Phil Johnson, Justice.

This case requires us to interpret Texas's "tied house" statutes that prohibit overlapping ownership between the manufacturing, wholesaling, and retailing segments of the alcoholic beverage industry.

Fomento Económico Mexicano, S.A.B. de C.V. (FEMSA) owns 20% of the stock in two Heineken companies which in turn own breweries. The brewers hold non-resident manufacturer's permits in Texas. FEMSA also owns, through intermediate holding companies, 100% of Cadena Comercial USA Corp., a company formed to operate convenience stores in Texas. When Cadena sought a retailer's permit to sell alcohol, the Texas Alcoholic Beverage Commission (TABC) protested the permit's being granted on the basis that FEMSA's ownership interests in Cadena and Heineken would violate the tied house statutes if the permit were granted. The county judge in an administrative hearing agreed with the TABC. The district court judge did likewise, and the court of appeals affirmed.

We affirm.

I. Background

A. Underlying Facts

Petitioner Cadena is a Texas corporation and wholly owned subsidiary of FEMSA, a Mexican entity. Before 2010, FEMSA was directly involved in brewing beer. It transferred that part of its business to Heineken N.V. and Heineken Holding, N.V. (collectively, the Heineken Group) in exchange for more than 72 million shares of stock in Heineken N.V. and more than 43 million shares in Heineken Holding N.V. — a 20% combined interest in the Heineken Group. FEMSA's holdings make it the largest shareholder in the Heineken Group except for the parent companies that own the controlling shares. The Heineken Group, through a series of intermediary companies, owns three brewers (the Heineken Brewers).

When FEMSA obtained its interest in the Heineken Group, it entered into a Corporate Governance Agreement that entitles FEMSA to appoint one of Heineken Holding N.V.'s five directors and two of ten members of the Supervisory Board of Heineken N.V. The Agreement also specifies that FEMSA is not given "any right or control or influence or consultation right or other form of cooperation" relating to the Heineken Group. Similarly, L'Arche Green, a parent company of the Heineken Group, reserved all rights to make decisions in its management of the Heineken Group, "independently and at their sole discretion and without any requirement to consult or cooperate with ... FEMSA." The Agreement bars the Heineken Group from acquiring any stock in FEMSA.

FEMSA owns approximately 10,000 convenience stores concentrated in Mexico and Colombia that operate under the name OXXO, and it continues to open more regularly. FEMSA formed Cadena to extend FEMSA's retail convenience store business into Texas. Cadena wanted to sell wine and beer in its stores, which in Texas would require it to have a wine and beer retailer's off-premises consumption permit. When Cadena tried to obtain one of these permits from the TABC, routine financial disclosures it made during the application process revealed FEMSA's 100% ownership of Cadena as well as its significant ownership interest in the Heineken Group, which owns the Heineken Brewers that, in turn, hold Texas non-resident brewer's permits. The TABC protested Cadena's permit on grounds that granting it would result in a violation of the Texas tied house statutes, and rejected its application.

B. Texas's Tied House Statutes

The Texas tied house statutes are found in the Texas Alcoholic Beverage Code. See TEX. ALCO. BEV. CODE §§ 102.01-.82. The genesis of the provisions was the Liquor Control Act, which the Legislature adopted two years after the repeal of Prohibition. See Texas Liquor Control Act, 44th Leg., 2d C.S., ch. 467, §§ 1-23, 1935 Tex. Gen. Laws 1795. The Liquor Control Act's progeny were eventually codified into the Alcoholic Beverage Code. An Act Adopting the Alcoholic Beverage Code, 65th Leg., R.S., ch. 194, § 1, 1977 Tex. Gen. Laws 391 (codified as amended in TEX. ALCO. BEV. CODE §§ 1.01-251.82). The catalyst for the tied house provisions was a fear of returning to the state of affairs before Prohibition when tied houses played what was thought to be a substantial role in over-intoxicating society. The provisions are designed to prevent certain overlapping relationships between those engaged in the alcoholic beverage industry at different levels, or tiers. See TEX. ALCO. BEV. CODE § 102.01(a)-(b).

Pre-Prohibition tied houses generally developed from tavern owners selling their taverns to brewers and becoming the brewers' tenants. See generally, D. M. Knox, The Development of the Tied House System in London, 10 OXFORD ECON. PAPERS, NEW SERIES, no. 1, 1958, at 66-83. Financial conditions and other factors made these agreements a near-necessity for tavern owners to survive economically. Most of the agreements included a stipulation that the tavern would only sell the brewer-landlord's products. The brewer then had a vested interest in the tavern selling as much of the brewer's beer as possible, with little or no regard for the personal or societal effects. This tied house phenomenon contributed to the push for Prohibition.

When Prohibition ended, lawmakers started from a relatively clean slate with respect to regulating the alcoholic beverage industry, and their goal was to prevent a return to the pre-Prohibition status. See TEX. ALCOHOLIC BEVERAGE COMM'N, THE HISTORY OF THE TEXAS ALCOHOLIC BEVERAGE COMMISSION 1-2 (2005). One of the targets was tied house relationships. In an attempt to prevent these relationships from forming, the Code provides for "strict adherence to a general policy of prohibiting the tied house and related practices." TEX. ALCO. BEV. CODE § 102.01(b). The Code defines "tied house" as

Id. § 102.01(a). The Code contains numerous provisions designed to achieve this overarching goal by separating the industry into three independent tiers: manufacturing (brewing), distribution, and retail. See id. §§ 102.01-.82. It attempts to achieve this separation by prohibiting cross-tier relationships. Several of these provisions served as grounds for the TABC's protest of Cadena's application for a permit.

C. Procedural Background and Positions of the Parties

During the retail permit application process, an applicant such as Cadena must submit designated disclosure forms to the TABC. Id. §§ 26.03, 61.31(a). Based on information in these forms, the TABC either protests the application or grants the permit. Id. § 61.31(a). If the TABC finds there are reasonable grounds to protest the permit, then it is required to do so and reject the application. Id.

If the TABC rejects a permit application, the applicant may request an administrative hearing before the county judge in the county in which the applicant desires to conduct business. Id. §§ 61.31-.32. If the county judge finds no legal grounds to refuse the application, the judge orders the TABC to grant the permit. See id. § 61.32. If the judge denies the application, the applicant has thirty days to appeal to the district court. See id. § 61.34.

Here, the TABC determined that Cadena's connection to the Heineken Brewers through FEMSA's 100% ownership of Cadena and large ownership interest in the Heineken Group meant that granting the retail permit would result in Cadena having overlapping interests in the manufacturing and retail levels of the industry in violation of four separate provisions of the tied house statutes. Thus TABC protested and rejected Cadena's application.

The matter then proceeded to an evidentiary hearing before the county judge. At the hearing, the parties stipulated to the corporate relationships between Cadena, FEMSA, and the Heineken companies. The TABC further stipulated that the five statutory provisions at issue were the only grounds for its protesting Cadena's application. During the hearing, the TABC's licensing director and an expert in alcoholic-beverage industry laws testified that if an entity in one tier of the industry owned even one share of stock in a member of another tier, the overlapping ownership would violate the statutory tied house prohibitions. The TABC argued that FEMSA's interests in Cadena and the Heineken Brewers were prohibited "interests" under the Alcoholic Beverage Code under any interpretation. Although the TABC disputed that actual cross-tier control of entities is required to implicate the tied house restrictions, it nevertheless asserted that FEMSA could control the Heineken Brewers because of its ability to appoint directors to the Heineken Group's boards. It also argued the court should impute this connection to Cadena for regulation purposes.

Conversely, Cadena argued that the only "interest" sufficient to violate the tied house prohibitions is one allowing simultaneous actual financial or administrative control of entities in different tiers. Under Cadena's interpretation, its permit application should have been granted as a matter of law because FEMSA has no ability to manage or control either the Heineken Holding Companies or the Heineken Brewers. As a result, Cadena argued, granting its application would not violate the tied house statutes because no company within the business structure would have managing control over entities in more than one tier. Cadena further maintained that FEMSA's remote connection with the Heineken Brewers was too attenuated to implicate historical tied house concerns, and that the interest could not be imputed to Cadena without piercing the corporate veils of all the entities involved.

Following the administrative hearing, the county judge denied Cadena's application based on the statutory grounds cited by the TABC, finding that: (1) Cadena "has a real interest in the business or premises of the holder of a manufacturer's or distributor's license"; (2) "[f]or licensing purposes, as a subsidiary of FEMSA, [Cadena] is a manufacturer"; (3) "[f]or licensing purposes, as a subsidiary of FEMSA, [Cadena] has an interest in the business of a brewer"; and (4) issuing the requested permit "would violate Sections 102.01(c), (h), 102.07(a)(1), and 102.11(1) of the Code." Cadena appealed to the district court, which affirmed the administrative order. Cadena appealed again.

The court of appeals focused on section 102.07(a)(1), which provides that "no person who owns or has an interest in the business of a ... brewer ... may ... own or have a direct or indirect interest in the business ... of a retailer." 449 S.W.3d 154, 159, 171 (Tex. App.-Austin 2014); TEX. ALCO. BEV. CODE § 102.07(a). The court determined that the dispositive issue was whether FEMSA's shares in Heineken, paired with FEMSA's indirect ownership of Cadena, would violate section 102.07(a) if Cadena's permit application were granted. Id. It concluded "the term `interest,' as used in section 102.07(a)(1), broadly encompasses any commercial or economic interest that provides a stake in the financial performance of an entity engaged in the manufacture, distribution, or sale of alcoholic beverages." Id. at 166. The court further concluded that FEMSA owns an interest in the business of Cadena, a retailer, by virtue of its 100% indirect ownership interest, and rejected Cadena's corporate separateness arguments. Id. at 169-70. It reasoned that the definition of "business" includes more than merely having an interest in the assets of the business. The court concluded that even though FEMSA only indirectly owned Cadena and only indirectly held stock in the Heineken Brewers, the broad meaning of "interest" and "business" did not implicate the principles behind recognizing separate corporate identities, and under the statute FEMSA owned an interest in both. Id. at 168.

Finally, the court determined that Cadena's equal protection claim failed because Cadena failed to prove that the TABC granted permits to any similarly situated entities. Id. at 172. Although Cadena pointed to evidence of pervasive cross-tier ownership interests that violated the court's reading of section 102.07(a), Cadena did not show that any of the permitted entities held "similarly significant cross-tier investment interests." Id.

Before this Court, Cadena raises the same three claims it did in the court of appeals: (1) a plain reading of section 102.07(a) can only lead to the conclusion that the Legislature intended a control-based test when determining which interests come under the statute, and any other reading renders the statute unconstitutionally vague and unenforceable; (2) corporate separateness and veil-piercing principles are implicated and proper application of these rules would prevent Cadena's licensure from being a violation of the Code because FEMSA's interest in both Cadena and the Heineken Brewers is attenuated; and (3) the TABC's selective application of the statute to Cadena's permit application violates equal protection principles because of the pervasive cross-tier holdings by other entities across the State. Cadena also expresses concern throughout its briefing that the court of appeals' expansive interpretation provides the TABC authority to reject a permit application or cancel a permit based on a person's ownership of a single share of stock in two separate tiers.1

The TABC responds that (1) the court of appeals' reading of the interests implicated by 102.07(a) accurately conforms to the statute's plain meaning and the Legislature's desire to maintain strict separation between the three tiers of the alcoholic beverage industry; (2) Cadena's veil-piercing arguments rest on the faulty assumption that "interest in the business of a brewer" is tantamount to legal ownership; and (3) Cadena's equal enforcement claim is meritless because there is no evidence of permittees that are similarly situated to Cadena, nor of any suspect or improper motivation on the part of the TABC. The TABC also characterizes the one-share issue as a red herring designed to divert attention from FEMSA's significant interests in both Cadena and the Heineken Brewers.2

II. Standard of Review

The Alcoholic Beverage Code provides that a party whose application is refused, or whose permit is cancelled or suspended, may appeal that decision. TEX. ALCO. BEV. CODE § 11.67(a), 61.34(a). The Code requires courts to review challenged decisions under the substantial evidence rule pursuant to the Administrative Procedure Act (APA). Id. § 11.67(b). The APA dictates that "[t]he scope of judicial review of a state agency decision ... is as provided by the law under which review is sought." TEX. GOV'T CODE § 2001.172; State v. Pub. Util. Comm'n of Tex., 344 S.W.3d 349, 355 (Tex. 2011).

Statutory interpretation is the primary issue in this appeal, and that involves questions of law we review de novo. Sw. Royalties, Inc. v. Hegar, 500 S.W.3d 400, 404 (Tex. 2016). This is true even when we are reviewing agency decisions. See, e.g., State v. Shumake, 199 S.W.3d 279, 284-85 (Tex. 2006). An agency's interpretation of a statute it enforces "is entitled to `serious consideration,' so long as the construction is reasonable and does not conflict with the statute's language." R.R. Comm'n of Tex. v. Tex. Citizens for a Safe Future & Clean Water, 336 S.W.3d 619, 624 (Tex. 2011). The APA provides that cases should be reversed or remanded if the administrative decision is "in violation of a constitutional ... provision," is "not reasonably supported by substantial evidence," or the decision is "arbitrary or capricious or characterized by an abuse of discretion." TEX. GOV'T CODE § 2001.174(2)(A), (E), (F); see also Pub. Util. Comm'n, 344 S.W.3d at 356.

III. Analysis

The TABC originally asserted four statutory provisions as grounds for protesting Cadena's application. The county judge and district court determined that each of the provisions would support rejecting a permit. However, the court of appeals only considered section 102.07(a) because it concluded the provision was dispositive. 449 S.W.3d at 161-62, 171. In determining whether the court of appeals was correct, we consider three main issues: (1) the reach of section 102.07(a) and, in particular, the meaning of the phrase "an interest in the business of a brewer"; (2) whether the TABC properly disregarded the separate corporate statuses of the entities involved when deciding if issuing a permit to Cadena would result in a violation of section 102.07(a); and (3) whether the TABC's protesting of Cadena's application, and the subsequent sustaining of that protest by the lower courts, violated Cadena's equal protection rights.

A. "Interest in the Business of a Brewer"

Cadena's basic issue with the court of appeals' holding is summarized by the following language from its briefing:

Brief for Petitioner at 8 (alteration in original) (citation omitted).

