TROXEL V. GRANVILLE (99-138) 530 U.S. 57 (2000)
137 Wash. 2d 1, 969 P.2d 21, affirmed.
[ O’Connor ]
[ Souter ]
[ Thomas ]
[ Stevens ]
[ Scalia ]
[ Kennedy ]
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version

Stevens, J., dissenting


No. 99—138



[June 5, 2000]

Justice Stevens, dissenting.

The Court today wisely declines to endorse either the holding or the reasoning of the Supreme Court of Washington. In my opinion, the Court would have been even wiser to deny certiorari. Given the problematic character of the trial court’s decision and the uniqueness of the Washington statute, there was no pressing need to review a State Supreme Court decision that merely requires the state legislature to draft a better statute.

Having decided to address the merits, however, the Court should begin by recognizing that the State Supreme Court rendered a federal constitutional judgment holding a state law invalid on its face. In light of that judgment, I believe that we should confront the federal questions presented directly. For the Washington statute is not made facially invalid either because it may be invoked by too many hypothetical plaintiffs, or because it leaves open the possibility that someone may be permitted to sustain a relationship with a child without having to prove that serious harm to the child would otherwise result.


In response to Tommie Granville’s federal constitutional challenge, the State Supreme Court broadly held that Wash. Rev. Code §26.10.160(3) (Supp. 1996) was invalid on its face under the Federal Constitution.1 Despite the nature of this judgment, Justice O’Connor would hold that the Washington visitation statute violated the Due Process Clause of the Fourteenth Amendment only as applied. Ante, at 6, 8, 14—15. I agree with Justice Souter, ante, at 1, and n. 1 (opinion concurring in judgment), that this approach is untenable.

The task of reviewing a trial court’s application of a state statute to the particular facts of a case is one that should be performed in the first instance by the state appellate courts. In this case, because of their views of the Federal Constitution, the Washington state appeals courts have yet to decide whether the trial court’s findings were adequate under the statute.2 Any as-applied critique of the trial court’s judgment that this Court might offer could only be based upon a guess about the state courts’ application of that State’s statute, and an independent assessment of the facts in this case–both judgments that we are ill-suited and ill-advised to make.3

While I thus agree with Justice Souter in this respect, I do not agree with his conclusion that the State Supreme Court made a definitive construction of the visitation statute that necessitates the constitutional conclusion he would draw.4 As I read the State Supreme Court’s opinion, In re Smith, 137 Wash. 2d 1, 19—20, 969 P.2d 21, 30—31 (1998), its interpretation of the Federal Constitution made it unnecessary to adopt a definitive construction of the statutory text, or, critically, to decide whether the statute had been correctly applied in this case. In particular, the state court gave no content to the phrase, “best interest of the child,” Wash. Rev. Code §26.10.160(3) (Supp. 1996)–content that might well be gleaned from that State’s own statutes or decisional law employing the same phrase in different contexts, and from the myriad other state statutes and court decisions at least nominally applying the same standard.5 Thus, I believe that Justice Souter’s conclusion that the statute unconstitutionally imbues state trial court judges with “ ‘too much discretion in every case,’ ante, at 4, n. 3 (opinion concurring in judgment) (quoting Chicago v. Morales, 527 U.S. 41, 71 (1999) (Breyer, J., concurring)), is premature.

We are thus presented with the unconstrued terms of a state statute and a State Supreme Court opinion that, in my view, significantly misstates the effect of the Federal Constitution upon any construction of that statute. Given that posture, I believe the Court should identify and correct the two flaws in the reasoning of the state court’s majority opinion, and remand for further review of the trial court’s disposition of this specific case.


