Health Law
Nov 2021

Should Employment Division v Smith Be Overturned?

Jake Greenblum, PhD and Ryan Hubbard, PhD
AMA J Ethics. 2021;23(11):E864-868. doi: 10.1001/amajethics.2021.864.


Health professionals seeking religious exemption from caring for some patients or providing some interventions receive robust legal protection. Similarly, religiously affiliated organizations have great latitude in deciding which services to offer. These protections could soon become stronger, as the US Supreme Court considers 2 cases that revisit constraints on exemption claims established in Employment Division, Department of Human Resources of Oregon v Smith (1990). This article contends that overturning this case’s precedent might result in clinicians claiming more religious exemptions, which, barring acts of US Congress, would erode the rule of law and increase risk of harm to patients.

Smith as Precedent

In 1990, the US Supreme Court held in Employment Division, Department of Human Resources of Oregon v Smith that states can legally deny unemployment benefits to personnel terminated for using illicit drugs during religious ceremonies.1 Justice Antonin Scalia’s majority opinion states that persons should not be exempt from neutral laws of general applicability that conflict with their religious beliefs; exempting such persons “would open the prospect of constitutionally required religious exemptions from civic obligations of almost every conceivable kind.”1 The majority opinion employs the valid secular policy test, which requires laws’ neutrality and general applicability, and which was first articulated in Reynolds v United States (1879).2 According to this test, the government need only show that its actions serve a legitimate state interest and do not target particular religious groups when its actions burden those groups.

Smith was controversial among religious conservatives, although many liberals also opposed the ruling. Prior to Smith, the federal government used a strict scrutiny test to evaluate religious exemption claims. This test prohibits the government from substantially burdening a person’s free exercise of religion, unless doing so is the least restrictive means by which government can pursue its compelling state interest. Smith was important because it signaled a change in law governing free exercise of religion by using the secular policy test instead of the strict scrutiny test. This article contends that overturning this case’s precedent might result in clinicians claiming more religious exemptions, which, barring acts of US Congress, would erode the rule of law and increase risk of harm to patients.

Challenges to Smith

The outcry in response to Smith led Congress in 1993 to pass the Religious Freedom Restoration Act (RFRA),3 which restored the strict scrutiny test as a statutory standard for government actions that burden a person’s free exercise of religion. Since 1997, however, the RFRA has only applied to federal law.4 But this restriction has not stopped 21 states from using the RFRA as a model for state laws based on the strict scrutiny test.5 In states without their own laws, Smith, with its valid secular policy test, remains the relevant legal standard by which exemption claims are reviewed. Despite the presence of the RFRA and state laws modeled on it, Smith still serves as a powerful constraint on religious exemption claim proliferation. Notably, in some cases, courts appeal to Smith even in states with laws based on the strict scrutiny test.

Appellate courts have cited Smith to justify rejecting religious exemptions in Fulton v City of Philadelphia6 and Ricks v Idaho Contracting Board.7 Plaintiffs in both cases requested that the Court revisit Smith. In Fulton, Catholic Social Services (CSS), a foster care contracting agency, sought exemption from the city’s requirement to place children with all qualified families (eg, including lesbian, gay, bisexual, transgender, or queer families), which it stated would violate its religious beliefs. In June 2021, the Supreme Court held that Philadelphia violated CSS’ exercise of its First Amendment right by excluding CSS from the foster care program due to its refusal to place children with same-sex couples.6 In Ricks, Mr Ricks sought exemption from a law requiring contractors to register their social security numbers with the state; he believes social security numbers are “a form of the mark,”7 which violate his religious beliefs. The Supreme Court declined to review the case in June 2021.8

Overturning Smith

With a conservative majority’s record of expansive religious views,9,10,11,12Smith could likely be overturned by the Court, with numerous consequences in the health legal landscape.

Undermining the rule of law. The rule of law expresses general agreement among persons subject to law to behave according to public norms.13 Persons in positions of authority and power (eg, judges, school administrators, or clinicians) are subject to law, like all of us, and, also like all of us, can seek exemption from laws perceived as violating their personal preferences or ideologies. When we contemplate exempting persons with authority and power from a law’s applicability, it is nevertheless also necessary to consider such persons’ professional obligations. Health professionals specifically owe duties of care to patients and members of the public, many of whom are ill, injured, or otherwise reliant on those professionals’ responsiveness to their vulnerabilities. When health professionals seek exemptions, those whom they serve can be denied services or otherwise affected. Since overturning Smith would likely generate more religious exemptions, we argue that this consequence deserves ethical and legal consideration in terms of whether, when, and to what extent exemptions would violate the rule of law, exacerbate health inequity, or otherwise undermine the carriage of justice. Note that violating the rule of law might be legal (eg, a law that mandates presidential immunity from criminal investigation). Such a law would nevertheless be a violation of the rule of law.

When we contemplate exempting persons with authority and power from a law’s applicability, it is nevertheless also necessary to consider such persons’ professional obligations. 

There are legal means to mitigate some of these negative consequences. The Do No Harm Act, for example, would amend the RFRA to prohibit uses that harm third parties.14 The US Congress could also pass the Equality Act, which would broaden Title VII’s definition of public spaces to include “a good, service or program.”15 But even if Congress were to pass bills like these into law, the Court’s current makeup casts doubt on their long-term survival.