Our fundamental goal when reading statutes "is to ascertain and give effect to the Legislature's intent." Tex. Mut. Ins. Co. v. Ruttiger, 381 S.W.3d 430, 452 (Tex. 2012). To do this, we look to and rely on the plain meaning of a statute's words as expressing legislative intent unless a different meaning is supplied, is apparent from the context, or the plain meaning of the words leads to absurd or nonsensical results. Crosstex Energy Servs., L.P. v. Pro Plus, Inc., 430 S.W.3d 384, 389-90 (Tex. 2014). Words and phrases "shall be read in context and construed according to the rules of grammar and common usage." Id. (citing TEX. GOV'T CODE § 311.011). We presume the Legislature "chooses a statute's language with care, including each word chosen for a purpose, while purposefully omitting words not chosen." TGS-NOPEC Geophysical Co. v. Combs, 340 S.W.3d 432, 439 (Tex. 2011). In that vein, we take statutes as we find them and refrain from rewriting the Legislature's text. Entergy Gulf States v. Summers, 282 S.W.3d 433, 443 (Tex. 2009).

The relevant language in section 102.07(a) provides

TEX. ALCO. BEV. CODE § 102.07(a). These words and phrases are not to be considered in isolation, but rather in the context of the statute as a whole. Meritor Auto., Inc. v. Ruan Leasing Co., 44 S.W.3d 86, 90 (Tex. 2001). Put differently, our objective is not to take definitions and mechanically tack them together — as Cadena claims the court of appeals did — rather, we consider the context and framework of the entire statute and meld its words into a cohesive reflection of legislative intent. See Anheuser-Busch, L.L.C. v. Harris Cty. Tax Assessor-Collector, 516 S.W.3d 1, ___, 2016 WL 5920766 (Tex. 2016). Still, we are concerned with the definitions of the specific words because they provide the material that is refined with statutory context. The definitions of some terms are either provided by the Legislature or clear based on their common usage and meaning. For others, we must look to the statutory context to provide meaning.

We begin with definitions the Code provides. TEX. GOV'T CODE § 311.011; Hernandez v. Ebrom, 289 S.W.3d 316, 318 (Tex. 2009). The Code defines "person" as "a natural person or association of natural persons, ... corporation, organization, or the manager, agent, servant, or employee of any of them." TEX. ALCO. BEV. CODE § 1.04(6). Cadena, FEMSA, the Heineken Group, the Heineken Brewers, and each relevant intermediate holding company are, therefore, "person[s]" under the statute. Although the Legislature did not provide the definition of "brewer," its ordinary meaning is intuitive. Moreover, Webster's defines "brew" as "to prepare (as beer or ale from malt and hops) by steeping, boiling, and fermentation." WEBSTER'S THIRD NEW INTERNATIONAL DICTIONARY 275 (2002). The same dictionary defines "brewery" as "a building or plant where beer is manufactured." Id. The Heineken Brewers brew beer in breweries, thus they are brewers within the Code's meaning. And if Cadena were granted a wine and beer retailer's off-premise consumption permit, it would be a "retailer" as the term is used in section 102.07(a)(1). See TEX. ALCO. BEV. CODE §§ 26.01, .03. Finally, because the Legislature has not provided specific definitions for many terms, we look to their plain and ordinary meaning, unless a different meaning is apparent from the context of the statute. In re Ford Motor Co., 442 S.W.3d 265, 271 (Tex. 2014).

Section 102.07 is within Chapter 102 of the Code. This Chapter is entitled "Intra-Industry Relationships" and provides a comprehensive framework for regulating everything from overlapping ownership among the three tiers down to specific financial transactions and gifts and promotions. See TEX. ALCO. BEV. CODE §§ 102.01-.82. Subchapter A, which includes section 102.07(a), provides expansive and comprehensive statutes that prohibit a broad range of cross-tier relationships and influences. See id. §§ 102.01-.22. Further, the Legislature's express policy statements are indicative of legislative intent. Hebner v. Reddy, 498 S.W.3d 37, 41 (Tex. 2016); R.R. Comm'n of Tex. v. Tex. Citizens for a Safe Future & Clean Water, 336 S.W.3d 619, 628 (Tex. 2011). The Code expresses a policy of "strict separation between the manufacturing, wholesaling, and retailing levels" of the alcoholic beverage industry in Texas to prevent "the creation or maintenance of a `tied house.'" TEX. ALCO. BEV. CODE § 6.03(i). Section 102.01, which first defines a tied house as "any overlapping ownership or other prohibited relationship" among the three tiers of the alcoholic beverage industry, reiterates the State's "general policy of prohibiting the tied house and related practices." Id. § 102.01(a), (b) (emphasis added). Section 102.75(c) mandates "the independence of members of the three-tier system." Id. § 102.75(c). From all this, the Legislature clearly was concerned not only with preventing paradigmatic, pre-Prohibition tied houses, but also with "related practices" that might negatively affect public health and safety. See id. § 102.01(b). Although this Court has not previously addressed the Texas tied house provisions, courts that have are in accord with the foregoing. See Dickerson v. Bailey, 336 F.3d 388, 397 (5th Cir. 2003); S.A. Disc. Liquor, Inc. v. Tex. Alcoholic Beverage Comm'n, 709 F.2d 291, 293 (5th Cir. 1983); Neel v. Tex. Liquor Control Bd., 259 S.W.2d 312, 316 (Tex. Civ. App.-Austin 1953, writ ref'd n.r.e.); Tex. Liquor Control Bd. v. Cont'l Distilling Sales Co., 199 S.W.2d 1009, 1014 (Tex. Civ. App.-Dallas 1947, writ ref'd n.r.e.); see also Mayhue's Super Liquor Store, Inc. v. Meiklejohn, 426 F.2d 142, 147 (5th Cir. 1970) (noting that "the liquor business has been the subject of severe legislative restraints").

With this history and structure in mind, we turn to the meaning of "an interest in the business of a ... brewer." We start with the term "interest," which the Legislature did not define. At the time the tied house statutes were enacted, Black's Law Dictionary defined "interest" as "[t]he most general term that can be employed to denote a property in lands or chattels" or "a right to have the advantage accruing from anything." Interest, BLACK'S LAW DICTIONARY (3d ed. 1933). English language dictionaries at the time took a similarly broad view. For instance, the Oxford Dictionary defined "interest" to include a "[l]egal concern, title, right," "pecuniary stake," "advantage," "profit," and "party having a common interest." THE CONCISE OXFORD DICTIONARY OF CURRENT ENGLISH 427 (7th ed. 1919). Contemporary dictionaries confirm that "interest," standing alone, still has a broad meaning. See WEBSTER'S NEW WORLD COLLEGE DICTIONARY 758 (5th ed. 2016) ("[A] right or claim to something... as a business, in which one participates or has a share."); Interest, BLACK'S LAW DICTIONARY (10th ed. 2014) ("Collectively, the word includes any aggregation of rights, privileges, powers, and immunities....").

The court of appeals reasoned that "the term `interest' invokes many different definitions and, without a modifier, could in the abstract be so broad as to be vague and ambiguous." 449 S.W.3d at 165. We agree with that general proposition. Cadena contends that "interest" should be construed narrowly or it will be vague and unenforceable. With that, we do not agree. If an undefined word used in a statute has multiple and broad definitions, we presume — unless there is clear statutory language to the contrary — that the Legislature intended it to have equally broad applicability. See, e.g., Greater Hous. P'ship v. Paxton, 468 S.W.3d 51, 59 (Tex. 2015). When faced with a term that is so broad that it borders on being ambiguous, we look to the statutory context to limit the possible correct meanings. See, e.g., id. ("Not surprisingly, `supported,' the key term here, is subject to at least six disparate definitions in its verb form alone, with many of those including more nuanced sub-definitions. By reading the term in context, however, we can narrow the universe of possible definitions to the most apposite." (citations omitted)). Thus, when interpreting broad, context-sensitive terms such as "interest," we must be sensitive to the context.

Looking to the surrounding statutory environment for assistance in determining the meaning of "interest," we note that the Code contains instances of the term throughout, both with and without modifiers. "Interest" is variously referred to as a "pecuniary interest," "an ownership interest," "a financial interest," "a real interest," "an interest of any kind," and "any interest." See, e.g., TEX. ALCO. BEV. CODE §§ 5.05(a)(3), 61.44(a)(1), (b)(1), 61.71(a)(21), 102.01(c), 102.10(b). Further, the Code uses the specific terms "corporate stock," "affiliate," "director," "officer," and "control." See, e.g., id. §§ 11.48(a), 37.07(1), 102.14, 102.15(a), 102.18. Importantly, section 102.07(a)(1) refers to a "direct or indirect interest" in the business of a retailer. Id. § 102.07(a)(1). Cadena argues that the multiple references to different kinds of interests and this latter reference to a "direct or indirect interest" means that the Legislature intended for an interest in the business of a retailer to be viewed more broadly than an interest in the business of a brewer. We disagree. The term "interest," standing alone, necessarily subsumes the other modifiers that might limit the term. For example, either the term "an indirect interest" or the term "a direct interest," separately considered, is narrower than "an interest." "Interest" includes both of these, in addition to any other interest that is neither direct nor indirect. Further, "an interest" also subsumes corporate stock, affiliate-subsidiary relationships, and a level of control. But the term "interest," as used in the context of this statute, does not include any interest, as Cadena contends the court of appeals' holding requires. Rather, the court of appeals pointed to other modified uses of "interest" throughout the Code as having one thing in common: they all refer to commercial and economic interests. Consideration of this common theme leads to the conclusion that the Legislature was concerned with interests that result from the various business dealings among and between participants in the alcohol industry. And here, as evidenced by the lack of a narrowing modifier, the Legislature used the term "interest" broadly and intended to include all these interests.

However, the term is then narrowed by the phrase "in the business of a brewer." Having determined the plain meaning of "brewer," we must determine what "the business" of a brewer means. The definition of business is more finite than the definition of interest. "Business" generally refers to "[a] commercial enterprise carried on for profit; a particular occupation or employment habitually engaged in for livelihood or gain." Business, BLACK'S LAW DICTIONARY (10th ed. 2014). Combining this with the meaning of "brewer" — and with the contextualized definition of "interest" discussed above — leads to the conclusion that "an interest in the business of a brewer" means what the court of appeals said it did: the phrase "broadly encompasses any commercial or economic interest that provides a stake in the financial performance of an entity engaged in the manufacture ... of alcoholic beverages." 449 S.W.3d at 166. Cadena asserts it was error for the court of appeals to isolate the specific words, define them individually, and then combine them to produce this clause. While the court of appeals' approach might not be appropriate for every case, here it was. Its interpretation meshes with both the plain language and context of the statute's words, as well as the Legislature's policy of strict separation between the tiers of the industry.

Next, we turn to Cadena's other arguments, beginning with its argument that an "interest in the business of a brewer" should only extend to those actually engaged in the business of brewing beer, not a brewer's stockholders. We disagree for two reasons. First, this interpretation would effectively eliminate "interest" from the statute. At the very least, Cadena's reading would modify the term "interest" into something akin to "engaged." When the Legislature uses a word or phrase in one part of a statute but excludes it from another, the term should not be implied where it has been excluded. Safe Future & Clean Water, 336 S.W.3d at 628. A look to other Code provisions shows that the Legislature was careful to use the term "engage" when it sought to limit a statute's applicability to those who were directly participating in one of the three tiers of the industry. See, e.g., TEX. ALCO. BEV. CODE §§ 61.44(a)(2), 61.71(a)(29), 102.01(a), 102.09, 102.15(a). Had the Legislature intended to limit the meaning of a "person with an interest in the business of a ... brewer" to those engaged in brewing beer, we presume it would have said just that. See TGS-NOPEC Geophysical, 340 S.W.3d at 439; Laidlaw Waste Sys. (Dall.), Inc. v. City of Wilmer, 904 S.W.2d 656, 659 (Tex. 1995) ("When the Legislature employs a term in one section of a statute and excludes it in another section, the term should not be implied where excluded."). Our interpretation finds reinforcement in the broad meaning of "interest" established above. Second, we agree with the TABC that interpreting section 102.07(a) to extend only to brewers would subvert legislative intent. Holding that only a brewer, or someone in the shoes of a brewer, is a "person with an interest in the business of a ... brewer" would open the door for companies across the alcohol industry to circumvent the tied house provisions. Such a reading would significantly frustrate the Legislature's expressly stated purpose of "strict separation" and preventing "any overlapping ownership" among the three tiers of the industry. See TEX. ALCO. BEV. CODE §§ 6.03(i), 102.01(a).

Cadena also argues that the relationships prohibited in other provisions of the Code provide support for its argument that FEMSA's relationship with the Heineken Brewers is not included in section 102.07(a). For example, section 102.11 prohibits a "manufacturer or distributor" from "directly or indirectly, or through a subsidiary, affiliate, agent, employee, officer, director, or firm member," from owning "any interest in the business or premises of a retail dealer of beer." Id. § 102.11. An "affiliate" is "a person who controls, is controlled by, or is under common control with another person." TEX. BUS. ORGS. CODE § 1.002. Cadena asserts that this is evidence the Legislature intended only to prohibit typical tied houses and that the failure to use such terms as affiliate or subsidiary in section 102.07(a) shows the Legislature intended something more narrow. But this argument is fundamentally flawed because it fails to consider the Legislature's express policy of strict separation between the three tiers. A fair reading of section 102.07(a) in light of the Legislature's multiple policy statements shows the statutes are designed to prevent far more than the historical paradigm of a tied house, in which manufacturers directly owned retail outlets. When viewed in context, this statute manifests Legislative intent to prevent more tenuous relationships. See TEX. ALCO. BEV. CODE § 102.01(b) (providing for "strict adherence to a general policy of prohibiting the tied house and related practices" (emphasis added)). The sheer number of statutes the Legislature enacted and the different approaches it took in proscribing the prohibited relationships, both specific and broad, reinforce its clear intent. Compare, e.g., id. § 102.14 (specifically prohibiting manufacturers and wholesalers from providing fixtures and equipment to anyone selling brewery products for on-premises consumption), with id. § 102.11(1) (broadly prohibiting a manufacturer from directly or indirectly owning any interest in a retailer).

