The Establishment Clause: Its Original Public Meaning and What We Can Learn From the Plain Text

Modern times in church-state relations began in 1947 with the Supreme Court’s decision in Everson v. Board of Education.[1] The Justices in both the majority and the dissent said they were interpreting the Establishment Clause based on the intent of the founding generation. However, rather than examine the 1789 congressional lawmaking that directly led to the First Amendment, the Justices relied on the Virginia disestablishment experience from 1784 to early 1786, as well as the extended efforts of just two statesmen from among the founders, James Madison and Thomas Jefferson.[2] The devotees of this off-centered version of the pertinent history took on the appellation of “strict separationists.” Throughout the 1960s and 70s, the Virginia experience was more or less read into the Establishment Clause.

A rough collection of alternative theories concerning the relevant history and what it showed formed in opposition and travelled under the broad heading of “accommodationism.”[3] Although accommodationists never coalesced behind a single alternative interpretation of the Establishment Clause, they rallied behind allowing nonsectarian prayer to solemnize civic occasions, traditional government displays and other symbols with religious content, and funding for “nonsectarian” curricula and programs at religious schools, colleges, and charities. Accommodationism faced an uphill struggle. By the early 1980s, however, all but the most hardline separationists conceded that, as a matter of avoiding viewpoint discrimination that would violate the Free Speech Clause, private expression of religious content ought to be granted equal access to public forums.[4] And while accommodationists had only a few outright victories in the Supreme Court,[5] bright line separationism slipped to minority status upon the rise of Justice Sandra Day O’Connor’s no-endorsement test as an accretion to the three prongs of Lemon v. Kurtzman.[6]

These were the battle lines until the late 1990s when, as least as to government programs involving grants and other financial assistance, the presumption favoring separationists collapsed under the banner of government “neutrality” regarding religion.[7] The basic appeal of neutrality theory was that religious providers serve the common good (e.g., providing education, health care, or charitable services) and ought not to be discriminated against. Neutrality found historical support in the founding era’s upholding of “the right of private judgment” in matters of religious conscience. The relevant “private judgment” now being exercised lay with the parents of students who wanted to choose a religious school, welfare recipients freely selecting faith-based charities to deliver their welfare services, and patients desiring admission to religious clinics to secure government underwritten health services. It helped, of course, that neutrality theory could marshal to its cause the powerful rhetoric of “nondiscrimination” and “freedom of choice.”

To the extent the High Court relied on history to advance or oppose the foregoing interpretations of the Establishment Clause, it was professed to be a search for events and principles reflecting the original intent of the founding generation. This is now regarded as Old Originalism. It had a few basic problems. Lawmaking is a collective task, so there was no singular intent, and sometimes the various aims of the founders conflicted. Further, there was the problem of how long a timespan is relevant as one goes about collecting the lawmakers’ intent. For example, were the writings of Madison during an intense debate before the Virginia House of Delegates in 1784-1785 to be uncritically read into Madison’s work four years later as a member the U.S. Congress involved in composing what we now know as the First Amendment?

Jurisprudential conservatives have long urged an interpretation of the U.S. Constitution that is faithful to the time of its inception. As more defensible than original intent, they increasingly look to New Originalism. This is an interpretive principle that adheres to the original public meaning of the words on which the authors of the law finally settled. “Public” means that the final words of a law should mean today what literate people would have understood them to mean at the time the words were adopted. It is quite appropriate, for example, to consult dictionaries from the period of the lawmaking.[8]

It is now apparent that the American disestablishment story is far more multisided and complex than the story featuring only Madison and his Revolutionary Virginia.[9] And Jefferson was not even in the country when the Constitution and bill of rights were debated and adopted; he was at best a distant player, attending to American foreign interests in Revolutionary France as the First Amendment was birthed on this side of the North Atlantic.[10]

Reliance on the Virginia disestablishment experience, even when supplemented by Madison’s broader work, is inadequate to understanding the original public meaning of the Establishment Clause. Everson and its progeny are looking in the wrong place at the wrong time. The First Amendment is from an altogether different time (summer 1789) and source (Congress sitting in New York City) than the dramatic events in the Virginia legislature sitting in Richmond during 1784-1785. James Madison is the one common denominator, but his purposes and his power to shape the law emerging from these events were altogether different in the two instances. New Originalism moves the focus elsewhere.

Criticizing past mistakes is one thing. There is still the question before New Originalist interpreters: What was the original public meaning of “make no law respecting an establishment of religion”? And what evidence is informative about the events of the day that puts in context the meaning of these words when composed by members of the House and Senate during the First Federal Congress from June through September of 1789? And how does the state ratification of the federal amendments during 1789-1791 contribute to original meaning, especially given that states had to take the text of the bill of rights as presented and could only vote entire amendments up or down?

New Originalism looks at a narrower slice of the historical record, maintaining a laser-like focus on the September 1789 meaning of the final words of the Establishment Clause. Still, this interpretative theory requires some knowledge of a wider context to understand what the First Federal Congress was trying to do in settling on this text. Stated a little differently, the authors of a law choose their words to fit the task. How did those in control at the First Congress conceive of their task? And were there restraints on what they could do?

Answering these questions requires first going back and briefly exploring the task of the delegates to the 1787 Constitutional Convention, which is the topic of Part I. Then Part II takes up the task of the First Congress in composing and sending amendments to the states, and in particular the Establishment Clause of what became the First Amendment. In doing so, we look at some of the day-to-day debates in the House and Senate concerning the establishment question, not with the aim of determining the original intent of the framers, but with the aim of getting insight into the meaning of the words they chose to fit the task. Part III then hazards what we have learned concerning the original public meaning of the First Amendment text “respecting an establishment of religion,” and what we still do not know. Finally, no interpretive rule is required when the text alone is definitive. Thus, Part IV turns to consider what we can know from the grammar and plain text of the Religion Clauses. Whether one is an originalist or not, such a textual investigation allows us to put to bed some longstanding myths, such as the claim that the two clauses are in tension and sometimes conflict.

I.  The Overall Theory of (and Limitations to) the 1787 Constitution

As convention delegates began to gather in Philadelphia in May 1787, both religion and religious liberty were sensitive and sometimes contentious matters. But such disagreements were manageable if religion was left to the states. The Congregational church was still firmly established in all New England except Rhode Island, and South Carolina and Maryland were still working toward disestablishment of the Anglican church in the South. If the Constitution had granted to the federal government power over religion in any plenary sense, that could easily have prevented agreement in Philadelphia, and it certainly would have stirred enough trouble to prevent ratification of the Constitution by the designated minimum of nine states.

