This appeal was brought by the Court Clerk for Tulsa County, Oklahoma, asking us to overturn a decision by the district court declaring unenforceable the Oklahoma state constitutional prohibition on issuing marriage licenses to same-sex couples. It followed quickly on the heels of an analogous appeal brought by State of Utah officials requesting similar relief. Recognizing that the ruling in the Utah case would likely control the disposition of her appeal, the Oklahoma appellant asked that we assign these cases to the same panel. Our court did so.
Preliminary to reaching the merits, we are presented with two arguments challenging the plaintiffs’ standing. The first challenges whether plaintiffs may attack state constitutional provisions without simultaneously attacking state statutes to the same effect. The second challenges whether the Court Clerk is a proper defendant as to the non-recognition portion of the Oklahoma constitutional prohibition.
We hold that plaintiffs possess standing to directly attack the constitutionality under the United States Constitution of Oklahoma’s same-sex marriage ban even though their claim does not reach Oklahoma’s statutory prohibitions on such marriages. Under Oklahoma law, a constitutional amendment “takes the place of all the former laws existing upon the subject with which it deals.” Fent v. Henry, 257 P.3d 984, 992 n. 20 (Okla.2011) (per curiam) (quotation omitted). Because the statutory prohibitions are subsumed in the challenged constitutional provision, an injunction against the latter’s enforcement will redress the claimed injury.
An earlier appeal of this same case involving the standing inquiry led to a decision by a panel of our court that dismissed proceedings brought against the Governor and Attorney General of Oklahoma. That panel ruled that “recognition of marriages is within the administration of the judiciary.” Bishop v. Okla. ex rel. Edmondson, 333 Fed.Appx. 361, 365 (10th Cir.2009) (unpublished) (“Bishop I”).We conclude that the law of the case doctrine applies to Bishop I, but that the doctrine is overcome by new evidence demonstrating that the Tulsa County Court Clerk could not redress the non-recognition injury, thereby depriving Gay Phillips and Susan Barton (the “Barton couple”) of standing to sue.
Our merits disposition is governed by our ruling in Kitchen v. Herbert, No 13-4178, 755 F.3d 1193, 2014 WL 2868044 (10th Cir. June 25, 2014). In that companion case, we held that: (1) plaintiffs who wish to marry a partner of the same sex or have such marriages recognized seek to exercise a fundamental right; and (2) state justifications for banning same-sex marriage that turn on the procreative potential of opposite-sex couples do not satisfy the narrow tailoring test applicable to laws that impinge upon fundamental liberties. Exercising jurisdiction under 28 U.S.C. § 1291, and governed by our ruling in Kitchen, we affirm.
I
Mary Bishop and Sharon Baldwin are in a long-term committed relationship and seek to marry. They live together in Tulsa County, Oklahoma, where they both work for the Tulsa World newspaper. Bishop is a sixth-generation Oklahoman and Baldwin is “at least a fourth-generation Oklahoman.” They jointly own their home and other property.
In March 2000, the couple exchanged vows in a church-recognized “commitment ceremony.” They feel, however, that this ceremony fails to “signify the equality” of their relationship, and that marriage con*1075veys a “level of commitment or respect” that is not otherwise available. Bishop and Baldwin sought a marriage license from the Tulsa County Court Clerk in February 2009, but were denied because they are both women. The couple identifies several discrete harms they have suffered because of their inability to marry, including $1,300 in legal fees to prepare a power of attorney form and healthcare proxies. Moreover, they explain that their inability to marry under Oklahoma law is “demeaning” and “signals to others that they should not respect our relationship.”
Phillips and Barton have been in a committed relationship since 1984. They took part in a civil union ceremony in Vermont in 2001, were married in Canada in 2005, and wed again in California in 2008. The couple jointly owns a company that provides training and assistance to non-profit agencies that conduct youth out-of-home care. Barton also teaches classes at Tulsa Community College, including a course titled “Building Relationships.”
Phillips and Barton have suffered adverse federal tax consequences as a result of the Defense of Marriage Act (“DOMA”), as well as adverse state tax consequences stemming from Oklahoma’s refusal to recognize their marital status. They say that having their relationship recognized as a marriage “should have been a dream come true.” Instead, “the State of Oklahoma has said ours is not a real marriage, but something inferior to the relationships of married opposite sex couples.”
In November 2004, plaintiffs Bishop, Baldwin, Barton, and Phillips filed suit against the Oklahoma Governor and Attorney General, challenging Oklahoma’s state constitutional ban on same-sex marriage. The Oklahoma prohibition, known as State Question 711 (“SQ 711”), provides:
A.Marriage in this state shall consist only of the union of one man and one woman. Neither this Constitution nor any other provision of law shall be construed to require that marital status or the legal incidents thereof be conferred upon unmarried couples or groups.
B. marriage between persons of the same gender performed in another state shall not be recognized as valid and binding in this state as of the date of the marriage.
C. Any person knowingly issuing a marriage license in violation of this section shall be guilty of a misdemeanor.
Okla. Const, art. 2, § 35. The suit also named the United States President and Attorney General as defendants in a constitutional challenge to DOMA.
A motion to dismiss filed by the Governor and State Attorney General was denied by the district court in 2006. That decision was appealed to this court. In 2009, a panel of our court concluded that “[b]ecause the plaintiffs failed to name a defendant having a causal connection to their alleged injury that is redressable by a favorable court decision, ... the Couples do not have standing.” Bishop I, 333 Fed.Appx. at 364. The panel held that “recognition of marriages is within the administration of the judiciary,” and thus “the executive branch of Oklahoma’s government has no authority to issue a marriage license or record a marriage.” Id. at 365.
On remand, the district court permitted the plaintiffs to file an amended complaint naming as a defendant the “State of Oklahoma, ex rel. Sally Howe-Smith, in her official capacity as Court Clerk for Tulsa County.” The court granted Oklahoma’s motion to dismiss the state as a nominal party, leaving Smith as the sole state defendant. The amended complaint also asserted challenges to §§ 2 and 3 of DOMA against the United States ex rel. Eric Holder. However, in February 2011, the United States notified the district court that it would no longer defend § 3 of *1076DOMA on the merits. The Bipartisan Legal Advisory Group was permitted to intervene to defend the law. The case then progressed to the summary-judgment stage. Smith submitted an affidavit describing her duties as they related to the plaintiffs’ allegations. In that affidavit, Smith swore that she had “no authority to recognize or record a marriage license issued by another state in any setting, regardless of whether the license was issued to an opposite-sex or a same-sex couple.”
After the Supreme Court issued its decision in United States v. Windsor, — U.S. -, 133 S.Ct. 2675, 186 L.Ed.2d 808 (2013), the district court entered an opinion and order disposing of the United States’ motion to dismiss, as well as Oklahoma and plaintiffs’ cross-motions for summary judgment. See Bishop v. United States ex rel. Holder, 962 F.Supp.2d 1252, 1263 (N.D.Okla.2014) (“Bishop II”).The district court concluded that: (1) Phillips and Barton lacked standing to challenge § 2 of DOMA because state law, rather than that provision, resulted in non-recognition of them marriage, id. at 1263-68; (2) any challenge to § 3 of DOMA was moot in light of the Windsor decision, id. at 1269-72; (3) Phillips and Barton lacked standing to challenge the non-recognition portion of the Oklahoma amendment, Part B, because Smith is not involved in the recognition of out-of-state marriages, as established by her summary-judgment affidavit, id. at 1272-73; and (4) Part A of SQ 711 violates the Equal Protection Clause, id. at 1281-96. The court permanently enjoined enforcement of Part A. Id. at 1296. The decision, however, was stayed pending final disposition of any appeal. Id.
Smith timely appealed the district court’s merits ruling as to Part A. Phillips and Barton cross-appealed the district court’s conclusion that they lack standing to challenge Part B. The DOMA challenges are not at issue in this appeal.
II
A
Smith contends that Bishop and Baldwin (the “Bishop couple”) lack standing to challenge Part A of SQ 711 because they did not simultaneously contest the constitutionality of a state statute that bars same-sex couples from marrying. We review a district court’s standing determinations de novo. See Cressman v. Thompson, 719 F.3d 1139, 1144 (10th Cir.2013). To establish standing, a plaintiff must show:
(1) it has suffered an “injury in fact” that is (a) concrete and particularized and (b) actual or imminent, not conjectural or hypothetical; (2) the injury is fairly traceable to the challenged action of the defendant; and (3) it is likely, as opposed to merely speculative, that the injury will be redressed by a favorable decision.
Friends of the Earth, Inc. v. Laidlaw Envtl. Servs. (TOC), Inc., 528 U.S. 167, 180-81, 120 S.Ct. 693, 145 L.Ed.2d 610 (2000). Although the Bishop couple’s standing was not raised below, a party may “raise the issue of standing for the first time at any stage of the litigation, including on appeal.” New Eng. Health Care Emps. Pension Fund v. Woodruff, 512 F.3d 1283, 1288 (10th Cir.2008).
The Bishop couple has not established redressability, Smith argues, because a second, unchallenged legal obstacle bars their marriage. Under Okla. Stat. tit. 43, § 3(a), which was not properly put at issue below, “[a]ny unmarried person who is at least eighteen (18) years of age and not otherwise disqualified is capable of contracting and consenting to marriage with a person of the opposite sex.” Id. *1077Although the district court enjoined enforcement of Part A, it did not enjoin operation of the statute. See Bishop II, 962 F.Supp.2d at 1296. Because the statute permits marriage only between members of the opposite sex, Smith argues that the Bishop couple’s injury — their inability to marry — will not be redressed by an injunction against SQ 711 alone.1 “[R]e-dressability is satisfied when a favorable decision relieves an injury,” but a decision does not need to relieve “every injury.” Consumer Data Indus. Ass’n v. King, 678 F.3d 898, 905 (10th Cir.2012) (emphasis omitted).
