Two recent Supreme Court decisions expose a fundamental instability in the constitutional framework governing parental medical decision-making for minors. In United States v. Skrmetti (2025), the Court upheld a categorical prohibition on gender-affirming treatments sought by fit parents and licensed physicians without requiring any individualized showing of harm to the affected minors. In Chiles v. Salazar (2026), the Court curtailed state regulation of conversion therapy, a therapeutic practice widely condemned by the medical profession, thereby subordinating child welfare concerns to the therapist’s First Amendment objections grounded in viewpoint discrimination.

Across both cases, the operative variable is not risk to the child but ideological valence–a doctrinal asymmetry that cannot be reconciled within existing constitutional principles without reinstating the harm principle as the governing threshold for state intervention. The harm principle—most prominently articulated by John Stuart Mill in On Liberty and adapted to pediatric decision-making by Douglas Diekema—permits state interference with parental authority only where a child faces a substantial risk of serious harm. It stands in deliberate tension with the dominant “best interest” framework, which invites courts to substitute their judgment for that of parents whenever an alternative course appears preferable. Although the best interest standard has legitimate applications in clinical guidance and intra-family disputes, it becomes constitutionally problematic when deployed as a basis for coercive state override. In that posture, it lacks a limiting principle: virtually any parental decision can be recast as suboptimal—a role that is both normatively indefensible and constitutionally suspect.

A harm-based framework reorients the analysis from outcome maximization to risk prevention. Parental decision-making is irreducibly pluralistic: reasonable parents may differ in values and in how they interpret contested medical evidence. A constitutional regime committed to pluralism cannot require convergence on a single state-sanctioned conception of optimal care. By limiting intervention to cases involving substantial harm, both parental autonomy and the diversity of family life are preserved while maintaining a principled basis for state action.

Operationalizing this framework requires a structured, tiered inquiry. At the first tier—where harm is imminent and life-threatening—state intervention is mandatory and consistent with Prince v. Massachusetts (1944), which establishes that parental authority does not extend to decisions that gravely endanger a child’s life. At the second tier—where harm is substantial but not imminent—intervention is permissible only upon an individualized, evidence-based showing that the particular child faces serious risk. This requirement precludes categorical prohibitions grounded in generalized risk assessments, which resemble the group-based reasoning long criticized in disability law as incompatible with individualized justice. At the third tier—where a parental decision may be clinically suboptimal but does not create substantial harm—the state is categorically required to defer. Under Troxel v. Granville (2000), fit parents are entitled to a presumption that their decisions serve their children’s interests.

Reframing the doctrine in these terms resolves the apparent tension between the Court’s recent decisions and existing constitutional principles. The statute upheld in Skrmetti imposes a categorical prohibition without individualized risk findings. Similarly, the analysis in Chiles is distorted by its exclusive focus on viewpoint discrimination. As Jennifer Bard observes, Chiles marks the first time the Court has used the First Amendment to subject a state’s efforts to protect children from allegedly harmful therapeutic practices to strict constitutional scrutiny. The proper inquiry is not whether the regulation targets a particular viewpoint, but whether the regulated practice poses a substantial risk of harm to the child. Professor Luke Boso similarly argues that the decision obscures the lived harms experienced by LGBTQ youth by reducing conversion therapy to mere speech while disregarding evidence that such practices can inflict significant psychological injury on minors.

By centering the analysis on harm rather than expression, courts avoid the doctrinal instability that arises from importing First Amendment categories into questions of child welfare. Legal theorists such as Martha Minow and Deborah Brake have long cautioned that a harm-based framework avoids that problem by grounding doctrine in empirically verifiable risk.

The coherence of this approach is reinforced by existing doctrines governing medical accountability. Malpractice liability attaches where a practitioner deviates from professional standards and causes patient injury, focusing squarely on harm rather than the content of professional judgment. Informed consent doctrine likewise regulates disclosure obligations by reference to material risk and patient injury, not by restricting permissible viewpoints. Anchoring legal regulation in harm rather than expression provides a more stable and constitutionally defensible framework. That framework would enhance the human rights of LGBT children without interfering with the constitutional commitment to parental autonomy and pluralism while ensuring that the state retains authority to intervene in the limited set of cases where intervention is genuinely justified.

The costs of doctrinal incoherence are not merely theoretical. In the wake of Chiles, states have been forced to shift from prospective regulation to ex post tort remedies, a posture that protects children only after the harm is complete. Without a uniform, content-neutral threshold, judicial outcomes will continue to track ideological preferences rather than principled reasoning. Restoring the harm principle—requiring a showing of substantial, individualized risk—provides the necessary doctrinal anchor.

Acknowledgment

The author thanks Professor Anne Lofaso for her helpful suggestions and editorial assistance.