Shannon MacDonald v. President NJ State Board of Medical Examiners

majority opinion

PRECEDENTIAL

UNITED STATES COURT OF APPEALS

FOR THE THIRD CIRCUIT

No. 25-2090

SHANNON MACDONALD, M.D.; PAUL GARDNER,

M.D., J.A., a minor, by and through guardian and next friend Michael Abell; MICHAEL ABELL, HANK JENNINGS,

Appellants

PRESIDENT OF THE NEW JERSEY STATE BOARD OF

MEDICAL EXAMINERS

On Appeal from the United States District Court

for the District of New Jersey

(D.C. No. 1:23-cv-23044)

District Judge: Hon. Edward S. Kiel

Argued March 25, 2026

Before: HARDIMAN, SCIRICA, ∗ and AMBRO, Circuit

∗

The Honorable Anthony J. Scirica was unavailable to participate in the decision in this case after argument before the

Judges

(Filed: August 31, 2026 )

OPINION OF THE COURT

HARDIMAN, Circuit Judge.

This appeal raises important and thorny issues arising under the First Amendment to the United States Constitution. New Jersey requires any doctor who wishes to practice telemedicine with a patient located in the state to first obtain a license from its Board of Medical Examiners. Three New Jersey residents and two doctors licensed outside New Jersey claim that this requirement infringes their free speech rights. We disagree. And because Plaintiffs’ other constitutional claims are insubstantial, we will affirm the District Court’s order dismissing all claims, with one modification.

On pain of criminal and civil sanction, New Jersey requires that “[a]ll persons commencing the practice of medicine or surgery in th[e] State shall apply to the board [of medical examiners] for a license to do so.” N.J. Stat. Ann. §§ 45:9-6; 45:1-18.2(b)(2), -25(a); 2C:21-20, 2C:43-3(b)(1), -merits panel. This opinion is filed by a quorum of the panel pursuant to 28 U.S.C. § 46(d) and 3d Cir. I.O.P. 12.1(b).

6(a)(3). New Jersey defines the practice of medicine to include “any method,” except those contained in exceptions not relevant here, “of treatment of human ailment, disease, pain, injury, deformity, mental or physical condition,” id. § 45:9-5.1, and “offer[ing] or undertak[ing] by any means or methods to diagnose, treat, operate or prescribe for any human disease, pain, injury, deformity or physical condition,” id. § 45:9-18. The state’s licensure requirement applies to medicine by virtual modality too: a physician “who uses telemedicine or engages in telehealth while providing health care services to a patient” must “be validly licensed, certified, or registered . . . to provide such services in the State of New Jersey.” Id. § 45:1-62(b).

Physicians first licensed to practice in New Jersey must, among other things, pass an examination, complete postgraduate training, and submit to a background check. See id. § 45:9-6; N.J. Admin. Code §§ 13:35-3.1, -3.11A(b), -3.13. A physician who is already licensed and in good standing in another state with substantially equivalent licensure standards need not be reexamined but must submit an application and associated forms, undergo a background check, and pay processing fees. N.J. Stat. Ann. § 45:1-7.5; N.J. Admin. Code § 13:35-6.13. According to Plaintiffs, the fees amount to $550 and the average processing time is three months. The process may be expedited for physicians whose state of licensure participates in the Interstate Medical Licensure Compact. Rather than fill out the typical application forms, physicians applying under the Compact can obtain a “letter of qualification” from their “state of principal license” attesting eligibility to practice, which the principal state will issue upon verifying the applicant’s qualifications and conducting a background check. N.J. Stat. Ann. § 45:9-6.2(5)(b). Plaintiffs

aver that the fees for this process are $700 and that it can be completed in “weeks.” App. 48.

The licensure requirements for out-of-state physicians seeking to provide telemedicine and telehealth services to patients in New Jersey allegedly were relaxed during the COVID-19 pandemic. Those physicians could engage in any kind of telemedicine without a New Jersey license if they had a preexisting doctor-patient relationship with the recipient, and they could undertake COVID-19-specific telemedicine regardless of whether there was a preexisting relationship. New Jersey also “waived application fees and allowed qualified applicants to become licensed within 24 hours of applying” under a “Temporary Emergency Reciprocity Licensure program.” App. 53.

Shannon MacDonald, M.D., is a radiation oncologist at Massachusetts General Hospital who specializes in treating rare pediatric cancers. She lives in Massachusetts and is licensed to practice medicine there. Dr. MacDonald used proton therapy to treat her patient J.A. for a rare childhood cancer when he was 18 months old. J.A. requires annual scans to monitor for anomalies, as the cancer often recurs. An anomaly was detected previously when J.A. lived in New York, and Dr. MacDonald was able to provide a telemedicine consultation to address the matter. J.A. is now a teenager living in New Jersey. He and his father, Michael Abell, would like to consult virtually with Dr. MacDonald in the future if any new anomalies appear on his annual scans.

Paul Gardner, M.D., is a neurosurgeon at the University of Pittsburgh Medical Center and an expert in skull base

surgery. He lives in Pennsylvania and is licensed there. He too would like to speak with his patients located in New Jersey to “discuss treatment options” without requiring them to incur travel expenses. App. 43. Hank Jennings is a New Jersey resident who underwent surgery and treatment by specialists in Pittsburgh when he was nineteen. Like J.A., Jennings needs periodic follow-up consultations and would like to do them via telemedicine to mitigate financial and time burdens.