In support, Smith asserts that several courts have concluded that plaintiffs lack standing under circumstances somewhat similar to the present matter. In White v. United States, 601 F.3d 545 (6th Cir.2010), a group of plaintiffs challenged the federal Animal Welfare Act (“AWA”), which restricted “various activities associated with animal fighting that involve interstate travel and commerce, but did not (and does not) itself prohibit animal fighting, including cockfighting.” Id. at 549. All fifty states, however, have prohibited cockfighting under state law. Id. The plaintiffs claimed that they had suffered economic injuries as a result of the federal statute’s ban, including a decreased market for fighting birds. Id. at 549-50. The Court concluded that these allegations did not support standing:
Cockfighting is banned to a greater or lesser degree in all fifty states and the District of Columbia. Thus, while economic injuries may constitute an injury-in-fact for the purposes of Article III standing, the plaintiffs’ alleged economic injuries due to restrictions on cockfighting are not traceable only to the AWA. Nor would these injuries be redressed by the relief plaintiffs seek, since the states’ prohibitions on cockfighting would remain in place notwithstanding any action we might take in regard to the AWA.
Id. at 552 (citations omitted).
We are referred to numerous sign ordinance cases holding that “a plaintiff whose sign permit applications were denied on the basis of one provision in a county’s sign ordinance, but which could have been denied on the basis of some alternate, but unchallenged regulation, does not have a redressable injury.” Maverick Media Grp., Inc. v. Hillsborough Cnty., 528 F.3d 817, 820 (11th Cir.2008) (collecting cases). In Maverick, for example, the court ruled that a court order barring enforcement of a county’s ban on billboards would not aid the plaintiff because the signs it sought to build were also prohibited by unchallenged height and size limitations. Id. at 821, 823.
We need not decide whether the cases cited by Smith are consistent with our circuit precedent because they are readily *1078distinguishable from the case at hand. Courts have concluded that plaintiffs fail to establish redressability only when an unchallenged legal obstacle is enforceable separately and distinctly from the challenged provision. In White, the federal statute meaningfully differed from the state cockfighting prohibitions and was enforced by a different sovereign. See 601 F.3d at 549. Similarly, the sign cases rest on the existence of an “alternate” regulation addressing a distinct issue. See Maverick, 528 F.3d at 820.
Unlike the statutes and regulations at issue in the cases upon which Smith relies, Okla. Stat. tit. 43, § 3(a) is not enforceable independent of SQ 711. Under Oklahoma law:
A time-honored rule teaches that a revising statute (or, as in this case, a constitutional amendment) takes the place of all the former laws existing upon the subject with which it deals. This is true even though it contains no express words to that effect. In the strictest sense this process is not repeal by implication. Rather, it rests upon the principle that when it is apparent from the framework of the revision that whatever is embraced in the new law shall control and whatever is excluded is discarded, decisive evidence exists of an intention to prescribe the latest provisions as the only ones on that subject which shall be obligatory.
Fent, 257 P.3d at 992 n. 20 (quoting Hendrick v. Waiters, 865 P.2d 1232, 1240 (Okla.1993)). This rule suggests that SQ 711 “takes the place of’ § 3(a), and only the provisions of the constitutional amendment “shall be obligatory.” Fent, 257 P.3d at 992 n. 20.
Fent, Smith informs us, stands for the opposite proposition because another portion of the opinion notes the general rules that “repeals by implication are never favored,” that “it is not presumed that the legislature, in the enactment of a subsequent statute intended to repeal an earlier one, unless it has done so in express terms,” and that “all provisions must be given effect unless irreconcilable conflicts exist.” Id. at 991. But the quoted passage clarifies that when a constitutional amendment addresses the same subject as a statute, replacement is “not repeal by implication” and occurs even absent “express words.” Id. at 992 n. 20.
Fent did not involve a constitutional amendment replacing a statute; the court simply noted the rule in a footnote. The relevant quotation originates in Hendrick, which held that a constitutional amendment providing for a new oath of office for certain state positions superseded an existing statute prescribing a different oath. 865 P.2d at 1240-41. Smith is correct that the provisions at issue in Hendrick were arguably in conflict and the court found an “intent to abrogate.” Id. at 1240 n. 41. However, the broad language used in Hen-drick and quoted in Fent directs that if the “framework” of a constitutional amendment indicates “that whatever is embraced in the new law shall control and whatever is excluded is discarded,” courts should treat this framework as “decisive evidence” that the amendment is the only provision “on that subject which shall be obligatory.” Fent, 257 P.3d at 992 n. 20 (quoting Hendrick, 865 P.2d at 1240).
SQ 711 evinces such a framework. The Oklahoma Supreme Court cited Lankford v. Menefee, 45 Okla. 228, 145 P. 375 (1914), in support of its conclusion in Hendrick. See 865 P.2d at 1240 nn. 38-40. Lankford provides that “a subsequent statute revising the subject-matter of the former one, and evidently intended as a substitute for it, although it contains no express words to that effect, must operate to repeal the former” as long as “it is apparent that the Legislature designed a complete scheme *1079for the matter.” 145 P. at 376. It follows that SQ 711 provides a complete scheme for Oklahoma’s policy regarding same-sex marriage.
The statute identified by Smith has no effect beyond the restrictions on same-sex marriage imposed by SQ 711 because the two provisions are materially identical. Total eclipse of the function of the statute underscores our conclusion that the amendment provides a complete scheme. Further, it raises the concern that the statute could not be enforced without violating the district court’s injunction. Smith was enjoined from enforcing “Part A against same-sex couples seeking a marriage license.” Bishop II, 962 F.Supp.2d at 1296. If Smith were to deny the Bishop couple a marriage license because they are both women, she would simultaneously be enforcing both Okla. Stat. tit. 43, § 3(a) and Part A of SQ 711. There is no scenario in which Smith could enforce the statute but not enforce the amendment.2
Because the prohibition on same-sex marriage contained in Okla. Stat. tit. 43, § 3(a) is not enforceable independently of SQ 711, we conclude that the Bishop cou-pie has shown that their injury is redressa-ble in this suit.3
B
Our consideration of the merits of the Bishop couple’s appeal is largely controlled by our decision in Kitchen. As explained more fully in that opinion, we conclude that: (1) the Supreme Court’s summary dismissal in Baker v. Nelson, 409 U.S. 810, 93 S.Ct. 37, 34 L.Ed.2d 65 (1972) (per curiam), is not controlling, Kitchen, 755 F.3d at 1204-08, 2014 WL 2868044, at *7-10, 2014 U.S.App. LEXIS 11935, at *21-31; plaintiffs seek to exercise the fundamental right to marry, id. at 1208-10, 2014 WL 2868044, at *11-12, 2014 U.S.App. LEXIS 11935, at *33-63; and (3) state arguments that same-sex marriage bans are justified by the need to communicate a conceptual link between marriage and procreation, encourage parenting by mothers and fathers, and promote sacrifice by parents for their children fail to satisfy the narrow tailoring requirement of the applicable strict scrutiny test, id. at 1218-27, 2014 WL 2868044, at *21-29, 2014 U.S.App. LEXIS 11935, at *63-87.
*1080Facts and arguments presented in this case differ in some respects from those in Kitchen. But our core holdings are not affected by those differences. State bans on the licensing of same-sex marriage significantly burden the fundamental right to marry,4 and arguments based on the procreative capacity of some opposite-sex couples do not meet the narrow tailoring prong. In addition to the issues explicitly discussed in Kitchen, we address two other arguments raised by Smith.
She contends that lower federal courts are not free to reject on-point summary dismissals of the Supreme Court regardless of doctrinal developments. Thus, Smith argues, Baker remains controlling. Her focus is on the Court’s statement that a summary disposition “is not here of the same precedential value as would be an opinion of this Court treating the question on the merits.” Tully v. Griffin, Inc., 429 U.S. 68, 74, 97 S.Ct. 219, 50 L.Ed.2d 227 (1976) (quotation omitted, emphasis added). This statement, Smith contends, indicates that, although they may have diminished precedential value for the Supreme Court, summary dispositions are identical to merits decisions when considered by lower courts. She also cites the Court’s direction that summary dispositions “prevent lower courts from coming to opposite conclusions on the precise issues presented and necessarily decided by those actions.” Mandel v. Bradley, 432 U.S. 173, 176, 97 S.Ct. 2238, 53 L.Ed.2d 199 (1977).
Her argument that doctrinal developments do not allow a lower court to reject the continued applicability of a summary disposition is undermined by the explicit language of the case creating that rule. In Hicks v. Miranda, 422 U.S. 332, 95 S.Ct. 2281, 45 L.Ed.2d 223 (1975), the Court stated that “inferior federal courts had best adhere to the view that if the Court has branded a question as unsubstantial, it remains so except when doctrinal developments indicate otherwise.” Id. at 344, 95 S.Ct. 2281 (quotation omitted, emphases added); see also Shaw v. Delta Air Lines, Inc., 463 U.S. 85, 94 n. 11, 103 S.Ct. 2890, 77 L.Ed.2d 490 (1983) (noting circuit court’s holding that a doctrinal development warranted departure from precedent set by Supreme Court’s summary dispositions); Okla. Telecasters Ass’n v. Crisp, 699 F.2d 490, 495 (10th Cir.1983) (“[A] summary disposition is binding on the lower federal courts, at least where substantially similar issues are presented, until doctrinal developments or direct decisions by the Supreme Court indicate otherwise.” (emphases added)), rev’d sub nom. Capital Cities Cable, Inc. v. Crisp, 467 U.S. 691, 104 S.Ct. 2694, 81 L.Ed.2d 580 (1984). Thus, contrary to Smith’s position, the doctrinal developments statement is explicitly directed toward lower courts. And as explained in Kitchen, nearly every lower federal court to have considered the issue has concluded that Baker has been undermined by doctrinal developments. Kitchen, 755 F.3d at 1205-07, 2014 WL 2868044, at *8-9, 2014 U.S.App. LEXIS 11935, at *25-26.
