The City of Dallas v. the Employees' Retirement Fund of the City of Dallas

Supreme Court of Texas

══════════

No. 22-0102

══════════

The City of Dallas,

Petitioner,

v.

The Employees’ Retirement Fund of the City of Dallas,

Respondent

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

On Petition for Review from the

Court of Appeals for the Fifth District of Texas

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

Argued October 4, 2023

JUSTICE YOUNG delivered the opinion of the Court.

We must decide whether a city ordinance may bestow on a third

party the perpetual right to “veto” categories of future lawmaking. We

hold that such an alienation of lawmaking authority is impermissible.

The court of appeals, by contrast, relied on principles of trust law to

reach the opposite conclusion, holding that the City of Dallas cannot

amend Chapter 40A of its own code of ordinances unless the board of

trustees of the Employees’ Retirement Fund agrees to the amendment.

We reverse the court of appeals’ judgment.

I

In 1935, the legislature authorized larger Texas cities “to

formulate and devise a pension plan for the benefit of all employees in

the employment of such city.” Tex. Rev. Civ. Stat. art. 6243d, § 1. A

city’s “governing body” would develop such a plan, which “shall be

submitted in ordinance form by said governing body to the qualified

electors of such city” and “be approved by said qualified electors at an

election duly held.” Id.

Eight years later, the Dallas City Council announced that “there is

hereby established, subject to the approval of the electorate of the City at

an election to be called for that purpose, ‘The Employees’ Retirement Fund

of the City of Dallas.’” Dallas, Tex., Ordinance No. 3470 (Nov. 24, 1943).

The voters so approved and the Fund came into existence. The ordinance

creating and governing it is codified, as amended, as Chapter 40A of the

Dallas City Code.

Chapter 40A describes the Fund as a “trust fund” and a “public

entity” that is “established for the exclusive purpose of providing benefits

to members and their beneficiaries.” Dallas, Tex., Code of Ordinances

§ 40A-2(a)–(b).1 A seven-member board of trustees administers the Fund:

three city-council appointees, three City employees “who are elected by

members of the retirement fund and who are members of the retirement

fund,” and the city auditor. § 40A-2(c)(1). As of 2018, the Fund’s assets

exceeded $3.6 billion, benefiting more than 16,000 families.

The 1943 ordinance creating the Fund expressly provided that the

1 Citations in this opinion reference the Dallas Code of Ordinances unless

otherwise indicated.

2

city council could make unilateral amendments until 1945. See infra

note 16 and accompanying text. The city council approved at least one

amendment during that period. See Dallas, Tex., Ordinance No. 3577

(Oct. 26, 1944). Since 1945, the Fund says, Chapter 40A has mandated

that any amendments to Chapter 40A be placed on the ballot. We assume

that representation to be true.2 The record reflects, at least, that by 1977,

Chapter 40A incorporated the following amendment procedure:

This chapter may not be amended except by ordinance

adopted by the city council and approved by a majority of

the voters voting at a general or special election.

Dallas, Tex., Ordinance No. 15414 (Feb. 7, 1977) (then codified as § 40A-34;

recodified as amended as § 40A-35(a)).

In 1991, however, a city ordinance granted authority to the Fund’s

board that both parties here have described as a “veto” power. It did so

by adding these underlined words:

This chapter may not be amended except by ordinance

recommended by the board, adopted by the city council and

approved by a majority of the voters voting at a general or

special election.

Dallas, Tex., Ordinance No. 20960 (June 12, 1991) (then codified as

§ 40A-35; recodified as amended as § 40A-35(a)).3

In 2004, another amendment (to which the board unsurprisingly

2 See infra note 17 and accompanying text.

3 Two years later, the provision was renumbered as § 40A-35(a),

accompanied by a new § 40A-35(b) that is not at issue here but that (for

reference) concerned amendments “determined by the board” as “necessary to

comply with federal law.” Dallas, Tex., Ordinance No. 21582 (Feb. 24, 1993).

At the same time, subsection (a) was amended with the following underlined

language added and the stricken character deleted: “Except as provided in

Subsection (b) of this section, Tthis chapter may not be amended . . . .” Id.

3

consented) further strengthened the board’s veto power. The following

underlined language was added and the stricken language was deleted,

modifying § 40A-35(a) from its immediately preceding form:

Except as provided in Subsection (b) of this section, this

chapter may not be amended except by a proposal initiated

by either the board or the city council that results in an

ordinance approved recommended by the board, adopted by

the city council, and approved by a majority of the voters

voting at a general or special election.