Cadena further claims that the court of appeals was wrong to consider section 102.01(a)'s definition of tied house and use that language to support its conclusion that any overlap between tiers is forbidden. Cadena's reasoning is two-fold: (1) section 102.01(a) contains a definition and not a prohibition, and (2) the phrase "tied house" does not appear anywhere in section 102.07. Both of these statements are true. But it would be nonsensical to read the particular sections Cadena references without considering them in concert with section 102.01, which provides overarching context for fairly reading the entire statute. The statutory definition of tied house is not a narrow provision found in an unrelated statute; it undergirds and frames the purpose of Chapter 102. Each subsequent provision is informed and given context by it. At any rate, Cadena's application was not denied because FEMSA's cross-tier interests violated section 102.01(a). It was denied because the relationship violated 102.07(a). Even without considering section 102.01, section 102.07 — by its language, structure, and the multiple other policy statements found throughout the Code — requires strict separation. Thus, section 102.07(a), by its own terms, prohibits a person who has an interest in the business of a brewer from also having an interest in the business of a retailer.

Cadena claims that the court of appeals applied the statute "backwards" because the structure of 102.07 is such that it only applies to those with an interest in the business of a brewer who are trying to meddle in a retailer's business — i.e., the typical tied house relationship. Cadena correctly points out that there is a great deal more that a brewer is prohibited from doing with a retailer than the reverse. Section 102.07(a) provides that no person who meets a certain condition may meet any of the eight separate conditions set out in subsections (a)(1) through (a)(8). See TEX. ALCO. BEV. CODE § 102.07. While there are more things a brewer cannot do with respect to a retailer than vice versa, the statute does not mandate that we apply it in any particular manner. The condition-triggering verbs in subsections (a) and (a)(1) are both in the present tense and contemplate current and future holdings. If the Legislature only intended to prevent those with preexisting interests in brewers from acquiring interests in retailers, it could have, and presumably would have, said so. See TEX. GOV'T CODE § 311.012(a) ("Words in the present tense include the future tense."); see also TEX. LEGISLATIVE COUNCIL, TEXAS LEGISLATIVE COUNCIL DRAFTING MANUAL § 7.35 (Jan. 2017) (directing drafters to "[u]se present tense whenever possible"). To hold that this statute only flows in one direction would prevent those with interests in brewers from gaining an interest in retailers, but not retailers from gaining interests in brewers. Although a brewer controlling a retailer is the typical exemplar of a tied house, the statute's language, as well as the various policy statements, demonstrate legislative intent to provide strict separation going both ways. The reading Cadena proposes is out of step and completely at odds with that policy.

Additionally, Cadena claims that the court of appeals' interpretation of the statute effectively gives the TABC a new power that provides it carte blanche to grant or reject an application for a permit with no guiding principles on which applicants can rely. We disagree. First, this is not a new power conferred on the TABC; it is simply an interpretation of the power the TABC has always had under the Code. Second, the TABC does not have carte blanche to grant or reject any application. It only has the power to reject applications that violate section 102.07 or some other provision of the Code. See Tex. Nat. Res. Conservation Comm'n v. Lakeshore Util. Co., 164 S.W.3d 368, 377 (Tex. 2005) (holding that, despite petitioner's arguments to the contrary, "the Commission has statutory authority to pursue an enforcement action in district court"). Again, section 102.07 is only violated when a person with a direct or indirect interest in the business of a retailer also has a "commercial or economic interest that provides a stake in the financial performance" of an entity engaged in the manufacturing or distributing of alcoholic beverages. See 449 S.W.3d at 165. While we agree that the statement is broad, this does not render the statute unenforceable.

Cadena contends that the court of appeals read into the statute a "potential for influence" standard that finds no basis in the statute's language. Regardless of whether the court of appeals did so, we do not, as we have explained.

Cadena also raises the argument that the multiple references to "a permittee covered under Subsection (a)" in the exceptions in section 102.07 mean that section 102.07(a) only applies to permittees. See TEX. ALCO. BEV. CODE §§ 102.07(b), (d), (e), (g). This is wrong for at least two reasons. First, a person with an interest in the business of a brewer can be a permittee, but the language's reach extends well beyond permittees. Both permittees and entities like FEMSA are covered by subsection (a), but only a permittee can avail itself of the exceptions in the statute. Second, the term "permittee" appears over 500 times in the Code. If the Legislature had intended section 102.07(a) to include only permittees, it is safe to say it would have used that term.

Finally, Cadena claims that the court of appeals' use of the word "significant" to describe FEMSA's interest in the Heineken Brewers without further guidance renders section 102.07(a) impermissibly vague. But we view the court of appeals' consideration of "significant" in reaching its conclusion in light of the facts of this case. See Fin. Comm'n of Tex. v. Norwood, 418 S.W.3d 566, 594 (Tex. 2013) (citing TEX. CONST. art. I, § 13, art. II, § 1). We conclude that under the statute, an interest in the business of a brewer exists when a person has a commercial or financial interest — significant or otherwise — that provides a stake in the financial performance of an entity or person engaged in brewing. Further, Cadena can hardly make a credible argument that FEMSA's 115 million shares in the Heineken Group is not significant when it is unwilling to give it up so Cadena can obtain the permit it seeks.

FEMSA, by its stock ownership in the Heineken Group, has a commercial or economic interest that provides a stake in the financial performance of an entity engaged in brewing alcoholic beverages. This interest, coupled with FEMSA's indirect ownership interest in Cadena, who would be a retailer of alcoholic beverages if the permit were granted, would violate section 102.07. Thus, we agree with the court of appeals' interpretation of section 102.07(a) and now turn to Cadena's corporate separateness arguments to determine if FEMSA's cross-tier interests must be excluded from consideration because the corporations involved are separate entities.

B. Corporate Separateness

Cadena argues that the doctrine of corporate separateness applies in the regulatory context and cannot be ignored in determining whether FEMSA has an interest in the business of the Heineken Brewers or a direct or indirect interest in the business of Cadena. Cadena cites cases holding that subsidiaries are distinct from parent companies because parent companies are simply shareholders that do not own an interest in the business of the companies in which they hold stock. Thus, Cadena argues the TABC erred by, in effect, piercing the corporate veils and imputing the Heineken Brewers', Cadena's, and each intermediate holding company's interests to FEMSA, absent evidence of abuse of the corporate form, or use of it to circumvent a statute. See SSP Partners v. Gladstrong Invs. (USA) Corp., 275 S.W.3d 444, 451 (Tex. 2008).

While we have no dispute, generally, with Cadena's reading of the cases it references, we disagree with the application of the law it proposes. As the TABC points out, each of the cases Cadena cites regarding the principle of corporate separateness addresses that doctrine in the context of tort or contract liability, or other similar circumstances. E.g., Miles v. Am. Tel. & Tel. Co., 703 F.2d 193, 197 (5th Cir. 1983) (refusing to impose liability on parent company for tortious invasion of privacy); SSP Partners, 275 S.W.3d at 451-52 (refusing to hold subsidiary liable for parent company's failure to indemnify retailer); S. Union Co. v. City of Edinburg, 129 S.W.3d 74, 89 (Tex. 2003) (refusing to collapse corporate identity for taxing purposes); BMC Software Belg., N.V. v. Marchand, 83 S.W.3d 789, 799 (Tex. 2002) (refusing to collapse corporate separateness for jurisdictional purposes); Castleberry v. Branscum, 721 S.W.2d 270, 272 (Tex. 1986) (reinstating a jury finding on sham entity status and not allowing the corporate fiction to insulate individuals from their liability on the corporation's promissory note); Gentry v. Credit Plan Corp. of Hous., 528 S.W.2d 571, 573 (Tex. 1975) (affirming the principle of alter ego in tort liability); Bell Oil & Gas Co. v. Allied Chem. Corp., 431 S.W.2d 336, 341 (Tex. 1968) (holding parent corporation not liable for contract obligations of affiliated corporation); Auto. Mortg. Co. v. Ayub, 266 S.W. 134, 135 (Tex. 1924) (holding that shares of stock in a corporation are entirely separate and distinct from the corporation's property); but cf. R.R. Comm'n of Tex. v. Lone Star Gas Co., 844 S.W.2d 679, 690 (Tex. 1992) (upholding the separate corporate existence but because the Railroad Commission rule in question contemplated such, not because of common-law principles that would require that result).

But, as the TABC correctly points out, corporate separateness principles are different in the regulatory context. For example, in Beneficial Financial Co. of Midland v. Miskell, we were faced with a similar issue. 424 S.W.2d 482, 483-84 (Tex. Civ. App.-Austin 1968). Beneficial Financial Company of Midland was a wholly owned subsidiary of Beneficial Finance Company of Delaware. Id. at 483. Beneficial Delaware owned all of the stock in sixty Texas corporations that all held Regulatory Loan Licenses issued under the Texas Regulatory Loan Act of 1963. Id. When Beneficial of Midland filed an application for a Regulatory Loan License, it was denied because the Regulatory Loan Commissioner determined that the license would cause a violation of a provision of the Act that prohibited any person, "directly or indirectly, or through subsidiaries or holding companies, to hold or have an interest in more than sixty (60) licenses, the business thereof, or any interest in such license." Id. (citing Texas Regulatory Loan Act of 1963, 58th Leg., R.S., ch. 205, § 10(c), 1963 Tex. Gen. Laws 556, revised by Act of 1967 Revising the Texas Regulatory Loan Act of 1963, 60th Leg., R.S., ch. 274, § 3.06(3), 1967 Tex. Gen. Laws 617).

In upholding denial of the application, this Court held that the statutory provision directed that "the corporate fiction separating the parent, Delaware Corporation, and its subsidiaries, the sixty-one Texas corporations — the stock of which is wholly owned by the parent corporation" — should be ignored in that situation. Id. at 484. The Court implicitly recognized that a statute could authorize regulatory agencies to look beyond the corporate veil. See id. Doing so prevents corporations from circumventing statutes and frustrating legislative intent by using a legislatively authorized corporate form to avoid a statute's reach and allow harms the Legislature set out to prevent. See id. The same rationale applies here. Cadena argues that Miskell is distinguishable because the statute at issue in that case included references to subsidiaries and holding companies. But the language in section 102.07(a) is sufficiently broad to encompass subsidiary corporate relationships. In our view, by enacting broad language providing that "no person who owns or has an interest in the business of a ... brewer" may also have "a direct or indirect interest in the business of a ... retailer," the Legislature intended that the TABC and courts look beyond corporate separateness status in enforcing the tied house provisions.

Cadena points to various attorney general opinions that it claims support its view. We do not find these persuasive for several reasons. First, the two most applicable opinions — the only two that discuss the tied house statutes — question whether the phrase "or the business thereof," in a now-repealed tied house statute, requires corporate separateness to be observed. But notably, both of these were issued before Miskell was decided. Compare TEX. ATT'Y GEN. OP. No. O-7039 (1946), and TEX. ATT'Y GEN. OP. No. O-4750 (1942), with Miskell, 424 S.W.2d at 484. Our holding in Miskell brings the Attorney General's reasoning into question. Second, courts are not bound by attorney general opinions. In re Smith, 333 S.W.3d 582, 588 (Tex. 2011). Instead, we are bound by the principle outlined in Miskell: a statute, by its terms, controls an administrative agency's authority to ignore corporate separateness. 424 S.W.2d at 484. Even if the statutory language in the referenced attorney general opinions tracked the exact language in section 102.07(a), our interpretation of section 102.07(a) — and its reach — would still control. And as discussed above, section 102.07 is extremely broad and prohibits even attenuated interests so long as those interests are rooted in the financial performance of the entities in question. This is sufficient statutory authority to allow the TABC to look past the corporate fiction when enforcing this statute.

Cadena also argues that even if the statute permits the TABC to disregard corporate separateness, it does not have authority over FEMSA's interest. Rather, Cadena posits that the TABC's authority extends only to Cadena's interest, which does not violate section 102.07(a). Because the TABC has authority over the permit applicant, Cadena reasons, its authority does not extend to FEMSA or its interests. Cadena contends that allowing the TABC's exercise of authority to prevail in this case will give it authority over every person anywhere in the world with a remote financial interest in a permit holder. But this argument confuses both the power of the TABC and the reach of section 102.07(a).

Section 26.03(a) provides that the rules governing a wine and beer retailer's off-premise license are the same as those governing a retail dealer's off-premise license. TEX. ALCO. BEV. CODE § 26.03(a). If the TABC protests a permit application, the county judge holds a hearing. Id. §§ 61.31-.32. Various other provisions then give the county judge power to grant or refuse the application. Id. §§ 61.42-.44. Cadena points to the various statutes' use of "applicant" to prove that the TABC and county judge's authority only extends to applicants. This is true. State agencies are statutory creatures and have no inherent authority other than those powers the Legislature expressly confers. Tex. Mun. Power Agency v. Pub. Util. Comm'n of Tex., 253 S.W.3d 184, 192 (Tex. 2007); Lakeshore Util. Co., 164 S.W.3d at 377. Section 61.43 only grants the TABC and the county judge authority to refuse an applicant's permit when granting it would cause the retailer to "conduct business in a manner contrary to law or in a place or manner conducive to a violation of the law." TEX. ALCO. BEV. CODE § 61.43(a)(9). But section 102.07(a), by its very terms, is much broader and applies to every person who meets the requirements to establish a prohibited interest. The Code gives the TABC authority to enforce provisions like 102.07(a) against permit applicants. Thus, if this permit were approved, the TABC would not have authority to force FEMSA to sell one of its interests in the retailing and manufacturing tiers based on 102.07(a). But the TABC does have authority, based on the power conferred by the Legislature, to reject Cadena's permit if granting it would cause Cadena to operate in a "manner conducive to a violation of the law." See id. Thus, we do not read 102.07(a) to provide the TABC with authority over every person who might have a prohibited interest.

In sum, the Code authorized the TABC to refuse to grant Cadena's permit application when granting it would have resulted in a violation of the tied house statutes. As it relates to this case, section 102.07(a)'s language permits the TABC to disregard the corporate separateness of entities when determining whether an entity has a cross-tier direct or indirect interest in the business of a retailer or an interest in the business of a brewer. FEMSA's indirect 100% ownership of Cadena provides it with an interest in the business of a retailer. Likewise, FEMSA's large stock ownership and ability to appoint members to the board of directors in the Heineken Brewers' holding companies provides it with an interest in the business of a brewer. Thus, the TABC and county judge were acting within their authority when they refused to grant Cadena's application for a permit upon finding that granting it would result in a violation of section 102.07(a).

C. Equal Protection

Cadena claims its equal protection and due process rights were violated by the TABC's arbitrary and discriminatory refusal to grant it a permit. As evidence, Cadena references an expert report it introduced into evidence which reflects significant and pervasive cross-tier holdings by publicly traded companies throughout the State of Texas. Cadena argues that its expert report also proves that the State of Texas holds billions of dollars of cross-tier investments and is itself in violation of the tied house statutes.