In addition to her Baker argument, Smith also contends that children have an interest in being raised by their biological parents. Assuming that serving this interest is a compelling governmental goal, we *1081nevertheless conclude that a prohibition on same-sex marriage is not narrowly tailored to achieve that end. See Reno v. Flores, 507 U.S. 292, 301-02, 113 S.Ct. 1439, 123 L.Ed.2d 1 (1993) (stating strict scrutiny test). Oklahoma has enacted numerous laws that result in children being raised by individuals other than their biological parents. See, e.g., Okla. Stat. tit. 10, § 554 (“Any child or children born as a result of a heterologous oocyte donation shall be considered for all legal intents and purposes, the same as a naturally conceived legitimate child of the husband and wife which consent to and receive an oocyte pursuant to the use of the technique of heterologous oocyte donation.”); § 556(B)(1) (“Any child or children born as a result of a human embryo transfer donation shall be considered for all legal intents and purposes, the same as a naturally conceived legitimate child of the husband and wife that consent to and receive a human embryo transfer.”); § 7501-1.2(A) (“The Legislature of this state believes that every child should be raised in a secure, loving home and finds that adoption is the best way to provide a permanent family for a child whose biological parents are not able or willing to provide for the child’s care or whose parents believe the child’s best interest will be best served through adoption.”). And Oklahoma permits infertile opposite-sex couples to marry despite the fact that they, as much as same-sex couples, might raise non-biological children.
The State thus overlooks the interests of children being raised by their biological parents in a wide variety of contexts. Yet Smith does not explain why same-sex marriage poses a unique threat such that it must be treated differently from these other circumstances. See Zablocki v. Redhail, 434 U.S. 374, 390, 98 S.Ct. 673, 54 L.Ed.2d 618 (1978) (“grossly underinclu-sive” statute did not satisfy narrow tailoring requirement). As the Court explained in Eisenstadt v. Baird, 405 U.S. 438, 92 S.Ct. 1029, 31 L.Ed.2d 349 (1972), if “the evil, as perceived by the State, would be identical” with respect to two classes, the state may not impinge upon the exercise of a fundamental right as to only one class because “the underinclusion would be invidious.” Id. at 454, 92 S.Ct. 1029. As we explained in Kitchen, such divergence between the characteristic claimed to be relevant and the classification contained in the challenged provision is inconsistent with the narrow tailoring requirement. See Kitchen, 755 F.3d at 1218-22, 2014 WL 2868044, at *21-25, 2014 U.S.App. LEXIS 11935, at *64-75.
Moreover, Oklahoma’s ban on same-sex marriage sweeps too broadly in that it denies a fundamental right to all same-sex couples who seek to marry or to have then-marriages recognized regardless of their child-rearing ambitions. As with opposite-sex couples, members of same-sex couples have a constitutional right to choose against procreation. See Eisenstadt, 405 U.S. at 453, 92 S.Ct. 1029 (“If the right of privacy means anything, it is the right of the individual, married or single, to be free from unwarranted governmental intrusion into matters so fundamentally affecting a person as the decision whether to bear or beget a child.” (emphasis omitted)). But Oklahoma has barred all same-sex couples, regardless of whether they will adopt, bear, or otherwise raise children, from the benefits of marriage while allowing all opposite-sex couples, regardless of their child-rearing decisions, to marry. Such a regime falls well short of establishing “the most exact connection between justification and classification.” Gratz v. Bollinger, 539 U.S. 244, 270, 123 S.Ct. 2411, 156 L.Ed.2d 257 (2003) (quotation omitted); see also Frisby v. Schultz, 487 U.S. 474, 485, 108 S.Ct. 2495, 101 L.Ed.2d 420 (1988) (“A statute is narrowly tailored if it targets *1082and eliminates no more than the exact source of the ‘evil’ it seeks to remedy.”).
In summary, none of the arguments presented by Smith that were unaddressed in Kitchen persuade us to veer from our core holding that states may not, consistent with the United States Constitution, prohibit same-sex marriages.
Ill
I am grateful to Judge Holmes for his authorship of this, Part III of the majority opinion. Judge Holmes was on panel for our earlier decision in Bishop I. His authorship of this section is acknowledged with thanks.
Because Smith lacks “authority to recognize any out-of-state marriage and therefore [lacks the] ability to redress the Barton couple’s non-recognition injury,” Bishop II, 962 F.Supp.2d at 1273, the district court held that the Barton couple lacked standing to challenge Part B of SQ 711 as against Smith. We conclude that although the law of the case doctrine applied to Bishop I, Smith’s affidavit constituted new evidence sufficient to overcome the doctrine. We further conclude that the Barton couple’s argument that Part B is inseverable from Part A — and that both must therefore fall together — was forfeited.
A
“Under the ‘law of the case’ doctrine, when a court rules on an issue of law, the ruling ‘should continue to govern the same issues in subsequent stages in the same case.’ ” United States v. Graham, 704 F.3d 1275, 1278 (10th Cir.2013) (quoting Arizona v. California, 460 U.S. 605, 618, 103 S.Ct. 1382, 75 L.Ed.2d 318 (1983)) (quotation omitted). The doctrine pertains both to rulings by district courts, see, e.g., Clark v. State Farm Mut. Auto. Ins. Co., 590 F.3d 1134, 1140 (10th Cir.2009), and-as relevant here — by previous panels in prior appeals in the same litigation, see, e.g., United States v. Wardell, 591 F.3d 1279, 1300 (10th Cir.2009). Importantly, “[w]e have routinely recognized that the law of the case doctrine is ‘discretionary, not mandatory,’ and that the rule ‘merely expresses the practice of courts generally to refuse to reopen what has been decided, not a limit on their power.’ ” Kennedy v. Lubar, 273 F.3d 1293, 1299 (10th Cir.2001) (quoting Stifel, Nicolaus & Co. v. Woolsey & Co., 81 F.3d 1540, 1544 (10th Cir.1996)) (quotation omitted); accord Haynes Trane Serv. Agency v. Am. Standard, Inc., 573 F.3d 947, 963 (10th Cir.2009). Even so, it takes “exceptionally narrow circumstances” for the court not to follow the law of the case when the doctrine applies. United States v. Alvarez, 142 F.3d 1243, 1247 (10th Cir.1998).
In Bishop I, a panel of this court found that neither the Barton couple nor the Bishop couple had standing to challenge SQ 711. 333 Fed.Appx. at 365. It determined that the couples could not demonstrate redressability, reasoning as follows:
The Couples claim they desire to be married but are prevented from doing so, or they are married but the marriage is not recognized in Oklahoma. These claims are simply not connected to the duties of the Attorney General or the Governor. Marriage licenses are issued, fees collected, and the licenses recorded by the district court clerks. [A] district court clerk is judicial personnel and is an arm of the court whose duties are ministerial, except for those discretionary duties provided by statute. In the performance of [a] clerk’s ministerial functions, the court clerk is subject to the control of the Supreme Court and the supervisory control that it has passed down to the Administrative District Judge in the clerk’s administrative district. Because recognition of marriages is within the administration of the *1083judiciary, the executive branch of Oklahoma’s government has no authority to issue a marriage license or record a marriage.
Id. (alterations in original) (quotation and citations omitted). Taking this passage at face value, it is most logically construed as the panel’s determination that the Barton couple should have sued a district court clerk on their non-recognition claim. The panel: (1) prefaced its discussion with a reference to both the ban and the nonrecognition claims; (2) found standing on neither; (3) reasoned that the Attorney General and the Governor were improper defendants; (4) explained that judicial personnel were proper defendants; and (5) informed the plaintiffs that court clerks represented the judiciary and carried out many of the branch’s duties relating to marriage. Collectively, these points lead to but one interpretation: the correct defendant for the Barton couple’s nonrecognition claim was a court clerk.
One possible counterargument is that when the panel wrote that “recognition of marriages” was “within the administration of the judiciary,” id., it meant in the broader sense of recognizing a couple’s right to get a marriage license in Oklahoma. That argument makes little sense when one considers the context: the first sentence of the paragraph describes the complaint of the couples (more specifically, the Barton couple) as alleging that “they are married but the marriage is not recognized in Oklahoma,” id. (emphasis added), and the order consistently uses some form of the word “recognize” to describe the Barton couple’s claim, see id. at 362-63.
Another potential counterargument is that the panel determined only that the Barton couple should look for a defendant in the judicial branch, not that they should necessarily select a court clerk. See id. at 365 (“Because recognition of marriages is within the administration of the judiciary, the executive branch of Oklahoma’s government has no authority to issue a marriage license or record a marriage.” (emphasis added)). Again, though, context belies this interpretation. Why mention the role of the court clerks in administering the marriage statutes, and why describe their relationship to the rest of the court system, if not to express the opinion that they are appropriate defendants?
That the panel concluded that a court clerk was the proper adversary for the Barton couple does not necessarily mean that this conclusion became the law of the case. There are three potential reasons to hold that it did not: (1) the conclusion was dicta; (2) the conclusion dealt with recognition of an older marriage entered into by the Barton couple, not their current marriage; and (3) as a jurisdictional determination, the conclusion was not subject to the law of the case doctrine. None of these reasons are persuasive.
