Religious Exemption, LGBT Rights, and the Social Construction of Harm and Freedom

Abstract

In this article, we examine how courts make decisions in religious exemption cases that implicate LGBT rights in a wide range of contexts including education, employment, and medical care. Through an in-depth qualitative analysis of 50 federal cases decided between 1990 and 2020, we demonstrate a shift in how anti-LGBT sentiment is expressed by parties bringing religion-based claims—from a broad condemnation of LGBT identity to a narrow condemnation of same-sex marriage—and find that courts are more likely to rule in favor of the latter. We show how courts construct competing understandings of harm and religious freedom depending on the context of the case and whether the setting is deemed public or private. Our analyses shed light on the shifting and competing meanings about religion, sexuality, and public life in the United States.

INTRODUCTION

In June 2021, the United States Supreme Court announced its ruling in Fulton v. City of Philadelphia (2021), a case that centers on Catholic Social Services (CSS), an agency that brought a lawsuit after the city barred CSS from its foster placement program. The city barred the agency after learning of its policy to refuse to license same-sex couples to be foster parents, which, the city pointed out, violated Philadelphia’s nondiscrimination laws. In the lawsuit against city officials, however, CSS claimed the city’s actions violated its constitutionally protected rights. Under the Free Exercise Clause of the First Amendment, CSS claimed that it has the right to act in accordance with its sincerely held religious beliefs that oppose same-sex marriage. The federal district court disagreed, ruling in favor of the city. After the United States Court of Appeals for the Third Circuit affirmed the lower court’s ruling, CSS appealed to the Supreme Court. The Supreme Court ruled in favor of CSS, but issued a narrow ruling that has limited implications for future cases (Lane-Steele 2021).

Fulton is just one of the legal battles making headlines in recent decades where the courts have decided on LGBT rights in the United States. Opponents of such rights have argued that they should be exempt from adhering to nondiscrimination laws related to sexual orientation and gender identity based on the claim that to do so would violate their religious beliefs. Exemption claims for wedding-related services have garnered significant media attention, such as the highly publicized legal battle regarding a baker’s right to refuse to bake a wedding cake for a same-sex couple and a graphic designer’s right to refuse to design wedding websites for same-sex couples. ¹ The Fulton case focuses on foster care, but as our analysis will show, other litigation has considered religious exemptions related to mental and physical health care, employment, education, and a wide range of scenarios centering around public life, including sharing a bathroom with a transgender person or attending a company’s diversity seminar about gays and lesbians in the workplace.

Scholars and popular commentators alike have observed that these lawsuits measure the pulse of social change as American society moves to increasing acceptance of LGBT identity and relationships and as more Americans leave formal religious institutions (Fetner 2008; Baker and Smith 2009; Cooper et al. 2016; Lewis et al. 2017). Despite the many lawsuits that have been brought, the Supreme Court has yet to provide a definitive answer to whether religious exemptions must be granted to litigants opposed to adhering to nondiscrimination laws related to sexual orientation and gender identity (Minow 2007; Velte 2021). Amidst this legal uncertainty, empirical data can demonstrate how courts have navigated the broader social tensions over religion and LGBT identities and competing understandings over the meaning of discrimination, harm, and difference in public life. Drawing from content analysis of 50 court opinions, this article examines all federal cases that involve a religion-based claim used to oppose LGBT people or rights from 1990—the year the Supreme Court significantly limited constitutional religious exemption claims in its landmark Employment Division v. Smith (1990) decision—to 2020.

Through a qualitative analysis of how courts justify their decisions in cases where a religion-based claim is brought in opposition to LGBT rights, we answer a series of pressing questions: How is anti-LGBT sentiment expressed by parties bringing religion-based claims? How do courts articulate and interpret harm? And finally, how do courts set parameters for religious freedom in the context of anti-LGBT sentiment?

Our findings suggest three main themes in court opinions that grapple with conflicts over religious beliefs and LGBT people, relationships, and rights. First, since 1990 courts have confronted claims that conservative Christian beliefs fundamentally oppose LGBT identity, but more recent versions of this claim distinguish same-sex marriage as a unique violation of religious expression. We find that courts are more likely to rule in favor of religious litigants when the issue concerns same-sex marriage. Second, our analysis suggests competing definitions of harm across the cases in our sample, and decisions often turn on whether courts perceive religious people or LGBT people to be at risk. Finally, we find variation in how the court discusses religious freedom depending on the context of the case and whether the setting is deemed public or private.

Our findings illustrate how changing and competing discourses related to religion, sexuality, and discrimination have significant implications within American legal systems where sociocultural constructions of both harm and religious freedom can influence court decisions. Whereas in other cases implicating LGBT people, including litigation contesting sodomy laws, the courts have decidedly rejected arguments that sought to distinguish conduct from identity, our findings show that the courts appear to be more willing to accept that argument in religious exemption litigation. In other words, although courts ultimately rejected the argument in sodomy litigation that opposing same-sex sexual activities does not mean opposing gay people, our work indicates that some courts accept the argument in religious exemption litigation that opposing same-sex marriage does not mean opposing gay or lesbian couples. Moreover, our analysis indicates how judges construct competing understandings about harm insofar as there exists disagreement as to whether granting religious exemptions to nondiscrimination laws related to sexual orientation and gender identity results in any harm for LGBT people. Some courts even suggest that such religious exemptions may benefit LGBT people. Finally, our analysis sheds light on the dominance of Christianity, and in particular conservative Protestantism and Catholicism, in litigation involving sexuality, gender, and opposition to LGBT rights.

Religious Freedom in Context

Questions about the meaning of being able to freely exercise one’s religion in relation to sexuality have long been implicated in American law. One of the earliest religious liberty cases centered on the constitutionality of the federal government’s outlawing of polygamy in the Utah territory. ² The Supreme Court ruled unanimously that the law did not violate the First Amendment’s Free Exercise Clause, thus setting a precedent echoed in future court decisions, including Smith that religiously motivated conduct in public life can be regulated and restricted in order to promote the collective societal good (Hamilton 2009, 4). The landmark 1963 decision in Sherbert v. Verner reflected a Supreme Court that was more protective of free exercise rights. The court ruled that the Free Exercise Clause required states to grant religious exemptions to regulations that place burdens on religion, unless the state could meet the most demanding level of judicial scrutiny.

