JUSTICE BOYD delivered the opinion of the Court, in which Chief Justice Hecht, Justice Devine, Justice Busby, Justice Bland, and Justice Huddle joined. JUSTICE YOUNG filed a dissenting opinion, in which Justice Lehrmann joined. Justice Blacklock did not participate in the decision. This case involves the regulation of health-care providers who participate in the federal Medicaid program. The State, acting through the Attorney General, seeks to enforce a statute that imposes substantial penalties against a provider who submits a claim for payment and knowingly fails to indicate the type of professional license “and” the identification number of the person who actually provided the service. The defendant—a dentist—contends the statute applies only if a claim fails to indicate both the license type “and” the identification number of the actual provider. The State contends it applies if a claim fails to indicate either the license type “or” the identification number. Considering the statute’s text, grammatical structure, context, and purpose, we agree with the dentist’s construction. And to the extent any ambiguity exists, we construe such penal statutes strictly in favor of the party against whom the State seeks to impose the penalties. We reverse the court of appeals’ judgment and render judgment in the dentist’s favor. I. Background Dr. Richard Malouf co-founded All Smiles Dental Center in 2002 and began providing orthodontic services to Medicaid patients in 2004. The practice soon grew to employ several dentists at locations around the Dallas–Fort Worth area. Malouf bought out his partner in 2007 and retained full control until he sold most of his interest in 2010. During the period of Malouf’s ownership, the front-office staff at each of All Smiles’ locations relied on dentists’ chart notes to prepare bills for services rendered to Medicaid orthodontic patients and transmitted those bills to the company’s corporate office. The corporate-office staff reviewed the bills and submitted payment claims to the Medicaid office either electronically or on paper using a specific Medicaid-authorized form. A completed form must state the provider’s name and Texas Provider Identifier (TPI) number, which is a unique number assigned to each provider. The form need not separately state the provider’s type of professional license, license number, or other identification number. Instead, because providers must submit proof of their professional license to obtain a TPI number, a provider’s license type and license number are affiliated with the TPI number. As All Smiles’ owner, Malouf was responsible for ensuring the practice followed Medicaid’s requirements and policies. In that role, he periodically reviewed the Texas Medicaid Provider Procedures Manual and attended conferences and meetings to remain informed about Medicaid policies. In 2012, two former employees filed qui tam actions alleging that Malouf and All Smiles committed numerous violations of (what was then called) the Texas Medicaid Fraud Prevention Act. See TEX. HUM. RES. CODE §§ 36.001–.132 (amended 2023); see also id. § 36.101 (authorizing private persons to bring actions on behalf of themselves and the State).[1] The Attorney General, acting on the State’s behalf, intervened in both actions, which were then consolidated. See id. §§ 36.102 (authorizing State’s intervention), .107 (authorizing Attorney General to take “primary responsibility for prosecuting the action”). The State asserted several claims against Malouf and others, including a claim under Section 36.002(8), which provides that a person “commits an unlawful act if the person . . . makes a claim under the Medicaid program and knowingly fails to indicate the type of license and the identification number of the licensed health care provider who actually provided the service.” Id. § 36.002(8).[2] The State alleged that, under Malouf’s direction, All Smiles submitted 1,842 claims that stated Malouf’s TPI number even though a dentist other than Malouf actually provided the billed-for services. Based on this claim, the State sought to recover the amount Medicaid paid for those services plus prejudgment interest, statutory penalties, attorney’s fees, and expenses. See id. §§ 36.007, .052(a) (authorizing such recoveries). The State filed a motion for partial summary judgment on only that claim. Malouf did not dispute that All Smiles submitted 1,842 claim forms stating his TPI number for services a different dentist actually provided. He insisted, however, that he did not “knowingly” fail to indicate the actual provider’s information. Specifically, he testified he believed based on information provided to him by Medicaid that he was supposed to use his TPI number whenever (1) he personally supervised the dentist who provided the service or (2) Medicaid’s system suffered a “glitch” that prevented his staff from properly submitting a claim. He asserted that, except for those two circumstances, he had no knowledge that his staff submitted claims using his TPI number for services another dentist provided. This testimony, he argued, created fact issues as to which, if any, of the 1,842 claims actually constituted an “unlawful act.” In addition, Malouf argued that none of the 1,842 claims constituted an unlawful act under Section 36.002(8) because they all correctly indicated the license type of the provider who actually provided the billed-for services. In each case, Malouf explained, the services were actually provided by someone who—like Malouf—was a licensed dentist, so a form bearing Malouf’s TPI number in fact indicated the type of license held by the person who actually provided the service. And because all the claim forms indicated the actual provider’s license type, Malouf argued, none of them constituted an unlawful act under Section 36.002(8) because they did not fail to “indicate the type of license