Dallas, Tex., Ordinance No. 25695 (Aug. 11, 2004) (codified as amended

as § 40A-35(a)). The Fund contends and the City does not dispute that,

from 1991 onward, all amendments to Chapter 40A were adopted in

accordance with § 40A-35(a)—that Chapter 40A was never amended

without the board’s consent.

But then came the City’s desire to impose term limits on members

of city boards. The city council could unilaterally impose limits on every

board except (because of § 40A-35(a)) the Fund’s. The City passed a

general “board member” term-limit provision in 1994, when “member”

was defined as “a duly appointed member of a board.” See Dallas, Tex.,

Ordinance No. 22259 (Nov. 9, 1994) (codified as amended as § 8-1.5(a)).

The Fund did not object to term limits for board members “appointed” by

the city council. Its views were quite different as to the elected board

members, who declined to observe the term-limit provision, thus creating

an apparent “loophole.”

In 2017, without securing board approval, the city council amended

Chapter 8 of the City Code once more, this time expressly reaching the

Fund’s elected board members:

A person who has served on the board of the employees’

retirement fund pursuant to Section 40A-3(a)(1) of this code,

4

as amended, for three consecutive terms, of whatever length

of time, will not again be eligible to serve on that same board

until at least one term has elapsed, whether service was as

a member, chair, or other position on the board.

Dallas, Tex., Ordinance No. 30555 (Aug. 9, 2017) (codified as amended as

§ 8-1.5(a-1)). Included with this amendment was a definitional change

to the word “member,” which now included “a duly appointed or elected

member of a board.” Id. (codified as amended as § 8-1(8)).

The City then informed the Fund that § 8-1.5(a-1) rendered all

three elected board members ineligible for reelection. The City notified

the two members whose terms were set to expire at the end of 2018 of

their ineligibility and threatened to sue them if they sought another term.

The Fund disagreed with the City’s position. Invoking its authority

to interpret Chapter 40A, see § 40A-4(a)(18), (h),4 the Fund adopted

Resolution No. 2018-1 regarding board term limits. The Fund resolved

that: (1) Chapter 40A imposes no term limits on its elected board

4 See § 40A-4(a)(18) (“In addition to other powers and duties it may have

under state or federal law, the board shall have the power and duty to . . .

interpret this chapter as necessary to resolve any problems created by any

ambiguities, inconsistencies, or omissions that might be found in this

chapter . . . .”), (h) (“If the board, in good faith, is in doubt as to the construction

or interpretation of any provision of this chapter, or has any other question that

may arise during the administration of the retirement fund, the board may

resolve all such doubts and questions without obtaining a judicial construction.

All constructions and interpretations made by the board are binding and

conclusive.”).

We express no view regarding the general meaning or enforceability of a

delegation of interpretive authority of this sort and do not rely on it or defer to

the board when we conclude, as the board did, that § 8-1.5(a-1) operates as an

amendment to Chapter 40A. To the extent that the board’s authority to interpret

a city ordinance is invoked as a basis to impede the city council’s authority to

revise the law, however, our decision today necessarily rejects that contention.

5

members and (2) imposing such limits “would constitute an amendment

to Chapter 40A that would be required to comply with the procedure set

forth under Chapter 40A-35(a).” The Fund therefore contended that

§ 8-1.5(a-1) was ineffective. Accordingly, the two purportedly ineligible

elected board members ran for and won reelection to new terms.

The Fund sued the City and the City brought counterclaims. Both

sides sought declaratory relief regarding § 8-1.5(a-1)’s validity and

enforceability in light of § 40A-35(a). The parties cross-moved for

summary judgment on that issue and the trial court rendered judgment

for the City.

The court of appeals reversed and rendered judgment for the Fund.

636 S.W.3d 692 (Tex. App.—Dallas 2021). According to that court,

Chapter 40A was a codified “Trust Document” that may not be amended

except as that document provides. Id. at 694. Citing U.S. Term Limits,

Inc. v. Thornton, 514 U.S. 779 (1995), and Burroughs v. Lyles, 181 S.W.2d

570 (Tex. 1944), the court reasoned that imposing term limits on the

elected board members “added a substantive qualification for office” that

improperly “effected a fundamental change to the Trust Document.” Id.

at 697. The “City’s attempt to impose term limits on the elected members

of the Fund’s board by amending Chapter 8” was “invalid,” it explained,

so § 8-1.5(a-1) was “void and unenforceable.” Id. at 698.