Risk of harm to patients, including discrimination. Incursion of harm violates the principle of nonmaleficence, of key importance to health professionalism.16 We argue that the plaintiffs in Fulton harmed children (by adoption delays) and prospective parents with the suit’s implicit messaging that sexuality or gender identity confers parental fitness. Although the Supreme Court declined to overturn Smith in Fulton because the city’s law was outside the scope of Smith,17 there are petitions currently pending before the Supreme Court requesting that it overturn Smith,17 which, if successful, could alter the health legal landscape for the worse through harms of service denial, inferiority messaging,18 and discrimination.

Professional complicity as a source of harm. Depending on one’s point of view, overturning Smith could be positive. Nonmaleficence, after all, cuts both ways: if persons with authority and power are harmed by a legal requirement to act in a way they deem immoral, how should this be considered? We respond again with a focus on clinicians’ professional obligations. Credentialed and licensed by states to offer legal, clinically indicated, and publicly regulated health services to persons in need, clinicians are obliged to prioritize the interests of people they serve.19 We suggest that even when carrying out one’s professional duties poses a risk of harm, ethically and legally, the interests of the most vulnerable patients should be prioritized.20

Securing access to care. Individuals’ experiences of complicity matter and should have ethical and legal heft but not at the expense of patients’ access to legal, clinically indicated, and publicly regulated health services, which the federal government has a strong state interest in securing and protecting. Because we can plausibly expect that a post-Smith world would significantly reduce many patients’ access to care, state interest in securing patients’ access to care should become stronger “as the size and the number of businesses seeking exemption expands.”21

Liberty, autonomy, and respect. Smith is seen by some as disrespectful of persons with religious views that should be regarded as freely expressible in liberal pluralistic societies. But requiring health professionals to act against deeply held beliefs (eg, religious beliefs) does not necessarily violate their personal or professional autonomy.22 Expressing respect for fellow citizens’ views that differ from one’s own requires tolerance for those views and civility toward persons who hold those views. We suggest that it is neither intolerant nor uncivil to require clinicians to execute professional roles granted in fiduciary trust by states’ processes of credentialing and licensure. As John Stuart Mill argued long ago,23 government can justifiably limit individuals’ liberty to prevent harm or to protect others’ liberty; liberty rights are not, nor should they be, absolute.

Overturning Smith would very likely do more harm to many than good to a few if it results in preventing people who are ill, injured, or otherwise vulnerable from accessing legal, clinically indicated, and publicly regulated health services. A post-Smith landscape would likely exacerbate health inequity. This implication deserves attention from all who make law, either through legislation or jurisprudence.


  1. Employment Division v Smith, 494 US 872 (1990).

  2. Reynolds v United States, 98 US 145 (1879).

  3. Religious Freedom Restoration Act, Pub L No. 103-141, 107 Stat 1488 (1993).

  4. City of Boerne v Flores, 521 US 507 (1997).

  5. State religious freedom restoration acts. National Conference of State Legislatures. Accessed July 30, 2021.

  6. Fulton v City of Philadelphia, 593 US ____ (2021).

  7. Ricks v Idaho Contractors Board, 164 Idaho 689, 692, 435 P3d 1, 4 (Ct App 2018).

  8. Ricks v Idaho Contractors Board, cert denied (2021).

  9. Burwell v Hobby Lobby, 573 US 682 (2014).

  10. Our Lady of the Guadalupe v Morrissey-Berru, 591 US ____ (2021).

  11. American Legion v American Humanist Association, 588 US ____ (2019).

  12. Espinoza v Montana, 591 US ____ (2020).

  13. Waldron J. The rule of law. Stanford Encyclopedia of Philosophy. 2016. Accessed February 25, 2021.

  14. Do No Harm Act, HR 1450, 116th Cong (2019-2020). Accessed February 25, 2021.

  15. Equality Act, HR 5, 116th Cong (2019-2020). Accessed February 25, 2021.

  16. Beauchamp T, Childress J. Principles of Biomedical Ethics. 6th ed. Oxford University Press; 2009.

  17. Uddin AT, Slugh H. Why Supreme Court’s ruling in Fulton case is critical win for religious minorities. USA Today. June 21, 2021. Accessed July 30, 2021.

  18. Anderson E, Pildes, R. Expressive theories of law: a general restatement. Univ PA Law Rev. 2000;148(5):1503-1575.
  19. Goodin RE. Protecting the Vulnerable: A Reanalysis of Our Social Responsibilities. University of Chicago Press; 1985.

  20. American Medical Association. Opinion 8.3 Physicians’ responsibilities in disaster response and preparedness. Accessed June 28, 2021.

  21. Laycock D. Religious liberty and the culture wars. Univ Ill Law Rev. 2014;(3):839-880.

  22. Greenblum J, Kasperbauer TJ. Forget evil: autonomy, the physician-patient relationship, and the duty to refer. J Bioeth Inq. 2018;15(3):313-317.
  23. Mill JS. On Liberty. Cambridge University Press; 2011.

Editor's Note

Background image by Samantha Welker.


AMA J Ethics. 2021;23(11):E864-868.



Conflict of Interest Disclosure

The author(s) had no conflicts of interest to disclose.

The viewpoints expressed in this article are those of the author(s) and do not necessarily reflect the views and policies of the AMA.