In the 1990 Employment Division v. Smith decision, the court rejected the Sherbert framework, holding that laws that are neutral and generally applied do not violate the Free Exercise Clause, even if they burden religion, and thus religious exemptions to such laws are not required. The case centered on Native American litigants who brought a lawsuit after the state denied their unemployment benefits because they had violated Oregon’s state drug laws. They argued that their religious-based belief and practice of smoking peyote as part of religious ceremonies necessitated an exemption from adhering to the state’s drug-testing requirement. The court disagreed. As Justice Scalia wrote in the majority opinion that ruled against the Native American litigants, the court reasoned that allowing exceptions to every state law or regulation affecting religion “would open the prospect of constitutionally required exemptions from civic obligations of almost every conceivable kind.” ³

In addition to the Free Exercise Clause, religious litigants also rely on the Establishment Clause and Free Speech Clause of the First Amendment to advance claims of their constitutionally protected rights being violated. Most relevant to the current analyses is the fact that the Supreme Court has interpreted the First Amendment to mean that the state should not interfere with the ability for religious organizations and places of worship to exist, with the ability for people to worship as they choose, or with the ability for religious organizations and places of worship to make decisions within their institutions that align with its religious (Lupu and Tuttle 2009; Heise and Sisk 2013; Levy 2017; Witte, Nichols, and Garnett 2022).

Religious exemption and discrimination claims are also articulated in relation to statutory laws, including Title VII of the federal Civil Rights Act of 1964, which outlawed employment discrimination based on religion (along with race, color sex, and national origin), and the federal Religious Freedom Restoration Act of 1993 (RFRA). Congress passed RFRA with near unanimous support in response to the Smith decision, and although the court struck down the law as applied to state policy (leaving intact the parts of the law that applied to federal policy), subsequently 25 states enacted similar religious freedom laws. The Smith decision frustrated and galvanized both progressive and conservative political actors and animated bipartisan support for RFRA. William P. Marshall (2018, 72) argues that those on the political left supported for the RFRA out of the desire to protect religious “minorities from majoritarian actions that unfairly disadvantaged them,” such as the Native American plaintiffs in Smith. Those on the political right were arguably less interested in the rights of Native Americans wanting to smoke peyote as part of a religious ritual but, rather, supported RFRA out of the desire to protect the right to practice religion, more broadly—in other words, “the substance of religion itself.”

Political and religious conservatives have subsequently used RFRA and state-level religious freedom laws to oppose a variety of civil rights (Gedicks and Van Tassell 2014; Corvino, Anderson, and Girgis 2017; Scott-Railton 2018). The success of LGBT activists during the 1990s and 2000s in passing state and local laws prohibiting discrimination based on sexual orientation and gender identity was met with religious-based opposition in and outside the courtroom, as our findings underscore (Fetner 2008; Stone 2012; Edmonson 2019). Since the Supreme Court legalized same-sex marriage in 2015, some states have passed religious freedom laws with explicit reference to sexuality and gender. One such law is Mississippi HB 1523, which protects persons who have “the sincerely held religious belief” that marriage “should be recognized as the union of one man and one woman” (sec. 2) to decide whether or not to provide services, including housing and employment, to LGBT people. The law draws from explicitly conservative Protestant beliefs to put forth specific regulations surrounding gender and sexuality. It defines, for example, “a man” and “a woman,” according to law as “an individual’s immutable biological sex as objectively determined by anatomy and genetics at the time of birth” (sec. 2).

In sum, while RFRA garnered bipartisan support in 1993, recent religious freedom legislation has been drafted and supported by conservative religious politicians (Hamilton 2015). Moreover, in Bostock v. Clayton County, the Supreme Court ruled in 2021 that Title VII of the 1964 Civil Rights Act, which protects against employment discrimination on the basis of sex, applies to sexual orientation and gender identity (Baumle, Badgett, and Boutcher 2020; Valenti 2021). Yet Justice Neil Gorsuch referenced RFRA in the majority opinion, suggesting that it might act as a “super statute” and necessitate religious exemptions for employers (Burke and Kazyak 2020). Courts will continue to determine whether nondiscrimination laws protecting LGBTQ people in not only employment but also public accommodations, housing, and health care must offer religious exemptions and what those exemptions might be.

LGBT Rights, Christian Ideology, and the Law

As sociolegal scholarship has demonstrated, social activists look to law as an important site to enact social transformation and seek legal changes in terms of both tangible material benefits and symbolic cultural meanings (Barclay, Bernstein, and Marshall 2009; McCammon and McGrath 2015; Lehoucq and Taylor 2020). Given the symbolic power of law, as groups contest their legal rights in court, they also are contesting their imagined place in American society. Scholars have shown how activists’ success in changing law can depend on the way they frame the issue and the discourses they use to discuss the issue in their advocacy (McCammon et al. 2007; McCammon et al. 2008; McCammon 2012). Scholars have also analyzed processes within the courts that point to a myriad of factors that can influence judicial decisions including cause lawyers and the solicitor general (McGuire 1998; Bailey, Kamoie, and Maltzman 2005; Barclay and Fisher 2006; Boutcher 2013; Cummings and Sandefur 2013) judges’ political ideology (see Segal and Spaeth 2002; Pacelle, Curry, and Marshall 2011; Harris and Sen 2019), and legal precedent (Hansford and Spriggs 2006). Scholars have also analyzed how judges advance their own frames and discourses (Ziegler 2010; McCammon et al. 2018; Murray 2019; Akbari & Vogler 2021; McCammon and Beeson-Lynch 2021).