Turning to the first, it is well-settled that “[d]icta is not subject to the law of the case doctrine.” Homans v. City of Albuquerque, 366 F.3d 900, 904 n. 5 (10th Cir.2004); accord Octagon Res., Inc. v. Bonnett Res. Corp. (In re Meridian Reserve, Inc.), 87 F.3d 406, 410 (10th Cir.1996). Statements which appear in an opinion but which are unnecessary for its disposition are dicta. See United States v. Manatau, 647 F.3d 1048, 1054 (10th Cir.2011); United States v. Villarreal-Ortiz, 553 F.3d 1326, 1329 n. 3 (10th Cir.2009) (per curiam). One could argue that Bishop I held only that the Governor and the Attorney General were the wrong defendants, not that Smith was the right one. But it is not so easy to separate the two propositions as a logical matter, and the “law of the case applies to issues that are resolved implicitly.” Rishell v. Jane Phillips Episcopal Mem’l Med. Ctr., 94 F.3d 1407, 1410 (10th Cir.1996). Bishop J’s holding that the Governor and Attorney *1084General were improper defendants was tethered closely to the panel’s view of who the right defendant was. That is, the panel’s rationale for finding no standing was that the Governor and Attorney General were not responsible for administering marriage laws and the court clerks were. See Bishop I, 333 Fed.Appx. at 365 (“The Couples claim they desire to be married but are prevented from doing so, or they are married but the marriage is not recognized in Oklahoma. These claims are simply not connected to the duties of the Attorney General or the Governor. Marriage licenses are issued, fees collected, and the licenses recorded by the district court clerks.”). Therefore, the panel held, if only implicitly, that the court clerk was the correct defendant to name for the Barton couple’s non-recognition claim.
The second potential reason to rule that Bishop I created no law of the case on standing to sue on the non-recognition claim is that the panel never ruled on such a claim with reference to the Barton couple’s California marriage, upon which the claim is now based; rather, it ruled only on their Canadian marriage and Vermont civil union, since the California marriage was solemnized after briefing in the appeal was complete. See id. at 363 (mentioning the events in Vermont and Canada but not the California marriage). This is a distinction without a difference. The holding in Bishop I had nothing to do with what sovereign conferred the status that the Barton couple wished to have recognized; it had only to do with which state officials were responsible for offering or withholding that recognition. See id. at 365 (noting that “the executive branch of Oklahoma’s government has no authority to issue a marriage license or record a marriage”).
Lastly, it is Smith’s view that the law of the case doctrine is per se excluded from consideration on this point because the standing issue is jurisdictional. Smith’s stance is squarely foreclosed by Supreme Court precedent. In Christian-son v. Colt Industries Operating Corp., 486 U.S. 800, 108 S.Ct. 2166, 100 L.Ed.2d 811 (1988), the Court took up a dispute in which the Seventh Circuit and the Federal Circuit had each disclaimed jurisdiction and had each transferred the case to the other. Id. at 803-04, 108 S.Ct. 2166. The Supreme Court admonished the feuding circuit courts of the importance of “adhering strictly to principles of law of the case.” Id. at 819, 108 S.Ct. 2166. In so doing, the Supreme Court did not tailor its articulation of the law of the case doctrine to the jurisdictional context. Quite to the contrary, it explicitly declared that “[t]here is no reason to apply law-of-the-case principles less rigorously to transfer decisions that implicate the transferee’s jurisdiction.” Id. at 816 n. 5, 108 S.Ct. 2166. Christianson thus makes clear that the law of the case doctrine is never off the table solely because an issue is jurisdictional. The circuits have agreed that this rule applies to a situation, like the one present today, where a prior panel of the same court resolved a jurisdictional matter in an earlier appeal. See Alexander v. Jensen-Carter, 711 F.3d 905, 909 (8th Cir.2013); Sierra Club v. Khanjee Holding (US) Inc., 655 F.3d 699, 704 (7th Cir.2011); Free v. Abbott Labs., Inc., 164 F.3d 270, 272-73 (5th Cir.1999); Ferreira v. Borja, 93 F.3d 671, 674 (9th Cir.1996) (per cu-riam); LaShawn A. v. Barry, 87 F.3d 1389, 1394 (D.C.Cir.1996) (en banc); Oneida Indian Nation of N.Y. v. New York, 860 F.2d 1145, 1151 (2d Cir.1988).5
*1085For the proposition that the law of the case doctrine has no applicability to jurisdictional matters, Smith relies chiefly on Baca v. King, 92 F.3d 1031 (10th Cir.1996). Baca cannot support that weight. In the crucial passage from that case, we stated that “[o]ne application of the ‘law of the case’ doctrine gives an appellate court discretion to refuse to reconsider an issue decided at an earlier stage of the litigation” and that doctrine “is not a fixed rule that prevents a federal court from determining the question of its own subject matter jurisdiction in a given case.” Id. at 1035. Far from carving out an exception to customary law-of-the-case practices in the jurisdictional context, Baca was actually applying the classic law-of-the-case approach to a jurisdictional question. That is, the law of the case is never “a fixed rule,” id., but rather always a “discretionary ... practice of courts generally to refuse to reopen what has been decided.” Kennedy, 273 F.3d at 1299 (quotation omitted). Utilizing that well-established framework, the Baca court determined that the law of the case did not dictate the result of the jurisdictional question presented under the circumstances in that dispute. Baca did not foreclose the possibility that the law of the case might, in other controversies, control a jurisdictional issue.6
By emphasizing the jurisdictional nature of the issue, Baca reflected the longstanding rule that while there is no categorical exclusion from the law of the case doctrine for jurisdictional issues, a slightly more flexible methodology is called for in the jurisdictional context. In this regard, we have indicated that “[i]ssues such as subject matter jurisdiction ... may be particularly suitable for reconsideration,” even where the doctrine might otherwise counsel against it. Kennedy, 273 F.3d at 1299 (quotation and citation omitted). Our law on that point is consistent with respected secondary authority and with the pronouncements of our sister circuits. See Am. Canoe Ass’n v. Murphy Farms, Inc., 326 F.3d 505, 515 (4th Cir.2003) (“Law of the case, which is itself a malleable doctrine meant to balance the interests of correctness and finality, can likewise be calibrated to reflect the increased priority placed on subject matter jurisdictional issues generally, and Article III standing in particular which represents perhaps the most important of all jurisdictional requirements.” (emphasis added) (quotation omitted)); Shakman v. Dunne, 829 F.2d 1387, 1393 (7th Cir.1987) (“[C]ourts are significantly less constrained by the law of the case doctrine with respect to jurisdictional questions.” (emphasis added)); 18B Charles Alan Wright et al., Federal Practice and Proce*1086dure § 4478.5, at 790 (2d ed.2002) (henceforth Federal Practice) (noting that “[t]he force of law-of-the-case doctrine is affected by the nature of the first ruling and by the nature of the issues involved” and then ranking subject-matter jurisdiction as one of the issues “most likely to be reconsidered because of [its] conceptual importance”); id. at 798-800 (“Although a federal court is always responsible for assuring itself that it is acting within the limits of subject-matter jurisdiction statutes and Article III, this duty need not extend to perpetual reconsideration. A court may accept its own earlier determination supporting subject-matter jurisdiction or justiciability; a denial of subject-matter jurisdiction or justiciability is easily adhered to. Reconsideration of these matters is particularly appropriate nonetheless .... ” (emphases added) (footnotes omitted)).
In sum, the law of the case doctrine does apply to prior jurisdictional determinations by merits panels, but it applies in a somewhat weaker fashion such that the court can consider with special care whether an exception to the doctrine permits reassessment of jurisdiction. That more flexible form of the doctrine will be brought to bear in the following section.
B
Applying the law of the case doctrine with the foregoing considerations in mind, Bishop I does not require a finding of standing to sue on the non-recognition claim.
As a practice rather than a rigid rule, the law of the case is subject to three narrow exceptions: (1) when new evidence emerges; (2) when intervening law undermines the original decision; and (3) when the prior ruling was clearly erroneous and would, if followed, create a manifest injustice. See Irving, 665 F.3d at 1192 n. 12; Clark, 590 F.3d at 1140.
Although Smith focuses on the third exception, the first provides a better framework for the analysis. This is so because Smith does not make a case for why invocation of law of the case would work “a manifest injustice,” which the clearly-erroneous exception requires.7 See, e.g., Zinna v. Congrove, 755 F.3d 1177, 1181-83, 2014 WL 2523759, at *3-4, 2014 U.S.App. LEXIS 10460, at *11 (10th Cir.2014); Irving, 665 F.3d at 1192 n. 12; Rimbert v. Eli Lilly & Co., 647 F.3d 1247, 1251 (10th Cir.2011). Further, Smith is relying in her law-of-the-case argument on a document— her affidavit — that was not presented to the courts until after Bishop I’s issuance. If the affidavit shows Smith to be an improper defendant, as she maintains, then the Bishop I panel could not have clearly erred in finding to the contrary, as it did not have the benefit of that affidavit. Substantively, then, the new-evidence exception is the more appropriate exception to consider here.
Having located the relevant exception, we confront two questions: (1) whether the affidavit qualifies as new evidence for purposes of the exception; and (2) whether the affidavit proves the absence of standing. Both questions demand an affirmative answer.
1
Turning to the first question, there can be no serious argument that the affidavit is *1087anything other than new evidence within the meaning of the exception. Smith Machinery Co. v. Hesston Corp., 878 F.2d 1290 (10th Cir.1989), is a helpful place to begin. In that case, a district court at summary judgment reconsidered a previous ruling despite the law of the case, relying in part on the proposition that “the law of the case doctrine does not ... apply in cases in which new evidence is presented to a court.” Id. at 1292. We affirmed, noting that the district court had before it “depositions and affidavits presented by both parties” attesting to new and relevant facts. Id. at 1293. Tacitly, Smith Machinery endorsed the district court’s use of the summary-judgment affidavits in its new-evidence analysis.