Dr. MacDonald, Dr. Gardner, J.A., Abell, and Jennings filed suit in federal court and sought a permanent injunction prohibiting New Jersey from enforcing N.J. Stat. Ann. § 45:1-62(b) to prevent them from consulting by telemedicine. In this as-applied challenge, Drs. MacDonald and Gardner contend that New Jersey’s requirement that they undergo the burdens of New Jersey licensure violates the First Amendment, dormant Commerce Clause, and Privileges and Immunities Clause. The patients assert similar First Amendment and Commerce Clause challenges, and J.A.’s father also contends that the law violates his substantive due process right to make decisions about J.A.’s medical care. The District Court rejected each of these arguments and granted New Jersey’s motion to dismiss the complaint for failure to state a claim. This timely appeal followed.

II 1

We begin with Plaintiffs’ strongest contention: that

We have jurisdiction under 28 U.S.C. § 1291, and except as we explain below in our discussion of Article III standing for Abell’s substantive due process claim, the District Court had

New Jersey’s telemedicine law violates their First Amendment rights to free speech. The threshold question is whether the challenged law regulates speech. See Veterans Guardian VA Claim Consulting LLC v. Platkin, 133 F.4th 213, 219 (3d Cir. 2025). It does.

Drs. MacDonald and Gardner would like to use virtual modalities to “discuss treatment options” with patients, App. 43, and make “specific, nuanced . . . recommendations,” App. 42. The complaint focuses exclusively on the “diagnoses,” and “expert advice” communicated by these doctors through their spoken words to their patients. Pls. Br. 11, 25. That advice is speech. Veterans Guardian, 133 F.4th at 219; see also Chiles v. Salazar, 146 S. Ct. 1010, 1023 (2026) (“While the First Amendment protects many and varied forms of expression, the spoken word is perhaps the quintessential form of protected speech.”). And the telemedicine law conditions the doctors’ right to engage in that speech on holding a valid New Jersey license. N.J. Stat. Ann. § 45:1-62(b). Requiring a license to speak is, of course, a regulation of speech. See, e.g., Riley v. Nat’l Fed’n of the Blind of N. Carolina, Inc., 487 U.S. 781, 802 (1988); Billups v. City of Charleston, S.C., 961 F.3d 673, 683 (4th Cir. 2020).

Plaintiffs argue that New Jersey lacks good reasons for this speech restriction, at least as it applies to speech by specialists with national practices like Drs. MacDonald and

jurisdiction under 28 U.S.C. § 1331. We exercise plenary review of the District Court’s dismissal order under Rule 12(b)(6) of the Federal Rules of Civil Procedure and view the factual allegations in the light most favorable to Plaintiffs. Santiago v. Warminster Twp., 629 F.3d 121, 128 (3d Cir. 2010).

Gardner. The force of that argument depends on how closely the First Amendment requires us to scrutinize the law. As we will explain, New Jersey’s law implicates two strands of the Supreme Court’s First Amendment jurisprudence that are in tension. One strand holds that the First Amendment is highly suspicious of content-based speech restrictions, so courts must evaluate them under the most demanding form of

constitutional review: strict scrutiny. The other strand holds that the First Amendment can often accommodate speech restrictions that are consistent with a long and robust regulatory tradition. After describing these separate approaches and the telemedicine law’s place within them, we conclude that strict scrutiny is inappropriate notwithstanding the law’s contentbased character and that New Jersey’s law passes muster.

As Plaintiffs rightly note, laws that “target[] speech based on its communicative content” are usually subject to strict scrutiny. City of Austin, Texas v. Reagan Nat’l Advert. of Austin, LLC, 596 U.S. 61, 69 (2022) (citation modified). The most “egregious form” of content-based laws are those that discriminate based on a speaker’s viewpoint, i.e., those that “dictate what particular opinion or perspective individuals may express on [a] subject.” Chiles, 146 S. Ct. at 1021 (citation modified). All agree that the New Jersey law is not a viewpoint-based restriction. But while less inimical to the free speech right than viewpoint-based laws, a law whose application turns on “the topic discussed or the idea or message expressed” still threatens free speech and is normally subject to strict scrutiny. City of Austin, 596 U.S. at 69 (citation

modified).

Here, New Jersey’s telemedicine law is a content-based restriction because its application turns on the message a speaker conveys. A license is required before a physician may engage in speech that is part of the practice of medicine in New Jersey. See N.J. Stat. Ann. §§ 45:9-6, 45:1-62(b). And New Jersey’s definition of the practice of medicine covers speech with specified messages. The definition includes “offer[ing] or undertak[ing] by any means or methods to diagnose, treat, operate or prescribe for any human disease, pain, injury, deformity or physical condition,” N.J. Stat. Ann. § 45:9-18. So the very messages that Drs. MacDonald and Gardner wish to convey to their patients, namely diagnoses and recommended courses of medical treatment, are prohibited without a license. See, e.g., Pinkus v. MacMahon, 29 A.2d 885, 885–87 (N.J. 1943); State v. Jeannotte-Rodriguez, 261 A.3d 1005, 1021 (N.J. App. Div. 2021).

To be sure, some speech regulations are content-neutral even though determining whether there has been a violation requires knowing the content of a speaker’s speech. One example comes from City of Austin, which involved a municipal law that restricted signs unless the subject of the sign’s speech was physically located on the same premises as the sign. 596 U.S. at 65–66. The Supreme Court held that the law was content-neutral because it made no distinctions based on topic, message, idea, or subject matter as such. Id. at 71. No matter what communicative content a sign displayed, the restriction applied if the physical location of the sign was on different premises from the physical location of the thing it discussed. Id. And the only reason a regulator needed to know the sign’s content was to make that location-based determination. Id. True, that law effectively prohibited certain

billboard messages, but not because of any communicative feature of the message itself.