In administrative proceedings, the "rudiments of fair play" must be observed. Austin Chevrolet, Inc. v. Motor Vehicle Bd. & Motor Vehicle Div. of Tex. Dep't of Transp., 212 S.W.3d 425, 438 (Tex. App.-Austin 2006) (quoting Office of Pub. Util. Counsel v. Pub. Util. Comm'n, 185 S.W.3d 555, 576 (Tex. App.-Austin 2006, pet. denied)). An administrative "licensing authority acts arbitrarily and unlawfully if it treats similarly situated applicants differently without an articulated justification." Id. To establish an equal protection claim, a deprived party must show (1) it was treated differently from other similarly situated persons, and (2) no reasonable basis exists for the disparate treatment. See City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 439, 105 S.Ct. 3249, 87 L.Ed.2d 313 (1985); Mayhew v. Town of Sunnyvale, 964 S.W.2d 922, 939 (Tex. 1998). Cadena fails to meet the first element.

While Cadena points to evidence that cross-tier holdings are pervasive across the State, it does not show that any of the entities involved are similarly situated to itself. The primary evidence of overlapping ownership is the expert report submitted during the hearing before the county judge. The report is based on a review of the portfolios of four Texas Public Pension Funds and the defined contribution and benefit plans of publicly traded companies in the alcohol retailing and manufacturing industries. The experts reported that each Texas Public Pension Fund had cross-tier ownership interests totaling over $5 billion in the aggregate. According to the report, the State of Texas is a licensed retailer of alcoholic beverages through several state universities, which ostensibly causes it to be in violation of the tied house statutes because of the interest it holds in the manufacturing tier through the Texas Pension Funds. The report also provides evidence of significant indirect overlapping ownership, most often through mutual funds that own equity in alcohol-related companies, as well as direct overlapping ownership in multiple tiers by the defined retirement and benefit plans.

Despite the significant overlapping interests, the expert report does not provide evidence of any applicants that are similarly situated to Cadena. Each of the entity's cross-tier interests in alcoholic-beverage-related companies as outlined by the expert report is demonstrably and qualitatively different from those of Cadena, FEMSA, the Heineken Group, and its brewers. First, with respect to the defined employee contribution and benefit plans' cross-tier ownership, Cadena does not point to any entities holding cross-tier stakes that come close to FEMSA's multi-million share interest in the Heineken Group. Second, mutual fund managers generally have a duty to diversify the fund's portfolio. See RESTATEMENT (THIRD) OF TRS. § 227 cmt. m (AM. LAW INST. 1992) ("[A] vast array of pooled investment vehicles are now available to investors ... includ[ing] the shares of mutual funds ... [, which] offer trustees diversified holdings...."). FEMSA, of course, has not been shown to have such a duty and nothing suggests it would. And the Texas Public Pension Funds are heavily regulated statutorily created entities. See TEX. GOV'T CODE §§ 811.001-815.515 (Employees Retirement System of Texas); id. §§ 821.001-830.205 (Teacher Retirement System of Texas); TEX. EDUC. CODE §§ 43.001-.020 (Texas Permanent School Fund); id. §§ 66.01-.84 (Permanent University Fund). And the Texas Constitution caps the Permanent University Fund's security investments in a single corporation at 1% and stock ownership of a single corporation at 5%. TEX. CONST. art. VII, § 11a. In contrast, FEMSA is a non-governmental corporation with a far greater interest in both the Heineken Group and its subsidiaries — culminating in its indirect 100% ownership of Cadena — than the cap that the Texas Constitution places on the Permanent University Fund's investments.

The court of appeals determined that denial of Cadena's application was not an equal protection violation "because there is no evidence that the TABC has granted a permit or license to an applicant with a similarly significant cross-tier investment interest." 449 S.W.3d at 172. We agree. Although Cadena provides evidence that some entities have cross-tier holdings — both public entities and state pension funds — it provides no evidence that any of them are in a situation significantly, or even materially, similar to Cadena's. See Mayhew, 964 S.W.2d at 939; see also City of Dallas v. Jones, 331 S.W.3d 781, 787 (Tex. App.-Dallas 2010, pet. denied). We agree with the court of appeals' determination that Cadena failed to establish an equal protection claim based on the TABC's treatment of its permit application.

D. One Share Theory

Finally, we come to the one share theory addressed by both the parties and the amici. The dissent and one share theory advocates argue that if the court of appeals' decision is upheld, the TABC will have unlimited discretion to reject a permit application based on the applicant's ownership of single shares of stock in two of the three tiers. The essence of the argument is that this would produce an absurd result and give the TABC unbridled discretion to suspend almost any permit or reject almost any application. The amici seek "clarity" regarding how much, if any, cross-tier ownership is permissible under the statutes.

We are not unsympathetic toward the industry's desire for clarity. But as was recently said in Hall v. McRaven, going beyond the limits of our jurisdiction "is no trifling matter." 508 S.W.3d 232, 250 (Tex. 2017) (Guzman, J., concurring). And our lack of jurisdiction to issue advisory opinions such as the amici seek could not be more certain. Id. at n.18; Greene v. Farmers Ins. Exch., 446 S.W.3d 761, 767 (Tex. 2014); Rusk State Hosp. v. Black, 392 S.W.3d 88, 95 (Tex. 2012); In re Gen. Elec. Co., 271 S.W.3d 681, 693 (Tex. 2008); S. Tex. Water Auth. v. Lomas, 223 S.W.3d 304, 307 (Tex. 2007); Brooks v. Northglen Ass'n, 141 S.W.3d 158, 164 (Tex. 2004); Sw. Elec. Power Co. v. Grant, 73 S.W.3d 211, 222-23 (Tex. 2002); McAllen Med. Ctr. v. Cortez, 66 S.W.3d 227, 232 (Tex. 2001); Tex. Ass'n of Bus. v. Tex. Air Control Bd., 852 S.W.2d 440, 444 (Tex.1993); see also TEX. CONST. art. IV, §§ 1, 22 (empowering the attorney general, as part of the executive department of government, to issue advisory opinions to the governor and other officials); Valley Baptist Med. Ctr. v. Gonzalez, 33 S.W.3d 821, 822 (Tex. 2000) ("Under article II, section 1 of the Texas Constitution, courts have no jurisdiction to issue advisory opinions."). Addressing the one share issue head on would unquestionably be advisory in this case that is not about an application for a permit being denied because of one share of stock as the overlapping interest. This case is about FEMSA's 100% ownership of Cadena, its ownership of over 100 million shares of stock in the Heineken Group, and the resulting financial interests.

Cadena's application is about business. It is hard to conceive of a business enterprise that would refuse to divest itself of one share of overlapping stock in order to obtain a permit. And it is equally hard to conceive of the TABC expending the resources necessary to litigate a one-share cross-tier holding. But if it does happen, then the applicant will have opportunity to contest the TABC's decision based on the facts of that case and the statutes then in effect.

E. Response to the Dissent

Our role as a court is limited to determining legislative intent through the words the Legislature selected. The dissent's analysis and proposed result would significantly erode the Legislature's express mandate of strict separation between the three tiers of the alcoholic beverage industry, and would undermine the statutory construct designed to "assure the independence of the members of the three-tier system." TEX. ALCO. BEV. CODE § 102.75(c).

The dissent agrees that this case turns on the language "person who owns or has an interest in the business of a brewer," and that more specifically, it turns on the meaning of "an interest." In construing that language, the dissent does what we recently held to be error in ExxonMobil v. Coleman, 512 S.W.3d 895, 900-01 (Tex. 2017). That case concerned the construction of a provision in the Texas Citizens Participation Act (TCPA). Id. at 897-98. We reversed the court of appeals' judgment because it read language into the TCPA that narrowed its application. Id. at 897-98. What we said there applies here. First, we recited the unremarkable, but foundational, principle that "[a] court may not judicially amend a statute by adding words that are not contained in the language of the statute. Instead, it must apply the statute as written." Id. at 900 (alteration in original) (quoting Lippincott v. Whisenhunt, 462 S.W.3d 507, 508 (Tex. 2015)). Next, we noted that "the court of appeals improperly narrowed the scope of the TCPA by ignoring the Act's plain language and inserting the requirement that communications involve more than a `tangential relationship' to matters of public concern." Id. at 900 (citing Lippincott, 462 S.W.3d at 509 ("We presume the Legislature included each word in the statute for a purpose and that words not included were purposefully omitted.")). We expressly noted that "[e]ach of Coleman's arguments constitute[d] an effort to narrow the scope of the TCPA by reading language into the statute that is not there." Id. at 901. So it is here with Cadena's argument and the dissent's position.

In determining the plain language meaning of "interest," the dissent looks to legal and ordinary dictionaries in use at the time section 102.07(a)(1) was enacted. Post at 356-57. Those indicate that "interest" had a broad array of meanings. It "could mean anything from a mere concern or advantage to participation, a right, a share, or title." Post at 355. But rather than apply the presumption that the Legislature meant what it said and meant to use "interest" in its ordinary, broad sense, the dissent would narrow the meaning of the statute by adding words and engrafting into the statute a participation, control, or influence element that has no basis in the statutory language and is in direct contradiction of the express legislative directive regarding tied houses. It would significantly narrow the application of the statute. In essence, the dissent would manufacture a definition not found in the statute by adding words the Legislature did not enact.

The dissent further concludes that "`[i]nterest' must, then, mean something less than any interest, as the TABC maintains." Post at 360. It supports that statement by simply saying that the specific language at issue is narrower than the breadth of the Code as a whole. See post at 360-61. But this matter must be decided within the context of the Code as a whole, which is all about prohibiting tied houses. See Cont'l Cas. Ins. Co. v. Functional Restoration Assocs., 19 S.W.3d 393, 398 (Tex. 2000). After all, if the Legislature did not intend the tied house statutes to limit what alcoholic beverage industry participants could and could not do, what purpose would the statutes serve? And for the dissent to say that a major part of the statutory scheme is narrower than the whole, without pointing to any basis in the statutory language, is disingenuous.

The dissent further says that its interpretation is the only one that renders the statute enforceable. Post at 365-66. That is another way of saying that our construction, and that of the court of appeals, which is the construction urged by the TABC, is absurd or nonsensical. Indeed, the dissent expressly says just that. Post at 363-64. We disagree. The dissent attempts to show absurdity by stating "the logically inescapable extension of interpreting `interest' to mean any financial interest is the so-called single-share theory." Post at 363. It goes on to provide examples of the consequences of our reading of the statute. As discussed above, this Court is precluded by the Constitution from issuing advisory opinions. Advisory opinions are prohibited because they purport to bind future parties based on a "hypothetical injury," rather than "actual or imminent harm." Tex. Air Control Bd., 852 S.W.2d at 444. While any court interpreting a statute should consider "the consequences that result from each possible interpretation," the dissent relies on consequences it conjures that transcend the boundaries of realistic situations. We presume "the Legislature is bound to know the consequences" of its actions and only ask if "the Legislature intended those consequences." Cf. City of Desoto v. White, 288 S.W.3d 389, 395 (Tex. 2009). In this case, the Legislature provided a broadly inclusive statute but pared its reach by leaving its enforcement to the TABC. The TABC, an administrative agency, is afforded a great deal of discretion and deference. See State v. Malone Serv. Co., 829 S.W.2d 763, 767 (Tex. 1992). Indeed, that we are only now interpreting this statute for the first time — more than eighty years after it was enacted — suggests that the TABC has to date reasonably and effectively used that discretion and avoided pursuing the attenuated scenarios that trouble the dissent.

We recognize that under our interpretation, there likely are or will be violations of section 102.07 that will not come to light. But that does not render the statute meaningless or incapable of being enforced. There is a difference between enforceability and perfect enforceability. Experience teaches that no regulatory or police agency has the resources to enforce every statute against every violation. But laws against speeding, jaywalking, even murder, are not invalid because they are not perfectly enforceable. And it is not for courts to undertake to make laws "better" by reading language into them, absent the necessity to do so to effect clear legislative intent or avoid an absurd or nonsensical result that the Legislature could not have intended. Union Carbide Corp. v. Synatzske, 438 S.W.3d 39, 52 (Tex. 2014). That is not the circumstance here.

IV. Conclusion

FEMSA's indirect ownership interest in the Heineken Group and its breweries, together with its indirect ownership interest in Cadena, triggers the prohibitions outlined in section 102.07 as to Cadena and its application for a permit.

The judgment of the court of appeals is affirmed.

Justice Willett filed a dissenting opinion, in which Chief Justice Hecht joined.

Justice Boyd did not participate in the decision.

Justice Willett, joined by Chief Justice Hecht, dissenting.

That was the losing party's condemnation of England's pervasive tied-house regime in what is remembered as the "one man one drink" case. At the turn of the nineteenth century, an Englishman named Elliott Downs Till retired in the village of Eynsford, nestled within the farmland and woods of Kent, England. Till acquired considerable property in Eynsford and became invested in the value of his new surroundings. But one Eynsford establishment concerned Till: a small "public house" or beershop called "the Harrow," owned by a brewery, which Till would later describe as "a place [that] encouraged drunkenness."2 Till believed the establishment was better suited as "a proper place of accommodation" for travelers, so he contracted with the brewery to transform the Harrow into a country hotel called the Castle Inn.3 After completing renovations, Till entered into a 50-year lease with the brewery, promising to keep the Castle Inn open "on the greatest number of days and hours that the law allowed," one of the "usual tied-house conditions."4 He additionally pledged to "deal exclusively" with the brewery "for all porter, stout, beer, ale, or other malt liquors, whether draught or bottle, which should be sold or consumed" at the Castle Inn.5 But Till abhorred promiscuous drinking and wanted the Castle Inn to focus less on plying liquor and more on welcoming hotel guests. So Till tried to limit consumption at the Inn, imposing a semi-teetotaling "one man, one drink" policy. The brewery successfully sued, contending the liquor limits breached the lease.

Till's story was one that would be replicated throughout England, and ultimately America. Tied houses, saloons owned by or under exclusive contracts with producers, saturated pre-Prohibition America. But once the Twenty-First Amendment ended what President Hoover called the "noble experiment," the federal government and the states enacted sweeping legislation aimed at eliminating tied houses.