This implicit holding is in keeping with general principles of law. As In re Antrobus, 563 F.3d 1092 (10th Cir.2009) (per curiam), intimated, an affidavit is properly categorized as new evidence under the law of the case where it constitutes “admissible evidence,” id. at 1099 n. 3, and affidavits are plainly competent evidence at summary judgment, see Fed.R.Civ.P. 56(c)(1)(A) (providing that a party moving for summary judgment may support its motion by pointing to affidavits); Hansen v. PT Bank Negara Indon. (Persero), 706 F.3d 1244, 1250 (10th Cir.2013) (“[A]ffida-vits are entirely proper on summary judgment. .. ,”).8
Nor is there any apparent reason why an affidavit at summary judgment would not be regarded as a proper piece of new evidence such that the exception is satisfied. That is presumably why the Fifth Circuit has accepted such affidavits as new evidence in evaluating whether the law of the case controls or not. See United States v. Horton, 622 F.2d 144, 148 (5th Cir.1980) (per curiam) (finding that the law of the case did not preclude the entry of summary judgment despite an earlier contrary ruling “because the production of reports, admissions, affidavits, and other record material during the course of the proceedings had clarified and resolved questions of material fact on several of the [relevant] issues”).
It is true that previously-available evidence often cannot be used to unsettle the law of the case. See In re Antrobus, 563 F.3d at 1099 (“The difficulty is that the Antrobuses have not demonstrated that they were unable to present evidence along these very same lines over a year ago, when this litigation began.”); United States v. Monsisvais, 946 F.2d 114, 117 (10th Cir.1991) (“The ‘different or new evidence’ exception does not apply because ... the additional evidence provided by the government at the supplemental hearing was evidence it had in its possession, but failed to produce, at the time of the original hearing.”). But neither Smith nor any other court clerk was a party to the case at the time of Bishop I. Smith consequently did not have an opportunity to introduce the evidence earlier, and no party had any reason to seek it out. As demonstrated by the quotes recited above from Antrobus and Monsisvais, this previously-available-evidenee bar is applied when the party seeking to circumvent the law of the case had a chance to introduce the evidence in the prior proceedings and failed to exploit that chance. See In re Antrobus, 563 F.3d at 1099 (“The difficulty is that the Antrobuses have not demonstrated that they were unable to present evidence along these very same lines over a year ago, when this litigation began.” (emphases added)); Monsisvais, 946 F.2d *1088at 117 (“The ‘different or new evidence’ exception does not apply because ... the additional evidence provided by the government at the supplemental hearing was evidence it had in its possession, but failed to produce, at the time of the original hearing.” (emphases added)). That is not the case here. Smith did not fail to do anything during Bishop I because she was not participating in Bishop I. Accordingly, this bar does not apply, and Smith’s affidavit does qualify as new evidence within the meaning of the new-evidence exception to the law of the case doctrine.9
2
The next question is whether the affidavit demonstrates a lack of standing. It does.
Article III standing is a prerequisite to every lawsuit in federal court. See Petrella v. Brownback, 697 F.3d 1285, 1292-93 (10th Cir.2012); Jackson v. Volvo Trucks N. Am., Inc., 462 F.3d 1234, 1241 (10th Cir.2006). “Each plaintiff must have standing to seek each form of relief in each claim.” Bronson v. Swensen, 500 F.3d 1099, 1106 (10th Cir.2007); accord Meyer v. Christie, 634 F.3d 1152, 1157 (10th Cir.2011). In order to demonstrate “Article III standing, a plaintiff must show: (1) that [she] has suffered a concrete and particular injury in fact that is either actual or imminent; (2) the injury is fairly traceable to the alleged actions of the defendant; and (3) the injury will likely be redressed by a favorable decision.” Kerr v. Hickenlooper, 744 F.3d 1156, 1163 (10th Cir.2014); accord S. Utah Wilderness Alliance v. Palma, 707 F.3d 1143, 1153 (10th Cir.2013). The issue at hand turns on the third requirement — that of redressability — which “is not met when a plaintiff seeks relief against a defendant with no power to enforce a challenged statute.” Bronson, 500 F.3d at 1111. As established by her affidavit, that is the case with Smith and Part B.
In the affidavit, Smith swore that she had “no authority to recognize or record a marriage license issued by another state in any setting, regardless of whether the license was issued to an opposite-sex or a same-sex couple.” The plaintiffs have offered nothing of substance to contradict that statement.10 With the new affidavit, *1089the uncontroverted summary-judgment record shows that Smith had no power to recognize the Barton couple’s out-of-state marriage, and therefore no power to redress their injury.11 Since Smith was the only state defendant named in the operative complaint, the Barton couple had no standing to sue on their non-recognition claim. See Cressman, 719 F.3d at 1147 (finding that a plaintiff had no standing to sue a defendant because the plaintiff “provided no basis to conclude that the district court could order [the defendant] to do anything in her official capacity to redress [the plaintiffs] alleged injuries”); Nova Health Sys. v. Gandy, 416 F.3d 1149, 1159 (10th Cir.2005) (dismissing a claim in part for lack of redressability where a favorable “judgment would likely do nothing to prevent [the harm], and thus would not be substantially likely to redress [the plaintiffs] injury in fact”).
There are various potential counterarguments that resist this conclusion, but they all fail.
First, an argument could be made that the Barton couple was entitled to sue Smith as the face of the judiciary despite the undisputed fact that she has no personal involvement in recognizing foreign marriages. Granted, there are scenarios in which a plaintiff is permitted to seek relief against a defendant who would only be indirectly implicated in any harm suffered by the plaintiff. Notably, however, these scenarios frequently arise when a plaintiff fearing prosecution sues a state attorney general and other law enforcement officials to challenge a criminal statute. See, e.g., Doe v. Bolton, 410 U.S. 179, 188-89, 93 S.Ct. 739, 35 L.Ed.2d 201 (1973); Wilson v. Stocker, 819 F.2d 943, 946-47 (10th Cir.1987). An attorney general is the chief law enforcement officer of his or her jurisdiction. See Mitchell v. Forsyth, 472 U.S. 511, 520, 105 S.Ct. 2806, 86 L.Ed.2d 411 (1985). As such, he or she is charged with enforcing all of the criminal statutes on the books. See, e.g., Gandy, 416 F.3d at 1158. It is therefore logical to name that person in his or her representative capacity when one is concerned about a potential criminal prosecution. See id. (“[A]n official who is charged with enforcing a state statute on behalf of the entire state is a proper defendant, so long as the plaintiff shows an appreciable threat of injury flowing directly from the statute.”).
It is less logical to sue a court clerk as the face of a non-recognition regime. Far from being delegated the responsibility to enforce that regime, the court clerk has a very tenuous relationship to the non-recognition provision. To be sure, Oklahoma courts apply the State’s laws regarding the validity of marriages. See Copeland v. Stone, 842 P.2d 754, 755 (Okla.1992) (deciding a case involving a prohibition on remarriage within six months of divorce); Mueggenborg v. Walling, 836 P.2d 112, 112 (Okla.1992) (deciding a case involving the existence vel non of a common-law marriage); Allen v. Allen (In re Estate of Allen), 738 P.2d 142, 143 (Okla.1987) (deciding a case posing the question of whether a marriage had been properly dissolved for estate-distribution purposes); see also Oral Arg. at 15:08-29 (pointing out that Oklahoma’s judicial branch makes the “ultimate determination” of marriage validity *1090with respect to matters like divorce, child custody, inheritance, and bigamy). But all laws are applied by the courts, and all laws are ultimately given their binding meaning by the judiciary. See Clajon Prod. Corp. v. Petera, 70 F.3d 1566, 1571 n. 9 (10th Cir.1995) (“ ‘[I]t is, emphatically, the province and duty of the judicial department to say what the law is.’ ” (quoting Marbury v. Madison, 5 U.S. (1 Cranch) 137, 176, 2 L.Ed. 60 (1803))). If the judiciary’s responsibility to interpret Part B when disputes over its meaning arose were enough to confer standing, one could always sue the court clerk in any challenge to any state law. Standing, “perhaps the most important of the Article III justiciability doctrines,” id. at 1572, would then become little more than an empty formality, easily satisfied in every case.
The plaintiffs seek standing, moreover, on the basis of their bald assertion that Smith is statutorily responsible for deciding whether to recognize out-of-state marriages in the sense that if a couple with an out-of-state marriage attempts to obtain an Oklahoma marriage license, Smith’s office ascertains whether the out-of-state marriage is valid for purposes of determining whether the couple is qualified to receive an Oklahoma license. At oral argument, counsel for the plaintiffs elaborated on the point, explaining that if the ban is nullified in this litigation, same-sex couples in Oklahoma who were validly married in other states, like the Barton couple, would seek Oklahoma marriage licenses, and the eourt clerks would then determine the validity of those foreign marriages. This, however, is a strained argument. And, in light of the burden that the plaintiffs were obliged to carry at the summary-judgment stage, it is patently unavailing.