We granted the City’s petition for review. The City defends

§ 8-1.5(a-1) as valid because that provision does not amend Chapter 40A.

Under § 40A-35(a), the board may veto only amendments to Chapter 40A,

not Chapter 8, so the City argues that the board had nothing to veto. But

even if § 8-1.5(a-1) does amend Chapter 40A, the City claims that it did

6

not need board consent because § 40A-35(a)’s veto power is unenforceable.

Otherwise, the City contends, § 40A-35(a) would unconstitutionally

delegate lawmaking authority to the Fund’s board.

The Fund responds that § 8-1.5(a-1), despite its location in the

Code, plainly amends Chapter 40A. The Fund also defends Chapter 40A

as invulnerable to that or any other revision absent compliance with

§ 40A-35(a)’s procedure, which requires the board’s consent. This result,

the Fund says, is simply the natural consequence of Chapter 40A’s status

as a “trust document.” The Texas Trust Code thus shields Chapter 40A

from any unwelcome amendments, the Fund claims.

II

Under the Texas Constitution, cities may “adopt or amend their

charters,” provided that “no charter or any ordinance passed under said

charter shall contain any provision inconsistent with the Constitution of

the State, or of the general laws enacted by the Legislature of this State.”

Tex. Const. art. XI, § 5(a). Whatever power the City exercises is delegated

by the State through our Constitution. As is typical, the City of Dallas

exercises this delegated power through its city council. See Dallas City

Charter, ch. III, § 1. The city council, in turn, adopted § 8-1.5(a-1), the

ordinance that the Fund has challenged. We conclude that the consent

of the Fund’s board has no bearing on whether the ordinance is valid and

enforceable.

A

The City regards the case as easy because § 40A-35(a) purports to

give the board a veto only over amendments to Chapter 40A, and

§ 8-1.5(a-1) amends Chapter 8, not Chapter 40A. According to the City,

7

we therefore can reverse and render judgment solely on this ground.

We reject this argument. Section 8-1.5(a-1) expressly references

Chapter 40A and creates a term limit specifically applicable to the board

of trustees under that chapter. The ordinance that imposes term limits

on the board’s members necessarily amends Chapter 40A’s governance

of the board’s composition. Whether the City chooses to codify this

amendment within Chapter 40A or Chapter 8 or anywhere else within

its code of ordinances is immaterial.

Moreover, § 8-1.5(a-1)’s amendment of Chapter 40A goes beyond

adopting previously absent term limits. It also amends Chapter 40A

because it acts as a partial repeal by implication of § 40A-35(a). See

Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of

Legal Texts 278 (2012) (repealability canon). This Court long ago

explained that “the rule is well settled, that though the law does not favor

repeals by implication, yet a subsequent statute, revising the subject

matter of a former one, and intended as a substitute for it, although it

contains no express words to that effect, will operate a repeal of the

former, to the extent to which its provisions are supplied or repealed.”

Stirman v. State, 21 Tex. 734, 736 (1858); see also, e.g., Gordon v. Lake,

356 S.W.2d 138, 139 (Tex. 1962); Cole v. State, 170 S.W. 1036, 1037 (Tex.

1914). “The law makes no distinction between express and implied

repeals,” so courts are not “authorized to give an effect to one different

from that attached to the other.” Stirman, 21 Tex. at 736. Once an

implied repeal is established, in other words, its consequence is the same

as if the statute had expressly provided for that result. These principles

fully apply here because we “construe municipal ordinances the same way

8

we construe statutes.” Powell v. City of Houston, 628 S.W.3d 838, 842

(Tex. 2021).