Texas enacted its own tied-house prohibitions in the 1930s and has amended them repeatedly since then. The key provision in today's case is section 102.07(a)(1) of the Alcoholic Beverage Code, which states "no person who owns or has an interest in the business of a ... brewer ... may ... own or have a direct or indirect interest in the business of a ... retailer."6

The Texas Alcoholic Beverage Commission denied Cadena Comercial's application for a retailer's permit because Cadena's publicly traded parent company, FEMSA — several levels of intermediate ownership removed from Cadena — owns a 20 percent stock interest in and can appoint 20 percent of the board members of two holding companies — several levels of intermediate ownership removed from FEMSA — that, in turn, own three foreign Heineken breweries. TABC claimed that granting Cadena's application would violate section 102.07(a)(1) because FEMSA would have an interest in both the business of a retailer (Cadena) and the business of a brewer (Heineken).

The court of appeals sided with TABC, holding that "interest" as used in section 102.07(a)(1) "broadly encompasses any commercial or economic interest that provides a stake in the financial performance of an entity engaged in the manufacture, distribution, or sale of alcoholic beverages."7 Applying this expansive, zero-tolerance standard, the court of appeals held the Code is violated any time any person owns any interest — even one share of stock, apparently — in two companies engaged in different tiers of the industry. Cross-ownership of "any financial interest" and any "potential for influence" violates section 102.07(a)(1), said the court of appeals,8 concluding that FEMSA has a forbidden interest overlapping the retailer (Cadena) and brewer (Heineken) tiers.

Distilled down, the issue is simply stated: What is a prohibited "interest" under section 102.07(a)(1)? Does Texas law forbid any degree of commercial connectedness, however trifling and attenuated, or are only certain cross-tier relationships prohibited, namely those that raise the specter of marketplace influence or coercion?

I would reverse and hold that FEMSA, Cadena's far-removed parent, does not have an "interest" in the business of a brewer within the meaning of section 102.07(a)(1). Reading the statute neither nonliterally nor hyperliterally, but contextually, as we must, it is apparent that "an interest" cannot mean "any interest" or "an interest of any kind" — two all-encompassing formulations used elsewhere in the Code. Texas tied-house law expressly proclaims the overriding objective of the three-tier system: "to assure the independence of members of the three-tier system."9 This is the textually manifest "public policy," enshrined into the Code itself. The Legislature used a specific term with a specific meaning: "independence," defined by Black's Law Dictionary as "[t]he state or condition of being free from dependence, subjection, or control."10 We need not deduce when we can derive. The most textually plausible interpretation is that "interest" connotes the ability to control, coerce, or influence business operations in another tier.

This record is devoid of any such cross-tier subjection. The corporate actors here remain independent, incapable of flexing monopolistic tendencies. The Corporate Governance Agreement expressly denies FEMSA "any right or control or influence or consultation right or other form of cooperation" relating to the Heineken Holding Companies.11 There is no verboten cross-tier coercion, or otherwise-illicit retailer-manufacturer overlap that amounts to a tied-house violation — not unless corporate structures are discounted, company agreements are disbelieved, and contextual statutes are disregarded.

Because the Court holds otherwise, I respectfully dissent.

I. THE HISTORY OF LIQUOR REGULATIONS LEADING TO TIED-HOUSE STATUTES

A. England's Licensing Scheme and the American Pre-Prohibition Era

Tricking, treating, and prostitution.12 History has regarded saloons and public houses with no small measure of disdain, deeming them places of ill repute where society's seediest subjects ply their trades on innocent passersby. Regardless, the industry has been around for millennia. And so, the industry has been regulated for millennia.

Some of the first recorded alcohol regulations took the form of criminal penalties for intoxication.13 In the year 1000, governments began taxing the import of liquor.14 But the real uniformity in regulation began in the 16th and early 17th centuries during the reigns of English monarchs Edward VI and Elizabeth I. In 1552, Edward decreed that ale makers must receive permission to sell their wares within the empire.15 The new system was rife with corruption, however, and led to monopolies that the justices of the peace had little authority to curb.16 In the early 1600s, Parliament, fed up with the corruption, passed laws giving the justices power to enforce the regulations. Incremental legislation eventually led to the Licensing Act of 1627, which required makers and sellers of "spirits" to obtain a bona fide license from the governing authority.17 "An irrevocable monopoly was now being brought more and more into subservience to royal authority."18 These licenses were typically only granted to taverns and inns and were for the sole purpose of offering refreshment to travelers.19 Retailers were brought into the licensing scheme in 1737.20 Under the first tied-house-type statute, the "Gin Act of 1751," monetary penalties were assessed against distillers found to also be selling their own product.21

As the regulatory noose tightened on consumers and producers alike, justices of the peace began requiring proof of suitable premises before issuing licenses.22 Eventually, this forced would-be tavern owners to seek loans to underwrite these "suitable premises"; breweries were all too eager to fill in this financial gap, buy the property, and lease it to the license applicant.23 The price for this good deed? Not much. The lessee need only agree to sell their benefactor's products — and only those products.24 "This was the purchase of public houses. When one big firm suddenly realized the profit that it would be possible to make by buying public houses and keeping them under managers of their own for the sale of their own beer only the others rushed in madly."25 And "madly" well describes the speed with which privately owned public houses were transformed into tied houses. As one scholar noted in 1895, under the tied-house system, "It is safe to say that not more than twenty-five per cent of the inns and beerhouses are free from the brewers."26 And thus the regulatory effort designed to dispose of monopolies in the alcohol trade led to the creation of the English tied-house system.

In America, too, the tied-house system figured prominently in the pre-Prohibition era. In the early 1900s, beer consumption reached a then-historic high of 21 gallons per capita annually.27 More than 150,000 saloons existed in the United States to facilitate that consumption.28 Even so, Americans exhibited an air of superiority and contempt regarding England's tied-house system. A 1905 article in the San Francisco Call reported on and lambasted an Englishman's "Turn[] on [the] Purity Movement He Supported to Lease Them to Brewers."29 The man had previously worked with a movement that promoted the end of the tied-house system — in the author's words, a scheme whereby public houses "should not be owned by the large breweries, whose interests necessarily lie in selling as much beer as possible."30 The man later replaced tenants at one of his own public houses with a brewery, effectively transforming those houses into tied houses. This, the article disdainfully noted, was "a striking example of the sacrifice of principle to profit by no means rare in the British aristocracy."31 An article published in 1892 entitled "To Stop Beer Wars" gave one of the first glimpses into what Chicago could experience under a tied-house system.32 The article explained that two English brewing companies had invested $6 million in new Chicago enterprises. According to a "director" (presumably a representative of one of the companies), "In England 99 per cent of the places selling Bass or Bullard's or Guinness beer are controlled by the brewer ... and we propose to stop beer wars by owning saloons."33 The most important part, in the director's view, was "that all the `tied houses' we secure will be our customers for all time; [they] will be unaffected by beer wars and cannot be taken away from us by keener competition."34

B. American Prohibition and Its Aftermath

Concerns grew and public sentiment eventually turned against saloon owners. The temperance movements waged a focused and organized war — and would settle for nothing less than complete abolition of liquor. One of the movement's chief tactics was to put up members of pro-temperance groups as candidates for office. This was particularly helpful on a national scale. For example, in 1913, the Congress overrode President Taft's veto of the Webb-Kenyon Act by a vote of 246 to 95.35 The Act had prohibited the importation of alcohol into states that had already decided to prohibit alcohol within its borders. The Temperance Movement was suddenly much more than a religious movement out to save souls — it was a powerful political machine with a sudden big victory in its pocket.

Emboldened by the movement's sudden success, Wayne Wheeler, leader of the powerful lobbyist group the Anti-Saloon League, moved to Washington for the express purpose of convincing members of Congress to enact a prohibition amendment.36 Bankrolled by John D. Rockefeller, Jr. and other pro-temperance industrialists, the League dominated the movement in the early 1900s, achieving prohibition in 30 states by 1919. And in 1913 the League declared support for a federal constitutional prohibition. Amending the Constitution is a heavy lift, requiring a two-thirds majority vote from each house of Congress and then adoption of the amendment by 36 of the then 48 state legislatures.37 In 1914, a resolution calling for an amendment to the Constitution passed out of the House committee and reached the floor for the first time in history.38 Most House members voted for the amendment, but supporters failed to muster the two-thirds necessary to pass it out of the chamber.39 The close vote, however, reinvigorated the movement and steeled its leaders' determination to win the war by electing enough members to Congress so that next time, the amendment would pass. And pass it did. In 1917, after the 1916 election swept temperance members into office across the country, Texas Senator Morris Sheppard led the newly elected Congress to the overwhelming approval of the prohibition resolution that would become the Eighteenth Amendment.40 The necessary state ratification followed in quick succession.

But the ingenuity of the American people would eventually defeat Senator Sheppard, Wheeler, the temperance movement, and the Eighteenth Amendment. Citizens resorted to any means — legal or illegal — to gain access to alcohol. The gains of the 1920s eroded as organized-crime syndicates rose in power, in part by providing illegal liquor to the masses.41 And then came the Great Depression. By the early 1930s, the country was in disarray, and enforcement of Prohibition laws was not the foremost priority of the federal and state governments. The mass disobedience and rampant lawlessness spurred Rockefeller, a teetotaling member of the temperance movement, saying:

President Roosevelt urged passage of the Twenty-First Amendment, promising that taxes from alcohol sales would benefit the nation.43 The amendment sailed through both houses of Congress and was ratified by the states in a flurry reminiscent of the Amendment it repealed.

Rockefeller, while favoring repeal, believed the end of Prohibition would not be the end of America's "liquor problem,"44 saying, "If carefully laid plans of control are not made, the old evils against which prohibition was invoked can easily return."45 But he saw the writing on the wall regarding the Twenty-First Amendment and commissioned a report on the nation's impending liquor regulation.46 Toward Liquor Control, written by scholars Raymond Fosdick and Albert Scott, attempted to be the first study to get in front of the problem of how states should grapple with Prohibition's repeal.47 Toward Liquor Control was highly influential, leading to model legislation for states regarding the manufacture and sale of alcoholic beverages.48 Fosdick and Scott suggested two post-Prohibition systems of regulation: (1) a monopoly approach (which they strongly preferred) that managed all alcohol sales through the government, or (2) licensing the alcohol industry under the auspices of a state regulatory board.49

As Prohibition's repeal took hold, President Roosevelt encouraged the states to enact sensible liquor legislation.50 Most states adopted some version of state licensure, abiding Toward Liquor Control's admonition to guard against tied houses by keeping the tiers of the industry separate from one another.51 States implemented what has come to be known as the "three-tier system,"52 licensing manufacturers, distributors, and retailers separately, and barring a firm in one tier from owning a firm in another. Businesses must operate only within their assigned tier. Today, 17 states use the monopoly-control model, while 33 states plus the District of Columbia have a licensing system.53

As in England, America's tied-house legislation took aim at the monopolistic tendencies of the brewer-retailer relationship. Connecticut's courts recognized the design of its tied-house laws, remarking the law codified an intention "to circumvent the concentration of tremendous power and inordinate control in the hands of wholesalers and manufacturers, who, by reason of economic superiority and the extension of generous business credit, might well be so circumstanced as to throttle the retail dealer and monopolize the retail market."54

The resulting tied-house legislation, however, was not solely the product of concerned legislators. To the contrary, brewers drafted the initial federal tied-house legislation on which much of the subsequent state legislation would be based. Self-interest existed on all sides. The refrains of the temperance movement and a general desire to cure society of the evils of alcohol may have prompted the legislation. But the brewers, too, understood that strict tied-house prohibitions would reduce competition among brewers — no more would there be endless vying for new real estate and saloon acquisitions because tied-houses would be prohibited. The brewers also knew that tied-house prohibitions would allow them some degree of plausible deniability regarding the negative public perception of saloons.55 Brewers that owned saloons in a tied-house system were, in a sense, directly connected to and responsible for any perceived drunkenness and debauchery. But under a tied-house-less system, brewers can sell their goods, while disclaiming any participation in or responsibility for the happenings in the saloons. Indeed, the U.S. Brewers Association attempted to use this precise argument regarding many brewers' lack of ownership of saloons. The Association disclaimed "a popular misconception" that brewers often owned all the rights in saloons.56 To the contrary, said the Association, brewers typically only lend money to saloon owners or take mortgages on saloon property; they don't own the saloons themselves. Therefore, the Association concluded, brewers largely cannot be considered responsible for saloons.57

C. The Promulgation of Liquor Regulations in Texas

The temperance movement was as prevalent in Texas as it was elsewhere in the nation. Some sources place the Texas membership of groups such as the Sons of Temperance at 3,000 by the late 1840s.58 Concerned citizens urged their representatives to pass laws providing for the regulation and elimination of intoxicating beverages. In 1854, the Legislature introduced a ballot measure that ordered the governor to hold an election in which the individual counties would vote on the prohibition of saloons or any other entity selling liquor in quantities less than a quart.59 Though a majority of the counties voted not to issue licenses to liquor salesmen, we held the law unconstitutional, and it was never enforced.60 Undaunted, anti-liquor groups like the United Friends of Temperance, Bands of Hope, the Women's Christian Temperance Union, and the Grange, among others, flocked to Texas and waged war on Texas saloons. Collectively, these temperance groups were called "Drys," in reference to their desire to turn the State of Texas "dry" from alcohol.61

Despite losing several referendum elections in which the Drys attempted to convince Texas citizens and legislators to prohibit alcohol, their numbers grew. And with each battle, the pro-Prohibition crowd drew support. Incremental changes came to the state's liquor laws. Though not outright bans, regulations greatly reduced the role of the saloon in Texas.62 The new Constitution of 1876 adopted local-option laws, where individual communities and counties could permit or prohibit the sale of alcohol. Finally, in 1918, the Texas Legislature adopted the language of the proposed national prohibition amendment. As across the nation, sweeping Prohibition legislation was enacted, including the Dean Law, which banned the manufacture of any liquor for any purpose.63

The year 1933 brought the end of Prohibition across the nation. In response to the repeal of the Eighteenth Amendment, the Texas Legislature passed the Texas Liquor Control Act and Texas voters adopted an amendment to the Texas Constitution legalizing the sale of beer. Nevertheless, 199 of the state's 254 counties chose to utilize the local-control option and maintained a complete prohibition of the sale of alcohol; only 10 counties were free from any form of regulation. After the Twenty-First Amendment repealing Prohibition took effect in December 1933, the Texas Legislature submitted another amendment to the Texas Constitution that would completely repeal the vestiges of statewide prohibition. Texas voters approved the amendment in 1935,64 restoring the right of individual communities and counties to decide whether alcohol could be sold within their respective borders. Also in 1935, the Legislature created the Texas Liquor Control Board (renamed the Texas Alcoholic Beverage Commission in 1970) to administer and enforce the Liquor Control Act.65 The Act governed all alcohol-related transactions from its inception in 1933 until 1977, when it was superseded by the Texas Alcoholic Beverage Code.66

The State adopted the three-tier system to ensure the independence of retailers, manufacturers, and distributors and to prevent the re-creation of tied-houses that Prohibition had largely eliminated. The Liquor Control Board was tasked with supervising and regulating every phase of the state's alcoholic-beverage industry. As one court put it shortly after the Board's creation, the Legislature delegated to the Board "certain functions, among which are determining in the first place to whom and when shall certain privileges be extended to persons to sell liquors, and second, whether or not such persons so favored have breached the conditions under which the privilege has been granted."67 The Board's overall mission and purpose mirrored the stated purpose of the Liquor Control Act, which preceded our current Alcoholic Beverage Code.68 Currently, TABC's mission includes a charge to "ensure fair competition within the alcoholic beverage industry [and] ensure consistent, predictable, and timely enforcement of [the Code]."69

II. FACTUAL OVERVIEW

Petitioner Cadena Comercial USA Corp. ("Cadena"), a Texas corporation, organized to own and operate convenience stores in the state under the Mexican brand "OXXO," wants to sell beer and wine. Cadena is a wholly owned subsidiary of Fomento Económico Mexicano, S.A.B. de C.V. ("FEMSA"), a publicly traded company that owns Cadena through several intermediary Mexican holding companies. FEMSA also, via various intermediary holding companies in the United Kingdom, holds a combined twenty-percent stock interest in Heineken NV and Heineken Holding NV ("Heineken Holding Companies"). The two publicly traded Heineken Holding Companies, through its own series of intermediary companies, own three foreign Heineken brewers ("Heineken Brewers"). The TABC issued non-resident permits to each of the Heineken Brewers, allowing them to manufacture beer at foreign breweries, but none of the Heineken Brewers has a brewery in Texas. The business structures are complicated, to put it mildly, as the detailed chart attached to this opinion illustrates.