The Smith affidavit was presented to the district court as an attachment to her motion for summary judgment. To show standing on non-recognition in the face of Smith’s unequivocal disavowal of any involvement in marriage recognition, the plaintiffs were not entitled in responding to the affidavit to depend on “ ‘mere allegations’ ” regarding standing; rather, they were required to “ ‘set forth’ by affidavit or other evidence ‘specific facts,’ which for purposes of the summary judgment motion will be taken to be true.” Lujan v. Defenders of Wildlife, 504 U.S. 555, 561, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992) (citation omitted) (quoting Fed.R.Civ.P. 56(e)); accord Bronson, 500 F.3d at 1111 10.12 Despite Smith’s affidavit, the plaintiffs produced no such evidence indicating that Smith would in fact inquire into the validity of their California marriage in the event they sought an Oklahoma license, and no evidence that they ever even intended to seek an Oklahoma marriage license. In short, they produced no evidence generating even a 'possibility that Smith would ever be called upon to evaluate the validity of their California marriage.
Even assuming that the Barton couple had sought a marriage license from Smith, or intended to do so, it is implausible to *1091imagine that Smith would have inquired into the validity of their California marriage. Looking at the state of the world at the time the suit was filed, as the law instructs, see Jordan v. Sosa, 654 F.3d 1012, 1019 (10th Cir.2011); Utah Ass’n of Cntys. v. Bush, 455 F.3d 1094, 1099 (10th Cir.2006), the standing inquiry must be predicated on the existence of a valid ban on same-sex marriage in Oklahoma. If the Barton couple had sought an Oklahoma marriage license in the face of the ban, it would have been odd, to say the least, for Smith to investigate the validity of their California marriage rather than denying them a license outright pursuant to the unambiguous mandate of a law that she was duty-bound to follow. That being the case, the plaintiffs have no believable hypothetical under which Smith would even be considering the validity of the Barton couple’s marriage, and hence no believable hypothetical rendering her a source of relief for their non-recognition injury. This theory is simply too conjectural to warrant a finding of redressability. See Kerr, 744 F.3d at 1171 (reiterating that “an injury is redressable if a court concludes it is ‘likely, as opposed to merely speculative, that the injury will be redressed by a favorable decision.’ ” (quoting Lujan, 504 U.S. at 561, 112 S.Ct. 2130)); accord Petrella, 697 F.3d at 1294.
There are other state officials with a much closer and more concrete relationship to the withholding of recognition than any courthouse staff, including Smith. The most salient example lies in the area of taxation. In Oklahoma, the Tax Commission presides over the State’s tax system. See Okla. Stat. tit. 68, § 203. One of the Commission’s responsibilities is to accept or deny joint tax returns mailed in by couples. See Grasso v. Okla. Tax Comm’n, 249 P.3d 1258, 1261 (Okla.Civ. App.2011). With that scheme in place, a non-recognition plaintiff could file a joint tax return, have that status denied, and then sue the members of the Tax Commission. See, e.g., Baskin v. Bogan, — F.Supp.3d -, -, -, 2014 WL 2884868, at *4, *16, 2014 U.S. Dist. LEXIS 86114, at *15, *50 (S.D.Ind.2014) (finding the commissioner of the state department of revenue a proper party and ordering him to permit same-sex couples to file joint tax returns); cf. Rott v. Okla. Tax Comm’n, No. CIV-13-1041-M, 2014 WL 2560710 at *1, 2014 U.S. Dist. LEXIS 77173, at *2-4 (W.D.Okla. June 6, 2014) (describing an action brought against, inter alia, members of the Oklahoma Tax Commission for wrongfully assessing and attempting to collect income taxes from the plaintiff in violation of his federal constitutional rights).
Other equally straightforward paths to redressability are easy enough to imagine, and several have in fact been taken in similar challenges being litigated elsewhere. See, e.g., Tanco v. Haslam, 7 F.Supp.3d 759, 763-64, 770-72, 2014 WL 997525, at *2, *8-9, 2014 U.S. Dist. LEXIS 33463, at *9, *33-34 (M.D.Tenn.2014) (sustaining a non-recognition challenge where the plaintiffs sued the commissioner of the department of finance and administration after they were prevented from using a family health insurance plan provided by a public university); Bostic v. Rainey, 970 F.Supp.2d 456, 461-63, 484 (E.D.Va.2014) (sustaining a non-recognition challenge where the plaintiffs sued the state registrar of vital records to obtain a birth certificate so that they could legally adopt the daughter they raise together); Obergefell v. Wymyslo, 962 F.Supp.2d 968, 972-73, 1000 (S.D.Ohio 2013) (sustaining a nonrecognition challenge where the plaintiffs sued the director of the state department of health to obtain a death certificate listing the couple as married).13
*1092The distinction between Smith and a proper defendant, moreover, is not a distinction between discretionary decisions enforcing the non-recognition provision and ministerial decisions doing so. In all relevant respects, a tax commissioner’s decision to withhold joint-filing status is, as a practical matter, just as ministerial as Smith’s decision to withhold recognition. Both officials are responsible for faithfully applying Oklahoma law, and Oklahoma law clearly instructs both of them to withhold marital status from same-sex couples. If the Barton couple had expressed a wish to file joint taxes and named a tax official responsible for authorizing that filing, there would be no doubt that a court order to the official would remedy the couple’s non-recognition injury: the official would then accept the joint return. See Baskin, 12 F.Supp.3d at 1152-53, 1165-66, 2014 WL 2884868, at *4, *16, 2014 U.S. Dist. LEXIS 86114, at *15, *50 (finding the commissioner of the state department of revenue a proper party and ordering him to permit same-sex couples to file joint tax returns). There is no analogue with respect to Smith. The supposition that Smith will have any specific involvement in recognizing or declining to recognize the Barton couple’s marriage lacks any demonstrated foundation in the record or in Oklahoma law.14
Unable to demonstrate standing on their principal non-recognition injury — the refusal of the State to recognize their marriage — the plaintiffs seek to rely upon a different injury. Specifically, the plaintiffs insist they have standing because “the injury of shutting the state courthouse doors *1093on Plaintiffs — on top of the injuries of ... non-recognition — would be redressed by an injunction against [Part B].” As Smith correctly points out, though, the Barton couple did not challenge Part B on the grounds that it foreclosed their right to access the state court system. Rather, they challenged it on the grounds that it violated their equal-protection and due-process rights to have their marriage recognized. Crucially, the district court never heard a contention from the Barton couple that Part B visited upon them an access-to-the-courts injury,15 and it was their obligation to show standing. See Kerr, 744 F.3d at 1163; Petrella, 697 F.3d at 1293. The district court could not have entertained jurisdiction over a claim on the basis of redressability for an injury that the Barton couple never alleged.
In sum, the Barton couple had no standing to sue, and the district court properly dismissed their non-recognition challenge as a result.
C
In a final attempt to nullify Part B along with Part A, the plaintiffs submit — for the first time on appeal — that the non-recognition provision must be struck down under severability law as soon as the ban is struck down, no matter whether there was standing to challenge the non-recognition provision or not. For her part, Smith asks for a finding that the plaintiffs forfeited their severability theory by failing to raise it in the district court. The plaintiffs do not deny that they omitted the argument from their summary-judgment filings, and a review of those filings finds no trace of severability doctrine. Nevertheless, the plaintiffs request that we take account of severability if the ban falls, regardless of the issue’s preservation, because — in their view — a severability analysis is required whenever a court declares invalid part of an enactment.
At the outset, it is necessary to determine the controlling source of law. The question of whether an unconstitutional provision of state law is severable from the remainder of the enactment is a matter of state law. See Leavitt v. Jane L., 518 U.S. 137, 139, 116 S.Ct. 2068, 135 L.Ed.2d 443 (1996) (per curiam); accord Am. Target Adver., Inc. v. Giani, 199 F.3d 1241, 1250 (10th Cir.2000). So too is the question of whether a severability analysis is triggered in the first place by the facts of the case, i.e., whether the type of judicial ruling at issue calls for a severability inquiry. See Local 514 Transp. Workers Union of Am. v. Keating, 66 Fed.Appx. 768, 779 (10th Cir.2003) (certifying to the Oklahoma Supreme Court the question of whether severability analysis applied to certain state constitutional provisions if they were declared preempted by federal law); Local 514 Transp. Workers Union of Am. v. Keating, 83 P.3d 835, 839 (Okla.2003) (answering that severability analysis would not apply and holding that “whether to apply severability analysis ... [was] a matter of state law”); see also Local 514 Transp. Workers Union of Am. v. Keating, 358 F.3d 743, 744 n. 1 (10th Cir.2004) (subsequently deciding the appeal on the basis of the Oklahoma Supreme Court’s answer and incorporating the certification into the published opinion).
*1094Unlike substantive severability law, though, the matter of whether an argument has been forfeited by a party’s failure to raise it in the district court is decided by federal procedural law. That proposition is underscored by the fact that when we have found an argument forfeited by its omission in district court proceedings in a diversity case — where we are applying substantive state law — we have supported our forfeiture ruling with citations to Tenth Circuit decisions that are either applying substantive federal law or the substantive law of a different state. See, e.g., Elm Ridge Exploration Co. v. Engle, 721 F.3d 1199, 1213 (10th Cir.2013); Brecek & Young Advisors, Inc. v. Lloyds of London Syndicate 2003, 715 F.3d 1231, 1234 n. 1 (10th Cir.2013); Emp’rs Mut. Cas. Co. v. Bartile Roofs, Inc., 618 F.3d 1153, 1176 n. 20 (10th Cir.2010).