Drawing from these bodies of literature, we theorize judicial decisions as important sites of institutional meaning insofar as judges are responding to—and either legitimating or discrediting—the claims about an individual or social group’s legal rights heard in court (Ferree 2003; Hall and Wright 2008). Specifically, given these bodies of literature, we theorize that the cases in our study provide insight into institutional understandings about the place of religion, gender, and sexuality in American life. Looking at discourses that appear in religious exemption cases tells the story of when religious litigants and their lawyers are successful in convincing the court that they have been wronged. In the context of our sample, it is religious (mainly conservative Christian) individuals claiming they have been wronged by adhering to nondiscrimination laws or policies. Thus, looking at discourses that appear in religious exemption cases can also shed light on the meaning that individuals experience in their daily life in relation to religion and sexuality (see Ammerman 2022).

When it comes to LGBT people and rights, Americans have experienced a dramatic shift toward acceptance (Whitehead 2010; Fetner 2016; Kaufman and Compton 2021). Most of those affiliated with the largest religious traditions in the United States are accepting of homosexuality, same-sex marriage, and nondiscrimination protection for LGBT people (Fuist, Stoll, and Kniss 2012; Murphy 2015; Bibi 2020). In virtually every major Christian denomination active in the United States there have been efforts to implement church policies and practices that welcomed gay and lesbian members and support gay rights under the banner of Christian social justice (Bean 2014; Moon and Tobin 2018; Coley 2020). Though the broad arc bends toward acceptance, it happens unevenly (Burke, Kazyak, and MillerMacPhee 2022; Young 2021). For instance, integrating religious identities is often not possible for those whose gender identity is nonbinary (Sumerau, Cragun, and Mathers 2016; Darwin 2020). Support of white evangelical Protestants, a group that historically opposed LGBT rights, for LGBT rights remains lower compared with that of other religious groups (Fetner 2008; Burke 2016; Herman 2000; Public Religion Research Institute 2020; Jones et al. 2021). In articulating their opposition to LGBT rights, this group has shifted their rhetoric away from biblical sources in favor of nonreligious language such as science and medicine and have reduced references to the personal morality of being LGBT (i.e., homosexuality is sin) and increased references to tolerance (i.e., love the sinner; Thomas and Olson 2012).

Apart from shaping individual beliefs, religion plays a broader ideological role for political conservatives. Braunstein and Taylor (2017, 35) find, for example, a large number of nonreligious members of the Tea Party Movement who still believe that America is a Christian nation. They argue that instead of understanding the Tea Party as a religious movement, it is more accurate to understand Christian nationalism as a way for conservatives—regardless of religiosity—to create boundaries between themselves and political others. As Whitehead, Perry, and Baker (2018) have established, the intersection of race (whiteness) and religion (conservative Protestantism) construct Christian nationalism through an ethnoracial and religiously based hierarchy (see also Butler 2021). Miller (2020) offers an empirical example of this in the courts in observing that Jewish and Muslim groups have less success than Christian congregations when challenging zoning laws in New York City (see also Aziz 2021).

Legal scholars argue that court decisions often reflect how Protestant values are taken for granted as universal and widely shared (Richardson 2006; Carbone 2007; Barringer-Gordon 2010; Dubler 2011; Johnson 2011; Pruitt and Vanegas 2015; Sullivan 2020). This is especially apparent in cases related to sexuality, gender, and LGBT rights (Jakobsen and Pellegrini 2004; Canaday 2009; Cott 2009; Adam and Cooper 2017; Frank, Moreton, and White 2018; Sullivan 2020). In upholding sodomy laws in the 1986 Bowers v. Hardwick decision, the Supreme Court noted that failure to do so would be to “cast aside millennia of moral teaching.” Even cases that appear to affirm sexual freedoms and LGBT rights validate religious opposition. For instance, in Obergefell v. Hodges, the Court declared “[i]t must be emphasized that religions … may continue to advocate with utmost, sincere conviction that … same-sex marriage should not be condoned” (sec. 4, para. 9). The decision in Bostock v. Clayton County (2020) notes that extending Title VII protections to include sexual orientation and gender identity “may require some employers to violate their religious convictions” and further underscores how taken-for-granted Christian beliefs inform legal discourses around gender and sexuality (see Jakobsen and Pellegrini 2004; Jakobsen 2020; Sullivan 2020).

METHODS

This article draws from a database that includes 1,281 federal court cases between 1990 and 2020 that involves a claim of religious discrimination or religious exemption. We limit our sample to federal level decisions from either US Circuit Courts or the US Supreme Court. To identify our sample, our research team gathered case summaries and opinions from Nexis Uni using a series of search terms: “religious exemption,” “religious discrimination,” “free exercise,” and “RFRA.” ⁴ Our search terms allow us to capture a wide range of scenarios in which people find their interactions in public life somehow at odds with their religious beliefs. We analyze cases where plaintiffs base their claims on a range of constitutional and legal provisions—including the First Amendment’s Free Exercise Clause, Title VII, the federal Religious Freedom Restoration Act, and state religious freedom laws. From our database, we selected a subsample of cases involving LGBT people or rights using the search terms “gay,” “homosexual,” “lesbian,” “transgender,” “bisexual, “sexual orientation,” and “gender identity.” The first two authors then reviewed each case to determine whether it was appropriate for inclusion in the sample. For this paper, we report on findings from 50 cases involving a free-exercise claim or religious discrimination claim that opposes LGBT people or rights (we describe these with the shorthand, “anti-LGBT cases”). These make up the majority of those cases involving LGBT people, where the overall database includes only 12 cases where a religion-based claim is used to affirm or defend LGBT people or rights. ⁵