A republic’s constitution can do three things. First, it can organize the government’s frame, form offices and assign them competencies, and carefully diffuse authority among multiple departments to avoid concentrations of power. Second, it can define the relationship between the government, on the one hand, and the people and their nongovernmental organizations, on the other, including the vesting of select rights in the latter. And third, it can declare those first principles around which the body politic is drawn together and the nation-state is founded. However, a constitution need not do all of these things, nor do any of them in a comprehensive way. Certainly, the U.S. Constitution of 1787 sought primarily to accomplish only the first of these objectives in a thorough manner. A bill of rights was added two years later in apparent response to the second of these three tasks. However, the original Constitution’s near silence with respect to the nation’s founding principles was in large part calculated. In significant measure, the gap reflects the difficulty of achieving agreement on first principles at the Philadelphia convention. In the face of such disagreement, a common way to get contending parties to sign a single document is to avoid topics on which there is no hope of consensus. Religion was one of those topics. Moreover, religion was doubly easy to avoid when constituting the federal government because otherwise disputing parties agreed that it was a matter for each state.

What is most apparent from the Constitution as agreed to on September 17, 1787, is that the frame of the new central government was a constitutional federalist republic of limited, delegated powers. The atom of sovereignty had been split, creating a new national government of enumerated powers with the preexisting states retaining their residual sovereignty. A republic had never in history succeeded for very long. James Madison’s solution was for the republic to unite states that spanned an extended geographic area. He believed that volatile factions in one part of the county would be dissipated over this vast expanse of land.[11] To achieve the Madisonian vision, the constitutional design was to prevent the concentration of power in any one branch. Rather, power was balanced and checked by others, with an underlying assumption that unmitigated power invites abuse and corruption.

Beyond these features, the first principles on which the government is founded are not altogether evident from the text.[12] It is true that the Preamble famously says that “We the people” are the ones who “do ordain and establish” this new government. But elsewhere the operative provisions of the 1787 Constitution indicate that matters of U.S. citizenship and who gets to vote in federal elections were left to each state to decide. Which is to say that giving definition to “the people” who are doing all this “ordaining” and “establishing” is a power vested not in the central government but residing in the several states. This was no small matter for female citizens denied the right to vote or slaves denied citizenship, voting rights, and even recognition of their full inclusion in the human race.

Historian Richard Beeman attributes the Preamble’s silence on religion and first principles more generally, other than republicanism and federalism, to Edmund Randolph of Virginia. He was chair of the Committee of Detail and the initial drafter of a provisional Constitution during the Philadelphia Convention’s recess from July 27 to August 6, 1787.[13] The Committee of Detail was given the task of assembling all the decisions the Convention had made to date into a coherent document. Within the Committee, to Randolph fell the task of putting pen to paper and producing a first draft. Concerning Randolph’s view on the proper role of the Constitution as a whole, and the Preamble in particular, Beeman writes:

The other notable aspect of Randolph’s approach to the task of constitution writing was his insistence that a lengthy preamble similar to that contained in the Declaration of Independence was not necessary. He considered the Constitution to be a legal, rather than a philosophical, document, and by his reasoning, “a preamble seems proper not for the purpose of designing the ends of government and human polities.” Randolph believed that elaborate displays of theory, though perhaps necessary in the drafting of the state constitutions, were inappropriate to the task now at hand. For Randolph, the business of constitution making was not an excursion back to fundamental principles or an articulation of the natural rights of man. Rather, it was a matter of taking those fundamental principles and natural rights already articulated in the Revolutionary state constitutions and interweaving them with the delegated powers written into a federal constitution. . . . Although what we call the “preamble” . . . went through several different transformations . . . [,] in the end, the framers of the Constitution supported Randolph’s fundamental premise.[14]

We learn from this that the absence of any explicitly stated religious presupposition undergirding the new nation’s charter was in character with the quietude concerning first principles generally. This goes a long way to dispelling the “Godless Constitution” thesis.[15]

Religion did get some explicit acknowledgment in the Constitution’s body.[16] The absence of any comprehensive mention of religion makes sense when one appreciates that in 1787, church-government relations (as distinct from safeguarding private religious conscience) were highly divisive, were widely regarded as a state-level matter, and varied considerably from state to state.[17] One can easily imagine the delegates thinking, “Why take up religious establishments when we have more than enough to disagree about when it comes to the basic frame and powers of the new central government?”

In the waning days of the Philadelphia Convention, there was an effort by George Mason of Virginia and a handful of others to add a bill of rights.[18] This was strongly resisted for multiple reasons, including that it was feared that such a bill could not be agreed upon and that the delegates were exhausted and wanted to return home.[19] Both reasons were understandable, but the decision came back to haunt proponents of the 1787 Constitution halfway through the process of state ratification.

Between December 1787 and July 1788, eleven of the thirteen states did ratify the Constitution (North Carolina and Rhode Island declined). Delaware, Pennsylvania, New Jersey, Georgia, and Connecticut quickly did so. Momentum slowed with Massachusetts, although eventually the Bay State did narrowly ratify. So did Maryland, South Carolina, New Hampshire, Virginia, and finally New York. However, to secure these latter votes, James Madison and others were forced (starting with Massachusetts) to promise that the federal government would adopt a bill of rights.[20]

The refusal by the Philadelphia Convention to take up the addition of a bill of rights shows that its members understood their task was limited. They were not about to attempt a declaration of the fundamental rights of humankind, and in particular they were not about to undertake the task of defining religious liberty and the proper scope of church-government relations. As we shall see below, that consensus concerning a limited federal role in matters of religion carried into the thinking of the First Federal Congress as it worked to create a bill of rights.

II.  The First Federal Congress, May – September 1789

As directed by the Confederation Congress, national elections of presidential electors and representatives in the House were held in the winter of 1788. Senators were chosen by the legislature in each state. The implementation of the new government was set to begin in April 1789 as the First Federal Congress and George Washington’s administration congregated at a temporary capital in New York City.