Another example of a content-neutral law that might require a regulator to know the content of infringing speech is a prohibition of solicitation. See id. at 72 (discussing Heffron v. International Soc. for Krishna Consciousness, Inc., 452 U.S. 640 (1981)). Those laws essentially forbid the statement “buy my wares,” but not because of anything about the content of that message. The laws take issue with only the speech’s function and purpose—to solicit. Id.

The Second Circuit invoked this line of cases recently in Brokamp v. James, 66 F.4th 374 (2d Cir. 2023), where it held that a licensing law for mental-health counselors was content-neutral because the law required a license for speech with certain enumerated “therapeutic purposes” if made in a particular “context.” Id. at 397. The content of a counselor’s speech, the court explained, was immaterial: “it matters not at all whether a counselor speaks to a client about personal relationships, professional anxieties, medical challenges, world events, planned travel, hobbies, sports, favorite movies, or any other subject.” Id. To the extent a regulator would ever need to examine a speaker’s speech, it would only be to determine whether the speech was uttered in service of an enumerated therapeutic purpose. See id.

The New Jersey law here is different because a

speaker’s message is itself the relevant criterion. If Drs. MacDonald and Gardner convey “diagnoses,” Pls. Br. 11, or give their “expert advice” to a patient, Pls. Br. 25, the communicative content of their speech is what places them in violation of the law. The content is not merely a tool for ascertaining some other content-neutral criterion like purpose

or physical location.

Of course, there are other necessary conditions for a violation of the telemedicine law, most obviously that the speaker not be licensed in New Jersey. But that fact does not alter the content-based nature of the law. Indeed, many if not most content-based laws have conditions other than speech content that are necessary for a violation. Take, for example, “a law banning the use of sound trucks for political speech.” Reed v. Town of Gilbert, Ariz., 576 U.S. 155, 169 (2015). Undoubtedly that law is content-based even though speech with political content is a necessary but not sufficient condition of a violation. Or take a recently considered law that prohibited robocalls unless made to collect a debt owed to the United States. See Barr v. Am. Ass’n of Pol. Consultants, Inc., 591 U.S. 610, 619 (2020) (plurality opinion). That law is contentbased even though the robocall modality is required for a violation, in addition to non-debt-collection content. See id.; see also id. at 636–37 (Sotomayor, J., concurring) (agreeing the law is content based); id. at 639–44 (Breyer, J., with whom Ginsburg, J. and Kagan, J. join, concurring in part and dissenting in part) (same); Id. at 650–51 (Gorsuch, J., concurring in part and dissenting in part) (same). Or, of particular relevance here, take laws that restrict speech with certain content by certain speakers—like a law restricting political speech by corporations, see Citizens United v. Fed. Election Comm’n, 558 U.S. 310, 340–41 (2010), or a law “limiting the content of newspapers,” see Reed, 576 U.S. at 170 (considering that example). Those laws, like New Jersey’s here, are content-based even though they might also “be

characterized as speaker-based.” Id. 2

For these reasons, we hold that New Jersey’s law is content-based.

Before it turns to history and tradition, New Jersey contends that its telemedicine law implicates exceptions to the typical rules for content-based laws already recognized in the caselaw. Its two arguments miss the mark. First, New Jersey argues that the law is a regulation of professional conduct that burdens speech only incidentally. See Nat’l Inst. of Fam. & Life Advocs. v. Becerra, 585 U.S. 755, 769–70 (2018) (NIFLA); Chiles, 146 S. Ct. at 1025–26. But the state’s argument is indistinguishable from ones that have been rejected many times, including recently in Chiles: the state contends that its licensing scheme limits the practice of medicine as a whole, and since the practice of medicine is mostly conduct, the law’s speech limitations are only incidental to its limitations on conduct. That is not what courts mean when they refer to “incidental” regulations of speech. That label refers to laws that “restrict[] speech only because it is integrally related to unlawful conduct.” Chiles, 146 S. Ct. at 1026 (emphasis

Even a law that discriminates against certain speakers without facially discriminating against any content might still be considered content-based if the speaker-based distinction appears to be a proxy for content. See Am. Ass’n of Pol. Consultants, 591 U.S. at 619–20 (plurality opinion) (discussing how a court might evaluate a law that prohibited robocalls by authorized debt collectors).

added). 3 It does not refer, as New Jersey suggests, to “a law [that] mostly addresses conduct and only sometimes sweeps in speech.” Id. The speech that Drs. MacDonald and Gardner want to express is not necessarily related to any conduct; they may well discuss diagnoses with and give medical advice to a patient located in New Jersey without ever performing physical treatment on the patient. Cf. NIFLA, 585 U.S. at 770 (holding that a mandatory notice for clinics was not incident to conduct because it was “not tied to a procedure at all”). And even if the speech were incident to some physical treatment by the doctors, that treatment would not be unlawful conduct because Drs. MacDonald and Gardner make clear they will only perform physical treatment in states where they are licensed. So the speech-incident-to-conduct exception does not apply.

New Jersey’s next argument is more sweeping. It argues that licensing laws for traditionally regulated professions like medicine are exempt from any form of heightened scrutiny, even where the law prohibits unlicensed individuals from giving professional advice. It invokes, as the District Court did, our decision in National Association for the Advancement of Multijurisdiction Practice v. Castille, 799 F.3d 216 (3d Cir. 2015). But the state misreads Castille, and its misreading would bring that case into conflict with more recent Supreme Court precedent.

