Saints, statues, and church-state separation

Rights and Responsibilities is a recurring series by Richard Garnett on legal education, the role of the courts in our constitutional structure, and the law of religious freedom and free expression.

I have observed that, when it comes to religious freedom and church-state relations – and contrary to some commentators’ complaints – the Roberts court has been conducting a “doctrinal clean-up,” moving the law “toward coherence and clarity, and better align[ing] it with American history, tradition and practice and with an appropriate understanding of judges’ capacities and the judicial role in a democracy.” In one particular case category, though – namely, disputes having to do with public religious symbols and officials’ religious expression – confusion and uncertainty remain. An ongoing controversy in the Massachusetts state courts provides an interesting and instructive illustration.

Briefly: The city of Quincy is building a new public-safety headquarters for its first responders. The city’s mayor commissioned two large statues of Saint Michael the Archangel and Saint Florian – widely regarded by Christians as the special patrons of police officers and firefighters, respectively – to be displayed on the new building’s facade. Eventually, after media coverage of the mayor’s plan, and of the statues’ price tag, several residents and activist groups filed a lawsuit contending that the statues’ purchase and display would violate the religious-freedom provision of the Massachusetts Constitution by, as the ACLU of Massachusetts put it, “imposing religious symbols upon all who work in, visit, or pass by the public safety building; by conveying the message that Quincy is exclusively a Catholic community and that non-Catholics do not belong or are less valued; and by excessively entangling the City with religion.” The local trial court agreed, and the litigation continues.

This is not a federal case, and it is not likely to reach the Supreme Court. State courts, generally speaking, do not need federal courts’ approval for their interpretations of state laws. Still, Quincy’s saint-statues fight is the kind of speech-and-symbols dispute that the justices have struggled with for more than four decades. How should the Roberts court deal with a case like this? What are the questions to ask, the tests to apply, the interests to balance, and the goals to pursue when judges – or, for that matter, officials and citizens – set out to determine whether decorating the police-and-firefighters building with the patron saints of police and firefighters violates the First Amendment’s rule against establishments of religion?

It should be clear that the mere facts that Saints Michael and Florian are, well, saints, and that saints are commemorated and venerated more in Roman Catholicism than in some other religious traditions do not compel a conclusion that Quincy has unlawfully set up an official faith. Public museums, for example, display religious images – often, images of saints – all the time. The city’s mayor is not purporting to, say, select the local bishop or mandate a local analogue to the King James Bible. Those who drafted and ratified the First Amendment knew what religious establishments looked like. The statue-display does not involve what some justices, following law professors Michael McConnell and Nathan Chapman, have called the “historical hallmarks of an established religion,” such as political control over religious doctrine, government-mandated church attendance, or legal punishments of religious dissenters.

There are no longstanding, canonical Supreme Court or other precedents suggesting that religious imagery necessarily amounts to an establishment of religion. For most of the nation’s history, the Supreme Court had nothing to say about the content and contours of the establishment clause’s prohibition, and it only got into the business of enforcing that prohibition in the context of our country’s longstanding and often bitter contests over the funding of Catholic schools. There is no evidence that, say, anyone in 1804 thought to bring a federal case about making “St. Louis” the capital of the Louisiana Territory. To be sure, there were debates and arguments concerning Sunday mail delivery, church incorporations, and Thanksgiving proclamations, but these did not end up in federal court.

That said, the Massachusetts trial court was candid about the limited relevance of history and hallmarks. To defer to the founders’ understanding, that court said, “would perpetrate the petty bigotries of the past.” Similarly, and for many years, the Supreme Court was reluctant to let either the original understanding of the text or the pervasiveness of a practice strictly limit the reach of the establishment clause’s commands. (The school-prayer decisions of the early 1960s illustrate this reluctance.) Instead, the court announced, developed, and for a time purported to apply the so-called “Lemon Test,” named for a 1971 parochial-school-aid case. According to this doctrinal gauntlet, government actions must have a “secular purpose”; they may not have the “principal or primary effect” of “advanc[ing]” religion; and they are invalid if they create or contribute to “excessive government entanglement” with religion.