It is undisputed that Cadena has no direct interest in the Heineken Brewers (or in any other entity involved in the alcoholic beverage industry). Similarly, the Heineken Brewers have no interest in Cadena (or in any other entity within the FEMSA corporate family). When FEMSA obtained its twenty-percent stock interest in the two Heineken Holding Companies, it entered into a Corporate Governance Agreement that, among other things, entitles FEMSA to appoint one of Heineken Holding, N.V.'s five directors and two of ten members of the Supervisory Board of Heineken N.V.70 L'Arche Green, the controlling-interest holder of the Heineken Holding Companies, was obligated to sponsor FEMSA's choice for board members, but the Corporate Governance Agreement specifies that FEMSA is not given "any right or control or influence or consultation right or other form of cooperation" relating to the Heineken Holding Companies.71 Similarly, L'Arche Green and Heineken Holding reserved all rights to make decisions in its management of the Heineken Holding Companies, "independently and at their sole discretion and without any requirement to consult or cooperate with ... FEMSA."72 The Agreement also bars Heineken from acquiring any stock in FEMSA.

FEMSA's stores under the OXXO brand must be authorized to sell beer and wine in order to thrive in the convenience-store market. Consequently, Cadena applied for a wine and beer retailer's off-premises permit from TABC. In applying the tied-house provisions, TABC disregarded the legal separateness among the various entities in both the FEMSA and Heineken corporate families. For example, TABC collapsed the elaborate corporate structure separating Cadena from its parent, FEMSA, and considered them a single business enterprise. TABC then collapsed the corporate structure separating the Heineken Brewers from their parent holding companies, and considered them a single enterprise, also. TABC's expert testified that corporate separateness can be disregarded in the regulatory area, arguing "the layers of entities here ought to be disregarded for regulatory purposes." According to TABC, any FEMSA shareholder would have an interest in a Heineken brewer. TABC's licensing director testified that a tied-house violation would exist if someone owned FEMSA stock and also stock in a TABC-permitted retailer. Indeed, TABC interprets the Code to bar overlapping ownership of even one share of permittees at different tiers, notwithstanding multiple layers of intervening ownership. TABC then reviewed the relationship between FEMSA and the Heineken Holding Companies and determined there was a forbidden retailer-brewer overlap that violated five different tied-house statutes.

FEMSA refused to divest itself of its indirect shareholder's interest in the Heineken Brewers, and TABC's denial proceeded to an administrative hearing. At the hearing, the parties stipulated to the corporate relationships between Cadena, FEMSA, and the Heineken companies.73 During the hearing, a witness — TABC's licensing director and an expert in alcoholic-beverage industry laws — testified that even one overlapping share of stock ownership would violate the statutory tied-house prohibitions. TABC argued that FEMSA's overlapping interest in Cadena and the Heineken Brewers was sufficient to be considered a prohibited "interest" under the Texas Alcoholic Beverage Code under any interpretation. TABC disputed that actual control is required to implicate the pertinent tied-house restrictions, but asserted FEMSA could control the Heineken Brewers because of its ability to appoint directors to the Heineken Companies' boards. It also argued the court should impute this connection to Cadena for purposes of regulation.

Conversely, Cadena argued that the only "interest" sufficient to violate tied-house prohibitions is one that would allow actual financial or administrative control among at least two of the three tiers. Under Cadena's interpretation, its permit application should have been granted as a matter of law because FEMSA has no ability to manage or control either the Heineken Holding Companies or the Heineken Brewers. As a result, Cadena argued, granting its application would not violate the purpose of the tied-house statute because no company within the business structure would have managing control over more than one tier. Cadena further argued that FEMSA's connection with the Heineken Brewers was remote and far too attenuated to implicate historical tied-house concerns and that this interest could not be imputed to Cadena without piercing the corporate veils of all the entities involved.

At the administrative hearing, the judge denied Cadena's application based on the statutory grounds cited by TABC, finding: (1) Cadena "has a real interest in the business or premises of the holder of a manufacturer's or distributor's license"; (2) "[f]or licensing purposes, as a subsidiary of FEMSA, [Cadena] is a manufacturer"; (3) "for licensing purposes, as a subsidiary of FEMSA, [Cadena] has an interest in the business of a brewer"; and (4) issuing "the requested permit would violate Sections 102.01(c), (h), 102.07(a)(1), and 102.11(1) of the Code."74 Cadena appealed to the trial court, which affirmed the administrative order denying Cadena's permit.75

The court of appeals noted it could affirm on any of the grounds cited in the administrative order, but focused specifically on section 102.07(a)(1), which provides that "no person who owns or has an interest in the business of a ... brewer... may ... own or have a direct or indirect interest in the business ... of a retailer."76 In analyzing this section of the Code, the court defined several terms and phrases within the statute, including, "person," "brewer," "retailer," "owns or has an interest in the business," and "own or have a direct or indirect interest in the business."77 The Code provides a definition for "person," so the court then held that all parties — FEMSA, Cadena, and the Heineken Brewers — were "persons" under the statute.78 The court also held the Heineken Brewers fell into the category of "brewer," just as Cadena would qualify as a "retailer" if its permit were granted.79 Thus, the court's main point of analysis became, "whether FEMSA owns or has an interest in the business of the Heineken Brewers and also owns or has a direct or indirect interest in the business of Cadena."80

According to the court of appeals, the issue was "the extent to which section 102.07(a)(1) implies a requirement that a disqualifying `interest' carry with it some degree of cross-tier control and whether implying such a requirement is essential to avoid rendering the statute unconstitutionally vague."81 Observing the Code failed to define the terms "own," "interest," and "business," the court of appeals attempted to "apply [the terms'] common meaning to the extent consistent with the context in which they are used and the statute's objective."82

The court of appeals, like TABC, contracted the entities' corporate separateness, then expanded the definition of "interest," concluding it "broadly encompasses any commercial or economic interest that provides a stake in the financial performance of an entity engaged in the manufacture, distribution, or sale of alcoholic beverages."83 The court rejected Cadena's argument for a control-focused test, stating it lacked a "foundation in the statutory text" and held FEMSA's financial and economic interests in both the Heineken Brewers and Cadena qualified as "interests" for the purposes of section 102.07(a)(1).84

The court then turned to the term "business," deciding its use in section 102.07(a)(1) was intended to be broad and include any "commercial enterprise carried on for profit."85 Under these sweeping definitions of "interest" and "business," the court held that section 102.07(a)(1)'s "plain language applie[d] to FEMSA's own relationships, not relationships ... imputed to FEMSA."86 The court then addressed Cadena's veil-piercing arguments, but held even if the principles of corporate law applied "in some regulatory contexts ... those principles are not implicated by the broad language the legislature employed in section 102.07(a)(1)."87 Because it determined TABC granting Cadena's permit would result in FEMSA having an interest in the business of a brewer and a direct or indirect interest in the business of a retailer, the court affirmed the order denying Cadena's permit.88

III. ANALYSIS

no person who owns or has an interest in the business of a ... brewer ... may... own or have a direct or indirect interest in the business ... of a retailer89

The parties agree that the Heineken Brewers are "brewer[s]," and that Cadena would be a "retailer" if granted a permit. They disagree, however, over (1) whether FEMSA is a "person," and even if so, (2) what "the business of a ... brewer" encompasses, and (3) whether FEMSA has an "interest" in the business of a brewer. For purposes of this dispute, I assume without deciding that FEMSA is a "person" and that "brewer" covers the Heineken Holding Companies that own the Heineken Brewers. The interpretive focus is thus narrow: What is a prohibited "interest" under section 102.07(a), and does FEMSA have such an "interest" in the Heineken Brewers' business?

The parties' arguments are straightforward. Cadena insists the word "interest" connotes a sense of control — i.e., FEMSA must exert control over the Brewers' business. TABC rejects a control-based test and endorses the court of appeals' view that "interest" in section 102.07(a) "broadly encompasses any commercial or economic interest that provides a stake in the financial performance of an entity engaged in the manufacture, distribution, or sale of alcoholic beverages."90

Two significant points merit mention at the outset. First, TABC reaffirmed both at oral argument and in a post-argument letter to the Court that the Code recognizes "no de minimis exception," a Latinized locution of the single-share theory — that even one overlapping share of stock constitutes a tied-house violation. There is no practical difference between "no de minimis exception" and the "single-share theory." The terms are functionally indistinguishable. Second, the Corporate Governance Agreement makes clear that FEMSA lacks the ability to control or manage, either directly or indirectly, any aspect of any Heineken entity.

As explained below, a contextual reading of section 102.07(a)(1) forecloses TABC's crabbed "one share" interpretation, which by including all excludes all. This hyperliteral construction is not fair-reading textualism, which is not allergic to interpretive aids like context. In short, the court of appeals rejected a test it says lacks a textual basis in favor of one lacking a contextual basis, a zero-tolerance test divorced from the reality of existing permittees who, as TABC concedes, hold billions of dollars in cross-tier holdings. The State of Texas, for example, through its public universities, is a retail permittee that sells alcohol at certain sporting events and mixed beverages at other university events. It also owns billions of dollars in cross-tier investments. Had TABC treated the State's application as it treated Cadena's, it would have rejected it. In other words, the State of Texas (as regulator) says the State of Texas (as regulated), is operating illegally and thus at risk of forfeiting its permit.91

A rational, fair-reading test cannot arbitrarily depend on who is being tested — strict for some, loose for others. Laws must be applied consistently, giving fair notice to what conduct is prescribed and proscribed. The Court's interpretation vests TABC with enormous power — rewriting statutes, collapsing separate corporate entities without an evidence-based veil-piercing inquiry, selectively applying standardless criteria in a manner that treats similarly situated applicants dissimilarly, thus picking winners and losers in the marketplace. Virtually all applicants are implicated by such a sweeping reading of "interest," a reading that bans any indirect interest of any degree — except when it doesn't. Such an arbitrary and selective permitting regime cannot be squared with Texas law, particularly the Legislature's explicitly stated public purpose of ensuring "independence," i.e., the absence of out-sized cross-tier influence or coercion.92

The most reasonable interpretation is this: A forbidden "interest" under section 102.07(a) connotes control, coercion, or influence over business activities in another tier, participation that imperils the Legislature's codified objective of no vertical strong-arming. No such interest is present here.

A. Judges Must Read Statutes Neither Literally Nor Liberally, But Commonsensibly, Discerning Words' Accepted Contextual Meaning

First things first. In statutory-interpretation cases, we are to begin (and almost always end) with the Legislature's chosen text, the surest index of lawmakers' collective will.93 Interpreting statutes is the bread and butter of modern appellate judging, and this Court has stated its view simply: The truest manifestation of what lawmakers intended is what they enacted.94 As an interpretive method, textualism has a singular objective: ascertaining words' accepted contextual meaning when they were enacted. No concern with abstract (and thus manipulable) purposes. No concern with wished-for (and thus preference-imposing) consequences. Just an unremitting focus on giving words their contextual meaning — not literal and not liberal, but commonsensible.

And by commonsensible, I mean communal, what the enacting community understood their words to mean. This is the second point: "Words must be given the meaning they had when the text was adopted."95 On this important principle, we recently quoted Justice Frankfurter: "Words must be read with the gloss of the experience of those who framed them."96 Our cases agree. When interpreting language, both statutory and constitutional, we aim to determine original public meaning, what the words meant to those who wrote and ratified them.97

Third, text cannot be divorced from context. It is said that text without context is pretext. This is a straightforward, well-defined interpretive principle, one we have asserted frequently and applied assiduously. The law, after all, begins with language, and one cardinal rule of language — not just legal language but all language — is this: "Language cannot be interpreted apart from context."98 For judges to play their lexicographic role in the legislative project, we must be attentive not just to words standing alone, but to structure and historical architecture. On this fateful point, our precedent is clear: A judge scrupulously concerned with giving legal texts their honest meaning must always consider "the surrounding statutory landscape" and welcome, not resist, interpretive context like linguistic usages and sound interpretive conventions that help illuminate meaning.99 Yes, this case is about the legal interpretation of one word, "interest," but that task requires us to understand the meaning of surrounding words and how they are put together.