and the identification number of the licensed health care provider who actually provided the service.” Id. § 36.002(8) (emphasis added). Based on these arguments, Malouf filed a no-evidence-summary-judgment motion. The trial court denied Malouf’s motion and granted the State’s motion for partial summary judgment. The State then nonsuited its remaining claims and moved for entry of a final judgment. The trial court rendered a final judgment awarding the State more than $16,500,000, consisting of about $538,000 for the amount Medicaid paid on the 1,842 claims, twice that amount (almost $1.1 million) as a civil penalty, a little over $9.2 million as an additional penalty of $5,000 for each of the 1,842 unlawful acts, and about $5.7 million for attorney’s fees and expenses the State and private plaintiffs incurred. The trial court denied Malouf’s new-trial motion, and Malouf appealed. The court of appeals disagreed with the amount of attorney’s fees and expenses but otherwise affirmed the trial court’s judgment. 656 S.W.3d 402, 418 (Tex. App.—El Paso 2022). We granted Malouf’s petition for review. We review the trial court’s summary judgment de novo. Provident Life & Accident Ins. Co. v. Knott, 128 S.W.3d 211, 215 (Tex. 2003).Summary judgment is appropriate when there is no genuine issue of material fact and judgment should be granted in the movant’s favor as a matter of law. Diversicare Gen. Partner, Inc. v. Rubio, 185 S.W.3d 842, 846 (Tex. 2005). We take as true all evidence favorable to the nonmovant, indulging every reasonable inference and resolving any doubts in the nonmovant’s favor. Knott, 128 S.W.3d at 215. II. Construing Penal Statutes This case requires us to construe Section 36.002(8). A statute’s meaning presents a question of law that we review de novo. Tex. Health Presbyterian Hosp. of Denton v. D.A., 569 S.W.3d 126, 131 (Tex. 2018). Any time we endeavor to construe statutory language, well-established rules guide our analysis. Fundamentally, we look to the statute’s text— to the words it actually uses—and apply the common, ordinary meaning of those words “unless the text supplies a different meaning or the common meaning leads to absurd results.” Id. We construe the words in light of their statutory context, considering the statute as a whole. Silguero v. CSL Plasma, Inc., 579 S.W.3d 53, 59 (Tex. 2019). If the text’s meaning is unambiguous, we do not resort to extrinsic aids or special rules of construction. Crosstex Energy Servs., L.P. v. Pro Plus, Inc., 430 S.W.3d 384, 389 (Tex. 2014). When possible, we construe the language in a way that does not render any of it meaningless. Whole Woman’s Health v. Jackson, 642 S.W.3d 569, 581 (Tex. 2022). In some cases, however, special rules of construction may apply. Malouf contends this is such a case. Specifically, he asserts that Section 36.002(8) is a penal statute, and we must construe penal statutes strictly against the State and in his favor. We thus begin by addressing this special rule of construction before turning to the statutory language. “All political power” in Texas “is inherent in the people.” TEX. CONST. art. I, § 2. Exercising that power, the people have established a republican form of government, granting all “Legislative power” to their elected representatives in the Senate and the House of Representatives. Id.; see also id. art. III, § 1. The people thus permit the Legislature to regulate the people’s own conduct, so long as the regulation is “justified by a rational legislative purpose and does not violate a specific constitutional provision.” Tex. Boll Weevil Eradication Found., Inc. v. Lewellen, 952 S.W.2d 454, 461 (Tex. 1997). And the power to regulate includes the power to impose “pains, penalties, and remedies” to enforce the regulations. Ex parte Hughes, 129 S.W.2d 270, 276 (Tex. 1939).[3] But in exercising that penal power, the Legislature must act clearly and specifically. A law that imposes penalties must be “plain enough to advise persons affected by it when and under what circumstances their acts and conduct would breach its terms.” State v. Int’l & Great N. Ry. Co., 179 S.W. 867, 868 (Tex. 1915). In other words, courts must strictly construe penal laws that suffer from uncertainty. Id. As Malouf correctly asserts, we “have consistently held that penal statutes should be strictly construed.” City of Houston v. Jackson, 192 S.W.3d 764, 770 (Tex. 2006).[4] This longstanding common-law rule, often referred to in the criminal-law context as the “rule of lenity,”[5] applies not only to many criminal statutes[6] and to statutes that impose tax liabilities,[7] but also to those that impose civil penalties. See id. (strictly construing statute imposing civil penalties); Brown, 156 S.W.3d at 565 (same).[8] We have articulated at least three related reasons for this rule. First, because the Legislature operates under a limited grant of authority from the people, we will not “presume” that the Legislature intends to impose a substantial “punishment” on the people and will instead insist that “that purpose is clearly manifested by the language employed in the statute.” Campbell, 45 S.W. at 4. Second, we insist that the Legislature give the people sufficient notice of the conduct that will subject them to statutory penalties. A statute imposing penalties must be couched in such explicit terms that the party upon whom it is to operate may with reasonable certainty ascertain what the statute requires to be done, and when it must be done; otherwise, there would be no opportunity for a person charged with the duty to protect himself by the performance of it according to the law. Mo., Kan. & Tex. Ry., 100 S.W. at 767.[9] And third, we have recognized that enforcing penalties exacted through ambiguous penal statutes risks denying citizens their constitutionally protected right to “due process of law, in violation of the principles of right.” Id. Like all common-law construction rules,[10] however, the rule of lenity applies only to the extent the statute at issue is unclear or ambiguous.