In Castille we asked whether Pennsylvania infringed the First Amendment rights of lawyers licensed in certain other states by requiring them to take and pass the Pennsylvania bar

The label can also refer to laws that “restrict[] expressive conduct only for reasons unrelated to its content.” Chiles, 146 S. Ct. at 1026 (emphasis added). But nobody contends that any expressive conduct is at issue here.

exam. Id. at 218. We held that it did not because the bar-exam requirement neither “pass[ed] judgment on the content” of an applicant’s speech nor on its time, place, or manner. Id. at 221 (citation omitted). In other words, we held that the bar-exam requirement did not employ speech-based criteria in restricting who could practice law in Pennsylvania. Id. That is a different kind of First Amendment issue than the one raised by the Plaintiffs here, who dispute the scope of a state’s power to require a license before speaking. Castille did not discuss whether there are aspects of legal practice that constitute speech (of course there are, cf. Veterans Guardian, 133 F.4th at 219), nor whether a licensure requirement to engage in that speech is a regulation of speech (of course it is). So Castille does not bear on the level of scrutiny we must apply here.

In any event, even had we held in Castille that licensure requirements for the speech aspects of professions always receive mere rational-basis scrutiny, the Supreme Court’s intervening decision in NIFLA would require us to revisit it. NIFLA held that there is no category of speech called “professional speech” that receives diminished First Amendment protection. 585 U.S. at 767–68; see also Chiles, 146 S. Ct. at 1024 (reiterating that principle). So we generally must treat speech uttered in the course of professional practice the same as any other speech. NIFLA, 585 U.S. at 767–68; see also Veterans Guardian, 133 F.4th at 220 (“With few exceptions, the same First Amendment principles apply when professionals speak to clients as when anyone else talks.”). And ordinarily, strict scrutiny applies to content-based restrictions on speech like the one here.

Our determination that New Jersey’s law is content13

based is not the end of the story, however. The Supreme Court has consistently instructed that “a long (if heretofore unrecognized) tradition” of a particular kind of speech restriction can show that the traditionally restricted speech warrants “diminished” protection. Chiles, 146 S. Ct. at 1026 (citation modified); see also Houston Cmty. Coll. Sys. v. Wilson, 595 U.S. 468, 474–75 (2022) (“‘[A] regular course of practice’ can illuminate or ‘liquidate’ our founding document’s ‘terms and phrases’” (quoting Letter from J. Madison to S. Roane (Sept. 2, 1819), in 8 Writings of James Madison 450 (G. Hunt ed. 1908)).

For example, the First Amendment permits wholesale proscriptions of certain “exceptional categories” of speech that “share a long and well-recognized historical pedigree” like fraud and defamation. Chiles, 146 S. Ct. at 1021; see also Free Speech Coal., Inc. v. Paxton, 606 U.S. 461, 472–73 (2025) (summarizing obscenity prohibitions’ historical pedigree). And even when those categories aren’t implicated, evidence of a long tradition of a particular kind of speech regulation might tell a court to relax its guard when evaluating First Amendment challenges to similar regulations. See Houston Cmty. Coll. Syst., 595 U.S. at 474–77 (historical use of censures by assemblies against members for their speech suggested such censures do not constitute First Amendment retaliation); City of Austin, 596 U.S. at 75 (tradition of regulating on- and offpremises signs differently counseled against conclusion that such distinctions were really content-based regulations); Vidal v. Elster, 602 U.S. 286, 301 (2024) (collecting cases where the Supreme Court has “consider[ed] [a restriction’s] history and tradition”). See also Williams-Yulee v. Fla. Bar, 575 U.S. 433, 462 (2015) (Scalia, J., dissenting) (“Our cases hold that speech enjoys the full protection of the First Amendment unless a

widespread and longstanding tradition ratifies its regulation.”); Chiles, 146 S. Ct. at 1031 (Kagan, J., concurring) (opining that “experience and reason alike” might show that content-based and viewpoint-neutral restrictions on speech in the medical field do not warrant strict scrutiny)

The Supreme Court’s recent opinion in Vidal v. Elster is particularly instructive. There, the Court reasoned that the long history of trademark law—an “inherently content-based” endeavor, 602 U.S. at 296—showed that it could “play well with the First Amendment,” id. at 300 (citation modified). Specifically, the Court upheld the Lanham Act’s prohibition on including a person’s name in a trademark without the subject’s permission—a content-based restriction on speech. See id. at 294–95. The Court began by observing that trademark law existed only in nascent form at the founding and “developed slowly,” id. at 296, perhaps because a modern trademark regime would have had little utility in the localized commercial markets of that period, see id. at 297 (observing that “for most of our first century, most commerce was local and most consumers therefore knew the source of the goods they purchased”); accord id. at 312 (Barrett, J., concurring) (“American trademark law did not develop in earnest until the mid-19th century.”).

But from the time trademark law developed in earnest, the Court explained, it has been little cause for First Amendment concern, and that “longstanding, harmonious relationship suggest[ed] that heightened scrutiny need not always apply” to content-based trademark rules. Id. at 299. The Court concluded its analysis by canvassing authorities from the late 19th and early 20th centuries showing that trademark law has traditionally restricted the ability to trademark names. See id. at 301–05; see also id. at 311, 323–24 (Barrett, J.,

concurring). That tradition, the Court held, was “sufficient to conclude that the [Lanham Act’s] names clause . . . is compatible with the First Amendment.” Id. at 301.