Under § 40A-35(a), the City must follow the multi-step amendment

procedure—which includes obtaining the board’s consent—to modify even

comparatively modest matters, like how many terms board members may

serve. Unless it consents, the board is deemed to veto an amendment

such that it could never become part of the City’s law at all. Enacting

§ 8-1.5(a-1), in other words, implies rejecting § 40A-35(a)’s procedure and

replacing it, at least for the term-limit purposes of § 8-1.5(a-1), with the

traditional method of amending ordinances.5

Courts do not readily find implied repeals. If we can reasonably

harmonize two seemingly inconsistent enactments of the same level of

authority—like two constitutional provisions, two statutes, or two

ordinances—we will do so. But “as a matter of statutory construction, if

statutes are irreconcilable, the statute latest in date of enactment

prevails.” Jackson v. State Off. of Admin. Hearings, 351 S.W.3d 290, 297

5 Legislative procedural rules—even codified ones—are not immune to

implied repeals, as multiple supreme courts across the country have observed:

“It is generally recognized that a ‘council may abolish, suspend, modify or waive

its own rules. This also may be done by implication, when action is had not in

accordance therewith.’ ” Smith v. City of Dubuque, 376 N.W.2d 602, 605 (Iowa

1985) (quoting 4 E. McQuillan, The Law of Municipal Corporations § 13.42,

at 749–50 (3d ed. 1985)). See also, e.g., State ex rel. La Follette v. Stitt, 338

N.W.2d 684, 687 (Wis. 1983) (stating that “the failure to follow such procedural

rules amounts to an implied ad hoc repeal of such rules”); Patterson v. Dempsey,

207 A.2d 739, 745 (Conn. 1965) (explaining how the “action” of passing a law

can impliedly repeal the “prohibitory part” of another law). We agree that this

principle necessarily flows from the bedrock rule that a legislative body cannot

bind its successor, including by erecting procedural hurdles. Only if a higher

source of law mandates the procedural hurdle would a legislature be bound to

follow it in exercising its lawmaking function. See infra note 11.

9

(Tex. 2011) (internal quotation marks omitted).

This principle is, in essence, a choice-of-law rule that requires

courts to apply a later-enacted provision that clearly contradicts a prior

one—which is another way to describe an implied repeal. See Scalia &

Garner, supra, at 279, 327.6 Because § 8-1.5(a-1) is inherently

incompatible and inconsistent with § 40A-35(a)’s procedure, § 8-1.5(a-1)

necessarily repealed § 40A-35(a)’s board-veto provision. We therefore

cannot avoid the parties’ dispute about whether, unlike other ordinances,

the board-veto provision was beyond the city council’s authority to modify,

supersede, or repeal.

B

To justify denying this legislative authority to the city council, the

Fund and the court of appeals point to trust law and the Texas Trust

Code (which is codified within the Texas Property Code, see §§ 111.001–

117.012). Even assuming for argument’s sake that the Trust Code governs

the Fund,7 we disagree that trust law affects the city council’s authority

to make city law.

The specific statute that created the Fund, article 6243d, expressly

6 The Code Construction Act expresses the legislature’s understanding of

this principle. Tex. Gov’t Code § 311.025(a) (“[I]f statutes enacted at the same

or different sessions of the legislature are irreconcilable, the statute latest in

date of enactment prevails.”).

7 The parties apparently agree that Chapter 40A is a trust or pension

trust “document” governed by the Trust Code. Chapter 40A itself refers to the

Fund as a “trust fund” “entity” instead of a “trust.” § 40A-2(a)–(b); cf. Ditta v.

Conte, 298 S.W.3d 187, 191 (Tex. 2009) (“A trust is not a legal entity; rather it

is a fiduciary relationship with respect to property.” (internal quotation marks

omitted)). In any event, as we explain, we need not resolve whether or to what

extent the Trust Code applies to the Fund or others like it and thus expressly

reserve that question as open in this Court.

10

and unambiguously insisted that every provision in Chapter 40A be first

and foremost a codified city “ordinance”—not part of a city charter, not a

distinct stand-alone “trust” (a word that does not even appear in the

statute), or anything else. Instead, the “ordinance” adopting the plan

“shall be so worded as to authorize the governing body of such city or

town” to select one of two stated means to fund it (annual appropriations

from “the general revenue” to “carry out” the plan or through “a general

ad valorem tax sufficient to provide for” it). Rev. Civ. Stat. art. 6243d,

§ 1. Chapter 40A is part of the code of ordinances because the statute

demanded that it be an ordinance.

Chapter 40A’s status as an ordinance entails the consequence

that § 8-1.5(a-1) is one ordinance amending another, not an ordinance

purporting to amend something that exists apart from city law or

something that has higher legal status than an ordinance. U.S. Term

Limits, on which the court of appeals relied, involved a provision of the

Arkansas Constitution that, the Supreme Court held, amounted to

modifying the minimum electoral requirements that the U.S.

Constitution prescribes for members of Congress. See 514 U.S. at 783.