In my view, the State Supreme Court erred in its federal constitutional analysis because neither the provision granting “any person” the right to petition the court for visitation, 137 Wash. 2d, at 20, 969 P.2d, at 30, nor the absence of a provision requiring a “threshold … finding of harm to the child,” ibid., provides a sufficient basis for holding that the statute is invalid in all its applications. I believe that a facial challenge should fail whenever a statute has “a ‘plainly legitimate sweep,’ Washington v. Glucksberg, 521 U.S. 702, 739—740 and n. 7 (1997) (Stevens, J., concurring in judgment).6 Under the Washington statute, there are plainly any number of cases–indeed, one suspects, the most common to arise–in which the “person” among “any” seeking visitation is a once-custodial caregiver, an intimate relation, or even a genetic parent. Even the Court would seem to agree that in many circumstances, it would be constitutionally permissible for a court to award some visitation of a child to a parent or previous caregiver in cases of parental separation or divorce, cases of disputed custody, cases involving temporary foster care or guardianship, and so forth. As the statute plainly sweeps in a great deal of the permissible, the State Supreme Court majority incorrectly concluded that a statute authorizing “any person” to file a petition seeking visitation privileges would invariably run afoul of the Fourteenth Amendment.

The second key aspect of the Washington Supreme Court’s holding–that the Federal Constitution requires a showing of actual or potential “harm” to the child before a court may order visitation continued over a parent’s objections–finds no support in this Court’s case law. While, as the Court recognizes, the Federal Constitution certainly protects the parent-child relationship from arbitrary impairment by the State, see infra, at 7—8 we have never held that the parent’s liberty interest in this relationship is so inflexible as to establish a rigid constitutional shield, protecting every arbitrary parental decision from any challenge absent a threshold finding of harm.7 The presumption that parental decisions generally serve the best interests of their children is sound, and clearly in the normal case the parent’s interest is paramount. But even a fit parent is capable of treating a child like a mere possession.

Cases like this do not present a bipolar struggle between the parents and the State over who has final authority to determine what is in a child’s best interests. There is at a minimum a third individual, whose interests are implicated in every case to which the statute applies–the child.

It has become standard practice in our substantive due process jurisprudence to begin our analysis with an identification of the “fundamental” liberty interests implicated by the challenged state action. See, e.g., ante, at 6—8 (opinion of O’Connor, J.); Washington v. Glucksberg, 521 U.S. 702 (1997); Planned Parenthood of Southeastern Pa. v. Casey, 505 U.S. 833 (1992). My colleagues are of course correct to recognize that the right of a parent to maintain a relationship with his or her child is among the interests included most often in the constellation of liberties protected through the Fourteenth Amendment. Ante, at 6—8 (opinion of O’Connor, J.). Our cases leave no doubt that parents have a fundamental liberty interest in caring for and guiding their children, and a corresponding privacy interest–absent exceptional circumstances–in doing so without the undue interference of strangers to them and to their child. Moreover, and critical in this case, our cases applying this principle have explained that with this constitutional liberty comes a presumption (albeit a rebuttable one) that “natural bonds of affection lead parents to act in the best interests of their children.” Parham v. J. R., 442 U.S. 584, 602 (1979); see also Casey, 505 U.S., at 895; Santosky v. Kramer, 455 U.S. 745, 759 (1982) (State may not presume, at factfinding stage of parental rights termination proceeding, that interests of parent and child diverge); see also ante, at 9—10 (opinion of O’Connor, J.).

Despite this Court’s repeated recognition of these significant parental liberty interests, these interests have never been seen to be without limits. In Lehr v. Robertson, 463 U.S. 248 (1983), for example, this Court held that a putative biological father who had never established an actual relationship with his child did not have a constitutional right to notice of his child’s adoption by the man who had married the child’s mother. As this Court had recognized in an earlier case, a parent’s liberty interests “ ‘do not spring full-blown from the biological connection between parent and child. They require relationships more enduring.’ Id., at 260 (quoting Caban v. Mohammed, 441 U.S. 380, 397 (1979)).

Conversely, in Michael H. v. Gerald D., 491 U.S. 110 (1989), this Court concluded that despite both biological parenthood and an established relationship with a young child, a father’s due process liberty interest in maintaining some connection with that child was not sufficiently powerful to overcome a state statutory presumption that the husband of the child’s mother was the child’s parent. As a result of the presumption, the biological father could be denied even visitation with the child because, as a matter of state law, he was not a “parent.” A plurality of this Court there recognized that the parental liberty interest was a function, not simply of “isolated factors” such as biology and intimate connection, but of the broader and apparently independent interest in family. See, e.g. . id., at 123; see also Lehr, 463 U.S., at 261; Smith v. Organization of Foster Families For Equality & Reform, 431 U.S. 816, 842—847 (1977); Moore v. East Cleveland, 431 U.S. 494, 498—504 (1977).