Against this backdrop, we consider the history and tradition of medical licensing. Medical-licensing laws have a lengthy pedigree in the Anglo-American legal tradition, appearing in England at least as early as 1518. That year, Henry VIII granted the Royal College of Physicians a charter to license those qualified to practice “physick” in London and to prosecute those who practiced without a license. Jeffrey Lionel Berlant, Profession and Monopoly: A Study of Medicine in the United States and Great Britain 134–35 (1975); see also Harold J. Cook, Policing the Health of London: the College of Physicians and the Early Stuart Monarchy, 2 Soc. Hist. Med. No. 1 at 6–7 (April 1989) (tracing the expansion of the College’s regulatory ambit). An Act of England’s Parliament confirmed the charter a few years later. See B. Abbott Goldberg, Horseshoers, Doctors and Judges and the Law on Medical Competence, 9 Pac. L.J. 107, 122 (1978). See also Dr. Bonham’s Case, 8 Co. Rep. 107a, 108a–109a, 77 Eng. Rep. 638, 639–41 (C.P. 1610) (describing the charter and statute). The practice of physicians in England at that time consisted almost exclusively of speech with specified content: “[a]s gentlemen,” they “declined to work with their hands and only observed, speculated, and prescribed.” Paul Starr, The Social Transformation of American Medicine: The Rise of a Sovereign Profession and the Making of a Vast Industry 37–38 (1982). Manual tasks were left to surgeons and apothecaries. Id. So prohibiting the unlicensed practice of “physick” was a content-based regulation of speech.

Two American colonies—New York and New Jersey—

also passed medical licensing laws before the revolution. 4 See Richard Harrison Shryock, Medical Licensing in America, 1650-1965, at 17 (1967). That was notable given the nascent state of medical practice and education on this side of the Atlantic. See Shryock at 3 (“In new communities, there was often a lack of men with any pretence to medical education.”); Joseph F. Kett, The Formation of the American Medical Profession; The Role of Institutions, 1780-1860, 10–12 (1968) (describing initial efforts in the late 1700s to establish medicine as a regulated profession). The pace of medical licensing picked up shortly after the founding, and by the 1830s nearly all the states in the Union required a license to practice. Shryock at 23, Kett at 13. Under these laws, one could practice medicine only after demonstrating competency by passing an exam or obtaining a qualifying degree. Shryock at 27.

Like their English precursors, the early American laws regulating medical practice restricted speech. After all, medicine at the time was heavily speech-based. For example, diagnoses were often reached only by drawing inferences from

An Act to Regulate the Practice of Physick and Surgery in the City of New York, in Laws of New York, From the 11th Nov. 1752, to 22d May 1762, at 188–89 (1762); An Act to Regulate the Practice of Physick and Surgery Within the Colony of New Jersey, in Acts of the General Assembly of the Province of New Jersey, From the Surrender of the Government to Queen Anne, on the 17th Day of April, in the Year of our Lord 1702, to the 14th Day of January 1776, at 376–77 (1776). Other colonies issued medical licenses as well but those were merely honorific; the unlicensed could still practice. Joseph F. Kett, The formation of the American medical profession: The Role of Institutions, 1780-1860, 7, 12 (1968).

a patient’s medical history rather than by physical examination. See Roy Porter, the Greatest Benefit to Mankind: A Medical History of Humanity 257–58 (1997); see also David A. Johnson & Humayun J. Chaudhry, Medical Licensing and Discipline in America: A History of the Federation of State Medical Boards 6–8 (2012) (explaining that doctors in early America performed all medical tasks, including the speechbased ones that physicians in England would perform). Doctor and founding father Benjamin Rush advised other doctors:

Begin to interrogate your patient. How long has

he been sick? When attacked and in what

manner? What are the probable causes, former

habits and dress; likewise the diet, etc., . . . Pay

attention to the phraseology of your patients, for

the same ideas are frequently conveyed in

different words. . . . Patients often conceal the

cause of their disease — therefore interrogate

them particularly when you suspect

intemperance as a cause of the disease.

Porter at 257–58. Simply put, conditions on the right to practice medicine were conditions on the right to speak about specific content and convey specific messages.

To be sure, these early English and American laws did not represent enduring consensus on medical licensure. In London, the scope of the Royal College’s authority was narrowed by exemptions from licensing for domestic practitioners, and apothecaries infringed physicians’ monopoly on medical advice with impunity. See Kett at 3–4; Berlant at 144. In the United States, the founding-era licensing laws were often weak on their own terms, merely prohibiting unlicensed practitioners from suing to recover payment. William G.

Rothstein, American Physicians in the Nineteenth Century: From Sects to Science 76 (1972). Unlicensed practitioners could circumvent that restriction by demanding payment up front. Lewis A Grossman, Choose Your Medicine: Freedom of Therapeutic Choice in America 16 (2021). And even the laws that provided for substantial punishment like fines or imprisonment were rarely enforced. Rothstein at 76–79. Most American licensing laws were then repealed altogether during the Jacksonian era. John Duffy, From Humors to Medical Science: A History of American Medicine 218 (2d ed. 1993).

This temporary swing in momentum was probably

attributable to various factors. For one, the practice of medicine was so inadequate (and often harmful) that the American people were perhaps reluctant to authorize states to declare who could practice. See Starr at 56, 58. For another, medicine was still viewed by many as more like religion than science, so religious-freedom sentiment often colored debates about licensing. See Grossman at 38–40. Licensure requirements in general also were inconsistent with the prevailing egalitarian attitudes of the Jacksonian era. See id. at 32–38. Finally, most Americans could not afford professional medical services, and the nature of travel presented serious impediments, so domestic practitioners were often the only realistic option. Starr at 65– 66.