Wholly apart from its shaky foundations in history, that three-part “test” was, from the beginning, a poor fit with the cases that started coming to the justices in the 1980s involving public displays and official expression. And so, in situations where the test’s straightforward application would produce a result about which the justices were squeamish – say, the outlawing of legislative chaplains or the abandonment of our nation’s motto – they simply ignored it.

Beginning in 1984, the late Justice Sandra Day O’Connor attempted an alteration that, she urged, would better tailor the test to cases involving expression rather than regulation or funding. Concurring in Lynch v. Donnelly – a 1984 case that rejected, after a light-touch, half-hearted application of Lemon’s test, an establishment clause challenge of a Rhode Island city’s Christmas display – she suggested that the right question to ask in these cases is whether a challenged government action “endorse[s]” religion and so “sends a message to nonadherents that they are outsiders, not full members of the political community[.]” And so, the justices’ – and in the hundreds of cases that followed, other courts’ – inquiry shifted to the semiotics and social meaning of manger scenes, candy canes, and Christmas carols.

This new “endorsement test” fared little better, despite additional refinements, adjustments, and dodges, than Lemon 1.0 had in terms of producing predictable and politically palatable results. In the 1989 case of Allegheny County v. ACLU, a badly fractured court eventually concluded that a menorah-and-tree pairing outside a county office building did not violate the First Amendment but a nativity scene inside a nearby county courthouse did. And in a 2005 pair of cases, decided on the same day, shifting majorities of the justices determined – with the nine-member court somehow producing 10 opinions – that a Ten Commandments display in a Kentucky courthouse was unlawful but one on the grounds of the Texas State Capitol was not. And so on. In the words of a 2022 amicus curiae, or “friend of the court,” brief, which surveyed dozens of endorsement-test applications, O’Connor’s charge to courts to enforce the Constitution’s ban on religious establishments by asking about subtle signals of exclusion and second-class citizenship is “unworkable and has sown disarray, confusion, and discrimination against protected religious speech and conduct.”

Today, it appears, a majority of the justices agree with this assessment. A few years ago, in American Legion v. American Humanist Association, the court rejected an establishment clause challenge to a century-old, 32-foot-tall war memorial cross on public property, without bothering or pretending to apply either the Lemon test or its no-endorsement makeover. That test, Justice Samuel Alito put it understatedly, “presents particularly daunting problems in cases . . . that involve the use, for ceremonial, celebratory, or commemorative purposes, of words or symbols with religious associations.” He emphasized, instead, the controlling relevance of “the passage of time.” And, most recently, in Kennedy v. Bremerton School District – the “praying football coach” case – Justice Neil Gorsuch wrote for the majority that, “this Court long ago abandoned Lemon and its endorsement test offshoot” for “[a]n analysis focused on original meaning and history.”

So, back to Saints Michael and Florian. The Massachusetts trial court opted, again, to stick, for state-law purposes, with Lemon and its progeny (although it remains to be seen if the state’s appellate courts will agree). But if we reject this approach, as the Roberts court has when it comes to interpreting and applying the First Amendment, do judges, officials, and citizens have anything more to go on than (something like) “old displays can stay but new ones are suspect”?

We should remember, for starters, that not all bad policies are unconstitutional, and constitutional litigation is not the only, or the ideal, way of aligning communities’ religious-display and public-expression policies in line with changing demographics and increasing diversity. Yes, the court is right to reject a judicial approach that privileges judicial ruminations and speculations about the psychological impact of symbols and speech. At the same time, all of us, not only judges, should – as Justice Elena Kagan once put it – treat “seriously the multiplicity of Americans’ religious commitments,” in keeping with the “practice of civic friendship” when we are making decisions about public symbols and spaces.

The court should continue its tidying up of establishment clause doctrine, and state clearly that the question for courts in cases involving symbols and expression is not simply whether the challenged display or practice is old or new. Keeping in mind the importance of humility when it comes to second-guessing decisions by communities and their politically accountable representatives and public servants, the judicially enforceable reach of the no-establishment rule should be limited to actions sharing the historical hallmarks of entangling religious establishments. Even if we recognize the importance of church-state separation, correctly understood, we thus shouldn’t think that this separation requires or authorizes federal judges to remove all religious elements from the public square.