A parent’s rights with respect to her child have thus never been regarded as absolute, but rather are limited by the existence of an actual, developed relationship with a child, and are tied to the presence or absence of some embodiment of family. These limitations have arisen, not simply out of the definition of parenthood itself, but because of this Court’s assumption that a parent’s interests in a child must be balanced against the State’s long-recognized interests as parens patriae, see, e.g., Reno v. Flores, 507 U.S. 292, 303—304 (1993); Santosky v. Kramer, 455 U.S., at 766; Parham, 442 U.S., at 605; Prince v. Massachusetts, 321 U.S. 158, 166 (1944), and, critically, the child’s own complementary interest in preserving relationships that serve her welfare and protection, Santosky, 455 U.S., at 760.

While this Court has not yet had occasion to elucidate the nature of a child’s liberty interests in preserving established familial or family-like bonds, 491 U.S., at 130 (reserving the question), it seems to me extremely likely that, to the extent parents and families have fundamental liberty interests in preserving such intimate relationships, so, too, do children have these interests, and so, too, must their interests be balanced in the equation.8 At a minimum, our prior cases recognizing that children are, generally speaking, constitutionally protected actors require that this Court reject any suggestion that when it comes to parental rights, children are so much chattel. See ante, at 5—6 (opinion of O’Connor, J.) (describing States’ recognition of “an independent third-party interest in a child”). The constitutional protection against arbitrary state interference with parental rights should not be extended to prevent the States from protecting children against the arbitrary exercise of parental authority that is not in fact motivated by an interest in the welfare of the child.9

This is not, of course, to suggest that a child’s liberty interest in maintaining contact with a particular individual is to be treated invariably as on a par with that child’s parents’ contrary interests. Because our substantive due process case law includes a strong presumption that a parent will act in the best interest of her child, it would be necessary, were the state appellate courts actually to confront a challenge to the statute as applied, to consider whether the trial court’s assessment of the “best interest of the child” incorporated that presumption. Neither would I decide whether the trial court applied Washington’s statute in a constitutional way in this case, although, as I have explained, n. 3, supra, I think the outcome of this determination is far from clear. For the purpose of a facial challenge like this, I think it safe to assume that trial judges usually give great deference to parents’ wishes, and I am not persuaded otherwise here.

But presumptions notwithstanding, we should recognize that there may be circumstances in which a child has a stronger interest at stake than mere protection from serious harm caused by the termination of visitation by a “person” other than a parent. The almost infinite variety of family relationships that pervade our ever-changing society strongly counsel against the creation by this Court of a constitutional rule that treats a biological parent’s liberty interest in the care and supervision of her child as an isolated right that may be exercised arbitrarily. It is indisputably the business of the States, rather than a federal court employing a national standard, to assess in the first instance the relative importance of the conflicting interests that give rise to disputes such as this.10 Far from guaranteeing that parents’ interests will be trammeled in the sweep of cases arising under the statute, the Washington law merely gives an individual–with whom a child may have an established relationship–the procedural right to ask the State to act as arbiter, through the entirely well-known best-interests standard, between the parent’s protected interests and the child’s. It seems clear to me that the Due Process Clause of the Fourteenth Amendment leaves room for States to consider the impact on a child of possibly arbitrary parental decisions that neither serve nor are motivated by the best interests of the child.

Accordingly, I respectfully dissent.


1. The State Supreme Court held that, “as written, the statutes violate the parents’ constitutionally protected interests.” In re Smith, 137 Wash. 2d 1, 5, 969 P.2d 21, 23 (1998).

