendations for

City Council’s consideration and that these provisions

control over other provisions in the City Charter?

(emphases added).

Proposition U seeks to bind the City to appropriate money for

public safety, dedicating certain percentages of annual revenue to wages

and other compensation for police officers and firefighters. But

Proposition K—as well as Proposition N discussed above—seeks to

protect the City from being bound by any such requirements, demoting

them to mere recommendations the council may follow and giving it

complete discretion over appropriations. Because Proposition K’s ballot

language does not state the chief feature that it proposes to nullify

Proposition U, which appears on the same ballot, we hold that the ballot

language of Proposition K is misleading.

In sum, the ballot descriptions for council-initiated

Propositions K, M, and N do not “identify [each] measure for what it

is” 12—a proposed charter amendment that conflicts with and proposes

to nullify a citizen-initiated charter amendment on the same ballot.

Simultaneously holding an election on contradictory propositions with

which the City cannot comply is confusing, and ballot language that fails

to address that contradiction or how it will be resolved does not

12 Dacus, 466 S.W.3d at 826.

14

“substantially submit[] the question . . . with such definiteness and

certainty that voters are not misled.” Petricek, 629 S.W.3d at 919.

IV. The appropriate remedy is to remove duplicative

Propositions K, M, and N from the ballot.

The parties disagree regarding how to remedy a Dacus violation

of this nature. Relators contend that Propositions K, M, and N should

be removed from the ballot, while the City contends that all this Court

can do before the election is order it to correct the ballot descriptions so

they are not misleading. Because Propositions K, M, and N are the

converse of other propositions on the ballot, which is the source of the

confusion, we conclude that removal of the duplicative propositions in

their current form is the appropriate remedy.

Given the separation of legislative and judicial powers, we have

refrained from enjoining elections before they are held. In re Morris,

663 S.W.3d 589, 593 (Tex. 2023). But we have exercised our statutory

and common-law authority to grant pre-election mandamus or

injunctive relief “forbidding the City’s use of a misleading ballot

proposition so long as the [relief] does not operate to delay or cancel the

called election.” Blum, 997 S.W.2d at 264; see also Williams, 470 S.W.3d

at 821.

Here, directing the City to remove current Propositions K, M, and

N does not interfere with or delay the upcoming election. Instead, it

recognizes that a city may not confuse its voters by submitting the

converse of citizen-initiated propositions that must appear on the ballot.

“[C]omplying with incompatible charter amendments” is an “impossible

task.” Hotze, 672 S.W.3d at 390. And a city has no discretion to submit

15

a proposition more than once. See TEX. ELEC. CODE § 52.072(b) (“A

proposition . . . may appear on the ballot only once.”).

As we have explained, the text of each challenged council-

initiated proposition demonstrates that its purpose is to nullify a citizen-

initiated proposition. In this circumstance, only one question is really

being asked of voters, so directing that it cannot be asked a second time

in a different way is not canceling an election. Rather, that remedy

“correct[s] deficiencies in the ballot” and “facilitates the elective

process,” Blum, 997 S.W.2d at 263, by ensuring that voters understand

what they are being asked and that votes in favor of one proposition do

not negate votes in favor of another. 13

CONCLUSION

For the foregoing reasons, and without hearing oral argument, see

TEX. R. APP. P. 52.8(c), we conditionally grant mandamus relief in part

and direct the City to remove Propositions K, M, and N from the ballot.

We are confident the City will comply, and the writ will issue only if it

does not.

J. Brett Busby

Justice

OPINION DELIVERED: September 11, 2024

13 To the extent the City wishes its council-initiated propositions to have

broader application, nothing in this opinion prevents the City from rewording

them in a manner that does not contradict the citizen-initiated propositions.

16