A state constitution, of course, may not amend the federal Constitution.

Likewise, this Court observed in Burroughs that a Texas statute could

not expand the Texas Constitution’s eligibility requirements for election

as a state senator. See 181 S.W.2d at 574. These principles do not

govern a city council’s ability to amend a city ordinance.

Article 6243d’s statutory mandate to enact an “ordinance”

implicates at least two relevant constitutional principles. The first “is

traditionally known as the nonentrenchment doctrine,” Scalia & Garner,

11

supra, at 278,8 which is an ancient maxim that applies as fully here as to

lawmaking bodies at any level of government: “[O]ne Legislature cannot

bind the hands of a subsequent Legislature by the enactment of laws

which may not be altered or repealed by a subsequent Legislature.”

Jefferson County v. Bd. of County & Dist. Rd. Indebtedness, 182 S.W.2d

908, 915 (Tex. 1944). In short, a “legislature cannot prevent future

legislatures from amending or repealing a statute.” Cent. Power & Light

Co. v. Pub. Util. Comm’n of Tex., 649 S.W.2d 287, 289 (Tex. 1983).9

Like any legislative act, § 40A-35(a) was “alterable when the

legislature shall please to alter it.” Marbury v. Madison, 5 U.S. (1

Cranch) 137, 177 (1803). The city council that adopted the veto

8 “Resting as it does on sheer logic, the principle dates from time

immemorial. As Cicero wrote to Atticus: ‘When you repeal the law itself, . . . you

at the same time repeal the prohibitory clause, which guards against such

repeal.’ ” Scalia & Garner, supra, at 278 (quoting 1 William Blackstone,

Commentaries on the Laws of England § 3, at 90–91 (1765)).

9 The Dallas City Charter, which of course is superior to an ordinance,

reaffirms the city council’s legislative power to pass ordinances. E.g., Ch. XVIII,

§§ 3–4. Section 3 contemplates the city council (not the board or any other body)

having final legislative authority over ordinances, except where “state law or

this charter provides for a different procedure.” Having received its authority,

the city council could not share it with the board absent some higher authority

to do so. Section 4 provides that every ordinance “shall require on final passage

the affirmative vote of a majority of the members present unless more is required

by state law, this Charter, or ordinance.” (Emphasis added.) Under the last-

quoted word, the city council could by ordinance change the number of votes

needed for passing a future ordinance. We need not resolve whether an

ordinance that purported to require a larger-than-majority vote for a different

future ordinance (as opposed to the very ordinance at issue) would bind a future

city council because this case does not involve an effort to require the city council

to pass an ordinance by a super-majority. It instead addresses a purported

transfer of authority to an entirely different body; whether and to what extent

the city charter could allow that is also something we need not address, in part

because Section 3 of the charter already disclaims it.

12

amendment in 1991, in other words, could not bind the city council in

2017 from further amending Chapter 40A to impose term limits on the

elected board members. Granting the board the authority to veto future

amendments to Chapter 40A is a self-evident impairment of a future city

council’s ability to amend city law.10 The veto, if it truly bound a future

city council, would clearly violate the nonentrenchment doctrine.

A second principle is that, to whatever extent cities may legislate,

they do so by drawing upon the State’s authority that has been vested in

them. See Tex. Const. art. XI, § 5(a). It is not within the power of a Texas

city to alienate to some third party the governmental authority that the

State allows the city alone to exercise. This basic principle applies in

multiple contexts—governmental immunity, for example. As we stated

in a case involving another city, “[e]ven if a governmental unit would be

happy to waive ‘its’ immunity, it is not the governmental unit’s immunity

to waive.” Rattray v. City of Brownsville, 662 S.W.3d 860, 867 (Tex. 2023).

Ordinances have the force of law—the ability to bind those resident

in or who come into a city. The source of such power is not the city, but

the State. After all, a city is not a sovereign entity—not even close. Cities

are municipal corporations (which is why Article XI’s title is simply

“Municipal Corporations”). See City of Brenham v. Brenham Water Co.,

4 S.W. 143, 149 (Tex. 1887) (“It is now universally conceded that powers

are conferred on municipal corporations for public purposes; and, as their

powers cannot be delegated, so they cannot be bargained or bartered

10 Notably, the voters can limit the city council’s ability to amend certain

ordinances by following procedures that the city charter specifies. See Dallas

City Charter ch. XVIII, § 11. Nothing would prohibit the voters from taking this

step in the future.