The First Congress was overwhelmingly comprised of Federalists, at this point meaning those who had supported ratification of the Constitution, as distinct from Antifederalists who had opposed ratification. The House had forty-nine Federalists and ten Antifederalists; the Senate had twenty Federalists and only two Antifederalists.[21] However, at the time there were no political parties in the formal sense, only tendencies to favor power in the central government or to desire retaining more power in the states. It was not until President Washington’s second term that parties calling themselves Federalists and Republicans began to coalesce. Accordingly, the congressional debates in the summer of 1789 over what would eventually be called the bill of rights were not partisan in the modern sense. The leading figure, James Madison, later a Republican and ally of Thomas Jefferson, was at this point in the forefront of those Federalists working to pass constitutional amendments to submit for state ratification.

Throughout the debates over the 1787 Constitution, Federalists had insisted that a bill of rights was unnecessary and that Antifederalist fears were overblown. As James Wilson, a convention delegate from Pennsylvania, argued early in the ratification period, the central government simply was not delegated enough power in the first place to disturb unalienable rights.[22] In April 1789, this was still the view of Federalists attending the First Federal Congress.

While James Madison was a major figure in shaping the coming deliberations, it would fall to Fisher Ames of Massachusetts and Samuel Livermore of New Hampshire to propose the determinative word choices for the final religious liberty phrases. Indeed, it is fair to say that Madison lost more debates than he won over the fate of the religious freedom amendments. Further, it would be a mistake to take views that Madison expressed in other venues and at other times and read them into the Religion Clauses, or to refer to the Religion Clauses as primarily the work of Madison.

As Congress assembled, Madison’s position had shifted. He still did not agree that a bill of rights was needed to thwart potential abuses by the national government. On the other hand, he now urged the adoption of a bill of rights to blunt the Antifederalist’s call for a second constitutional convention, to fulfill the demands of the five states that ratified the Constitution on the promise that a bill of rights be added, to entice North Carolina and Rhode Island to ratify and thus join the Union, and to fulfill his campaign promise to Baptists back in his congressional district.[23]

With Old Originalism, one looks for the intent of the lawmakers, and hence one places high importance on the unenacted drafts that lead up to a law, as well as the point-by-point debate over the various wordings culminating in the final text. Not so with New Originalism, which frees the interpreter from the greater subjectivity of the legislative process and focuses just on the final product. Still, what the lawmaking body thought it was doing (and what it was not doing because of outside restraints) does matter. This is because words are chosen to suit the agreed task. With that in mind, let us turn to the deliberations in the First Federal Congress that eventually yielded what we call the bill of rights, in particular the Establishment Clause.

On June 8, 1789, James Madison introduced his proposed amendments in the House. He said that the task before Congress was “to limit and qualify the powers of the Government, by excepting out of the grant of power those cases in which the Government ought not to act, or to act only in a particular mode.”[24] Accordingly, from the very start, the defined task was made politically feasible because the effort was not to agree on a comprehensive list of unalienable human rights.[25] Rather, the undertaking was the more modest task of agreeing on what powers were not vested (Federalists would have said, “were never vested”) in the national government by the 1787 Constitution.[26] That meant the amendments would be stating negatives, that is, identifying what the federal government had no power to do. This tack is further borne out by Madison seeking to interlineate the amendments into Article I, Section 9 of the Constitution, which is where negatives on national power are cataloged.

A.  Before the House of Representatives

Madison’s June 8 draft amendments dealt separately with religious pacifists and military service, as well as separately protecting religious conscience from the states. Neither effort survived in the Senate. But Madison’s central proposal stated, “nor shall any national religion be established.”[27] Madison’s amendment was referred to a Select Committee. On July 28, the Committee recommended, “no religion shall be established by law.”[28]

August 15 was the longest day for debate in the House over religious freedom. An Antifederalist from Massachusetts, Elbridge Gerry, attempted to narrow the sweep of the no-establishment phrase to merely, “no religious doctrine shall be established.”[29] The suggestion was ignored by the Federalists. Roger Sherman, a Federalist from Connecticut, said that the amendment was redundant because “Congress had no authority whatever delegated to them by the constitution to make religious establishments; he would, therefore, move to have it struck out.”[30] In response to a question from a member, James Madison gave his understanding of the Select Committee’s words to be

that Congress should not establish a religion, and enforce the legal observation of it by law, nor compel men to worship God in any manner contrary to their conscience. . . .
[T]he words . . . had been required by some of the State Conventions, who seemed to entertain an opinion that under the clause of the constitution, which gave power to Congress to make all laws, necessary and proper
. . . [, Congress was] enabled . . . to make laws of such a nature as might infringe the rights of conscience and establish a national religion; to prevent these effects he presumed the amendment was intended . . . .[31]

Benjamin Huntington, a Federalist from Connecticut, expressed concern that the amendment might upend the laws in his state levying religious taxes to pay ministers.[32] Huntington failed to understand that the amendments ran only against the federal government. A motion was made by Samuel Livermore, a Federalist from New Hampshire, to revise the text to, “Congress shall make no laws touching religion . . . .”[33] The revision made it clear the amendment was not directed at the states. This eased Huntington’s concern, but it created another over the wide sweep of “laws touching religion.” Many a facially neutral federal law could inadvertently touch religion.

This hazard was waylaid on August 20 when Fisher Ames, a Federalist from Massachusetts, suggested trimming the impossibly broad “no laws touching religion” to “no law establishing religion.” He did not elaborate on what he meant by “establishing.” Ames also introduced for the first time the “free exercise” phrase. Both features passed without comment. The Third Article now read: “Congress shall make no law establishing religion, or prohibiting the free exercise thereof, nor shall the rights of conscience be infringed.”[34]

On August 24, the Resolve of the House passed with seventeen Articles of Amendment—including Ames’s text—and was delivered to the Senate.[35]

B.  Before the Senate

The Senate met in secret. The motions and amendments from the Senate Journal are available, but the debate is not.

On September 3, numerous proposals bearing on religious liberty were entertained.[36] The Senate returned to the matter on September 9. The Senate sharply limited the restraint on an establishment to “articles of faith and a mode of worship.” The amendment now read: “Congress shall make no law establishing articles of faith or a mode of worship, or prohibiting the free exercise of religion . . . .”[37] The restraint on establishments was now so narrow it would not even prohibit a federal tax earmarked to pay the salaries of religious ministers. The Senate also reduced the number of Articles of Amendment to twelve from the House’s seventeen.