As medical science and economic conditions improved and prevailing political attitudes shifted, however, proponents of licensing regained momentum. See John S. Haller, American Medicine in Transition, 1840–1910, at ix (1981); Starr at 69–71; Rothstein at 20–21. Licensing laws reemerged even before the Civil War—as early as 1859. Johnson and Chaudhry at 23. The pace picked up after the war: by 1890, 35 states had passed medical-licensing laws. Haller at 223. By

1898, all had done so. Shryock at 54–55. Though some initially suffered similar weaknesses as their founding-era precursors, by the 1890s the laws were effective at limiting the profession to the competent and upright. Duffy at 219–20; see also Kett at vii (explaining that “between 1880 and 1910,” “the establishment of effective state licensing boards barred quacks from the profession.”). And in the century-and-a-half since medical licensing reemerged, the sophistication of medical knowledge and rigor of licensure conditions have grown in tandem. See Duffy at 313–14; Kett at 163–64; Rothstein at 20; Shryock at 67–68.

Throughout our history, providing advice has remained a core part of medical practice, and licensing schemes have continued to cover it. See, e.g., People v. Allcutt, 117 A.D. 546, 549, 553 (N.Y. App. Div. 1907) (upholding conviction for unlicensed practice where defendant did not administer drugs but purported to diagnose a disease and “prescribe[] diet and conduct remedies”), aff’d, 81 N.E. 1171 (N.Y. 1907); Pinkus, 29 A.2d at 886 (holding that a food-store owner engaged in unlicensed medical practice when he “diagnosed alleged ailments of the witnesses and expressed an opinion as to their cause”); Norville v. Miss. St. Med. Ass’n, 364 So. 2d 1084, 1089 (Miss. 1978) (chiropractor who “represent[ed] to a patient” that certain vitamins would “cure a disease or ailment” engaged in unlicensed practice of medicine); Wong v. Chappell, 773 S.E.2d 496, 497, 501 (Ga. Ct. App. 2015) (remanding for jury to determine whether the unlicensed defendant’s actions constituted the practice of medicine, defined to include “suggest[ing]” or “recommend[ing]”

treatment). 5

The upshot is this: for well over a century, the States have uniformly required a license to practice medicine, including when that practice takes the form of speech, and those licensure laws have always been content-based speech restrictions.

New Jersey’s telemedicine law is a viewpoint-neutral, content-based speech restriction consistent with the “long (if heretofore unrecognized) tradition” of state medical-licensing restrictions just described. Chiles, 146 S. Ct. at 1026 (citation omitted). No precedent establishes the mode of scrutiny for such cases.

We acknowledge that precedent does not take strict scrutiny entirely off the table for laws with strong historical pedigrees. Cf. New York State Rifle & Pistol Ass’n, Inc. v. Bruen, 597 U.S. 1, 36 (2022) (“[P]ost-ratification adoption or acceptance of laws that are inconsistent with the original meaning of the constitutional text obviously cannot overcome or alter that text.” (citation omitted)). Here, however, the history and nature of medical-licensing laws convince us that

The parties here don’t dispute that the speech Drs. MacDonald and Gardner wish to make is within the traditional scope of the practice of medicine; they dispute only the extent of New Jersey’s power to limit that kind of speech. Therefore, we need not and do not wade into the precise scope of what other speech has traditionally been restricted by medicallicensing laws.

strict scrutiny is inapt.

For starters, we see no reason to conclude that the original public meaning of the Free Speech Clause requires the most demanding scrutiny for medical-licensing laws. History shows that medical licensure predates the founding and has always been a content-based endeavor. To be sure, the founding generation was undecided on the wisdom and permissibility of those laws—widely enacting but rarely enforcing them. But that suggests, at worst, indeterminacy of original meaning. So we turn to the robust tradition that has now existed for well over a century, which strongly suggests that medical licensure can coexist with the Free Speech Clause. See Vidal, 602 U.S. at 295; Houston Cmty. Coll. Syst., 595 U.S. at 474–75.

Moreover, traditional medical-licensing laws do little to undermine the First Amendment values that precedent tells us to safeguard. Licensure laws do not “prescribe what shall be orthodox,” West Virginia Bd. of Ed. v. Barnette, 319 U.S. 624, 642 (1943), because they make no viewpoint distinctions, and among licensed doctors the laws do not even make content distinctions, see Chiles, 146 S. Ct. at 1028 (distinguishing the viewpoint-discriminatory law at issue from “licensing laws,” which “have traditionally addressed what qualifications an individual must possess before practicing a particular profession”); NIFLA, 585 U.S. at 772 (extolling the importance of debate within professions). And by limiting their restrictions only to speech falling within the bounds of medical practice, licensing laws do not exclude the unlicensed from the public marketplace of ideas. See Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting). The unlicensed may still participate in national conversations about medicine even if they cannot convey medical advice or diagnoses to

individual patients.

On the other end of the spectrum from strict scrutiny is rational-basis review—“the minimum constitutional standard that all legislation must satisfy.” Free Speech Coal., 606 U.S. at 471. That standard applies to viewpoint-neutral laws that restrict wholly proscribable categories of speech like obscenity. Id. It is unclear whether it also applies in cases involving speech that has traditionally been subject to lesser restrictions than total proscription. Compare Vidal, 602 U.S. at 307–08, with id. at 324 (Barrett, J., concurring). In between strict scrutiny and rational-basis review is intermediate scrutiny, which asks whether the law “further[s] an important Government interest unrelated to the suppression of free expression and do[es] not burden substantially more speech than necessary to further that interest.” TikTok Inc. v. Garland, 604 U.S. 56, 73–74 (2025). We need not decide which of those standards is more appropriate here because New Jersey’s telemedicine law survives even under intermediate scrutiny, the more demanding of the two.