2. As the dissenting judge on the state appeals court noted, “[t]he trial court here was not presented with any guidance as to the proper test to be applied in a case such as this.” In re Troxel, 87 Wash. App. 131, 143, 940 P.2d 698, 703 (1997) (opinion of Ellington, J.). While disagreeing with the appeals court majority’s conclusion that the state statute was constitutionally infirm, Judge Ellington recognized that despite this disagreement, the appropriate result would not be simply to affirm. Rather, because there had been no definitive guidance as to the proper construction of the statute, “[t]he findings necessary to order visitation over the objections of a parent are thus not in the record, and I would remand for further proceedings.” Ibid.

3. Unlike Justice O’Connor, ante, at 10—11, I find no suggestion in the trial court’s decision in this case that the court was applying any presumptions at all in its analysis, much less one in favor of the grandparents. The first excerpt Justice O’Connor quotes from the trial court’s ruling, ante, at 10, says nothing one way or another about who bears the burden under the statute of demonstrating “best interests.” There is certainly no indication of a presumption against the parents’ judgment, only a “ ‘commonsensical’ ” estimation that, usually but not always, visiting with grandparents can be good for children. Ibid. The second quotation, ante, at 11, “ ‘I think [visitation] would be in the best interest of the children and I haven’t been shown that it is not in [the] best interest of the children,’ ” sounds as though the judge has simply concluded, based on the evidence before him, that visitation in this case would be in the best interests of both girls. Verbatim Report of Proceedings in In re Troxel, No. 93—3—00650—7 (Wash. Super. Ct., Dec. 14, 1994), p. 214. These statements do not provide us with a definitive assessment of the law the court applied regarding a “presumption” either way. Indeed, a different impression is conveyed by the judge’s very next comment: “That has to be balanced, of course, with Mr. and Mrs. Wynn [a.k.a. Tommie Granville], who are trying to put together a family that includes eight children, … trying to get all those children together at the same time and put together some sort of functional unit wherein the children can be raised as brothers and sisters and spend lots of quality time together.” Ibid. The judge then went on to reject the Troxels’ efforts to attain the same level of visitation that their son, the girls’ biological father, would have had, had he been alive. “[T]he fact that Mr. Troxel is deceased and he was the natural parent and as much as the grandparents would maybe like to step into the shoes of Brad, under our law that is not what we can do. The grandparents cannot step into the shoes of a deceased parent, per say [sic], as far as whole gamut of visitation rights are concerned.” Id., at 215. Rather, as the judge put it, “I understand your desire to do that as loving grandparents. Unfortunately that would impact too dramatically on the children and their ability to be integrated into the nuclear unit with the mother.” Id., at 222—223. However one understands the trial court’s decision–and my point is merely to demonstrate that it is surely open to interpretation–its validity under the state statute as written is a judgment for the state appellate courts to make in the first instance.

4. Justice Souter would conclude from the state court’s statement that the statute “do[es] not require the petitioner to establish that he or she has a substantial relationship with the child,” In re Smith, 137 Wash. 2d 1, 21, 969 P.2d 21, 31 (1998), that the state court has “authoritatively read [the ‘best interests’] provision as placing hardly any limit on a court’s discretion to award visitation rights,” ante, at 3 (Souter, J., concurring in judgment). Apart from the question whether one can deem this description of the statute an “authoritative” construction, it seems to me exceedingly unlikely that the state court held the statute unconstitutional because it believed that the “best interests” standard imposes “hardly any limit” on courts’ discretion. See n. 5, infra.

5. The phrase “best interests of the child” appears in no less than 10 current Washington state statutory provisions governing determinations from guardianship to termination to custody to adoption. See, e.g., Wash. Rev. Code §26.09.240 (6) (Supp. 1996) (amended version of visitation statute enumerating eight factors courts may consider in evaluating a child’s best interests); §26.09.002 (in cases of parental separation or divorce “best interests of the child are served by a parenting arrangement that best maintains a child's emotional growth, health and stability, and physical care”; “best interest of the child is ordinarily served when the existing pattern of interaction between a parent and child is altered only to the extent necessitated by the changed relationship of the parents or as required to protect the child from physical, mental, or emotional harm”); §26.10.100 (“The court shall determine custody in accordance with the best interests of the child”). Indeed, the Washington state courts have invoked the standard on numerous occasions in applying these statutory provisions–just as if the phrase had quite specific and apparent meaning. See, e.g., In re McDoyle, 122 Wash. 2d 604, 859 P.2d 1239 (1993) (upholding trial court “best interest” assessment in custody dispute); McDaniels v. Carlson, 108 Wash. 2d 299, 310, 738 P.2d 254, 261 (1987) (elucidating “best interests” standard in paternity suit context). More broadly, a search of current state custody and visitation laws reveals fully 698 separate references to the “best interest of the child” standard, a number that, at a minimum, should give the Court some pause before it upholds a decision implying that those words, on their face, may be too boundless to pass muster under the Federal Constitution.