C. Committee of Conference

Because the House and Senate versions differed, the matter went to a Committee of Conference. The Conference faced a choice between a narrow Senate disempowerment (“no law establishing articles of faith or a mode of worship”) and a broader House disempowerment (“no law establishing religion”). No record exists of negotiations in the Committee of Conference. Without explanation, the Conference proposed that the Third Article read: “Congress shall make no law respecting an establishment of Religion, or prohibiting the free exercise thereof . . . .”[38]

The Conference Committee’s “no law respecting an establishment” was close to the broader-in-scope House version. However, the Conference favored the Senate when it came to adopting the stand-alone “free exercise” text rather than the broader House protection for both “free exercise” and “rights of conscience.”[39]

The Conference Committee alteration did expand the sweep of the no-establishment prohibition. Period dictionaries indicate that the introduction of the participle “respecting” meant “in relation to,” “concerning,” or simply “about.”[40] Hence, under the text, Congress was prohibited from making a law about an establishment. That is, in use of its powers delegated elsewhere in the Constitution, Congress could neither establish religion nor disestablish religion. Hence, the text prevented congressional authority not only from establishing a national religion, but also from hindering or disbanding the remaining state establishments[41]—a federalism feature.

This has led to speculation as to why “respecting” was added in Conference Committee. During the House debate, Huntington had expressed fear that the no-establishment phrase might be binding on states. The object of his fear was enforcement by the federal judiciary against states. But that was clarified to Huntington’s satisfaction by beginning the amendment with “Congress shall . . . .” However, the version still did not prevent Congress from exercising its enumerated powers in such a way as to do away with religious establishments in states like Connecticut. The use of “respecting” changed that. Yet, never in the debate in the House or Senate did anyone voice a fear that Congress would make such a bold move. Thus, the evidence that the Conference Committee’s addition of “respecting” was motivated by federalism is thin. That said, lack of evidence concerning the Conference’s intent does not matter under New Originalism. Rather, the mere fact that the plain text uses the word “respecting” means that the Establishment Clause as passed—whether intended or not—has a federalism feature.

In summary, as of September 1789, the plain text of the Establishment Clause restrained Congress in two directions. First, Congress had no power to disestablish religion in the five or six states that still had established churches. Second, Congress had no authority to use its powers otherwise enumerated in the 1787 Constitution to establish a national religion.[42]

D. Final Action, September 24-29

The House considered the Report of the Committee of Conference on September 24, and it passed by a vote of thirty-seven to fourteen.[43] The Senate concurred on September 25.[44] Two-thirds of both the House and the Senate had now agreed on twelve amendments to submit to the states.

On September 29, a preamble explaining the impetus behind passing the twelve proposed Articles of Amendment was inserted into the record of the Senate Journal:

The Conventions of a Number of States having, at the Time of their adopting the Constitution, expressed a Desire, in order to prevent misconstruction or abuse of its Powers, that further declaratory and restrictive Clauses should be added: And as extending the Ground of public Confidence in the Government, will best insure the beneficent Ends of its Institution— . . . .[45]

The preamble speaks to Congress’s limited purpose in producing the amendments. The stated task was to make it clear that the new central government had no power as to certain subject matters. This leaves it beyond doubt that the amendments vested no new powers in the federal government. On the contrary, the bill of rights was merely to reassure Americans that the 1787 Constitution was not to be misconstrued so as to impute powers that were not delegated. The Third Article would have been understood by the public as saying that the federal government was delegated no power to “make [a] law” about “an establishment of religion.” That left such power where it always had been: in the states.

E.  Summing Up: A Limited Task in the First Congress

Relations between government and church in the American states was a highly contested matter in 1789. Just a few years earlier, the states of New York and Virginia had thrown off their church establishments in widely reported contests,[46] as had other southern states with less notoriety.[47] On the other hand, despite well-organized resistance, in 1780 the Commonwealth of Massachusetts had decided by popular referendum to keep its Congregational church establishment.[48] In Connecticut and New Hampshire, the Congregational establishments were even more secure.[49] And in the south, Maryland and South Carolina still had their Anglican establishments, albeit both in slow decline.[50]

Because of the pressing business of setting up a new government, many Federalists did not want to take the time to pass a bill of rights.[51] Historians have puzzled over how the First Congress could reach agreement on the matter of establishment with so little effort and debate.[52] Indeed, there was no discussion in Congress on the merits or demerits of an established church, only modest concern that the Third Article should not be binding on the states. Huntington, a staunch establishmentarian, said he agreed with Madison, a staunch separationist, on the meaning of the text prohibiting an establishment. Huntington’s only worry was that others might attribute to the words some meaning other than what Madison and he agreed upon.[53] The obvious answer to this riddle is that the First Congress fully understood that the task before them was quite limited, to wit: to say only that the federal government had no authority concerning the subject of an establishment of religion. This could be done effortlessly, for it did not require the resolution of any existing disagreements. This was the thinking of not only the Federalists, who were firmly in control, but of the Antifederalists, who agreed for other reasons on this limited task.[54]

The general public appeared to have been aware of the limited object of the Third Article, and their quiescence on the matter shows that they apparently agreed. The public was informed of this modest purpose by Madison’s public remarks in the House and by the Senate’s preamble. As the First Congress did its work, there were no petitions or letters appearing in newspapers.[55] There was some private correspondence by Federalists in Congress that described the overall work on the bill of rights as tossing “a tub to the whale” or “frothy and full of wind.”[56] That is, the bill’s words were reputed to be chosen to fool the public into thinking that their concerns were being addressed. However, these private letters—while cynical—are fully consistent with a good faith effort to adopt a text leaving jurisdiction concerning church-state relations in the hands of the states, a placement of authority on which both Federalists and Antifederalists agreed. The non-cynics would say the amendments were to reassure the American people that the new central government was not a danger to their interests. In this the Federalists succeeded: the reporting out of the twelve proposed amendments put an end to any serious talk about convening a second constitutional convention.

We can also say a little more about what the words “an establishment” meant to the members of Congress (if not the general public) at that time. On August 15, the House ignored the Antifederalist Elbridge Gerry’s proposal to narrow their scope to merely “no religious doctrine.” Gerry’s line of thinking surfaced again in the final Senate version of September 9 that prohibited only legislation adopting “articles of faith or a mode of worship.” The establishment of the Church of England entailed Parliament prescribing as doctrine The Thirty-Nine Articles of Faith and for order of worship The Book of Common Prayer. The narrower disempowerment of the Senate version ultimately was rejected in favor of the broader House restraint on all federal “law[s] respecting an establishment of religion.” At least for Congress, if not for the wider public, there was more to the Church of England establishment than dogma and liturgy, just as it would seem there is more to the words “an establishment of religion” than “doctrine,” “articles of faith,” and “mode of worship.” The final text therefore constrained Congress from establishing a religion something like the Church of England in Great Britain.