[11] Indeed, to say that a statute will be “strictly construed” is simply to say that any uncertain or ambiguous provision will be construed in favor of a particular party or result. Int’l & Great N. Ry., 179 S.W. at 868 (explaining that strict construction affects provisions that “are vague and uncertain of meaning”).[12] “When the statutory language is unambiguous, we must apply the statute as written.” Jackson, 192 S.W.3d at 770. Statutes need only be “as definite in meaning as the nature of the subject would allow,” and we do not require the Legislature to “accomplish in expression of clearness that which is impossible.” Int’l & Great N. Ry., 179 S.W. at 868. All we require is that the statute be “sufficiently definite for those affected by it to understand its meaning so as to know under what circumstances they would be transgressing its provisions.” Id. at 869. But the degree of clarity required may vary in proportion to the severity of the penalty a statute imposes. As we have explained, “the more severe the penalty, and the more disastrous the consequence to the person subjected to the provisions of the statute, the more rigid will be the construction of its provisions.” Mo., Kan. & Tex. Ry., 100 S.W.at 767. “[V]ery great strictness has been observed in the construction of [statutes] of the most highly penal character.” Randolph v. State, 9 Tex. 521, 523 (1853). Indeed, courts construing “more highly penal statutes” have sometimes “resorted to what may seem to be even strained construction in cases of doubtful guilt to avert the terrible penalty denounced by the law.” Estes v. State, 10 Tex. 300, 309 (1853). We agree with Malouf that this rule applies to the Texas Medicaid Fraud Prevention Act. As we have explained elsewhere, the Act is “a powerful tool for targeting fraud against the Texas Medicaid program and securing the program’s integrity.” In reXerox Corp., 555S.W.3d518, 525 (Tex. 2018). “[R]eports about allegedly fraudulent dental and orthodontic schemes have been front-page news in Texas,” and the Act plays a crucial role in the State’s ongoing “efforts to deter, detect, and punish” those schemes. Id. By granting the Attorney General broad investigative and enforcement authority and permitting private citizens to sue on the State’s behalf, the Act works to identify and exclude bad actors while deterring others from following their lead. Id. But the Act accomplishes this important objective by authorizing civil remedies that—as we recently held—”are penalties, not damages.” Id. at 527. It allows the State to recover not just the amount of any payment the State makes in response to an “unlawful act,” but three times that amount, plus interest, plus a substantial additional penalty for each separate unlawful act, thus “imposing monetary liability far surpassing the amount of Medicaid funds the State may have actually expended due to an unlawful act.” Id. at 526–27. As this case illustrates, the Act imposes these significant penalties regardless of whether the State actually suffers any financial loss as a result of the unlawful act. The State does not dispute in this case, for example, that licensed dentists in fact provided, and Medicaid recipients in fact received, all the services for which Malouf submitted the 1,842 claims at issue. Yet because, in the State’s view, the claims failed to indicate the actual provider’s license type and identification number, Malouf must return over $500,000 worth of payments for those claims and pay penalties in excess of $10 million. We conclude the Act is the very type of penal statute we must construe strictly. Thus, the statute must define any conduct giving rise to such penalties “in plain language,” Duke, 137 S.W. at 665, and the State must show that Malouf engaged in conduct that falls “clearly within the terms of the statute,” Agey, 172 S.W.2d at 974. To the extent the Act’s requirements are uncertain or ambiguous, we will construe them “in favor of such person and against the enforcement of such law.” Mo., Kan. & Tex. Ry., 100 S.W. at 767. III. “Unlawful Act” under Section 36.002(8) Although Section 36.002 provides a laundry list of “unlawful acts,” the judgment here is based only on the act described in subsection (8). As explained, that subsection provides that a person commits an unlawful act if the person “makes a claim under the Medicaid program and knowingly fails to indicate the type of license and the identification number of the licensed health care provider who actually provided the service.” TEX. HUM. RES. CODE § 36.002(8) (amended 2023). Malouf first argues that the trial court erred by granting the State’s summary-judgment motion because his testimony created a fact issue as to whether he “knowingly” failed to provide the required information as to any particular claim. Alternatively, he asserts that none of the 1,842claims constitutes an unlawful act because all of them correctly indicated the actual provider’s license type and an unlawful act occurs only when a claim fails to indicate both “the type of license and the identification number” of the actual provider. Applying our fundamental statutory-construction rules and construing any ambiguity strictly in Malouf’s favor, we agree with his alternative argument.