6. It necessarily follows that under the far more stringent demands suggested by the majority in United States v. Salerno, 481 U.S. 739, 745 (1987) (plaintiff seeking facial invalidation “must establish that no set of circumstances exists under which the Act would be valid”), respondent’s facial challenge must fail.

7. The suggestion by Justice Thomas that this case may be resolved solely with reference to our decision in Pierce v. Society of Sisters, 268 U.S. 510, 535 (1925), is unpersuasive. Pierce involved a parent’s choice whether to send a child to public or private school. While that case is a source of broad language about the scope of parents’ due process rights with respect to their children, the constitutional principles and interests involved in the schooling context do not necessarily have parallel implications in this family law visitation context, in which multiple overlapping and competing prerogatives of various plausibly interested parties are at stake.

8. This Court has on numerous occasions acknowledged that children are in many circumstances possessed of constitutionally protected rights and liberties. See Parham v. J. R., 442 U.S. 584, 600 (1979) (liberty interest in avoiding involuntary confinement); Planned Parenthood of Central Mo. v. Danforth, 428 U.S. 52, 74 (1976) (“Constitutional rights do not mature and come into being magically only when one attains the state-defined age of majority. Minors, as well as adults, are protected by the Constitution and possess constitutional rights”); Tinker v. Des Moines Independent Community School Dist., 393
U.S. 503, 506—507 (1969) (First Amendment right to political speech); In re Gault, 387 U.S. 1, 13 (1967) (due process rights in criminal proceedings).

9. Cf., e.g., Wisconsin v. Yoder, 406 U.S. 205, 241-246 (1972) (Douglas, J., dissenting) (“While the parents, absent dissent, normally speak for the entire family, the education of the child is a matter on which the child will often have decided views. He may want to be a pianist or an astronaut or an oceanographer. To do so he will have to break from the Amish tradition. It is the future of the student, not the future of the parents, that is imperiled by today’s decision. If a parent keeps his child out of school beyond the grade school, then the child will be forever barred from entry into the new and amazing world of diversity that we have today… . It is the student’s judgment, not his parents’, that is essential if we are to give full meaning to what we have said about the Bill of Rights and of the right of students to be masters of their own destiny.”). The majority’s disagreement with Justice Douglas in that case turned not on any contrary view of children’s interest in their own education, but on the impact of the Free Exercise Clause of the First Amendment on its analysis of school-related decisions by the Amish community.

10. See Palmore v. Sidoti, 466 U.S. 429, 431 (1984) (“The judgment of a state court determining or reviewing a child custody decision is not ordinarily a likely candidate for review by this Court”); cf. Collins v. City of Harker Heights, 503 U.S. 115, 128 (1992) (matters involving competing and multifaceted social and policy decisions best left to local decisionmaking); Regents of the University of Michigan v. Ewing, 474 U. S 214, 226 (1985) (emphasizing “our reluctance to trench on the prerogatives of state and local educational institutions” as federal courts are ill-suited to “evaluate the substance of the multitude of academic decisions that are made daily by” experts in the field evaluating cumulative information”). That caution is never more essential than in the realm of family and intimate relations. In part, this principle is based on long-established, if somewhat arbitrary, tradition in allocating responsibility for resolving disputes of various kinds in our federal system. Ankenbrandt v. Richards, 504 U.S. 689 (1992). But the instinct against over-regularizing decisions about personal relations is sustained on firmer ground than mere tradition. It flows in equal part from the premise that people and their intimate associations are complex and particular, and imposing a rigid template upon them all risks severing bonds our society would do well to preserve.