13

away.” (internal quotation mark omitted)); Kirby Lake Dev., Ltd. v. Clear

Lake City Water Auth., 320 S.W.3d 829, 843 (Tex. 2010).

Like other corporations, municipal corporations depend on the

State for their existence. Before the 1912 adoption of the “home-rule

amendment” in Article XI, § 5, cities came into existence when the

legislature granted an individual charter—a tedious exercise that

exhausted considerable legislative time. See City of San Antonio v. City

of Boerne, 111 S.W.3d 22, 26 n.5 (Tex. 2003). What we call the home-rule

amendment streamlined the method of incorporating cities. But it

certainly did not purport to recognize authority in cities that in any sense

was independent of the State itself. To the contrary, Article XI, § 5 states

with marked clarity that “no charter or any ordinance passed under said

charter shall contain any provision inconsistent with the Constitution of

the State, or of the general laws enacted by the Legislature of this State.”

Tex. Const. art. XI, § 5(a).

A city’s authority over the law is not within a city’s power to give

away any more than the legislature could hand away its own power. Even

assuming that trust law or the Trust Code otherwise apply, that does not

change how ordinances are drafted and by whom, which is a matter of

constitutional import. State law may preempt the substantive reach of

city law, see, e.g., City of Laredo v. Laredo Merchs. Ass’n, 550 S.W.3d 586,

592–93 (Tex. 2018), just as federal law may preempt state law.11 But

11 Section 8-1.5(a-1) does not purport to divest Fund beneficiaries of any

vested pension rights, which would present a different question. See infra Part

II.C. Our opinion today addresses only the city council’s formal ability to amend

its own city code, and our holding that the city council retains that authority—

and never validly gave it up—does not mean, of course, that the substantive

content of any municipal enactment is immune to challenge by a proper plaintiff.

14

trust law no more stops a city from amending city law than it stops the

legislature from amending state law. Neither trust law generally nor the

Trust Code specifically purports to allow—much less require—cities to

grant authority to other entities over the content of city law. The board

never received that authority from the City because it was not the City’s

to give. Said differently, state law does not compel the City to retain the

arrangement that the Fund defends because state law never allowed the

City to bind itself to that arrangement in the first place. Until now, the

city council was willing to accommodate the board by not amending

Chapter 40A without the board’s consent. But that practice in no way

implies that the city council was legally bound to do so, even though it

may well have been prudent then—and in the future—for the city council

to closely involve the board before amending Chapter 40A.

The Fund is to be commended for its candor in arguing that trust

law entails the legal consequences that we reject. At oral argument, it

agreed that under its theory any number of unconstitutional alienations

of authority could be achieved simply by the artifice of creating a trust

with some rather tenuous link to the subject matter. It confirmed, for

example, that even our legislative process could be converted from a

bicameral into a tricameral system for broad swaths of our law. How?

Suppose that the legislature decided that future legislatures would likely

be benighted about higher-education policy. Imagine that it put the

property of state universities into a “trust” and, within a relevant “trust

document,” included the requirement that amendments to the law

governing those universities could not be presented to the governor for

his signature without the concurrence of the two houses of the legislature

15

and the University of Texas faculty senate. The Fund readily agreed that

such a result would be permissible and not even particularly troubling.

The contrary is true. Embracing the Fund’s approach would

represent an earthquake in our constitutional order. It would amount to

an escape hatch both from the principle that one legislature cannot bind

its successors and the principle that core lawmaking authority vested by

the State cannot be given away.12 The Fund’s candor and consistency are

helpful because they illustrate the underlying infirmity of the Fund’s

position. If trust law could provide so simple a way to “crack the code” of

the basic constitutional principles of self-government, it surely would

have been discovered long ago.13

12 The higher-education hypothetical differs from this case in one obvious

and significant respect: city councils cannot repeal state law (like trust law), but

the legislature can. For the Fund to contend that trust law could compel even

the legislature to honor a transmission of legislative power to the faculty senate,

the Fund still must (at minimum) reject the application of the implied-repeal

doctrine. Otherwise, the Fund could not deny a future legislature’s ability to

instantly revoke the faculty senate’s legislative power. The Fund’s theory also

requires it to reject the other core principle: that any entity vested with

constitutional authority to enact legislation cannot hand that power away. The

Fund’s view that trust law not only tolerates but mandates such a result triggers

the Fund’s ready acquiescence to farfetched hypotheticals. But—even if only as

a matter of constitutional avoidance—we cannot accept any view of trust law

that allows commandeering of the legislative process in this way.