[13] Specifically, we conclude that the 1,842 claims at issue here did in fact “indicate” the “type of license” of the “provider who actually provided the service,” and that an unlawful act occurs under Section 36.002(8) only if a claimant knowingly fails to indicate both “the type of license and the identification number” of the actual provider. Because the 1,842 claims indicated the actual providers’ license type, we conclude that none constitute an unlawful act under Section 36.002(8). A. Failure to indicate license type We begin with the question of whether the 1,842 claims “fail[ed] to indicate the type of license . . . of the licensed health care provider who actually provided the service.” Id. § 36.002(8). The State contends they did because they provided Malouf’s unique TPI number and thus indicated Malouf’s license type when Malouf did not actually provide the service. The court of appeals agreed, reasoning that the “claims at issue contain no information about the provider who [actually] rendered the services reflected because [Malouf] falsely represented he provided the service.” 656 S.W.3d at 414. Malouf contends the claims in fact indicated the actual providers’ license type because his TPI number indicates that the services were provided by a licensed dentist and (as the State concedes) all the actual providers were licensed dentists. The court of appeals rejected this argument, concluding that the “fact that [Malouf] and the performing provider are both dentists and therefore share the same license [type] is immaterial.” Id. But the statute’s language makes it difficult to reach that conclusion. In particular, Section 36.002(8) makes it a wrongful act if a claim fails to “indicate” the provider’s license type and identification number, not if it fails to “provide” or “state” them. While the verb “indicate” can mean to “state” or “express,” it can also mean to “point to,” “suggest,” or “demonstrate,” allowing for a far less direct means of disclosure. See Indicate, WEBSTER’S THIRD NEW INTERNATIONAL DICTIONARY (2002). In fact, as explained, the Medicaid-approved claim form never requires the provider to state or express his license type, license number, or any other identification number.[14] Instead, the form requires the provider to state his TPI number, and the license type and license number are affiliated with or embedded in that number. But the statute expressly refers separately to both the license type “and” the identification number. The TPI number belongs to and indicates a particular individual and location where that individual practices, but it also indicates a particular license type, and the statute refers separately to both. Under the State’s proposed construction, the statute’s reference to the license type becomes meaningless because every claim that fails to indicate the actual provider’s identification number necessarily also fails to indicate the actual provider’s license type. When possible, courts must construe a statute in a way that does not render any of it meaningless. Whole Woman’s Health, 642 S.W.3d at 581. To give meaning to this statute’s separate reference to the “type of license,” it must be possible that a claim could “indicate” the actual provider’s license type but not the actual provider’s identification number. Otherwise, the statute’s separate reference to “type of license” would be meaningless. That’s what the 1,842 claims do in this case: they all indicate the actual provider’s license type (a dental license), but they do not indicate the actual provider’s identification number. If the holder of a different license type—a licensed dental hygienist, for example—had actually provided the services, Malouf’s TPI number would have failed to indicate both the license type and the identification number of the actual provider. But no one disputes that a licensed dentist actually provided the billed-for services in each of those claims, and the claims indicated that a licensed dentist provided the services. We thus conclude that, although the 1,842 claims did not indicate the actual providers’ identification numbers, they did indicate the actual providers’ license type. B. Failure to indicate license type and identification number Having concluded that the 1,842 claims did not fail to indicate the actual providers’ license type, we turn to the question of whether they nevertheless failed to indicate “the type of license and the identification number of the licensed health care provider who actually provided the service.” TEX. HUM. RES. CODE § 36.002(8). Malouf argues that because the statute uses the conjunctive “and,” as opposed to the disjunctive “or,” it requires the State to prove he failed to indicate both, so he did not commit an unlawful act if he indicated either. The State contends the statute required him to indicate both, so he committed an unlawful act if he failed to indicate either. In addressing this issue, we find guidance in the United States Supreme Court’s recent decision in Pulsifer v. United States, 144 S. Ct. 718 (2024). Pulsifer involved “criminal history points” assigned to persons who are sentenced for a criminal conviction under federal law. Id. at 723. Under the federal sentencing guidelines, an offense resulting in a sentence of fewer than sixty days is a 1-point offense, an offense resulting in a sentence between sixty days and thirteen months is a 2-pointoffense, andan offense resulting in asentence exceeding thirteen months is a 3-point offense. Id. at 724. The statute at issue in Pulsifer provides that, for certain offenses, a court must impose a sentence pursuant to the federal sentencing guidelines and “without regard to any statutory minimum” if five requirements are met. 18 U.S.C. § 3553(f). Under the first requirement, the court must find that “the defendant does not have” three things: “(A) more than 4 criminal history points, excluding any criminal history points resulting from a 1-point offense . . . ; (B) a prior 3-point offense . . . ; and (C) a prior 2-point violent offense . . . .” Id. § 3553(f)(1) (emphasis added). Pulsifer had two prior 3-point offenses and thus had (A) and (B), but he did not have a prior 2-point violent offense and thus did not have (C). The issue was whether the requirement is met when the defendant “does not have” one of the listed elements but does have the other