To analyze our sample, we used quantitative and qualitative coding. For the quantitative portion, we used Qualtrics survey software for members of the research team to systematically code each case on 80 dimensions, including descriptive variables about the case, variables about the social identities of who was involved in the case, variables about the legal bases for the plaintiff and defendant’s claims, and variables about the outcome of the case. ⁶ We conducted an interrater reliability analysis (Cohen’s Kappa) of the data to assess coding consistency and reliability (Hallgren 2012). We did so by pulling a sample of 75 cases, recoding them (with a unique coder), and analyzing the level of agreement. Across variables, we find that agreement ranges from substantial to perfect (k = .61—1.0). To perform qualitative analysis, the first two authors used Atlas.ti to code each case and capture the interpretations that emerge in court opinions. We began with open coding by reading the cases and taking notes on interesting themes (Emerson, Fretz, and Shaw 1995). We generated a list of 25 codes that we then used to code each case. The research team met weekly to discuss the coding of cases. After coding, we wrote analytic memos that linked themes, which were developed into our findings section. Given that we analyzed federal court opinions written by judges, our findings reflect reference to judges’ descriptions and discourses. In other words, we reference judges’ summaries about the case details that appear in the court opinion, including the religious beliefs of the litigants and judges’ own reasoning and descriptions about their positions.

Table 1 displays the characteristics of the 50 federal cases. Most of these cases were decided between 2001 and 2010 (n = 27), and decisions were nearly evenly distributed across circuit courts (n = 7–10), with two decisions from the Supreme Court. The majority of cases involved plaintiffs who were individuals (n = 43) and defendants who were government agencies and/or parties representing the government in their official capacities (n = 36). Cases most often involved education (n = 15), employment (n = 19), and medical care (n = 9). In terms of the plaintiff’s religion, the majority are Protestant (n = 22) or Christian without further specification (n = 18). The other cases involved parties who are Catholic (n = 3) or whose religion was coded as “other,” a single case involving the new religious movement, Thelema. In six cases, religion is unknown (that is, not referenced by the court or publicly available). The majority of courts (n = 35) sided against the party bringing the religion-based claim, whereas in fifteen of the cases, the courts sided with the party bringing the religion-based claim.

TABLE 1. Description of federal cases involving a religion-based claim that is anti-LGBT

Note: Religion is for the party bringing the religion-based claim. In a small number of cases where both parties bring a religion claim, we code for the party initiating the lawsuit. Percentages do not add to 100%, as some cases had multiple arenas and categories for plaintiffs and defendants.

FINDINGS

Opposing LGBT Identities versus Relationships

Despite the relative uniformity when it comes to Christian representation within these cases, we find variation in how religious beliefs about LGBT identities and relationships are articulated in the courts. In a majority of the cases (n = 41), religious litigants disparage homosexuality writ large, for instance, explicitly condemning gay and lesbian sexuality and transgender identity in broad terms. We find consistent representation of this theme over time across the courts. In contrast, in a minority of cases, religious litigants narrow their focus to oppose same-sex marriage specifically (n = 9)—and in fact, also emphasize their acceptance of gays and lesbians and distance themselves from broad condemnations of same-sex sexuality. The narrow focus on opposition to same-sex marriage appears in cases starting in 2001, suggesting a new tactic in litigation that precedes the federal legalization of same-sex marriage but that also reflects changing discourse and acceptance of LGBT identity at the beginning of the twenty-first century (Fetner 2008; Dorf and Tarrow 2014; Zilis and Borne 2021).

Examples of broad condemnation emerge in cases that involve religious parties seeking to have nondiscrimination laws ruled unconstitutional. For instance, after the New Jersey legislature expanded its nondiscrimination law to include sexual orientation as a protected class in 1992, two churches and a pastor brought a lawsuit, Presbytery of New Jersey v. Whitman (1996), challenging the constitutionality of that expansion, arguing that their religious belief “requires them to condemn homosexuality” (4, para. 4). Among the activities that the churches and pastor feared they would no longer be able to do include “lobby against them,” “circulate literature condemning them,” “encourage, aid and abet discrimination … against homosexuals,” “refuse to knowingly buy from, contract with or otherwise do business with persons on the basis of that person’s homosexual … practices,” and “refuse to employ any individual who is practicing … homosexuality” (8, para. 4).

Other cases involve religious litigants who do not seek to strike down entire nondiscrimination statutes but who likewise condemn LGBT identity in broad strokes as evidenced in the following descriptions of people’s religious views about same-sex sexuality: “Homosexual activities violate the commandments contained in the Bible” ⁷ ; “immoral, sinful, perverse, and contrary to the teachings of the Bible” ⁸ ; “[t]he homosexual lifestyle is an abomination against God” ⁹ ; “God does not accept gays, they should not ‘be on earth,’ and they will ‘go to hell’ because they are not ‘right in the head’” ¹⁰ ; “[h]omosexuality is wrong and sinful” ¹¹ ; “[h]omosexuality is shameful” ¹² ; “[h]omosexuality as an abomination, loathsome, detestable, and an enormous sin” ¹³ ; “[h]omosexual behavior … [is] immoral and violate[s] God’s law.” ¹⁴

In contrast, in other cases religious litigants more narrowly focus their opposition to same-sex marriage and emphasize that their religious beliefs do not extend to how they think about LGBT individuals. For instance, in New Hope Family Services v. Poole (2020), a case involving an adoption and foster care agency bringing a Free Exercise Clause claim, we see evidence of this theme. Speaking about the Christian agency, the United States Court of Appeals for the Second Circuit wrote, “New Hope’s ministry is informed by its religious belief in the biblical model of marriage as one man married for life to one woman” (6, para. 2). The court noted that this religious belief informs their practices related to adoption: “Because of this belief, New Hope asserts that it ‘will not recommend or place children with … same-sex couples as adoptive parents’” (12, para. 3). However, the court also noted that the agency’s “religious views about marriage do not otherwise limit its ministry. In providing pregnancy counseling, New Hope routinely works with unmarried women and does so without regard to their sexual orientation” (12, para. 4). In other words, the court clearly delineates that although New Hope’s religious beliefs about same-sex marriage prohibit their working with same-sex couples seeking adoption (thus violating the state’s nondiscrimination law), it does not prohibit their working with single lesbian women seeking pregnancy counseling.