Plaintiffs do not (and could not) dispute that New Jersey’s interest in promoting health and safety is important, and that the state generally furthers that interest, as it traditionally has, by requiring doctors to demonstrate competence and good character before practicing medicine. Nor do they take issue with any of the substantive requirements (e.g., board certification, satisfaction of a background check) for obtaining a New Jersey license. They instead argue that the licensure process is an undue burden on specialists who, like Drs. MacDonald and Gardner, have national practices and are already licensed in states with essentially the same substantive conditions for licensure. The doctors say New Jersey’s procedures are “duplicative,” Reply Br. 2, 10, 12, of their home

state’s licensure process and do very little to protect patient health because they have been vetted already. Essentially, they propose that New Jersey treat their home-state licenses as licenses to give advice to and diagnose New Jersey patients.

The problem with this argument is that the burdens Drs. MacDonald and Gardner complain about are too “slight” for us to plausibly conclude that “substantially” more speech is burdened than if New Jersey simply authorized them to give advice on account of their home-state license. Bruni v. City of Pittsburgh, 824 F.3d 353, 372 n.20 (3d Cir. 2016). Plaintiffs concede in their complaint that New Jersey already “streamline[s]” the application process for out-of-state physicians like Drs. MacDonald and Gardner. App. 48. And though they complain that applying for and maintaining a license still entails some fees and administrative tasks, 6 nothing in their complaint suggests that those add substantially to the burdens that Drs. MacDonald and Gardner concededly must assume to maintain their home-state licenses. They do not contend that the fees are remotely prohibitive. They do not explain how tasks like “[m]onitoring renewal dates,” App. 48, create anything more than a de minimis burden. And their oblique reference to “continuing education requirements,” id., tells us nothing about the burdens New Jersey imposes beyond those imposed by their home states. Plaintiffs have plausibly pleaded only that Dr. MacDonald and Dr. Gardner desire not to assume additional burdens, and that is not enough for us to

New Jersey identifies some substantive requirements for maintaining a license that are more rigorous than the requirements in Dr. MacDonald and Dr. Gardner’s home states, but Plaintiffs only complain about the fees and the time burdens of administrative tasks, so that is all we consider.

conclude that the burdens are substantial. Because Plaintiffs’ suggested alternatives are not substantially less burdensome on speech, New Jersey’s policy survives intermediate scrutiny. See Bruni, 824 F.3d at 370–71, 372 n.20.

* * *

For all these reasons, we hold that New Jersey’s

requirement that one obtain a New Jersey medical license before providing medical advice by telemedicine to persons in New Jersey does not violate Plaintiffs’ First Amendment rights.

III

Unlike their First Amendment challenge, Plaintiffs’ remaining challenges to the telemedicine law either have no persuasive force or are not within our jurisdiction to resolve.

First, the law does not violate the so-called “dormant” Commerce Clause because it neither discriminates against outof-state commercial actors, see Nat’l Pork Producers Council v. Ross, 598 U.S. 356, 369 (2023), nor imposes burdens on interstate commerce that are “clearly excessive in relation to the putative local benefits.” Pike v. Bruce Church, Inc., 397 U.S. 137, 142 (1970); see Nat’l Pork Producers, 598 U.S. at 379–80.

The telemedicine law applies equally to in-state and out-of-state doctors; all who wish to practice medicine (virtually or in person) with a patient in New Jersey must be licensed, regardless of the doctor’s location. See N.J. Stat. Ann. § 45:1-62(b). Plaintiffs counter that the law “mak[es] it harder

for out-of-state physician-specialists to do business in New Jersey” because “maintaining multiple licenses” is burdensome, Pls. Br. 43, and they say the law “strips away” the advantages of running a national practice based in a single location, Pls. Br. 44 (citation omitted). But New Jersey’s law equally burdens physician-specialists with national practices based in New Jersey, so the discrimination Plaintiffs have identified is against multi-state practitioners, not against doctors located outside of New Jersey. Cf. Tolchin v. Sup. Ct. of N.J., 111 F.3d 1099, 1107–08 (3d Cir. 1997) (requirement that attorneys admitted to practice in New Jersey maintain offices in the state burdened attorneys with “small or sporadic practices in New Jersey,” and such attorneys could be based in New Jersey or elsewhere).

Plaintiffs also contend that the law has a protectionist “leveling effect” like the law challenged in Hunt v. Washington State Apple Advertising Commission, 432 U.S. 333 (1977). Pls. Br. 44. Not so. Hunt involved a state law mandating use of an inferior apple-grading system that in-state growers were already using and out-of-state growers were not. Id. at 340, 350–52. The so-called “leveling effect,” id. at 351, resulted from the requirement that out-of-state growers conform to what in-state growers were already doing, see id. at 340, 351– 52. Here, Plaintiffs have identified nothing in the New Jersey law that favors practices unique to physicians based in New Jersey. To the contrary, they spend other parts of their brief explaining that competency standards are nationally uniform.