two. Pulsifer, 144 S. Ct. at 723. Just as the State argues here that Section 36.002(8) applies if a claim “fails to indicate” either one of the two elements listed (the license type or the identification number), Pulsifer argued that Section 3553(f)(1) applies if a defendant “does not have” any one of the three elements listed in that section. See id. And just as Malouf argues that an unlawful act occurs under Section 36.002(8) only if a claim “fails to indicate” both of the elements listed (the license type and the identification number), the government argued in Pulsifer that Section 3553(f)(1)’s requirement is met only if the defendant “does not have” all three of the elements listed. See id. The Supreme Court ultimately agreed with the government, holding that Section 3553(f)(1) effectively “creates an eligibility checklist, and demands that a defendant satisfy every one of its conditions.” Id. at 725. Pulsifer differs from this case in several obvious respects. Section 3553(f)(1), for example, refers to three separate elements while Section 36.002(8) only refers to two. And when the elements are satisfied, Section 3553(f)(1) benefits a defendant by rendering statutory-minimum sentences inapplicable, while Section 36.002(8) subjects a defendant to harm by imposing liability for penalties. So in Pulsifer the defendant wanted the statute to apply and the government did not, while here the government wants the statute to apply and the defendant does not. But the cases are nearly identical in at least a few key respects: both statutes express a negative verb phrase (the defendant “does not have” and “fails to indicate”), both statutes then list more than one element as an object of that verb phrase, both statutes use the conjunctive “and” to connect the listed elements, and in both cases one, but only one, of the listed elements was not satisfied. Addressing Section 3553(f)(1), the Supreme Court considered the statute’s grammatical structure, statutory context, legislative purpose, and, ultimately, the rule of lenity. Id. at 726–38. We also consider these when construing statutes, see Tex. Health Presbyterian Hosp., 569 S.W.3d at 131, and will apply that same analysis here. 1. Grammatical structure The Court began its analysis in Pulsifer by looking to the statute’s grammatical structure primarily because Pulsifer contended it resolved the case. 144 S. Ct. at 726. In particular, Pulsifer focused on Section 3553(f)(1)’s use of the conjunctive term “and”—as opposed to the disjunctive term “or”—to connect the three listed elements. See id. Pulsifer argued that if Congress intended the requirement to be satisfied only when a defendant “does not have” all three of the listed elements, the statute would say “the defendant does not have A, B, or C,” instead of “A, B, and C.” Id. at 728–29. The dissenting Justices agreed with Pulsifer’s argument that “we wouldn’t be sitting here if Congress had used the word ‘or’ in paragraph (f)(1).” Id. at 747 (Gorsuch, J., dissenting) (cleaned up). The Court acknowledged that the word “and” is typically used as a conjunction meaning “along with or together with.” Id. at 726 (quoting And, WEBSTER’S THIRD NEW INTERNATIONAL DICTIONARY (1993)). But it rejected Pulsifer’s argument that the use of “and” instead of “or” establishes that the requirement is satisfied if the defendant “does not have” only one of the three elements. Id. at 729. In the Court’s view, if Section 3553(f)(1) used the word “or” instead of “and,” the parties would likely still make the same arguments: Pulsifer would still argue that the requirement is satisfied if the defendant does not have any of the three (“does not have A, does not have B, or does not have C”), and the government would still argue that it is only satisfied if the defendant “does not have” all three elements (“does not have (A, B, or C)”). Id. (emphases added). In the Court’s view, the appropriate grammatical question was not whether Section 3553(f)(1) uses “and” in the conjunctive or disjunctive sense, but whether it uses it in a joint or distributive sense. See id. at 726. In other words, the key question is “what the ‘and’ in Paragraph (f)(1) connects.” Id. If construed in the joint sense as Pulsifer argued, “and” serves to join the listed elements into a “single disqualifying characteristic,” making A, B, and C a “complete combo,” and the statute applies only if the defendant “‘does not have’ that full package.” Id. Read in that sense, “[i]t is as if Pulsifer inserted parentheses into the paragraph, so that it asks whether ‘the defendant does not have (A, B, and C),’” so the requirement is satisfied only if the defendant “does not have” the “combination” of A, B, and C. Id. But if construed in the distributive sense as the government argued, the verb phrase (“does not have”) “operates on each [element] seriatim,” carrying over “to every item on the ensuing list.” Id. at 727 (quoting BRYAN A. GARNER, GARNER’S DICTIONARYOF LEGAL USAGE 639(3ded. 2011)). Read in the distributive sense, “and” requires the verb phrase to operate “on A, and on B, and on C consecutively, rather than on the three combined.” Id. at 726. Pulsifer argued that “and” must be read in the joint sense when the verb phrase is framed in the negative (“does not have”), suggesting the phrase “don’t drink and drive”—meaning don’t do both together—as a common example. Id. at 727. But the Court rejected that argument, noting that other negative phrases, like “I’m not free on Saturday and Sunday”—meaning the speaker is not free on either day—use “and” in the distributive sense. Id. Ultimately, the Court concluded that “and” in Section 3553(f)(1) could reasonably be read in either the joint or the distributive sense. See id. at 730. In the end, the Court held, “[t]here are two grammatically permissible ways to read” the statute, so as to that statute, “grammar is not the primary determinant of meaning.” Id. at 726, 728. The decision between the two, the Court concluded, requires reviewing not just the grammar of the text, but “reviewing text in context.” Id. at 726. We generally agree with the Court’s analysis and conclusion. Certainly, “grammar rules can be crucial to proper construction” of statutory language. Tex. Health Presbyterian Hosp., 569 S.W.3d at 132. And we have long recognized that the conjunctive “and” is rarely interchangeable with the disjunctive “or.” See Bd. of Ins. Comm’rs v. Guardian Life Ins. Co. of Tex., 180 S.W.2d 906, 908 (Tex. 1944).