Other cases where these discourses emerge include Ward v. Polite (2012), where a therapist stressed “she had no problem counseling gay and lesbian clients” and only had requested “that she be allowed to refer gay and lesbian clients seeking relationship advice to another counselor” (7, para. 4). Others involved wedding-related services, for example where the plaintiff in Masterpiece Cakeshop v. Colorado Civil Rights Commission (2018, 4, para. 2), stressed, “I’ll make your birthday cakes, shower cakes, sell you cookies and brownies, I just don’t make cakes for same-sex weddings.” In all these cases, we see the framing of the religious belief in opposition to same-sex marriage, not gay identity.

Given the law, courts do not question the sincerity of religious-based beliefs about LGBT identities and relationships (see Sullivan 2020). Yet we find that courts are more likely to rule in favor of the religious side in cases that center only on narrow religious-based condemnation of same-sex marriage (courts ruled with the religious side in 50 percent of these cases) compared with cases that center on broad religious-based condemnation of gay, lesbian, or transgender identity (courts ruled with the religious side in only 24 percent of these cases). Moreover, as we describe in the next section, courts can discursively construct different understandings about harm—harm related to either nondiscrimination protections for LGBT people or religious people—depending on the way that the religious beliefs are articulated.

Shifting Threats, Shifting Harms

When religious beliefs are articulated as broad condemnations of same-sex sexuality or transgender identity, judges construct an understanding that nondiscrimination protections might be undermined, or harm might occur to LGBT people, should they rule in favor of the religious litigant. Yet when religious beliefs are articulated more narrowly, judges are more likely to suggest that LGBT people may actually benefit (or at the very least not be affected) should they rule in favor of the religious litigant. Additionally, in cases that involve same-sex marriage, we see some judges articulate an understanding of (Catholic and Christian) religious people opposed to same-sex marriage as a minority group out of sync with the majority and thus in need of protection.

Take for instance the case of EEOC v. R.G (2018), which involved a transgender woman, Aimee Stephens, who brought a Title VII employment discrimination claim after being fired from her job at a funeral home following her gender transition. The funeral home director used RFRA to argue that to employ a transgender person would be a substantial burden to his religion because it would “render him complicit” in supporting the idea that “sex is a changeable social construct rather than an immutable God-given gift” (12, para. 2). Here we see the articulation of a religious belief aligned with a broad rejection of transgender identity. The United States Court of Appeals for the Sixth Circuit ruled in favor of the transgender woman and argued that she “has and would suffer substantial harm if we exempted the Funeral Home from Title VII’s requirements” and that the government has an interest in “eradicating and remedying such discrimination” (29, para. 2). In contrast, the court did not believe that the funeral home director would experience harm as a result of being required to adhere to nondiscrimination laws (and employ a transgender woman) because “as a matter of law, tolerating Stephens’s understanding of her sex and gender identity is not tantamount to supporting it” (26, para. 2).

In other cases in which religious litigants condemn gay and lesbian identities broadly, the courts emphasized the harmful effect on gay and lesbian people. In Peterson v. Hewlett-Packard Co. (2004), the United States Court of Appeals for the Ninth Circuit rejected an employee’s Title VII claim against Hewlett Packard. Richard Peterson opposed the company’s campaign to promote diversity through a series of posters depicting current employees, including a gay employee. Because he believed that “that “homosexual activities violate the commandments contained in the Bible and that he has a duty ‘to expose evil when confronted with sin,’” he posted in his work cubicle biblical passages condemning homosexuality “that were large enough to be visible to those passing through an adjacent corridor” (4, para. 3). The court argued that allowing the biblical scriptures to be posted would harm LGBT workers and agreed that Peterson violated the company’s harassment policy as the messages were “intended to demean and harass” (9, para. 2).

In contrast to decisions that seek to protect LGBT people, some courts imagine religious litigants opposed to same-sex marriage as a minority group whose rights need protection. The courts reason that they are harmed by being required to adhere to nondiscrimination laws and policies. Courts assume the majority of Americans are progressive in their attitudes and treatment of LGBT people and thus granting religious exemptions has little effect or harm. In these decisions, religious beliefs are positioned as static and monolithic as opposed to societal beliefs that are positioned as ever changing.

In New Hope Family Services, the Second Circuit Court ultimately allowed New Hope to be exempt from adhering to the state’s nondiscrimination law. They reasoned that an exemption does not harm LGBT people wanting to adopt because New Hope refers married same-sex couples to other agencies. Moreover, the court also noted, “[t]he existing record reveals no complaint from any referred couple. Nor does it indicate that any couple was unable to adopt as a result of referral” (31, para. 3). In contrast, the court stressed the harm that the religious agency would incur should it be required to certify same-sex couples as adoptive parents: “It is no small matter for the State to order the closure of a privately funded, religious adoption ministry that has, over 50 years of authorized operation, successfully placed approximately 1,000 children in adoptive homes, particularly when there is no suggestion that any placement was not in the best interests of the adopted child” (31, para. 1). Here religious adoption agencies are positioned as threatened by the government. The court opinion also points to the fact that “since [the law] took effect, several voluntary faith-based authorized [adoption] agencies” could be implicated. These include “several Catholic providers, a Jewish provider, an LDS provider, and a Muslim provider” (17, para. 3).

The Meaning of Religious Freedom

We coded cases based on whether religious-based claims centered on the right of public expression or the right to an exemption or accommodation based on privately held religious beliefs. Although we found clear trends when it came to the outcome of court cases based on whether religious litigants opposed LGBT identity writ large or specifically same-sex relationships, we found no consistent pattern in whether courts are more likely to side with religious litigants who frame their case around exemption compared with expression. We coded fourteen cases as exemption, and in three of those cases (21 percent) the court ruled in favor of the party bringing the religion-based claim. We coded thirty-six cases as expression, and in eleven of those cases (30 percent) the court ruled in favor of the party bringing the religion-based claim. Although it does not appear that arguments related to expression or exemption correlate with specific outcomes in the courts, our qualitative analysis suggests that the public/private distinction informs how courts interpret the meaning of harm in the context of religious freedom and LGBT rights.