The interpretive process recognizes that lawmaker-drafters, not judge-interpreters, enact language. Judges' interpretive role is to discern, not dictate, how the Legislature uses language. The Legislature authors policy and its baked-in political bargains, and the judiciary, to avoid aggrandizing its confined-but-consequential role, must examine all the enacted text, not mere snippets. Yes, a statute's words reign supreme, but when seeking statutory meaning, ascetic literal parsing can sometimes cloak rather than clarify. Even when construing an ostensibly clear statute that seems intuitively obvious, we may consider related legislation plus other contextual cues to glean the text's semantic import — not extra-statutory materials like legislative history, but the full range of intra-statutory aids: grammatical conventions, dictionaries, specialized legal or technical usage, colloquial nuances, and so forth.100 Textualism is not literalism, and courts ought not adopt wooden constructions foreclosed by statutory context, as the Court has done here.101 The import of language, plain or not, must be drawn from its linguistic context, a self-evident rule rooted in common sense,102 Texas statutory law,103 and caselaw from both this Court104 and the United States Supreme Court.105

B. "One Share" Is One Dimensional — "An Interest" Should Not Mean Any Interest, However Attenuated, But Only an Interest That Threatens Cross-Tier Control or Influence

"Literalness may strangle meaning."106

When divining what enacted law means, the judge-interpreter's aim is not a myopic reading, but a sound one. Reading clinically does not mean reading under a microscope. Today's case concerns "interest," but that term (and its meaning) is found within a larger enactment. Accordingly, we must resist hyperliteralism — "a sterile literalism which loses sight of the forest for the trees"107 — in favor of "literal meaning in context."108

Sometimes the Legislature helpfully defines the terms it uses,109 imbuing even ordinary words with technical (and usually conclusive) meaning.110 The Alcoholic Beverage Code, for example, has a general definitional provision, defining 28 terms, everything from "minibar" to "wine cooler."111

1. The Legislature's Codified Purpose (to Ensure "Independence") and TABC's Own Internal Guidance Buttress the Determination That "An Interest" Connotes Something More Than De Minimis

Unfortunately, the Code nowhere defines "interest." When a statute is silent, judges often seek guidance in reputable dictionary definitions, particularly legal dictionaries from the enacting era, since semantic usage and nuances can shift over time. Not all dictionaries are created equal, however; some are richer and more explanatory. Unfortunately, the definition of "interest" in the 1910 edition of Black's Law Dictionary, available at the time Texas originally adopted its tied-house laws, isn't helpful in divining the common, preferred usage.112 Neither are the early-1900s definitions of "interest" in the Chambers and Oxford dictionaries, which provide little aid in narrowing the possible definitions relevant here.113 Unsurprisingly, the word "interest" can mean various things, depending on context. Using a more modern dictionary, "interest" is defined as "the power of influencing," or "persons effectively controlling an enterprise or dominating a field of activity," or "the dominating group of owners in a field of business, industry, or finance."114

In 2013, the Legislature expressly — and helpfully — declared the public policy aim of the three-tier system.115 When legislators articulate an explicit purpose in the very words of the statute, the Court need not — and should not — speculate.116 As two eminent legal lexicographers put it, "words are given meaning by their context, and context includes the purpose of the text."117 Here, the Legislature codified its objective rather plainly: "It is the public policy and in the interest of this state to assure the independence of members of the three-tier system."118 Maintaining cross-tier independence, then, is the manifest object of Texas tied-house laws. Lawmakers defined their purpose precisely, concretely, and succinctly — not to supplant text, but to give it meaning.

By contrast, it makes no interpretive difference that the Code provides, as statutes commonly do, that it should be "liberally construed," here to protect "the welfare, health, peace, temperance, and safety of the people of the state."119 On this point — whether the legislative branch can command the judicial branch to put a liberal (or nonliberal) gloss on language — I answer no.120 Yes, the Code's overall aim is legislatively (if nebulously) prescribed: to promote "welfare, health, peace, temperance, and safety,"121 akin to the Constitution's goal "to form a more perfect Union, establish Justice, insure domestic Tranquility...."122 But diaphanous, good-government precatory provisions ought not expand text beyond its contextual meaning, lest all gauzy phrases become judicial license to impose wished-for outcomes. The judicial goal must remain fixed — ascertaining fair meaning — "the meaning that causes it to make sense."123 Legislators can certainly supply text-specific definitions, and a purpose or preamble provision can clarify textual ambiguity, as section 102.75(c) does here, proclaiming cross-tier "independence" (read: no undue control/coercion/influence) as the goal of the three-tier system.124 But lawmakers, by commanding "liberal construction," cannot "instruct judges to put a thumb on the scale in this fashion."125 In our system of separated powers, interpreting laws is a quintessential judicial function.

What, then, did legislators mean by "independence"? As with "interest," the Code itself is silent. But the contemporaneous edition of Black's Law Dictionary offers on-point guidance, defining "independence" as "[t]he state or condition of being free from dependence, subjection, or control."126 Similarly, "independent" is described as "not subject to the control or influence of another."127 The Court contends independence can only be achieved by strict and total separation, but the Code says no such thing, and the Texas alcoholic-trade landscape, replete with TABC-permitted cross-tier investments, punctuates the point. Rather, as the definitions of "interest" and "independence" underscore, the Code's tied-house restrictions aim to rein in cross-tier coercion and subjection.

TABC's own regulations actually make this point, stating that under section 102.07(a)(1), a "direct or indirect interest in the business of a retailer" means an interest held by the manufacturer sufficient to "place retailer independence at risk."128 Likewise, TABC's Application Guide for Retailers centers on control, stating that applicants "cannot control, in any fashion, the interests of a licensee/permittee at a different level."129 TABC's license application form for new businesses, when discussing the prohibition of cross-tier interests, explains that applicants "cannot control in any fashion the interests of a licensee/permittee at a different level."130

Courts have long understood the fundamental purpose of Texas tied-house laws to be prohibiting "vertical integration" in the alcohol industry, to prevent anyone from "controlling" or "dominating" business operations in multiple tiers.131 This is not an outlier view but rather a familiar one shared by courts132 and state regulators133 throughout the nation. The term "interest" in section 102.07 must be construed in light of the legislatively declared purpose of Texas tied-house laws: "to assure the independence of members of the three-tier system."134 And "independence" connotes the absence of control, just as "interest" connotes the presence of control — or at least influence sufficient to portend forbidden cross-tier coercion.

2. The Code's Varying Use of "Interest" — Sometimes Narrow, Sometimes Broad — Supports an Interpretation That, As Used Here, the Term Means a Degree of Influence That Imperils Cross-Tier Independence

Courts must analyze statutes in their entirety — not cherry-picking individual words or phrases to discern meaning. The Court adopts the court of appeals' definition of "interest" as "encompass[ing] any commercial or economic interest that provides a stake in the financial performance of [a brewer]."135 But this considers the provision in a vacuum rather than in context, and adopts in practice TABC's "no de minimis exception" standard — a standard so broad as to be no standard at all. In doing so, the Court disregards the entirety of the text, separately defining discrete words and then cobbling together those separate definitions. The provision must be defined as a contextual whole, however, not merely as a sum of its stand-alone parts.136

The tied-house provisions focus intently on prohibiting certain relationships (and allowing others), using the word "interest" many times,137 and usually preceding "interest" with a modifier:

Section 102.07 is structured differently. Again, here's the relevant text:

It bars a person who has "an interest" in the business of a brewer from having "a direct or indirect interest" in the business of a retailer. The latter phrase, with its "direct or indirect" modifier, indicates that "interestedness" with a retailer invites a broader examination than with a brewer. The Code bars someone with an interest in a brewer from having not merely an "interest" in a retailer but something more expansive, "a direct or indirect interest." This looser "direct or indirect" formulation casts a wider net, expanding the scope of "interest" to thwart the classic tied-house arrangement of brewers controlling retailers.

The Court goes another route, holding that "interest" standing alone and unmodified necessarily includes every modifier sprinkled throughout the Code and "broadly encompasses any commercial or economic interest that provides a stake in the financial performance of an entity engaged in the manufacture, distribution, or sale of alcoholic beverages."145 All interests. Of any type. In any degree.

I disagree. This reading is far too broad, because "interest" here (1) is not modified by "direct or indirect," and (2) is modified by the phrase "in the business of a brewer," thus it cannot mean any interest of any kind in a brewer, but logically only a direct interest in a brewer's business.

Neither Cadena nor its parent FEMSA has any such prohibited interest in the Heineken Brewers. Neither has legal or equitable title to the Heineken Brewers' business or stock in the Heineken Brewers, or any other direct way to exert cross-tier influence or control. Moreover, the Corporate Governance Agreement makes clear that FEMSA lacks any ability to control or manage, either directly or indirectly, the business of the two Heineken Holding Companies, which in turn own the Heineken Brewers. As noted above, the Agreement states explicitly that FEMSA has no "right or control or influence or consultation right or other form of cooperation" relating to the Heineken Holding Companies.146 Similarly, Heineken Holding reserved all rights to make decisions "independently and at their sole discretion and without any requirement to consult or cooperate with ... FEMSA."147 The Code's tied-house provisions aim to promote "independence,"148 and FEMSA lacks any ability to direct Heineken Brewers.

The Legislature's overarching concern with "independence" — minimizing cross-tier control and coercion — permeates the Code, which bars relationships that exert undue influence. Section 102.07 itself contains several prohibitions to restrict how a manufacturer might strong-arm a retailer. For example, a manufacturer cannot be a retailer's guarantor, pay for a retailer's advertising, give a retailer aggressive discounts, etc.149 Similarly, section 102.01 restricts various cross-tier incursions. For example, permittees in one tier are prohibited from serving as officers in another tier.150 Additionally, permittees may not own fixtures or equipment, provide credit security, extend loans, agree to manage, or enter into profit-sharing arrangements with a permittee in another tier.151 Such activities are forbidden because they imperil independence and induce monopolistic industry practices, and lawmakers want to blunt manufacturer control/coercion/influence/subjection over retailers. But all potential connections are not prohibited; there is no absolute bar on even the most picayune cross-tier affiliations.152

The Legislature uses the terms "any interest" and "an interest of any kind" elsewhere in the Code.153 But not here. If "an interest" captures everything on its own, then why is the broadening "direct or indirect" needed in section 102.07(a)(1)? Why are other scope-expanding modifiers used elsewhere in the Code? The Code forbids a brewer from having a "direct or indirect interest" in a retailer, but does not forbid a retailer from having any interest whatsoever in a brewer. Many relationships forbidden by the Code affirmatively encompass "affiliate[s],"154 "subsidiar[ies],"155 and majority stockholders.156 The Legislature included no such language in section 102.07(a)(1).

"Interest" must, then, mean something less than any interest, as TABC maintains. The Court holds the definition of "tied house" in section 102.01(a) is applicable here, because the definition includes "any overlapping ownership."157 But this reading is too broad in relation to section 102.07(a)(1) because the phrase "interest in the business of a brewer" is narrower than the breadth of the Code as a whole.158 The independence of the three tiers was the Legislature's explicit objective, and tied houses were originally formed because of the intersection of retailers' financial pressures and brewers' corresponding financial benevolence. The prohibition of cross-tier interests, then, affects financial interests.

FEMSA's twenty-percent stock ownership in the Heineken Holding Companies does not mean it can step into the shoes of the Heineken Brewers or even flex influence to the point of affecting the Brewers' business. The publicly traded FEMSA is separated by at least three parent or holding companies from the Heineken Brewers. Two of these intermediate companies are internationally publicly traded companies. Additionally, though FEMSA's officers hold positions on the board of directors and a "supervisory board," these positions relate only to the Heineken Holding Companies, not the Heineken Brewers. Insofar as the holding companies control the actions of the Brewers, the Corporate Governance Agreement strips FEMSA of any ability to direct or control any aspect of the Brewers' dealings. FEMSA is both legally (by the Corporate Governance Agreement) and practically (by nature of its attenuation from the Brewers) barred from any attempt to act in place of or exert control over the Heineken Brewers.

The statutory context of section 102.07 strongly indicates that an "interest" must involve more than mere stock ownership. Elsewhere in section 102, for example, the Legislature expressly — and repeatedly — refers to interests in "corporate stock":

This treatment is not unique to section 102. Indeed, throughout the Alcoholic Beverage Code references to interests in corporate stock appear frequently and — at least three times — are distinguished from interests in a business:

This distinction is also true for statutes outside the Alcoholic Beverage Code; in many other instances "interest" encompasses even the ownership of a small percentage of stock.166 Nevertheless, and as articulated above, section 102.07(a) does not define interest as stock ownership and does not contemplate that even a small percentage of stock qualifies as an "interest." Texas law recognizes corporate separateness, the principle that subsidiaries have distinct legal identities from their holding or parent corporations.167 It is a fundamental principle that different corporate entities must be treated as legally distinct, even if one is owned by another. Thus a shareholder, as an investor, has a financial interest in a company's monetary value. But the shareholder has no cognizable legal interest in the company's assets.168

There is not even the assertion here that the complicated FEMSA/Heineken corporate structure, specifically the intermediate entities between FEMSA and the Heineken Holding Companies, are being finagled to circumvent Texas tied-house laws.169 Nevertheless, the Court disregards the entities' corporate separateness, collapsing all distinctions between parent companies, holding companies, and subsidiaries. I disagree that veil-piercing principles do not apply "in the regulatory context," absent evidence of furtiveness. Cadena and FEMSA are distinct corporate entities, separated by multiple layers of intermediate corporations. Likewise, the Heineken Holding Companies and Heineken Brewers are distinct corporate entities, separated by multiple layers of intermediate corporations. The burden should have rested on TABC to make the case for disregarding distinct corporate identities. But, despite TABC offering no evidentiary basis for ignoring corporate separateness, the Court adopts TABC's reasoning as its own.

Again, the Legislature knows well how to forbid specific relationships. If lawmakers had wished for section 102.07(a)(1) to disrespect corporate separateness, paying no mind to distinct legal identities, they could have done so, as they did in other Code provisions.170 Chapter 102 of the Code indeed specifies certain prohibited intra-industry relationships and different degrees of acceptable inter-tier connectedness. Some sections within Chapter 102 explicitly include "affiliates" and "subsidiaries" of a permittee when listing prohibited relationships.171 Other sections, like section 102.07(a)(1), do not.172 Some sections within the Code broadly define prohibited relationships and forbid "an interest of any kind," including stock ownership.173 Others, like section 102.07(a)(1), do not.174 All to say, the Legislature is adept at prohibiting specific relationships when it wishes to, and vice versa. We should presume the Legislature included words it wanted to include and excluded words it wanted to exclude.175

The Court should have rejected TABC's contention that corporate separateness can be blithely disregarded in the regulatory context. Our cases are precisely the opposite, resisting regulators' attempts to treat distinct legal entities as one, unless the record shows the parent controls the internal business operations and affairs of the subsidiary.176 In this case, TABC has never produced evidence of (or even alleged) "subterfuge ownership" or other abuse or circumvention that would justify veil-piercing.177 Cadena fully disclosed the complicated FEMSA-and Heineken-related corporate relationships, complexity driven by tax and corporate laws, and TABC never alleged, much less demonstrated, that these entities were anything other than distinct legal identities. The Legislature has repeatedly drafted language to forbid specific corporate relationships involving parents, subsidiaries, affiliates, etc. It did not so do in section 102.07, presumably on purpose.