F. State Ratification

Once reported out by Congress in late September 1789, the twelve proposed amendments were sent to the states for ratification. The first and second proposals did not receive the requisite three-fourths affirmation, thus the Third Article was renumbered as the “First Article of Amendment” or just First Amendment. The ratification process bears on how we understand the original public meaning of constitutional provisions. It should be acknowledged, however, that while states could approve some of the twelve amendments and not others, they could not alter the text of any given amendment.

Yet there is little surviving history of the debates in state legislatures that sheds light on the meaning of the Establishment Clause. There are records from only two states. Massachusetts did not record any discussion about the Establishment Clause and ultimately did not ratify the Third Article. The Virginia record, while scant and complex, is clouded by the posturing of Antifederalist state senators vaguely asserting that the amendments were inadequate to protect religious freedom. The criticism was never made any more specific. The senators were stalling, and the complaints are dismissed by historians as a last stand by Antifederalists disgruntled over the loss of state powers.[57] After some delay, Virginia ratified the Third Article without further comment.

The absence of popular pushback in the states to ratification of the Third Article is consistent with the understanding that the American public viewed the text as prohibiting any federal involvement in an establishment of religion, be it at the state or federal level.

III.  The Original Public Meaning of “Respecting an Establishment of Religion”

In the First Congress, Federalists from New England and Federalists from elsewhere differed sharply concerning establishmentarianism.[58] Accordingly, it is certainly true that the Federalists in control of the House and Senate from June to September of 1789 did not have in mind a unified theory of church-government relations which they were pouring into the amendment process. But they did not need one. That was never the task Congress had set for its members. The Establishment Clause would be binding only on the federal government, and Federalists did not envision the federal government having many dealings with the subject of religion. Given the religious pluralism across the thirteen states, there was no chance that a national religion would be established.[59] Sure, they were experienced enough to anticipate that religious liberty issues would occasionally surface—hiring congressional chaplains, pacifists refusing military service—but in the big picture, they thought such occasions would be few and minor. The real action concerning church-government relations was in the hands of the states, and all parties—especially the Federalists—anticipated that such relations would remain with the states. That is the only reason Congress could have agreed so easily on a text concerning church-government relations and reported it out as a constitutional amendment that Congress could realistically expect three-quarters of the states would ratify.

The Establishment Clause did two things, both thought to be rather modest at the time. First, it denied that there was power in the federal government to interfere with the remaining establishments in the states; structurally speaking, this made a vertical separation between the federal government and state-level religious establishments. Second, it denied that there was power in the federal government to make law about “an establishment of religion”; structurally speaking, this made a horizontal separation between the federal government and religious institutions it might otherwise establish. The latter was not controversial because no one wanted an established church at the national level. Even back then, there were too many religious differences across the thirteen states to make possible a full-on national establishment of religion.

In this understanding of the limited task taken up by the First Congress with respect to religious establishment, a search for a detailed set of substantive rules governing church-government relations in the text “an establishment of religion” is a fool’s errand. Nevertheless, there is substance in these words. The vertical restraint was federalist in character, telling Congress it could not disturb what the states did with respect to their church-state relations. That restraint was destroyed in 1947 when Everson incorporated the Establishment Clause. However, the horizonal restraint meant Congress cannot use its powers enumerated elsewhere in the 1787 Constitution to “make [a] law” about “an establishment.” This was (and, New Originalists would say, still is) a substantive restraint on Congress’s powers to regulate federal lands, territories, and the District of Columbia; to make treaties and conduct foreign affairs; to legislate with respect to the military; to relate with the Indian tribes; to oversee patents, post offices, bankruptcy courts; to regulate interstate commerce; and so on.

To determine original public meaning, New Originalism draws on the ordinary sense of the words at the law’s inception. Research professor Donald Drakeman conducted a search of American founding-era documents for the phrases “establishing religion” (the House version) and “an establishment of religion” (the Conference Committee version).[60] He found that, of the two, “establishing religion” was used far less frequently. Further, “establishing religion” was sometimes considered synonymous with “an establishment of religion,” both meaning a formal legal establishment. At other times “establishing religion” was broader in referring to laws auxiliary to but short of a formal establishment.

Drakeman’s quantitative research probes whether founding-era references to an establishment were confined to laws solely about a formal national church, or if such references also include laws that supported or were auxiliary to a formal establishment. There were many examples of supporting or auxiliary laws in England—laws that were not directly about ordering or governing the Church of England, but that supported it indirectly. For example, one had to be a member of the Church of England to receive a military commission or obtain a faculty appointment at Oxford or Cambridge. Certain Protestants outside the Church of England (nonconformists) were legally tolerated, but nonetheless were “second class”; clergy of these nonconforming sects had to secure a license to preach, and their meeting houses had to secure a license to hold worship services. Nonconforming clergy could not officiate at a marriage. These laws were clearly designed to support the Church of England establishment, but they were not part of the formal governance structure of the Church of England that entailed doctrine, liturgy, polity, administration of the sacraments, membership, sources of revenue, and property holdings.

Law professor Stephanie Barclay has taken a somewhat different approach to a corpus linguistics analysis of founding-era documents that referenced an “establishment of religion.” For each document that referenced religious establishment, she determined the salient characteristic of an establishment that was implicated by the document.[61] By sifting the results, Professor Barclay found that the most commonly occurring characteristics of the phrase “an establishment” were: (1) government compelling individuals to engage in a religious practice favored by the established church; (2) government interfering with the internal operations of the established church or of nonconforming churches; (3) government aiding the established church, particularly in the form of taxes earmarked for the state church; and (4) government imposing a religious test to hold public office, vote, or receive a post such as a faculty appointment.[62]

We have already acknowledged that because of the introduction of the participle “respecting” by the Conference Committee, the text prohibits Congress from using its enumerated powers either to make a law establishing religion or to make a law disestablishing religion in a state. “Like the Church of England, the colonial establishments were supported, not by a single act of comprehensive legislation, but by a network of laws and policies. And the full web of these laws was the object of resistance by America’s religious dissenters. In other words, the religious dissent was not just directed to the formal establishment of Congregational and Anglican churches, but also to the auxiliary laws that directly affected nonconforming Protestants. It was the full web of laws that prevented these Protestant clergy from preaching without a state license, preaching anywhere but at a government-approved meeting house, and officiating at a marriage of congregants.[63] The formal establishment together with the wider supports can be profiled as follows:

1. Government financial support of the state church, including assessments to pay ministers’ salaries and rent on glebe lands.