Plaintiffs also have failed to allege that the burdens of New Jersey’s law on interstate commerce are “clearly excessive” in relation to local health benefits. Plaintiffs begin at a disadvantage in the Pike-balancing analysis because medicine is a field traditionally subject to local regulation. See,

e.g., Zahl v. Harper, 282 F.3d 204, 211 (3d Cir. 2002). So New Jersey’s power to interfere with the interstate market is at its peak. Kassel v. Consol. Freightways Corp. of Delaware, 450 U.S. 662, 670 (1981). We thus “will not second-guess” New Jersey’s judgment that its local benefits are “important[t] in comparison with related burdens on interstate commerce” unless those benefits are “illusory.” Id. (citation omitted). Plaintiffs have failed to overcome that disadvantage.

Initially, the burdens imposed on interstate commerce by New Jersey’s telemedicine law appear straightforwardly proportional to the benefits it produces. The licensing conditions aim to protect New Jersey-based patients from unqualified and unfit practitioners. In service of that commonsense purpose, New Jersey requires practitioners who are already practicing in another state to undergo a streamlined licensing process to verify their qualifications. True, that process burdens cross-border practice, but we have deemed similar trade-offs uncontroversial. See, e.g., Castille, 799 F.3d at 225 (burden on interstate commerce of requiring attorneys barred elsewhere to take the Pennsylvania bar exam was “not clearly excessive in relation to Pennsylvania’s interests in regulating its bar and securing favorable treatment for Pennsylvania-barred attorneys” (internal quotation marks omitted)). If requiring out-of-state attorneys to take another bar exam is not an excessive burden, then requiring out-of-state doctors to complete a few administrative requirements and pay a reasonable fee should pass muster easily.

In response, Drs. MacDonald and Gardner do not

contend that New Jersey’s licensing scheme has any unique features that stretch its effects to the national market for physician-specialists or for medicine more broadly. Cf., e.g., Nat’l Pork Prods., 598 U.S. at 399–400 (Roberts, C.J.,

concurring in part) (reasoning that a California law was plausibly excessive in relation to local benefits because it would impact interstate transactions with no connection to California); Edgar v. MITE Corp., 457 U.S. 624, 643 (1982) (Illinois law was excessive in relation to local benefits because it had “nationwide reach”). Instead, they allege that the administrative tasks and fees New Jersey requires are unnecessary for the state’s patient-protection purposes. Specifically, they allege that no harm resulted during the twoyear period in which New Jersey waived fees for and expedited its licensure process and permitted out-of-state doctors to see COVID-19 patients and patients with whom they had a preexisting doctor-patient relationship. Even assuming that allegation was enough for us to second-guess New Jersey’s policy judgments, no allegations about the licensure process plausibly suggest that its burdens are excessive in light of the alleged lack of benefits conferred by the more stringent rules. The complaint merely alleges a few extra weeks of processing time, several hundred extra dollars in fees, and the unspecified time burdens of certain administrative tasks for maintaining the license. Nothing in the complaint suggests how these marginal burdens on individual doctors might result in any substantial burden on the interstate market.

Nor does New Jersey’s telemedicine law violate the Privileges and Immunities Clause of Article IV. That clause prohibits states from needlessly discriminating against residents of other states “on matters of fundamental concern,” United Bldg. & Constr. Trades Council of Camden Cnty. & Vicinity v. Mayor & Council of Camden, 465 U.S. 208, 220 (1984). Essentially for the reasons just explained, the telemedicine law does not discriminate against those who

reside outside of New Jersey. Any physician who wishes to conduct telemedicine with a patient located in New Jersey, no matter where he or she lives, must obtain a New Jersey medical license.

Finally, Michael Abell lacks standing for his claim that the telemedicine law deprives him of his substantive due process right to direct his son’s medical care. Article III standing “implicates our subject matter jurisdiction,” so “we must consider it independently” even though New Jersey never raised the issue. Cook v. GameStop, Inc., 148 F.4th 153, 163 (3d Cir. 2025). It is axiomatic that a plaintiff may sue for redress of only “actual or imminent” injuries and not “conjectural or hypothetical” ones. Lujan v. Defs. of Wildlife, 504 U.S. 555, 560 (1992) (citation modified).

Abell would like the option to consult with Dr.

MacDonald via telemedicine from New Jersey if an anomaly appears on one of his son’s future scans for reemergence of the cancer he was diagnosed with at 18 months old. But Abell’s son is a teenager now, and only one annual scan has ever required Dr. MacDonald’s consultation. And Abell has not alleged any other facts—like the percentage of patients whose cancer recurs or the likelihood that future scans will contain anomalies requiring expert consultation—suggesting that another anomaly is “certainly impending” or a “substantial risk.” Clemens v. ExecuPharm Inc., 48 F.4th 146, 152 (3d Cir. 2022) (citation modified). “[P]ossible future injury—even one with an objectively reasonable likelihood of occurring—is not sufficient.” Id. (internal quotation omitted). So the substantive due process claim must be dismissed for lack of subject-matter jurisdiction. GameStop, Inc., 148 F.4th at 162–63. We will

modify the District Court’s dismissal order to indicate dismissal without prejudice.

For the reasons stated, we will affirm the District Court’s order dismissing all claims, with the modification that Abell’s substantive due process claim will be dismissed without prejudice for lack of jurisdiction.

Jack E. Brown [Argued]

Caleb R. Trotter

PACIFIC LEGAL FOUNDATION

Counsel for Appellants

Matthew J. Platkin

Jeremy M. Feigenbaum

Stephen Ehrlich

Nathaniel F. Rubin [Argued]

Francis X. Baker

Michael Antenucci

OFFICE OF ATTORNEY GENERAL OF NEW JERSEY

Counsel for Appellee

Shannon MacDonald v. President NJ State Board of Medical Examiners

Authorities cited

Opinion · public domain