First Amendment precedent requires courts to treat churches and religious organizations as exceptional spaces when it comes to these disputes (Hamilton 2009; Lupu and Tuttle 2009). When cases involve a religious institution, the court is likely to emphasize harm in relation to religious expression and to advance a broad understanding of religious freedom. For instance, Bryce v. Episcopal Church in the Diocese of Colorado (2002) centered on a lesbian woman and her partner who brought a lawsuit against her employer, an Episcopal church, after one of them was fired from her job as a youth minister following their commitment ceremony in 1998. The lesbian couple alleged harm (harassment) as a result of church officials and members’ discussion of homosexuality including that “[h]omosexuals are promiscuous, suffer odious diseases, are engaged in sin, and are unfit to work with children” (7, para. 1). The United States Court of Appeals for the Tenth Circuit ruled in favor of the church, emphasizing the importance of church autonomy: “When a church makes a personnel decision based on religious doctrine, and holds meetings to discuss that decision and the ecclesiastical doctrine underlying it, the courts will not intervene” (13, para. 3). In other words, the church is an arena in which discussions of religious doctrine can occur without judicial oversight. For the court to “insert itself into a theological discussion about the church’s doctrine and policy towards homosexuals” (6, para. 2) would run counter to the importance of church autonomy.

Courts are more likely to find harm to LGBT people when religious litigants express their religious-based anti-LGBT beliefs in public settings. Given legal precedent, courts do not consider state regulation of religiously motivated public conduct as harmful to religious litigants, so long as that regulation serves the collective societal good (Hamilton 2015). For instance, in Piggee v. Carl Sandburg College (2006), Martha Louise Piggee brought a lawsuit including a Free Exercise Clause claim against Carl Sandburg College, a public community college where she worked as a cosmetology instructor. She was not rehired following the semester that she gave a gay student in her course two pamphlets with messages about why “homosexuality is an abomination” (3, para. 3) and the importance of repenting. The college determined that her “proselytizing in the hopes of changing Mr. Ruel’s sexual orientation and religious beliefs” (4, para. 3) had violated their anti-harassment policy. Although Piggee argued she had the right to express her religious views about homosexuality while at work, the United States Court of Appeals for the Seventh Circuit disagreed, writing, “[w]e see no reason why a college or university cannot direct its instructors to keep personal discussions about sexual orientation or religion out of a cosmetology class or clinic” (7, para. 4). The court framed discussions about religion as “personal” and more suited for conversations that occur outside of the public workplace. Moreover, the Seventh Circuit addressed the harm it saw resulting from Piggee’s expressions and noted the harassing effect they had on the gay student.

Similar discourses about harm in relation to distinctions between public and private emerge in expression cases centering on health care. For instance, in Walden v. CDC & Prevention (2012), a therapist alleged harm for not being able to express her religious-based opposition to same-sex sexual relationships when referring gay and lesbian clients to a different therapist. Walden brought claims including violations of the Free Exercise Clause, Title VII of the Civil Rights Act of 1964, and RFRA. After her supervisors advised her to simply refer clients without expressing why, Walden replied that would be “unfair” and “dishonest” explaining, “[i]t seemed unfair that [a client] was able to talk about being gay and lesbian, and yet I couldn’t freely talk about me and my religious beliefs, or being Christian…. To me, it’s about honesty. If she can be honest … I should be honest about why I’m transferring her” (11, para. 3). Although Walden argued she was fired because of her religious beliefs, the United States Court of Appeals for the Eleventh Circuit disagreed, noting that her employer did not take issue with her religious-based desire to refer gay and lesbian clients but only to the manner in which she made those referrals.

Likewise, in a case involving a challenge to a public high school’s antidiscrimination policies, Harper v. Poway (2006), the Ninth Circuit explicitly discussed the harm LGBT students would experience should individuals be able to express their religious-based opposition to LGBT identities in the public setting of schools. The case centers on a student, Tyler Chase Harper, who wore a shirt to school on the day that the school’s gay–straight alliance sponsored a “Day of Silence” that read as follows: “Be ashamed, our school embraced what God has condemned” on the front and “Homosexuality is shameful Romans 1:27” on the back. The school had Harper stay in the main office and not attend classes because wearing the shirt caused a “disruptive effect upon the education environment” and “effect on the other students” (14, para. 2). In response Harper brought a lawsuit that included claims of violation to the Establishment Clause and the Free Exercise Clause. The court ruled against him. It pointed to the harm that Harper’s shirt caused gay and lesbian students, who are “members of minority groups that have historically been oppressed, subjected to verbal and physical abuse, and made to feel inferior” (16, para. 2). Moreover, the court rejected the argument that because the school allowed students to voice affirmation of gay and lesbian identities (via a “Day of Silence”) that it should also allow students to voice condemnation of those identities because “public schools may permit, and even encourage, discussions of tolerance, equality and democracy without being required to provide equal time for student or other speech espousing intolerance, bigotry or hatred” (22, para. 1). The court argued that Harper’s ability to express his religious beliefs was not harmed given that he has “numerous locations and opportunities” to express those beliefs.

Yet in some cases involving public settings, the court does emphasize harm to religious litigants who feel unable to express their beliefs. For instance, in another case involving a public school, Saxe v. State College Area Sch. Dist. (2001), the United States Court of Appeals for the Third Circuit ruled in favor of families who sued their school district because they believed its nondiscrimination policy inhibited their ability to speak out against homosexuality. The families “openly and sincerely identify themselves as Christians. They believe, and their religion teaches, that homosexuality is a sin. Plaintiffs further believe that they have a right to speak out about the sinful nature and harmful effects of homosexuality. Plaintiffs also feel compelled by their religion to speak out on other topics, especially moral issues” (8, para. 1). The court determined that the school’s policy was “unconstitutionally overbroad” (16, para. 3).