More relevant to the case at bar, in Awad v. Ziriax, 670 F.3d 1111, 1132 n. 16 (10th Cir.2012), we applied a federal approach to a highly analogous situation. In Awad, a popular vote approved a proposal to add to the state constitution a provision that included, inter alia, language forbidding Oklahoma courts from considering Sharia law in rendering their decisions. Id. at 1117-18. The district court issued a preliminary injunction, ordering state officials not to certify the election result until the court had ruled on the merits of a federal constitutional challenge to the proposed amendment. Awad v. Ziriax, 754 F.Supp.2d 1298, 1308 (W.D.Okla.2010). On appeal, we affirmed the preliminary injunction. Awad, 670 F.3d at 1133. We attached the following footnote to the end of our substantive analysis:
Appellants raised the issue of severability of the Sharia law portions of the amendment for the first time to this court in post-oral argument supplemental briefing. Their argument consisted of one sentence and cited no authority, stating that if this court decides the Sharia law provisions in the amendment render the amendment invalid, “the court should simply treat the explicatory example as surplusage, and strike it.” Because this issue has not been adequately briefed, we do not address it. See United States v. Cooper, 654 F.3d 1104, 1128 (10th Cir.2011).
Id. at 1132 n. 16. In other words, in a federal constitutional challenge to an Oklahoma constitutional provision, we upheld, at least preliminarily, a decision striking down the provision and declined to consider severability because of a failure to adequately preserve the issue for review— specifically, a waiver of the issue through deficient briefing. The Awad footnote is only explicable if an appellate court has no inherent obligation to consider severability sua sponte, as it would with, say, a jurisdictional issue. See, e.g., United States v. Ramos, 695 F.3d 1035, 1046 (10th Cir.2012), cert. denied, — U.S. -, 133 S.Ct. 912, 184 L.Ed.2d 701 (2013); Columbian Fin. Corp. v. BancInsure, Inc., 650 F.3d 1372, 1375-76 (10th Cir.2011).
As in Awad, this court is upholding here a decision striking down a provision of the Oklahoma Constitution on federal constitutional grounds, and, as in Awad, the litigant failed to adequately preserve the issue for review — this time, by effecting a forfeiture through failure to present the issue to the district court. There is no apparent reason why the result the court reached in Awad should not be the same here. In other words, the same principle should have equal purchase in the forfeiture context: if there is no obligation to consider severability sua sponte where it has been waived,16 there is *1095no obligation to consider it where it has been forfeited.
Having thus resolved the issue of whether in a forfeiture context the court is obligated to consider severability, “the decision regarding what issues are appropriate to entertain on appeal in instances of lack of preservation is discretionary.” Abernathy v. Wandes, 713 F.3d 538, 552 (10th Cir.2013), cert. denied, — U.S. -, 134 S.Ct. 1874, 188 L.Ed.2d 916 (2014). Waiver through appellate-briefing omission and forfeiture through silence before the district court are admittedly distinct failures of preservation, and arguably there is more discretionary leeway to consider issues not preserved under the latter (forfeiture) than the former (appellate-briefing waiver). See Richison v. Ernest Grp., Inc., 634 F.3d 1123, 1128-30 (10th Cir.2011) (exploring the distinction between forfeiture and waiver, including waiver through omissions in appellate briefs); see also United States v. McGehee, 672 F.3d 860, 873 (10th Cir.2012) (“ ‘Unlike waived theories, we will entertain forfeited theories on appeal....’” (quoting Richison, 634 F.3d at 1128)). However, where a litigant attempts to rely upon a forfeited theory, ‘“the failure to argue for plain error and its application on appeal ... surely marks the end of the road for an argument for reversal not first presented to the district court.’ ” United States v. Lamirand, 669 F.3d 1091, 1100 n. 7 (10th Cir.2012) (omission in original) (quoting Richison, 634 F.3d at 1131). The plaintiffs are at the end of the road here.
In essence, in arguing for reversal, the plaintiffs are asserting that the district court erred in refusing to enjoin Part B in addition to Part A under sever-ability law, despite their alleged lack of standing to challenge the former. They offer no explanation as to how the district court plainly erred in this regard.17 In fact, the plaintiffs’ only response to Smith’s forfeiture argument is that a sev-erability theory is not susceptible to forfeiture. As noted above, that is incorrect— pursuant to Awad, the plaintiffs could in fact forfeit their severability argument, and they did.18 Therefore, absent any argument by the plaintiffs for plain error, much less a cogent one, it is appropriate to decline to exercise the court’s discretion to hear this forfeited severability issue.
*1096To recapitulate, a severability theory can be forfeited, the plaintiffs’ severability theory was forfeited, and the plaintiffs supply no argument for overlooking the forfeiture. As a consequence, they are not entitled to the benefit of any severability analysis, and the district court’s dismissal of the challenge to Part B must be affirmed.19
That the non-recognition claim is doomed to dismissal may seem a harsh result. The Barton couple first challenged Part B almost ten years ago. After the first appeal, the plaintiffs fairly understood Bishop I as a directive instructing them to name Smith as the lone defendant for all of their grievances. It was reasonable of the Barton couple to follow that perceived directive, and it is regrettable that their compliance has resulted in a lack of standing, especially after nearly a decade of complex, time-consuming, and no doubt emotional litigation.
No matter how compelling the equitable arguments for reaching the merits of the non-recognition claim, however, its fate must be determined by the law, and the law demands dismissal. The frustration that may be engendered by the court’s disposition today should be tempered, however. Although it would not be appropriate to definitively opine on the matter, it is fair to surmise that the court’s decision in Kitchen casts serious doubt on the continuing vitality of Part B. See 755 F.3d at 1199, 2014 WL 2868044, at *1, 2014 U.S.App. LEXIS 11935, at *4 (“A state may not ... refuse to recognize [a] marriage ... based solely upon the sex of the persons in the marriage union.”).
IV
For the foregoing reasons, we AFFIRM. We STAY our mandate pending the disposition of any subsequently-filed petition for writ of certiorari. See Fed. R.App. P. 41(d)(2); see also Kitchen, 755 F.3d at 1229-30, 2014 WL 2868044, at *32, 2014 U.S.App. LEXIS 11935, at *97-98.
. Smith also argues that the Barton couple does not have standing to contest Part B of SQ 711 because they did not challengé Okla. Stat. tit. 43, § 3.1, which provides that “[a] marriage between persons of the same gender performed in another state shall not be recognized as valid and binding in this state as of the date of the marriage.” We will refer above only to Part A in discussing plaintiffs' failure to challenge the statutory codifications of Oklahoma’s same-sex marriage policy as it relates to standing. As explained infra, the Barton couple lacked standing to sue because they named a defendant who could not redress their injury. Therefore, there is no need to consider whether they lacked standing for the alternative reason that they failed to challenge the statutory nonrecognition provision. See Niemi v. Lasshofer, 728 F.3d 1252, 1260 (10th Cir.2013) (noting that where there are multiple threshold issues that can be resolved without engaging in the merits a court has " 'leeway to choose among’ them and to ‘take[ ] the less burdensome course’ ” (alteration in original)) (quoting Sinochem Int’l Co. v. Malaysia Int’l Shipping Corp., 549 U.S. 422, 431, 436, 127 S.Ct. 1184, 167 L.Ed.2d 15 (2007)).
. If the court relies on the subjective motivations of lawmakers to determine the constitutionality of Oklahoma’s two provisions, Smith suggests that one might survive even if the other falls. However, as explained in Kitchen, 755 F.3d at 1229-30, 2014 WL 2868044, at *32, 2014 U.S.App. LEXIS 11935, at *97, we conclude that because state laws prohibiting same-sex marriage impinge upon a fundamental right without satisfying the strict scrutiny test, such provisions fail regardless of subjective intent.
. The remaining prongs of standing as to the Bishop couple’s ability to challenge Part A are not contested. We conclude nonetheless the couple has satisfied those prongs. See Alvarado v. KOB-TV, L.L.C. (Channel 4 News), 493 F.3d 1210, 1214 n. 1 (10th Cir.2007) (this court has authority to consider standing issues sua sponte). Having ruled that an injunction barring enforcement of Part A of SQ 711 redresses the Bishop couple’s injury— inability to marry — we have no trouble concluding that they satisfy the traceability requirement. See Cache Valley Elec. Co. v. Utah Dep’t of Transp., 149 F.3d 1119, 1123 (10th Cir.1998) (noting that in many cases, "re-dressability and traceability overlap as two sides of a causation coin” (quotation omitted)). The Bishop couple sought a marriage license from Smith’s office, but were denied because they are both women. See Papasan v. Allain, 478 U.S. 265, 282 n. 14, 106 S.Ct. 2932, 92 L.Ed.2d 209 (1986) (a defendant "responsible for general supervision of the administration by local ... officials” of a challenged provision is a proper defendant). And the Bishop couple has identified several negative financial consequences of that denial. See Singleton v. Wulff, 428 U.S. 106, 113, 96 S.Ct. 2868, 49 L.Ed.2d 826 (1976) (financial harm caused by challenged provision constitutes injury in fact).
. Although the district court declined to rule on whether the plaintiffs asserted a fundamental right, Bishop II, 962 F.Supp.2d at 1285 n. 33, and instead applied rational basis review, id. at 1295, we may "affirm on any ground supported by the record, so long as the appellant has had a fair opportunity to address that ground," Schanzenbach v. Town of Opal, 706 F.3d 1269, 1272 (10th Cir.2013) (quotation omitted). As in Kitchen, we do not address the question of whether a ban on same-sex marriage might survive lesser forms of scrutiny given our holding that such bans burden fundamental rights.
. The law of the case doctrine is inapplicable when a merits panel considers a jurisdictional issue that was addressed by a motions or mandamus panel. See Kennedy, 273 F.3d at 1299-1300 (mandamus panel); Stifel, Nicolaus & Co., 81 F.3d at 1544 (motions panel). Bishop I, however, was a fully-reasoned decision by a merits panel. The motions-panel *1085and mandamus-panel exceptions are therefore not germane here.