The Court's error here is two-fold: (1) it improperly ignores corporate separateness (collapsing distinct legal entities across multiple levels throughout the FEMSA and Heineken corporate families based solely on stock ownership), and then (2) interprets "interest" so open-endedly that it prohibits nonprohibited relationships. TABC erred in treating far-removed entities as one interconnected business enterprise and the Court compounds that error here. Cadena and FEMSA are not a single retailer entity, nor are the Heineken Brewers and Heineken Holding Companies a single brewer entity. The majority's free-wheeling interpretation, disregarding distinct legal identities, confers limitless power upon TABC. Unanchored in statutory text, the Court's interpretation grants TABC the authority to selectively and arbitrarily permit similar applicants dissimilarly.

3. TABC's Ascetic View That the Code Allows "No De Minimis Exception" — aka the Single-Share Theory — Invites Absurd and Arbitrary Results

The Court today adopts TABC's view that section 102.07(a)(1) covers not only every type of interest, but also every quantum of interest, however slight and trifling. The Court offers assurances that we aren't dealing with a single-share issue here, concluding we need not decide the issue.178 But the logically inescapable extension of interpreting "interest" to mean any financial interest is the so-called single-share theory, an interpretation that would yield nonsensical results — not just odd results, but preposterous ones. A single-share interpretation, as opposed to asking whether an entity in one tier exerts control or influence over an entity in another tier, is neither rational nor practical. No reasonable person could have intended it, which is the very definition of absurdity. Indeed, changing "an interest" to "any interest" — and applying it across the board — would require the revocation of many currently issued alcohol permits (those granted to brewers and retailers alike).179 The absurdity bar in this Court is a high one,180 but the single-share theory scales it — easily.

Under the Court's hyperliteral interpretation of section 102.07(a)(1), an individual or special-interest group could easily manipulate and potentially wreak havoc on the permitting system. Everyone could be affected — from the mom-and-pop general stores that dot our rural counties to the large retail chains with locations across the state and nation. Without the ability to provide alcohol to their guests and customers, many of these establishments would be run out of business. Under the Court's restrictive view — a practical application of the single-share theory — any person can unilaterally imperil the permit of a Texas business. How? Buy a share of stock in that company. Then buy a share in another company permitted within another tier. According to the Court and TABC, both companies are now in violation of section 102.07(a)(1) and must lose their permits.181 A disgruntled employee could potentially buy shares in a retailer and a manufacturer and ensure both will lose their permits. Motivated competitors could buy cross-shares of both up-and-coming and long-standing industry rivals and end their quests for economic success. The consequences of the Eighteenth Amendment could again be realized if modern-day teetotalers obtained interests in different tiers of the industry in order to shut down the Texas alcohol trade. A lone citizen convinced of the evils of alcohol could use today's holding to ensure his neighborhood grocery store was alcohol-free. This is a ridiculous interpretation. And it is not compelled by a fair, contextual reading of section 102.07(a)(1).

Even without a bad actor spurred by malicious intent, the single-share theory will still affect the management of mutual funds, insurance agreements, retirement plans, and nearly anyone with a diversified portfolio.182 If section 102.07(a)(1) applies to any financial interest, then:

If these situations seem impossible or far-fetched, it bears repeating: TABC insists there is "no de minimis exception." Their position, baked into today's holding, inevitably leads to the single-share rule, which, given the nature of modern stock ownership, mutual funds, and pension plans, would be tantamount to de facto Prohibition if enforced. TABC counsel agreed that under a single-share rule, "a large number of current permit holders... are violating the statute." And "if the statute does impose share ... prohibition," that would "cause[] a lot of problems and would result in a lot of cancellations and revocations." Nonetheless, the Court, while acknowledging TABC's post-argument view that there is no de minimis exception, adopts an expansive interpretation indistinguishable from the "one-share rule" and dictates a disquieting result: Countless current permittees, including State permittees, are operating illegally by TABC's own admission.

TABC says regardless of whether the Code prohibits de minimis cross-tier interests, "TABC need not enforce that prohibition against a de minimis interest." Today's case, though, is about selective permitting, not selective enforcement. This is a permitting case, not an enforcement case. The Legislature sets the permitting criteria, not TABC. The question is simply stated: Does Cadena meet the objective statutory criteria? Either Cadena qualifies for a retailer permit or it doesn't. And if Cadena qualifies, TABC has no discretion to deny the permit. Courts, including the U.S. Supreme Court, have repeatedly recognized that selective permitting, unlike prosecutorial discretion, is impermissible where someone is "treated differently from others similarly situated" with "no rational basis for the difference in treatment."187

TABC contends, "it is one thing to interpret a statute as being applicable to a certain factual scenario, and quite another to enforce it in that scenario," adding, "just because an agency interprets a statute to apply in a particular situation does not mean the agency must always enforce it in that situation." The rules are the rules, and I am unaware of any principled basis, certainly none required by the Code, for applying them differently to different companies. Alcohol laws are complex, but the Rule of Law requires uniformity, not selective enforcement and anticompetitive favoritism benefitting preferred permittees.

Every industry actor has an "interest" in knowing what Texas law does and does not require. Every industry actor has an "interest" in TABC adopting a consistent permitting approach that applies the law uniformly to incumbents and newcomers alike. It is not enough to cast the single-share situation as a purely hypothetical scenario and insist that TABC would refrain from expending its enforcement resources in such an audacious way. In this Court, TABC was unwilling to defend the single-share theory yet also unwilling to disavow it. It dismisses the single-share standard as a fictitious stalking horse that it refuses to dismount.

At oral argument, the State reaffirmed its view that the statute contains no de minimis exception and that numerous current permittees are in violation of Texas tied-house laws:

TABC asserts it is "not trying to walk away from" the one-share rule; it will just never enforce it. Startlingly, the Court seems untroubled by TABC's insistence on an admittedly unenforceable standard that arbitrarily favors some businesses and disfavors others.

Again, every Texan possesses a legitimate "interest" in knowing what Texas law prescribes and proscribes and that those in power will enforce it evenhandedly. TABC, when pressed for a principled basis for treating similar businesses dissimilarly, responds: fear not — lack of enforcement resources will prevent uniform enforcement of its "no de minimis exception" position. This is distressing.

Legal rules must apply consistently to everyone, meaning TABC treatment of industry upstarts must match TABC treatment of industry heavyweights. TABC's enforcement regime today can fairly be described as passive-aggressive: passive for some, aggressive for others. "[N]o de minimis exception" is synonymous with "no limit to our discretion." But equal treatment under the law means precisely that, not vesting regulators with standardless power to play legal favorites.

As noted above, TABC promises to exercise prosecutorial discretion in enforcing its expansive interpretation of the statute, partially by admitting its inability to enforce uniformly a single-share rule. Indeed, TABC's denial of Cadena's application clashes with its treatment of other similarly situated permittees. The record shows billions of dollars of cross-tier ownership by permittees at the retail and brewer tiers. Cadena's application was rejected while other retail permittees simultaneously own stock in publicly traded manufacturers.

Interpreting a statute in a manner that ratifies unequal enforcement is odd. Our focus should not be on line-drawing — isolating the lowest percentage below which ownership interests must fall to satisfy section 102.07(a)(1). Instead, we should read "interest" in the context of related statutes, particularly the codified purpose of ensuring actors' "independence." This non-ascetic reading yields a more contextual and thorough understanding of "interest": a financial stake sufficient to exert cross-tier influence or coercion.

A common-sense reading of the Code is especially warranted here given the Legislature's express admonitions against "subterfuge and related practices" by industry participants.188 When it comes to liquor regulation, the path of the law has been a rocky one. Perhaps because of the long, colorful history of liquor regulation and the difficulties governments invariably experience in attempting to regulate alcohol, the Legislature has inserted repeated statutory edicts into the Code that prohibit specific conduct as well as any "subterfuge" whereby industry participants engage in gamesmanship that technically complies with the Code but undermines its goals. In my view, prohibiting subterfuge should cut both ways, by seeing through attempts to avoid compliance but also by avoiding hyper-technical constructions that find violations defying common sense. The Legislature's unease with artifice should dissuade interpretations that invite non-sensical ripple effects that either permit proscribed conduct or proscribe permitted conduct.

IV. CONCLUSION

During Prohibition, Americans seeking alcoholic refreshments scarpered to the black market in mass disobedience. The "noble experiment" (President Hoover's term)189 had many ignoble consequences, however, and after the Twenty-First Amendment was ratified, states were urged to adopt protective measures to forestall pre-Prohibition licentiousness and lawlessness. President Roosevelt pleaded for temperance, condemning "the curse of excessive use of intoxicating liquors" and imploring states to regulate liquor purchases in a way that avoided the "repugnant conditions" that predated the Eighteenth Amendment: "I ask especially that no state by law or otherwise authorize the return of the saloon in its old form or in some modern guise."190

For 80-plus years, Texas's tied-house laws have mandated a three-tiered alcohol industry — producers, distributors, and retailers — to guard against the criminal element in alcoholic-beverage trafficking and to prevent cross-tier ownership overlaps that induce coercion, monopolies, and domination. With this in mind, in the tied-house context, "interest" should mean "control."

This Court has a well-earned reputation for textualism, an interpretive mooring that prizes clear interpretive rules, eschews legal dice-loading, and minimizes judicial lawmaking. Policy calls are for the political branches, not adventurist, reform-minded judges. Fidelity to text, by curtailing judicial discretion and prizing well-defined rules consistently applied, best secures certainty and thus the Rule of Law. Unlike other methods of interpretation, like purposivism, scrupulous textualism is politically agnostic, and proudly so — caring not whether a result scratches an ideological itch, but only whether interpretive principles are applied forthrightly to honor the political bargains of legislators, whatever color their jersey.

That said, interpretation is done by flesh-and-blood people, and as this case shows, textualism does not guarantee unanimity. Some disputes are hard, and avowed textualists will disagree. Chief Justice Marshall observed the slipperiness of words nearly two centuries ago: "Such is the character of human language, that no word conveys to the mind, in all situations, one single definite idea...."191 Words are slippery, and interpreting slippery words eludes robot-like precision. Textualist readers invariably read text variably. Judges seize upon different interpretive clues and balance them differently, not to reach a desired outcome but through earnest grappling. Determinations can be particularly indeterminate when excavating how an original interpretive community, perhaps generations or centuries ago, understood the language they enacted. That's why we're a nine-member Court, in hopes that collaborative (and hopefully collegial) jousting will sharpen our analysis.

The Court doubtless believes its hyperliteral interpretation — reading "an interest" as "any interest" — is more pragmatic and workable. I disagree with this consequentialist view, which, in any case, does not square with reality, given the irrationality of the single-share theory. Our commitment to judicial textualism and self-abnegation requires us to honor what has been written. Judges ought neither draft laws nor revise them under the guise of interpreting them. Rather, we should seek an objective basis for interpretation, namely the semantic, contextual meaning of the words themselves.192 The term "interest" may seem all-encompassing standing alone, but as explained above, context clarifies content. Courts are tasked every day with assigning meaning to imprecise language. Isolating the exact point when an "interest" becomes unduly influential or coercive resists mathematical certainty. But it is no more blurred as when judges determine under the U.S. Constitution when a search is "unreasonable"193 or bail "excessive"194 or cause "probable"195 or punishment "cruel and unusual."196 These are vexing, fact-laden inquiries, requiring judges to be judgmental.

Unfortunately, the Court adopts a construction of "interest" so stringent as to deform fair meaning. TABC does not dispute that numerous current permittees hold billions of dollars in so-called cross-tier investments, and nothing like the evils that plagued the classic tied-house paradigm of saloons controlled by breweries has befallen Texas. It blinks reality to grumble that the State's entire regulatory scheme hangs in the balance when TABC countenances scores of overlapping interests that, under its "no de minimis exception" position, are illegal and ripe for revocation. If anything threatens functional derailment of the three-tier system, it is strict, no-favorites enforcement of TABC's no-exception standard.

FEMSA's equity stake in the Heineken Holding Companies does not implicate section 102.07(a)(1)'s concern with retailer-manufacturer overlap. The Court seeks to enforce "strict separation" between the three tiers, but the Legislature does not proscribe all cross-tier relationships, only specifically enumerated ones. The Code is a hodgepodge of laws enacted since Prohibition's repeal, and lawmakers over time have tweaked the Code, inviting overlap through exceptions to the three-tier system. For example, wineries can operate across all three tiers;197 brewpubs can sell their product directly to retailers;198 small brewers and manufacturers can distribute wholesale and sell retail;199 package stores can also be local distributors.200 Strict in rhetoric is sometimes lenient in reality.201

Nothing in section 102.07 addresses, much less forbids, an applicant's parent company from having an indirect minority stock ownership in the parent companies of non-resident brewers. The Code's overriding goal of safeguarding "independence" — the absence of control, coercion, and subjection — is not imperiled by such attenuated, far-removed connectedness that doesn't portend vertical domination, or even trace influence.

The state vested TABC with authority to regulate the alcoholic-beverage industry, and history provides a backdrop for the wisdom of tied-house statutes. But no state agency should be able to discriminate indiscriminately. TABC's "no de minimis exception" standard confers vast autonomy and conjures an erratic system of constantly moving goalposts. Government must not treat similarly situated parties dissimilarly, playing regulatory favorites by applying different standards to different companies.

Such effects were in fact predicted by the authors of Toward Liquor Control in 1933:

Section 102.07(a)(1) prohibits one with an interest in a brewer from having a direct or indirect interest in a retailer. Neither Cadena (the applicant) nor its distant parent FEMSA influences or controls the business of the Heineken Brewers in such a way as to hazard the Brewers' "independence." It is fanciful to contend that Texas tied-house laws, explicitly aimed at preserving actors' independence, are remotely imperiled.

On this record — corporate separateness cemented by a governance agreement that denies FEMSA any form of influence that would imperil the Code's stated goal of "independence" — there is no prohibited, cross-tier "interest" under section 102.07(a)(1). Because the Court holds otherwise, I respectfully dissent.

Attachment