2. Government control over creeds, order of worship, polity, and clerical appointments of the state church; the licensure of dissenting or nonconformist clerics; and licenses tethered to particular meeting houses, which prevent itinerant preaching by nonconformists.

3. Mandatory attendance at worship services in the state church, prohibitions on services by others, and required licensure to open a meeting house for tolerated nonconformists.

4. Use of the state church to record births; to perform all marriages and funerals; and to administer tax revenues to care for the poor and widowed. Today we regard these tasks as civil functions, but in the British-like establishments of eighteenth-century America, these matters were within the jurisdiction of the established church.

5. Confining public office and voting rights to members of the state church, or using a broader religious test to include only select nonconformists. Religious tests and preferences for granting military commissions, government contracts, admission to university, and faculty appointments.[64]

To see how this understanding of the original public meaning would play out today, suppose we apply a few from this list of establishmentarian laws to a current setting. Assume Congress directed that federal lands in the Western U.S. be set aside as a glebe, and that the rents paid by ranchers grazing cattle on the land are earmarked to defray expenses incurred by the National Cathedral in the District of Columbia, an Episcopal church. Such a law would violate the Establishment Clause, even though the stand-alone glebe falls well short of fully establishing a national church.

Assume the U.S. military academies required cadets to attend worship services each Sunday morning at the campus churches. Or suppose that any student at an academy who sought permission to get married was required to have his or her wedding ceremony, including the content of the vows, conducted by a military chaplain. These two hypothetical policies are parallel to laws and practices that supported the state-level established churches in colonial America. It would follow that they ought to be struck down today as prohibited by the Establishment Clause, albeit these two policies at the military academies fall well short of fully establishing a national church.

Now consider a different set of hypotheticals. Assume the local postmaster general invited a rabbi to lead a prayer at the dedication of a new post office building. Or assume the Department of Defense allowed the erection of a stand-alone Latin cross as a war memorial in a cemetery for veterans of World War I. Or assume a city located in the Virgin Islands, a federal territory, celebrates Christmas by placing a stand-alone nativity of Jesus Christ in a municipal park. If the original meaning of the Establishment Clause was to prohibit the federal government[65] from making laws parallel to the laws and practices that supported the colonial-era established churches, then prayer at a public building dedication, a Latin cross memorial, and a Christmas nativity would not be prohibited by the Establishment Clause.[66] Given that the modern Supreme Court has struggled intensely with cases like these, it would seem that New Originalism would point to a realignment as to some of its case law.

IV.  How the Plain Text Gives Meaning to “an Establishment of Religion”

When the plain text is definitive, the courts need not resort to an interpretive rule, be it originalist or otherwise. In relevant part, the First Amendment text reads:

Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; . . . .

Although it ends with a semicolon (the first of two semicolons in the amendment), this first clause of the First Amendment can stand alone as a complete sentence. The longstanding convention is to refer to the two phrases in this first clause as the Establishment Clause and the Free Exercise Clause. However, this nomenclature mistakes phrases for clauses. While there is but one clause addressing religious freedom, there are two participial phrases (“respecting an establishment” and “prohibiting the free exercise”) modifying the object (“no law”) of the verb (“shall make”).

The grammar is such that the two participial phrases are of equal rank, and thus each carries a meaning independent of the other phrase. It is therefore entirely proper—as the convention has it—to think in terms of two separate disempowerments on the subject (“Congress”) of the sentence. This is not to say that the two restraints can never overlap. The government might transgress both participial phrases—much like a law might violate a person’s rights to both free speech and due process. However, notwithstanding an occasional overlap, the establishment restraint and the free exercise restraint give rise to separate and independent writs.

The First Amendment, along with the other provisions of the bill of rights, was meant to bind only the new national government—not the states.[67] The states were already bound by their own constitutions, which in post-Revolutionary times Americans thought sufficient. Moreover, the Religion Clauses restrain Congress as to all of its enumerated powers. That is, rights trump powers, as they teach in law school. For example, the two participial phrases restrain the federal government as it regulates the territories and District of Columbia, when it adopts treaties and conducts foreign relations, in military affairs, when it deals with Indian tribes, in regulating interstate commerce, in operating post offices and bankruptcy courts, in issuing patents, and so on.

A. The Vertical and Horizontal Dimensions to the Establishment Clause

As a result of a seemingly minor change in the Conference Committee text as reported September 24, 1779, the Establishment Clause reads, “Congress shall make no law respecting an establishment of religion.” Hence, by the phrase’s express terms, federal laws could neither establish nor disestablish religion. This meant that Congress could not abolish the remaining state establishments in Massachusetts, Connecticut, New Hampshire, soon-to-be-admitted Vermont, South Carolina, or Maryland.

The Establishment Clause therefore limits Congress in its use of power with respect to both the residual powers held by the several states and the enumerated federal powers. Congress is prohibited from interfering with state laws about “an establishment of religion” (the vertical dimension). And Congress is prohibited from enacting laws about “an establishment of religion” concerning matters within its enumerated powers (the horizontal dimension).

Because the operative words are the same, the scope of the disempowerment is the same with respect to both dimensions of congressional authority, vertical (state) and horizontal (federal). Accordingly, overblown claims that the Establishment Clause means that the federal government cannot interfere with existing establishments in the states (vertical) but that the Establishment Clause means only that Congress cannot formally establish a national church (horizontal), rely on an asymmetry that defies the plain text.

Consider a scenario in which Congress amends the Primary and Secondary Education Act by directing that federal funds be withheld from local public schools that have teacher-led prayer at the beginning of the classroom day. A religiously conservative advocate complains, arguing that the Establishment Clause denies authority to Congress to use its power to interfere with state and local practices about “an establishment of religion,” and that this is what the new amendment is doing: interfering with a local “establishment of religion.”[68] Our advocate would be entirely correct that the Supreme Court has deemed teacher-led prayer in public schools as falling within the definition of “an establishment,” and that here Congress is interfering with a state establishment via its Spending Power.