. Though worded somewhat more confusingly than Baca, Smith's other central authority for this jurisdictional argument — Public Interest Research Group of New Jersey v. Magnesium Elektron, Inc., 123 F.3d 111 (3d Cir.1997) — is to the same effect. There, the Third Circuit cabined the pivotal footnote from Christian-son to the transfer context, reasoning that the Supreme Court could not have “intended in one footnote to eviscerate, in all instances, the federal courts' prerogative to revisit important jurisdictional questions.” Id. at 118. But the very reason the Magnesium Elektron court reevaluated the jurisdictional issue there was that new evidence “was presented to the district court which had a direct bearing on the issue of standing.” Id. As explained at length below, new evidence of this sort is one of the established exceptions to the law of the case, United States v. Irving, 665 F.3d 1184, 1192 n. 12 (10th Cir.2011), and the new evidence in Magnesium Elektron was in fact the exact type of new evidence at issue in the present appeal. Magnesium Elektron is therefore consistent with the approach taken herein.
. Insofar as Smith is arguing, implicitly, that application of law of the case works a manifest injustice, that argument is unconvincing. If any party here can make a colorable claim of injustice, it is the Barton couple, who named as a defendant the official that the Bishop I panel told them to name and who find out today that they should have named someone else and, as a result, are denied the satisfaction of an explicit invalidation of Part B.
. The new-evidence exception is often set forth with reference to new evidence at a new trial. See, e.g., Irving, 665 F.3d at 1192 n. 12; Clark, 590 F.3d at 1140. As the authorities assembled in this section show, a new trial is not necessary for the production of new evidence — a summary-judgment affidavit can suffice.
. Had Bishop I been published, its force as law of the case would have been significantly strengthened by its status as law of the circuit as well. See LaShawn A., 87 F.3d at 1395 ("[W]hen both [the law of the case and the law of the circuit] are at work, the law-of-the-circuit doctrine should increase a panel's reluctance to reconsider a decision made in an earlier appeal in the same case.”). Because the order was unpublished, law-of-the-case principles are the only constraint here. See 10th Cir. R. 32.1(A) ("Unpublished decisions are not precedential, but may be cited for their persuasive value. They may also be cited under the doctrines of law of the case, claim preclusion, and issue preclusion.”); Federal Practice § 4478.2, at 731 ("If an unpublished opinion does not command prece-dential force under circuit rules, law-of-the-case rules hold full sway.”).
. The plaintiffs assert that Smith's affidavit is contradicted by her answer to the complaint, wherein she "admitted] that Defendants, and those subject to their supervision, direction and control, are responsible for the enforcement of the laws challenged by Plaintiffs’ First Amended Complaint.” In rebuttal, Smith notes that the challenged laws referenced in the answer did not include the nonrecognition provision, since the first amended complaint did not address that provision. Smith has the better argument. The parties apparently came to terms on this point in the district court, where a minute sheet reflected their consensus "that plaintiffs' motion for summary judgment [would] address [the nonrecognition provision], notwithstanding the absence of such language in the Amended Complaint." (Emphasis added). Although the complaint included some stray passages that appeared to attribute all of the plaintiffs' injuries to SQ 711 as a whole, it never explicitly mentioned the non-recognition provision and repeatedly suggested that it was the ban, in conjunction with DOMA, that caused the *1089nonrecognition injury. Smith’s "admission” in her answer is therefore irrelevant to this issue.
. The authorities cited by Bishop I for its standing determination either impose responsibilities on court clerks with respect to issuing marriage licenses, see Okla. Stat. tit. 28, § 31; id. tit. 43, § 5, or examine the general relationship between court clerks and the judicial branch, see Speight v. Presley, 203 P.3d 173 (Okla.2008). None of the authorities address the role court clerks play in regards to marriage recognition.
. Of course, if the Barton couple had been entitled to a finding of standing on the basis of law of the case, they would not have been required to demonstrate their standing before the district court, or here. That is to say, had there been no new evidence to sufficiently undermine the effect of the law of the case of Bishop I, then Bishop I would have been enough, without more, to establish standing. See Christianson, 486 U.S. at 816 n. 5, 108 S.Ct. 2166 (“There is no reason to apply law-of-the-case principles less rigorously to [a jurisdictional issue].”). But since there was new evidence that did effectively undermine Bishop I’s non-recognition standing holding, the Barton couple had to meet their summary-judgment burden in rebutting that evidence. See, e.g., Clark, 590 F.3d at 1140 (describing new evidence as a reason to “depart from the- [law of the case] doctrine” (emphasis added)); United States v. Parada, 577 F.3d 1275, 1280 (10th Cir.2009) (same).
. That the plaintiffs' action was in part for a declaratory judgment does not affect the *1092standing analysis. Like any lawsuit, a declaratory-judgment action must meet Article Ill’s standing criteria, including redressability. See Consumer Data Indus. Ass’n, 678 F.3d at 906; City of Hugo v. Nichols (Two Cases), 656 F.3d 1251, 1263-64 (10th Cir.2011). As part of the redressability requirement, a declaratory-judgment action must be brought against a defendant who can, if ordered to do so, remedy the alleged injury. See Coll v. First Am. Title Ins. Co., 642 F.3d 876, 892 (10th Cir.2011); Bronson, 500 F.3d at 1111. Since Smith cannot provide relief to the Barton couple on their non-recognition claim, they had no standing to sue her, regardless of whether the claim was brought in a declaratory-judgment form or not.
Similarly, the doctrine of actionable conduct capable of repetition yet evading review is not applicable here. As an initial matter, the doctrine creates an exception to mootness, not to lack of standing. See United States v. Juvenile Male, - U.S. -, 131 S.Ct. 2860, 2865, 180 L.Ed.2d 811 (2011) (per curiam); Buchheit v. Green, 705 F.3d 1157, 1160 (10th Cir.2012); see also Lucero v. Bureau of Collection Recovery, Inc., 639 F.3d 1239, 1242-43 (10th Cir.2011) (acknowledging that the capable-of-repetition-yet-evading-review class of cases constitutes an exception to mootness and noting that such "exceptions do not extend to the standing inquiry”). The Barton couple’s claim is plainly not moot, as they continue to desire recognition for their marriage and continue to be denied such recognition. See United States v. Alaska, 503 U.S. 569, 575 n. 4, 112 S.Ct. 1606, 118 L.Ed.2d 222 (1992) ("We agree that the controversy is not moot, since it involves a continuing controversy...."). At any rate, to the extent the capable-of-repetition-yet-evading-review test does go to re-dressability, the complained-of conduct, i.e., the denial of marriage recognition, does not evade review. Rather, as discussed above, a non-recognition couple could easily seek recognition from the State in some fashion, such as by filing a joint tax return, and when recognition was denied, the couple could then sue the official responsible for that nonrecognition decision.
. In the plaintiffs’ eyes, standing on nonrecognition can be found by virtue of the fact that Smith, and the court system that employs her, would not refuse to honor a court order enjoining enforcement of Part B. It is of no moment that Smith would presumably obey a judicial invalidation of Part B if she were directed to enforce the provision. The problem is there is no reason to believe that she enforces the provision at all, and thus no conceivable injunction for her to obey.
. In their response to Smith’s motion for summary judgment, the plaintiffs did submit in passing that Smith’s affidavit might create an injury in its own right, namely, the erection of “a barrier making it more difficult for members of a group to obtain a benefit.” However, the plaintiffs did not frame this argument in terms of access to the state court system, and it is more naturally read as a point about access to the federal court system. After all, a finding of no standing on the basis of Smith’s affidavit removes the Barton couple from federal court, not from state court.
. The parties in Kitchen did not address sev-erability in their appellate briefing, thereby rendering the issue waived in that case through briefing omission and relieving this court of any responsibility to discuss the matter in its opinion. See United States v. Bader, *1095678 F.3d 858, 894 (10th Cir.2012) (observing that a litigant's briefing omissions prompt the conclusion that he or she “has waived [the] argument”).
. A litigant may obtain relief under the plain-error doctrine upon a showing of "(1) an error, (2) that is plain, which means clear or obvious under current law, and (3) that affects substantial rights. If [she] satisfies these criteria, this Court may exercise discretion to correct the error if it seriously affects the fairness, integrity, or public reputation of judicial proceedings.” United States v. Goode, 483 F.3d 676, 681 (10th Cir.2007) (quotation omitted).
. The plaintiffs use Brockett v. Spokane Arcades, Inc., 472 U.S. 491, 105 S.Ct. 2794, 86 L.Ed.2d 394 (1985), and Panhandle Eastern Pipeline Co. v. State of Oklahoma ex rel. Commissioners of Land Office, 83 F.3d 1219 (10th Cir.1996), to bolster their view that a court has an obligation to consider severability even in the face of forfeiture. Cf. Acosta v. City of Costa Mesa, 694 F.3d 960, 974 n. 7 (9th Cir.2012) (relying upon Brockett, inter alia, to support the proposition that "severability is an inherent part of the process of constitutional adjudication” that is not subject to waiver by omission from appellate briefs), withdrawn, 708 F.3d 1122 (9th Cir.2013). Neither Brockett nor Panhandle nor any of the other Supreme Court cases cited by Acosta say anything about forfeiture or waiver, or anything about whether severability had been raised or argued to the trial or appellate courts. Given this silence, the explicit invocation of waiver by Awad in a comparable case is controlling here on the question of whether severability must be considered sua sponte.
. Because the plaintiffs’ severability theory is forfeited, there is no need to consider Smith’s argument that a severability analysis regarding Part B is foreclosed by the plaintiffs' lack of standing to challenge that provision. See Sinochem Int’l Co., 549 U.S. at 431, 127 S.Ct. 1184 (authorizing federal courts to choose at their discretion among alternative threshold grounds for disposing of a claim without reaching its merits); accord Niemi, 728 F.3d at 1260.