In a similar case, the Seventh Circuit ruled in favor of Christian Legal Society, a student organization at Southern Illinois University, which brought a Free Exercise Clause claim after the university revoked its official status as a registered student organization. The university determined that the group’s membership policy, which excludes anyone who “engages in or affirms homosexual conduct” (13, para. 1), violated the university’s nondiscrimination policy. The court disagreed. In their opinion in Christian Legal Society v. Walker (2006), the court discussed the harm that the religious student organization would suffer should they be required to adhere to the nondiscrimination policy and asked, “[w]ould having to admit homosexuals inhibit the group’s ability to send their message that condones homosexuality?” (8, para. 4). The court answered affirmatively and wrote, “[t]here can be little doubt that requiring CLS to make this change would impair its ability to express disapproval of active homosexuality” (9, para. 1). Here, restrictions on public expression are seen as detrimental to the religious student organization. The court sees little harm to LGBT people, because as they argued, the student organization’s policy is based on “behavior rather than status” and that “persons ‘who may have homosexual inclinations’” can in fact become members so long as they do not act or affirm those inclinations. Importantly, other court decisions have rejected this argument. ¹⁵

Courts that granted religious exemptions or accommodations have sometimes framed them as private and thus of little consequence to LGBT people. Take for instance the case of Ward v. Polite, which revolves around a graduate student, Julea Ward, who was expelled from a master’s counseling program after referring a client who sought counseling for same-sex relationship issues. The university stated that she was expelled for violating a code of ethics which prohibited students from discriminating against others. She claimed that the university had violated her First Amendment Free Exercise rights. The Sixth Circuit, which remanded the case for additional adjudication, took issue with the fact that the university did not have an exemption from its nondiscrimination code to allow students to make these kinds of referrals. Without such exemption, the university required Ward to “alter or violate her belief systems … as the price for obtaining a degree” (12, para. 2). In contrast, the court saw no harm inflicted on the gay client seeking therapy and in fact framed their decision as potentially beneficial to him, because “[t]he client never knew about the referral and perhaps received better counseling than Ward could have provided” (10, para. 2). Thus, the court signaled that religious exemptions could be viewed as a win–win for religious counselors and LGBT people alike.

In other accommodation or exemption cases, courts did not view denying accommodations as harmful to religious litigants. In Parker v. Hurley, parents sued their children’s school district for not allowing them to opt out of a curriculum that “recognized differences in sexual orientation,” (7, para. 6), a children’s book used in a second-grade classroom that depicted a gay married couple. They argued that the Free Exercise Clause protected their right to be excluded from the school’s recognition of diverse families. The United States Court of Appeals for the First Circuit rejected their claim: “There is no free exercise right to be free from any reference in public elementary schools to the existence of families in which the parents are of different gender combinations” (17, para. 1).

In other exemption cases, the courts not only rejected a claim that a refusal to offer an accommodation harmed religious litigants; they also emphasized the ways in which an accommodation undermined nondiscrimination protections for LGBT people. For instance, in Fulton, the Third Circuit rejected the Catholic Social Services argument that an exemption would cause no harm to LGBT people because “no same-sex couples have approached CSS seeking to become foster parents” (8, para. 2). The court wrote, “[t]he harm is not merely that ‘gay foster parents will be discouraged from fostering.’ It is the discrimination itself” (20, para. 2). The stakes were not the specific experiences of any particular same-sex couple but a more general harm caused by granting a religious accommodation to litigants opposed to adhering to a nondiscrimination policy.

DISCUSSION AND CONCLUSION

Our analyses shed light on the shifting and competing meanings about religion in relation to LGBT rights in the United States. We underscore how discourses related to religion, discrimination, gender, and sexuality that appear in courtrooms and judicial interpretations can be contradictory, nuanced, and changing (Baumle and Fossett 2005; Richman 2010; Volger 2016; Gash and Raiskin 2018; Kazyak et al. 2018; Phipps 2019). Specifically, our analyses illustrate a marked shift between 1990 and 2020 in how religious litigants articulate their beliefs about sexuality and gender. As opposition to same-sex marriage became a prominent issue for religious anti-LGBT litigants, they shifted the discourse from one based on a broad disparagement of gay and lesbian identity to one more narrowly framed around opposition to same-sex marriage. In their legal arguments, religious litigants often stressed that their beliefs about marriage did not translate to their beliefs about gays and lesbians more generally.

Importantly, we found that courts are more likely to rule in favor of cases when the religious belief is articulated narrowly and focused only on same-sex marriage. The distinction between marriage (i.e., being in a same-sex marriage) and identity (i.e., being gay) arguably parallels distinctions between sex (conduct, behavior) and identity (status) that emerged in litigation centered on other sexual rights, including sodomy laws. However, the Supreme Court eventually struck down that distinction. The court argued that discrimination based on sex (conduct, behavior) via criminalizing sodomy effectively resulted in discrimination on the basis of identity (status) for gay and lesbian people. In other words, the court reasoned it was not possible to be against same-sex sexual behavior and be supportive of gay and lesbian people. Yet in religious exemption cases involving beliefs about same-sex marriage, religious litigants have thus far had some success in arguing that it is possible to be against same-sex marriage and also be supportive of gay and lesbian people. Future research should assess whether this litigation strategy continues to be successful.