[15] We presume the Legislature used “and” instead of “or” in Section 36.002(8) for a reason. Xerox Corp., 555 S.W.3d at 527; DeQueen, 325 S.W.3d at 635. But we cannot tell solely from the word itself whether “and” is used in the joint sense or the distributive sense.[16] Under the joint sense, which the State proposes, “and” combines “the type of license” and “the identification number” into a single unit, so that an unlawful act occurs if the defendant “fails to indicate” either one. It’s like saying, “Don’t forget to put ham and cheese on my sandwich”: a server who fails to include either ham or cheese will violate the instruction. But under the distributive sense, which Malouf proposes, “and” connects each element to the verb phrase individually, so that an unlawful act occurs only if the defendant “fails to indicate” the license type and “fails to indicate” the identification number. Like the example “Don’t drink and drive,” a violation occurs only if the person does both. It may be that “and” is usually used in a distributive sense.[17] But we must decide whether it is used that way in Section 36.002(8). Like the Supreme Court in Pulsifer, we conclude we cannot make that determination solely by looking to the statute’s grammatical structure. 2. Statutory context Finding no clear answer in the statute’s grammatical structure, the Supreme Court looked in Pulsifer to Section 3553(f)(1)’s statutory context, considering its “text in its legal context” and how its provisions “fit with other pertinent law.” 144 S. Ct. at 724. Under this analysis, the Court identified two “statutory difficulties” resulting from Pulsifer’s proposed joint reading of the term “and.” Id. at 731. First, the Court observed that if it were to read the term “and” jointly as connecting the three elements into a single unit the defendant must “not have,” the first element—(A)—”would become superfluous— without any operative significance,” because a defendant who has a 3-point offense under (B) and a 2-point violent offense under (C) will necessarily have more than 4 criminal history points under (A). Id. If, to enjoy the benefit of Section 3553(f)(1), a defendant must only show that he “does not have” (A, B, and C) combined, there would be no reason to include (A) because he could never show that he “does not have” only (A). Id. By contrast, the government’s distributive reading of “and” produced no superfluity in the Court’s view because each of the three elements “does independent work, disqualifying defendants from relief even when the others would not.” Id. Second, the Court observed that, under Pulsifer’s joint reading, “defendants’ eligibility for relief would not correspond to the seriousness of their criminal records,” as “a defendant with numerous violent three-point offenses could get relief because he happens not to have a two-point offense.” Id. In the Court’s view, these contextual clues were sufficient to “answer[] the statutory puzzle here—reducing two grammatical possibilities to just one plausible construction.” Id. Based on the statutory context, the Court agreed with the government that Section 3553(f)(1) uses “and” in its distributive sense. See id. This Court also relies on statutory context when construing statutes. Our “text-based approach to statutory construction requires us to study the language of the specific provision at issue, within the context of the statute as a whole, endeavoring to give effect to every word, clause, and sentence.” Ritchie v. Rupe, 443 S.W.3d 856, 867 (Tex. 2014). We determine the meaning of the words a statute uses by “considering the context in which they are used, not in isolation.” Silguero, 579 S.W.3d at 59. The statutory context of “and” in Section 36.002(8) produces clarity in the same ways the Supreme Court found in Pulsifer. Like Section 3553(f)(1), construing the “and” in Section 36.002(8) in a joint sense results in superfluity, at least from the perspective of providers who face substantial penalties if they fail to indicate the correct identification number “and” license type. The Medicaid-approved claim form that a provider must use only requires the provider’s name andTPI number, and that number indicates both the provider’s identification number and license type. And as we have explained, a claim could correctly indicate the actual provider’s license type but not the actual provider’s license number, as the 1,842 claims do here. But it is not possible for a claim to correctly indicate the actual provider’s TPI number but not the actual provider’s license type because the actual provider’s TPI number will always correctly indicate that provider’s license type. If, as the State contends, an unlawful act occurs whenever a claim fails to indicate either the type of license or the identification number, there would be no reason for the statute to include the reference to the provider’s license type because the State could never show that a claim fails to indicate only the provider’s license type. In other words, to use the Pulsifer Court’s description, if to prove an unlawful act under Section 36.002(8) the State must only show that a claim fails to indicate A or B, there would be no reason to include A because the State could never show that a claim fails to indicate only A. In all cases, the State must