13 The Fund sought to persuade us to deny the petition by arguing that

the City had not preserved any argument that § 40A-35(a) would be

unconstitutional if deemed binding. Notably, the Fund did not even mention

that position at oral argument. We believe that the Fund was right to abandon

(or at least not press) that position. We have reviewed the City’s filings in the

trial court and, while the City obviously did not seek a declaratory judgment

against itself that one of its own ordinances was unconstitutional, it repeatedly

emphasized that § 40A-35(a) could not constitutionally impede the city council

from imposing term limits. The argument is not waived.

16

C

At the same time, we pause to note that our decision today does not

imperil the interests that seemingly animate the Fund’s arguments. The

Fund invokes trust principles to vaguely, but grimly, warn that allowing

amendments to Chapter 40A without board approval would destabilize

the rights of employee participants in the Fund. World history confirms

that the threatened image of craven politicians freely raiding the Fund is

less far-fetched than one might wish. That possibility is why our law

takes those concerns seriously. Important as they are, however, those

concerns have nothing to do with the question before us, which is only

whether the City may lawfully give the board veto power over city law.

For purposes of this case, we can readily accept the Fund’s

underlying principle: that the city council cannot abridge the vested

rights of the Fund’s participants. But that is true regardless of the board’s

consent. Said differently, individual rights are not subject to violation

even if the board agrees. Various sources of Texas law protect those

rights—maybe the Texas Constitution’s contracts clause, probably Texas

contract law, perhaps trust law, presumably the provisions of our

Constitution that address public-employee retirement, and possibly more

still.14 Article 6243d required, from the moment of the Fund’s creation,

that it be “for the benefit of all employees in the employment of [the] city.”

Rev. Civ. Stat. art. 6243d, § 1 (emphasis added). Whatever else that

14 The Fund invokes Article 16, § 67 of the Texas Constitution for the

proposition that Chapter 40A is a trust. If anything, § 67 bolsters the point

made by this footnote’s accompanying text, as it provides: “The assets of a

system are held in trust for the benefit of members and may not be diverted.”

Paramount constitutional protections over such assets neither depend on a

board’s veto nor would be sacrificed upon a board’s consent.

17

provision might mean, it does not authorize the City to raid the Fund at

will (and the City certainly does not claim otherwise).

Such protection should alleviate any concerns about the ordinance

“just” being an ordinance—part of Dallas law that can be amended in the

normal course without a bespoke “veto” over legislative enactments. The

Fund’s assets are protected, and its participants’ own rights are protected

in many other ways. In myriad contexts, the Constitution and our laws

protect all sorts of rights—but they have never been understood to

prevent a legislative body from legislating in the first place or to authorize

some third party to veto legislation based on that party’s special concern

with a given right. A unique calcification of the legislative process in

Dallas is not necessary to protect the Fund’s legitimate interests and

would contravene the constitutional principle forbidding the city council

from giving away its authority to legislate.15

* * *

We hold that the board’s veto in § 40A-35(a) is unenforceable and

cannot prevent an otherwise-valid ordinance from taking effect.

III

We decline, however, to resolve whether the City must hold an

election that submits § 8-1.5(a-1) to the voters before it may enforce that

15 The Dallas City Council could choose to refrain from amending

Chapter 40A absent the board’s consent—just like any member of the city

council could choose to not vote for any given measure absent some external or

internal sign of approval. As we have noted, it may be wise to amend

Chapter 40A only after at least consulting the board, and perhaps only with the

board’s agreement. The problem is not if a city council chooses to act in such a

way, but if it purports to legally bind itself, and successive city councils, to that

practice.

18

provision. At least two possible grounds may require such an election,

one of which is foreclosed while the other remains open. Specifically, to

the extent that the Fund argues that any amendment to Chapter 40A

requires an election because of § 40A-35(a), our decision today necessarily

rejects that contention. Our holding concerning the board’s veto relies on

the principle that the city council that adopted § 40A-35(a) could not bar

a subsequent city council from amending Chapter 40A, including by

impliedly repealing § 40A-35(a)’s procedural requirements. The adoption

of § 8-1.5(a-1) without board approval or popular election means that

§ 40A-35(a)’s election requirement has been repealed just like the board-

veto requirement, without any impediment from trust-law doctrines.