Our analysis of cases implicating LGBT people and religion illustrates how the courts construct two versions of American society. In one version, the courts see LGBT people primarily as victims of discrimination who need government protection. Thus, courts reason that offering religious exemptions or striking down nondiscrimination policies can be detrimental and harmful to advancing societal goals of tolerance and equality for people with minoritized sexual orientations and gender identities. The Ninth Circuit Court opinion in Harper v. Poway serves as an exemplar of this vision of American society. Here the court explicitly discussed gay and lesbian people as “members of minority groups that have historically been oppressed, subjected to verbal and physical abuse, and made to feel inferior” (16, para. 2). The other version of American society that courts construct sees the majority of Americans as progressive in their attitudes and treatment of LGBT people. Thus, courts reason that offering religious exemptions has little effect on LGBT people insofar as they can seek out service elsewhere, whether that is a different adoption agency or mental health counselor. Recall the Sixth Circuit decision in Ward v. Polite where the court speculated that the gay client would have “perhaps received better counseling” (10, para. 2) had the religious counselor been offered an exemption. Future research can address whether, why, and how these varied judicial framings translate and gain legitimacy outside of the courtroom (Ziegler 2010).

Across cases centering on expression and exemption, distinctions between public and private inform how courts construct meanings about harm and religious freedom. Two cases that involved a mental health counselor whose religious beliefs opposed same-sex marriage (Walden and Ward) exemplify this point. In Ward, the court construed the question of whether the therapist should be able to be exempt from working with gay clients as a private matter (“[t]he [gay] client never knew about the referral,” 10, para. 2). The court ruled in Ward’s favor. In Walden, the court considered whether the therapist should be able to express her religious beliefs to clients (recall that her employer had already given her an exemption and allowed her to refer gay and lesbian clients to other therapists). Walden’s employer raised concerns that such expression would make gay clients feel “judged and condemned” (7, para. 4). The court ruled in favor of her employer. In most expression cases involving individual plaintiffs, so long as regulations on public expression do not extend to restrictions on private beliefs, the court does not view those regulations as harming religious people. In cases involving religious organizations or institutions, however, courts are more likely to emphasize harm incurred by religious litigants—whether that be a church’s ability to discuss opposition to homosexuality in relation to firing a lesbian employee (Bryce) or a religiously affiliated adoption agency being able to adhere to its religious beliefs about marriage in relation to not working with same-sex married couples (New Hope).

Our findings underscore the discrepancy between the viewpoints of some religious litigants and the courts in relation to public expression of religious beliefs about LGBT identities and relationships. For many religious litigants in our sample, their beliefs are aligned with those of the plaintiffs in Saxe, who articulated that they “feel compelled by their religion to speak out … on moral issues” and they “believe that they have a right to speak out about the sinful nature and harmful effects of homosexuality” (8, para. 1). Thus, from their point of view, any restriction on their public expression is seen as a harmful attack on their religious beliefs. Yet courts often distinguish between public expression and private beliefs and do not see restrictions on the former as an attack on the latter. Future work can examine such discrepancies and tensions in other litigation with respect to how people might experience religion in their daily lives versus the courts’ conceptualizations (see Sullivan 2020; Ammerman 2022).

We also see courts construct religious litigants who are opposed to LGBT identities and relationships as a group in need of protection. Most notably, the court in New Hope Family Services lamented the threat that religious adoption agencies faced when they were required to adhere to nondiscrimination laws and work with same-sex married couples. The court painted this harm in broad strokes and implicated a range of religious agencies, including Christian, Catholic, Jewish, and Muslim organizations. This framing obscures how different religious groups garner drastically disparate levels of support both from the American public and from the courts when advocating for their religious freedom. Specifically, Jewish and Muslim Americans face lower levels of support compared with Protestant Americans (Miller 2020; Aziz 2021). Questions about the interplay between the framing and discourses that emerge in court compared with social movement groups warrant attention in future research. For instance, scholars should address whether, how, and why the claim that conservative Christians opposed to LGBT rights are a minoritized, oppressed group holds traction with the broader public and how LGBT activist groups respond (see Kazyak, Burke, and Stange 2018; Nash and Browne 2020). Likewise, future analyses can examine institutions beyond the courtroom to address how religious freedom claims are disputed and resolved outside of litigation (see Mayrl 2021). Future research can also analyze rulings on religious exemptions in relation to other protected categories codified in nondiscrimination laws. For instance, the Supreme Court has ruled against claims of religious exemption in the context of nondiscrimination on the basis of race, which raises questions about how courts reach different conclusions depending on the protected category in question (see Minow 2007; Velte 2021). Likewise, future work can address the degree to which religious exemption litigation opposing LGBT rights overlaps with the broader conservative legal movement (see Teles 2008).

Legal conflicts over religious freedom span far beyond LGBT politics and Christian identity (Volokh 2015). Yet our findings suggest that even for cases where the party bringing the religion-based claim is not Christian, their legal disputes typically implicate conservative Christianity. In other words, nearly all of the cases in our sample involve what is sometimes termed in the courts “Judeo-Christian” identity, a phrase invented in the early twentieth century to link American identity with conservative Christian beliefs (Gaston 2019). The dominance of Christianity, and in particular conservative Protestantism and Catholicism, across anti-LGBT religious freedom cases supports existing law and religion literature that finds Christian beliefs map onto legal decisions regarding gender and sexual minorities (Bernstein and Jakobsen 2010; Djupe, Lewis, and Jelen 2016; Coker 2018; Sullivan 2020).

At the same time, religion offers diverse strategies for litigation, some of which are progressive and pro-LGBT (Scott-Railton 2018; Burke, Kazyak, and Behrendt 2022). Future research would benefit from continuing to explore those cases that affirm and defend LGBT identities and rights and use religion-based claims to demonstrate how religion is not synonymous with conservative Christianity or conservative politics (Bernstein and Jakobsen 2010; Murphy 2015; Coley 2020; Sullivan 2020). Additional analyses could also compare religious freedom cases involving sexuality and gender rights and those involving other social issues to see what shared and divergent themes are present within the courts. Finally, future research should expand our focus on framing and discourses to address how other factors, such as the kinds of legal claims being brought and the relevant precedent (see Hansford and Spriggs 2006) and the ideology of the judges (see Pacelle, Curry, and Marshall 2011; Segal and Spaeth 2002; Harris and Sen 2019), also matter to these judicial outcomes. Answering such questions will continue to advance sociolegal scholarship related to religion and sexuality.