show that the claim fails to indicate B. By contrast, Malouf’s distributive reading of “and” produces no superfluity because it requires the State to show that a claim fails to indicate both A and B, such that each element does “independent work,” precluding the finding of an unlawful act “even when the other[] would not.” Pulsifer, 144 S. Ct. at 731. A second contextual clue also points in favor of construing “and” in the distributive sense. Section 36.002 provides a laundry list of unlawful acts that use the term “or” instead of “and” to connect two or more elements. Subsection (10)(A), in particular, offers a useful contrast to subsection (8). It has the same grammatical structure as subsection (8), using a negative verb phrase followed by two direct objects that are joined by a conjunction. It identifies an unlawful act that involves two or more elements but connects those elements by using “or” instead of “and.” Thus, following the same format as subsection (8), it states that it is an unlawful act to knowingly “fail[] to provide to an individual a health care benefit or service that the organization is required to provide under the contract.” TEX. HUM. RES. CODE § 36.002(10)(A) (emphasis added). The clear-cut understanding of this subsection is that it uses “or” to convey that a person commits an unlawful act if he fails to provide just one of the elements—either a health care benefit or a service. The State’s proposed construction would require us to conclude that the Legislature intended the word “and” in subsection (8) to bear the same meaning as the word “or” in subsection (10)(A). But like the Supreme Court, we generally presume the Legislature uses the same word consistently throughout a statute and uses different words to convey different meanings. See S.C. v. M.B., 650 S.W.3d 428, 445 (Tex. 2022); see also Pulsifer, 144 S. Ct. at 735 (discussing the interpretive principle that, “[i]n a given statute, the same term usually has the same meaning and different terms usually have different meanings” (citing ANTONIN SCALIA & BRYAN A. GARNER, READING LAW 170–71 (2012))).[18] Under that presumption, one could logically conclude that because subsection (10)(A), the only subsection that is structurally identical to subsection (8), uses “or” to convey a meaning that makes unlawful the failure to provide either one of the elements, subsection (8) must use “and” to convey a different meaning. We conclude the statutory context thus confirms that the term “and” in subsection (8) makes unlawful the failure to indicate both of the elements and is thus used in the distributive sense.[19] 3. Statutory purpose When discussing Section 3553(f)(1)’s statutory context, the Pulsifer Court also considered the government’s argument that, under Pulsifer’s joint reading, criminal defendants’ eligibility for relief “would not correspond to the seriousness of their criminal record,” while the government’s distributive interpretation “renders the provision capable of sorting more serious from less serious criminal records.” Pulsifer, 144 S. Ct. at 731, 737. Although the Supreme Court concluded in Pulsifer that the statutory context conclusively answered the question, it nevertheless proceeded to consider Pulsifer’s counterargument that Section 3553(f)(1) should be construed to accomplish the statute’s broader purpose, which he argued is to provide defendants more taking, possession, and sale of alligators,” because by using “or” Section65.003 governs all three. Id. § 65.002 (emphasis added). To be clear, we do not hold, as the dissent suggests, that Section 36.002(8) permits a claimant to simply “make up” an identification number or that it asks “for nothing more than a provider’s license type.” Post at 14, 25 (Young, J., dissenting). What we do—what we must do—is apply the section as written, and as written it does not “ask for” anything. Instead, it describes the conduct that will constitute a wrongful act and thus subject a provider to substantial statutory penalties. Perhaps, as the dissent is convinced, the Legislature would prefer to impose such penalties whenever a claim form fails to indicate the actual provider’s identification number or license type, for all the reasons the dissent attempts to explain. But what Section 36.002(8) does is impose penalties on a provider who fails to indicate his identification number “and” license type. We presume that “every word in a statute is used for a purpose,” Jessen Assocs. v. Bullock, 531 S.W.2d 593, 600 (Tex. 1975), and that “the Legislature included words that it intended to include and omitted words it intended to omit,” Union Carbide Corp. v. Synatzske, 438 S.W.3d 39, 52 (Tex. 2014). If the Legislature mistakenly used the term “and” instead of “or” in Section 36.002(8), it is up to the Legislature— not the courts—to amend the statute to fix that mistake. See DeQueen, 325 S.W.3d at 638. opportunities to avoid mandatory-minimum sentences. Id. at 736. The Court accepted Pulsifer’s description of the statute’s broader purpose but rejected the notion that it should construe the statute in a way that most fully promotes that purpose. See id. at 737. The Court observed that “[b]oth views of the paragraph—Pulsifer’s and the Government’s— significantly widen the opportunity for” defendants to avoid mandatory-minimum sentences, but “Pulsifer’s interpretation is not better just because it would go further than the Government’s.” Id. Malouf makes a similar purpose-based argument. He asserts that the purpose of the Medicaid Fraud Prevention Act is to prevent providers from deceptively claiming and obtaining payments for services they did not provide—in other words, to protect the Medicaid program against losses from fraud.