But law other than Texas trust law may compel a different result,

such as requiring an election here. The specific issue that we leave

unresolved today does not implicate § 40A-35(a) at all: whether article

6243d would obligate the City to seek voter approval of amendments to

Chapter 40A even if § 40A-35(a) had never required such elections.

As we have noted, article 6243d requires voter approval before a

city may enact a pension plan like Chapter 40A. Rev. Civ. Stat.

art. 6243d, § 1. The voters of Dallas so approved in 1943. The city council

had unilateral authority to amend what is now Chapter 40A until 1945,

and it approved at least one amendment during that period.16 The Fund

16 The ordinance approved by the voters provided as follows: “Governing

Body May Change Ordinance. With the exception of the sections relating to

contributions to the Fund by employee members and contributions by the City,

the City Council, or other Governing Body, shall have the power by ordinance to

amend all the terms and provisions of this ordinance without submitting such

amendments to a vote of the electorate, provided that the power of amendment

given herein shall expire on January 1, 1945.” Dallas, Tex., Ordinance No. 3470

(Nov. 24, 1943).

19

asserts that, after 1945, every ordinance amending Chapter 40A has been

put to the voters for their approval.17 Whether § 8-1.5(a-1) must be

submitted to a popular election turns on whether the practice was

commanded by the statute (and if so, whether the city council’s initial

authority to amend it without elections was unlawful or was excused for

some other reason).

We decline to address this question in the first instance. “As a

court of last resort, it is not our ordinary practice to be the first forum to

resolve novel questions, particularly ones of widespread import.” In re

Troy S. Poe Tr., 646 S.W.3d 771, 780 (Tex. 2022). The court of appeals

did not address whether the voters must approve this or other

amendments to Chapter 40A, in part because that court’s holding

regarding the board’s role made it unnecessary to further consider

whether an election was also required (much less whether article 6243d

would be the source of such a requirement). That court likely also

declined to reach that question because the briefing addressing it was

17 For purposes of this case, we assume the truth of this proposition, but

note that the record does not clearly support it. The Fund claims that each

amendatory ordinance after 1945 (the next one came in 1947) until 1991

expressly required voter approval to become effective. See, e.g., Dallas, Tex.,

Ordinance No. 19470 (Feb. 18, 1987) (“That this ordinance shall take effect

immediately from and after its approval by the voters of the City of Dallas in a

special election on April 4, 1987, and it is accordingly so ordained.” (emphasis

added)). But at least two amendments during that period textually omit the

requirement, leaving the extent of voter involvement in passing them unclear.

See Dallas, Tex., Ordinance No. 17713 (Mar. 2, 1983) (“That this Ordinance shall

take effect immediately from and after its passage and publication in accordance

with the provisions of the Charter of the City of Dallas, and it is accordingly so

ordained.”); Ordinance No. 18181 (Feb. 29, 1984) (similar). Indeed, similar

language omitting voter (and board) approval appears in amendments

postdating 1991. See, e.g., Dallas, Tex., Ordinance No. 28739 (Aug. 8, 2012).

Nothing in our opinion turns on these details.

20

insubstantial compared to the arguments surrounding the board’s

authority, which has always been the case’s central focus. The parties’

briefing in this Court gave even less attention to whether the voters must

approve amendments to Chapter 40A. The State as amicus briefly

addressed the question and the Fund briefly responded, but no party has

provided sufficient briefing for us to resolve the question in the first

instance. A decision by this Court that determines whether the statute

requires continuing voter approval would have “widespread import,” id.,

because it would provide a rule that would also bind every other city plan

created under article 6243d. If we must address such a question, we will

not do so until it is fully presented.

We instead have resolved the issues that were briefed and argued

to us, which are sufficient to reverse the judgment below and to remand

the case to the court of appeals. We leave it in the first instance for that

court to determine if the parties have adequately preserved any question

regarding the statutory basis for any continuing need for elections to

validate amendments to Chapter 40A. If the court concludes that the

issue was properly before it, that court should resolve the issue. If the

court concludes otherwise, then it should render judgment for the City of

Dallas and reinstate the trial court’s judgment.

The judgment of the court of appeals is accordingly reversed and

the case is remanded to that court.

Evan A. Young

Justice

OPINION DELIVERED: March 15, 2024

21