Now continue the scenario with the same advocate responding to a longstanding practice of collective prayer before meals at U.S. Military Academies. Assume that having considered complaints from a few cadets, military authorities announce that they will discontinue the collective prayer, citing the Establishment Clause. Our conservative advocate objects, arguing that when it comes to the federal government’s exercise of its own enumerated powers, the Establishment Clause only prohibits an act establishing a national church. The clause does not reach this century-long tradition of prayer at the nation’s military academies, argues our advocate, because the academies have nothing to do with the operation of an established national church.

Our advocate’s claims are asymmetrical. If teacher-led prayer at a local public school is protected from congressional legislation because it comes within the sweep of “an establishment,” then the prayer at federal military schools was rightly halted because also within the sweep of “an establishment.” The point is not that the Supreme Court’s school prayer decisions back in the early 1960s were right or wrong.[69] The point is that the words “an establishment” define an identical sweep of congressional disempowerment—here, with respect to prayer at government schools—in both the vertical and horizontal dimensions.

B. The Establishment Clause Permits Discretionary Regulatory Exemptions

Consider again just the text. It does not deny Congress power to “make [a] law” about religion. Rather, it more narrowly denies Congress the power to “make [a] law” about “an establishment of religion.” So legislation that touches on religion is allowed, except when it is respecting an establishment of religion. There is no command that government never take religion into account.

Assume, for example, that soon after the bill of rights was ratified Congress enacted a comprehensive act regulating the Army and Navy. In exercising its constitutional power to oversee the armed forces, Congress provides for a military draft but exempts religious pacifists. Nothing in the Establishment Clause prohibits such an exemption. In adopting the exemption Congress certainly does “make [a] law respecting” religion, but it does not more narrowly “make [a] law respecting an establishment of religion.” The draft exemption is designed to allow religious pacifists to follow certain practices born of their religious conscience. That is, the object of the exemption is not to advance religion (the affected pacifists are already religious) but to advance religious liberty.

This scenario raises a larger issue regarding the constitutionality of discretionary religious exemptions from regulatory and tax burdens. It is a categorical mistake to presume that a statutory religious exemption is a form of unconstitutional preference. Look again at the text. Although the government cannot “make [a] law” in support of “an establishment of religion,” it may “make [a] law” in support of religious freedom. Indeed, that would have to be so because the Free Exercise Clause is itself a law in support of religious freedom. The First Amendment would not make any sense if the Free Exercise Clause violates the Establishment Clause. Moreover, there are clauses in the 1787 Constitution that expressly exempt independent acts of religious observance. These are provisions permitting an affirmation in lieu of an oath,[70] specifically designed to accommodate Quakers and Anabaptists.[71] The Establishment Clause would not make any sense if it nullified the affirmation option that accommodates these religious minorities.

Another way of stating the matter is: Government does not establish religion by leaving its private exercise alone—which is exactly what a legislative religious exemption does.[72] Discretionary exemptions not only allow for private acts of religious exercise to continue, but they also reinforce the desired nonentanglement between church and state. Hence, it is entirely proper that the Supreme Court has held in the ten exemption cases to come before it that the act of the legislature in question did not violate the Establishment Clause.[73] The Court’s specific rationale in these cases has not always been a model of clarity, but the Justices have consistently reached the correct result—a result fully in harmony with the text of the Establishment Clause.

C. The Impossibility of Conflict Between the Religion Clauses

As the Senate concurred on September 25, 1789, in the House Resolution and the final draft of what became the bill of rights, a Preamble was added.[74] The Preamble makes clear that the proposed amendments did not vest any new power in the federal government. Rather, the amendments were designed to negate an assumption of power by the national government being wrongly implied from the 1787 Constitution. That is why provisions in the bill of rights are often referred to as “negative rights.” They tell the national government what it has no power to do. If you have two clauses from the bill of rights, you have two disempowerments. If the two overlap in their application, then you have a two-fold negation of governmental power. They reinforce each other. What you do not have, indeed cannot have, is a conflict between the clauses.

This has direct implications for correcting a present-day misunderstanding that is alarmingly widespread.[75] It is still common to find those who believe that the Establishment and Free Exercise Clauses are in unavoidable tension and that they occasionally conflict, as if the Free Exercise Clause is pro-religion and the Establishment Clause is there to hold religion in check. This reading of the text presumes that the Free Exercise Clause and the Establishment Clause sometimes run in opposing directions, and hence that they will clash. If this were true, it would be the Supreme Court’s task to determine whether the constitutionally questionable legislation is rightly “neutral”—neither too pro-religion nor too anti-religion. Not only is this contrary to the plain text as well as common sense, but it concedes too much power to the judiciary.[76]

Consider the Free Speech and Free Press Clauses. These two clauses negating federal power over speech and press sometimes overlap and thus reinforce one another, but they cannot conflict. Simply put, while the government can simultaneously violate both clauses, it is impossible for these two overlapping negatives on the government’s power to be in conflict. Similarly, the Free Exercise and the Establishment Clauses may occasionally overlap and thereby doubly negate the field of action otherwise permissible to the government, but they cannot conflict.[77] To be sure, the Religion Clauses, each in its own way, work to protect religious freedom. But when circumstances are such that the scope of the phrases overlap, they complement each other rather than conflict.

By way of illustration, consider a fourth-grade public school teacher who has twenty-five students in her classroom. Assume the teacher requires the students to recite in unison the Lord’s Prayer to begin the school day. A Muslim student sues under the Free Exercise Clause claiming that her rights are violated and offers evidence that reciting the Christian prayer is a violation of her conscience because its content contradicts the beliefs of Islam. The student will prevail, and the remedy will be that our Muslim fourth-grader may now opt out of the prayer while her classmates and teacher continue the daily recitation.[78] Assume a second suit is filed, this time alleging that teacher-directed prayer violates the Establishment Clause. Once again, our Muslim student will prevail, but this time the remedy will be to enjoin the recitation of the classroom prayer altogether.[79] While the remedies in the two lawsuits differ, both clauses are violated by the teacher-required prayer. The two clauses overlap and complement each other. They do not conflict.