[20] Malouf contends a failure to indicate either the provider’s license type or identification number, but not both, is merely a technical error that is not deceptive and does not result in losses to the program. Based on this argument, he urges us to construe Section 36.002(8) so that it only encompasses acts that “involve deception or misrepresentation to obtain an unauthorized benefit.” We proceed with great hesitation when asked to construe statutory text based on the statute’s purpose, particularly when the statute never expresses its purpose. On the one hand, the Legislature has said courts interpreting statutes may consider the “object sought to be obtained,” TEX. GOV’T CODE § 311.023(1), and we have acknowledged that courts may do so, Wal-Mart Stores, Inc. v. Forte, 497 S.W.3d 460, 466 (Tex. 2016). When a statute expressly states a legislative purpose, we have agreed that the statutory context “necessarily includes the Legislature’s codified purpose.” Hogan v. Zoanni, 627 S.W.3d 163, 175 (Tex. 2021). But we must “look to the statute’s text to determine the policy choices that the Legislature made when deciding how to achieve” its purpose, Jaster v. Comet II Constr., Inc., 438 S.W.3d 556, 570 (Tex. 2014) (plurality op.), and we “may not seek a different result by considering what unexpressed purposes, policy considerations, or interests the Legislature may have had in mind” but did not express, BonsmaraNat. Beef Co. v. Hart of Tex. CattleFeeders, LLC, 603 S.W.3d 385, 391 (Tex. 2020). Yet as we have previously recognized, Section 36.002 is “a powerful tool for targeting fraud against the Texas Medicaid program and securing the program’s integrity” and plays a crucial role in the State’s “efforts to deter, detect, and punish Medicaid fraud.” Xerox, 555 S.W.3d at 525. And we acknowledge that even a knowing failure to indicate an actual provider’s identification number in a claim that accurately indicates his license type may not sound a lot like fraud as the Government Code defines that term, at least when (as here) the State does not dispute that a licensed dentist actually provided, and a Medicaid recipient actually received, the services the State paid for. The State, in fact, has not identified any harm that affects Texas Medicaid when a person fails to indicate one element but not the other. We thus conclude that construing Section 36.002(8) as the State proposes would not correspond to the seriousness of the harm a violation may cause to the Texas Medicaid Program. A failure to indicate only a provider’s license type or only a provider’s identification number would not prevent Texas Medicaid from correctly processing a claim for dental services. As evidenced by the present case, even an incorrect indication of the identification number still allows Texas Medicaid to properly process claims and distribute the appropriate amount of funds to the provider as long as the license type embedded in the TPI number correctly reflects the license type of the performing provider. Under a distributive interpretation, providers are appropriately penalized only when they fail to correctly indicate both elements, which directly corresponds with the only scenario in which Texas Medicaid would be unable to accurately process the claim without over—or under— distribution of funds. We thus agree with Malouf that we should construe “and” in the distributive sense as he proposes because, as far as we can tell, the failure to indicate the license type and identification number is likely to be fraudulent and harmful to the State only in combination and each failure alone is not. See Pulsifer, 144 S. Ct. at 730 (explaining that we “interpret the injunction against drinking and driving in Pulsifer’s way—’do not (A and B)’—because the two activities are usually perilous onlyin combination,” but we “interpret the injunction against eating and drinking before surgery in the Government’s way—’do not A and do not B’—because each activity alone is likely to have adverse consequence”). The Act’s purpose of preventing fraudulent harm to the Medicaid program thus provides additional support to Malouf’s proposed construction. 4. The rule of lenity Pulsifer urged the Supreme Court to apply the rule of lenity and construe Section 3553(f)(1) in his favor because, he asserted, “the meaning of the criminal-history requirement is uncertain.” Id. The Court declined, however, because it did not view the statute “as gen-uinely ambiguous.” Id. Although it identified “two grammatically permissible readings of the statute when viewed in the abstract,” it concluded the statutory context eliminated Pulsifer’s proposed reading and “the two possible readings thus reduce to one—leaving no role for lenity to play.” Id. We reach the same conclusion regarding the statutory context of Section 36.002(8). The provision’s language, grammatical structure, statutory context, and general purpose leave Malouf’s construction of “and” in the distributive sense the only permissible reading. We are therefore convinced, as the Supreme Court was in Pulsifer, that the State’s alternative construction is unreasonable. But even if the State’s construction were also reasonable, the rule of lenity would require us to construe the statute in Malouf’s favor. IV. Conclusion and Disposition We conclude the State has failed to demonstrate in this case that Malouf committed unlawful acts under Section 36.002(8) by submitting the 1,842 claims at issue. When both sides move for summary judgment and the trial court grants one motion and denies the other, we review both sides’ summary-judgment evidence and determine all questions presented. FM Props. Operating Co. v. City of Austin, 22 S.W.3d 868, 872 (Tex. 2000). Here, the trial court denied Malouf’s summary-judgment motion and granted the State’s, and Malouf preserved his objection to the denial before both the court of appeals and this Court. We thus reverse the court of appeals’ judgment, and we render judgment in Malouf’s favor. Jeffrey S. Boyd Justice OPINION DELIVERED: June 21, 2024