Finally, assume that a third lawsuit is filed invoking the Free Exercise Clause. This claim is brought by three Christian students in the classroom who ask that the teacher-led recitation of the Lord’s Prayer be allowed to continue on a voluntary basis. With reference to the limits on the government’s power negated by the Establishment Clause, the court will deny relief to these three students. There is no right under the Free Exercise Clause to capture the engines of government and put its machinery to use advancing the students’ Christianity.[80] If the Christian faith is to be advanced, it must rely on the voluntary acts of Christians.[81] In this third lawsuit, there is no conflict between the clauses. Rather, the three students failed to state a claim under the Free Exercise Clause.

D. A Structural Establishment Clause

As we saw in the historical background to the bill of rights, the congressional drafters did not mean to imply that Congress, in the absence of the Establishment Clause, had power to “establish[] . . . [a] religion” under the 1787 Constitution. Alexander Hamilton warned against such inferences from negations.[82] Federalists were in complete control of the drafting process, and as we have seen from James Wilson’s speech forward, the Federalists (including Madison) repeatedly denied that the 1787 Constitution vested in Congress any such power. Rather, the drafters of the Establishment Clause meant only to reassure readers of the 1787 Constitution that Congress had no power to make laws concerning a religious establishment.

That said, it must be conceded that the text of the first participial phrase (“respecting an establishment”) is different in nature from the two rights-based participial phrases (“prohibiting the free exercise” and “abridging the freedom of speech, or of the press”). The latter two forbid “prohibiting” and “abridging” and thus disempower the government with respect to a person’s free exercise or free expression. The objects of these two phrases could be understood to acknowledge that people have self-evident, unalienable, or natural rights to free exercise and free expression. In any event, they imply a moral autonomy. On the other hand, the object of the participial phrase “respecting an establishment” is not about acknowledging an intrinsic human right, but is a reference to a discrete subject matter (“an establishment of religion”) that is being placed off limits to or outside the government’s authority. (It would sound silly to say that people have a natural right to a nation that does not have an established religion.) This difference in the nature of these participial phrases leads to a difference in their function: creating a structural relationship versus acknowledging an intrinsic right.

The Establishment Clause operates like a structural distancing of two centers of authority. Constitutional structure delegates, separates, and limits power. A happy consequence of good constitutional structure is the prevention of concentrations of power that lead to a loss of liberty. In the text of the Establishment Clause, we have a separation of the authority of government and the authority of organized religion. All persons within the jurisdiction of the federal republic benefit when the government cannot exercise power respecting “an establishment of religion.” A complainant cannot waive this separation of powers any more than she can waive a federal court’s lack of subject matter jurisdiction. Rather, the separation is there to benefit more than just the party-plaintiff before the court. This is much like the three-branch structuring we call “separation of powers”; the separation of the branches is there not just for the benefit of the complainant, but for all persons subject to the government’s jurisdiction.

Given the different natures of the Establishment Clause (structural in function) and the Free Exercise and Free Speech Clauses (rights-based in function), the modern Supreme Court is correct when it applies the Establishment Clause as structural, separating the two centers of authority we call church and government. The Court envisions the Establishment Clause as policing the boundary between church and state, and it understands its judicial task as keeping governmental power from trespassing in a space delineated as “an establishment of religion.”

This separation must not be exaggerated. This is a separation of the institutions of religion from the institutions of the republic. While the institutions of church and government can be separated, religion and politics cannot. The latter would be quite impossible, for it would mean requiring religious people to cleave themselves in two. And, of course, churches appropriately speak to how their teachings bear on social issues, consistent with their right to freedom of speech.[83]

That the Establishment Clause is regarded by the federal judiciary as structural explains several features in the case law.[84] For example, when it comes to the Establishment Clause, there are special rules concerning standing to sue because in structural cases there are often no parties with individualized harm.[85] In contrast to free exercise claims that remedy religious injuries, the Establishment Clause additionally provides a remedy for nonreligious harms such as economic damages and loss of academic freedom.[86] This also accounts for why federal courts sometimes frame the operation of the Establishment Clause as a limit on their subject matter jurisdiction.[87] Whereas free exercise lawsuits yield a conditional right that is subject to strict scrutiny, a prima facie Establishment Clause claim is not subject to a balancing test that weighs governmental interests against a claimant’s interests. Either the Establishment Clause is violated or it is not; no balancing. And both the prohibition on courts answering religious question and the ministerial exception are largely rooted in the no-establishment principle, as befitting rules that derive from church autonomy and the separating of matters of internal church governance from civil regulation.[88]

V. Conclusion

Without any analysis, the Supreme Court in Everson v. Board of Education held that the Establishment Clause should be incorporated through the Fourteenth Amendment and made applicable to state and local governments.[89] With Everson, the Supreme Court made the policing of the church-state boundary much more energetic than was ever contemplated in 1789. In part, this was because incorporating the Establishment Clause meant that thereafter the boundary keeping had to take place in a myriad of state and local governmental arenas, not just with respect to the limited precincts of the federal government. This line drawing became even more difficult with the modern increase in the size and regulatory activity of government, as well as how government money (with strings attached) increasingly marks out much of our common life together. Lastly, the boundary keeping has become harder because the American people and their religious allegiances have become far more pluralistic.

Everson’s incorporation of the Establishment Clause was a mistake if one follows the interpretive rules of original public meaning. Nonetheless, New Originalists should accept incorporation as settled law by virtue of stare decisis. Americans have already worked their way through incorporation’s many difficulties, many citizens have come to rely on it, and its reversal would be disruptive. Moreover, such a proposition is not a one-sided bargain; the Establishment Clause often serves to protect religious freedom when it comes to matters such as the prohibition on courts becoming entangled in religious questions and the resurgent protection of church autonomy.[90]

If we proceed with an Establishment Clause that is binding not only on the federal government but also state and local governments, where does that leave New Originalism? What remains is a disempowerment of government at all levels to “make [a] law about an establishment of religion.” The original public meaning of “an establishment” encompasses the full network of colonial laws, broken down into the five points in Part III.[91] This web of laws comprises more than just a formal religious establishment, but also those laws auxiliary to the Anglican and Congregational churches at the time of America’s disestablishment in the states that were contemporaneous with the work of the First Congress. This is an Establishment Clause of modest breadth, yet one that is well distant from the high watermark of strict separationism.

 

Note from the Editor: The Federalist Society takes no positions on particular legal and public policy matters. Any expressions of opinion are those of the author. Whenever we publish an article that advocates for a particular position, we offer links to other perspectives on the issue, including ones opposed to the position taken in the article. We also invite responses from our readers. To join the debate, please email us at [email protected].