INTRODUCTION & BRIEF SUMMARY

Religious liberty is a value of the highest order. In general, American public officials are sensitive to religious interests, and they often make commendable efforts to accommodate the needs of religious persons and practices. Nevertheless, there are undoubtedly times when officials--whether through prejudice, indifference, or misunderstanding--fail to show appropriate respect for the free exercise of religion. Congress has an important role to play in correcting these failures. If RLPA were a reasonable effort to discharge that responsibility, we would support it with enthusiasm.

Unfortunately, RLPA does something entirely different. By generating an extreme form of the "compelling state interest" test, and imposing it over a more sweeping range of cases than has ever been contemplated by the Supreme Court or by Congress, RLPA would undermine the government's capacity to pursue perfectly legitimate, even-handed, democratically chosen goals. In effect, RLPA would two classes of citizens: those who have religious reasons for their actions and who would thereby be privileged to defy otherwise perfectly valid governmental regulations, and those whose reasons for acting--however laudable and heartfelt--are not religious. RLPA's compelling state interest test goes far beyond protecting religiously-motivated people from hostility or insensitivity. Taken seriously, it would make religiously-motivated persons sovereigns among us.

Not surprisingly, Congress has no power to create the kind of special and arbitrary privileges that would result if RLPA were to become law. RLPA's peculiar statutory architecture amounts to a tacit admission of this problem: Even in an era when Congress retains broad license to act under its commerce clause and spending powers, RLPA stands out as depending upon a tenuous and improbable connection between those powers and the subject of religious liberty. Far from curing the constitutional vices of RFRA, RLPA's somewhat desperate hunt for constitutional authority proliferates such difficulties.

Specifically, RLPA manifests five distinct constitutional vices. First, RLPA's sweeping application of the "compelling state interest test" unconstitutionally privileges religion. Because RLPA defines "the exercise of religion" in novel and unprecedented terms, it would likely violate the Establishment Clause even if its predecessor, RFRA, did not do so. Second, Section 2(a)(1) invokes Congress' spending power for purposes unrelated to the goals of any particular spending program. As a result, it exceeds the scope of Congress' enumerated powers. Third, Section 2(a)(2) likewise invokes Congress' commerce power for purposes unrelated to any goal related to interstate commerce. It, too, exceeds the scope of Congress' enumerated powers, and so would be held unconstitutional. Fourth, Section 3(b) limits the land use authority of state and local governments in a way that bears no relationship to any plausible claims that such governments are discriminating against religion. RLPA attempts to justify these limits by relying upon Congress' authority to enforce the Fourteenth Amendment. That effort is starkly inconsistent with the Supreme Court's decision in Flores. Fifth, Section 3(a) attempts to alter the judiciary's interpretation of the Free Exercise Clause. It thereby compromises the separation of powers and exceeds the authority of Congress under Section Five of the Fourteenth Amendment.

ANALYSIS

I. ESTABLISHMENT CLAUSE ISSUES

I.1. The Compelling State Interest Test. Like RFRA before it, RLPA incorporates the compelling state interest test. That test appears in Section 2(b) of RLPA, and it is the heart of the proposed legislation. We have criticized this test extensively. Christopher L. Eisgruber and Lawrence G. Sager, Why the Religious Freedom Restoration Act is Unconstitutional, 69 N.Y.U. L. Rev. 437 (1994); see also Christopher L. Eisgruber and Lawrence G. Sager, Congressional Power and Religious Liberty after City of Boerne v. Flores, 1997 S. Ct. Rev. 79 (1997). If honestly applied, the "compelling state interest test" is the most demanding standard known to constitutional law. Accordingly, the test is suitable only where it is appropriate to entertain a broad presumption of unconstitutionality--where, in other words, almost all of the cases that trigger the test will be abhorrent to the best standards of government behavior. Such a presumption rightly applies, for example, to laws intended to censor speech or to discriminate against racial or religious minorities. This presumption is badly suited to religious exemption cases, however. Many perfectly sound, even-handed laws will impose incidental burdens on some religious practices. The breadth and variety of religious belief make such collisions inevitable; but this does not offer a reason for depriving ourselves of the capacity to govern. Nor does the mere fact that a person's conduct is motivated by religious belief offer a good reason for permitting that person to defy reasonable, even-handed laws.

As applied in RFRA and RLPA, the "compelling state interest test" offers religiously motivated persons a sweeping privilege to disregard the laws that others are obliged to obey. It indefensibly favors religious commitments over the other deep concerns and interests of members of our society--concerns and interests like the welfare and integrity of one's family, deep moral and political commitments not recognizably grounded in religious beliefs, and professional, artistic and creative projects to which individuals may be passionately committed. Under RLPA or RFRA, for example, a church charity might ignore rules that a secular charity, devoted to identical causes, would have to respect. This sweeping preference for religiously motivated projects is a violation of the Establishment Clause of the First Amendment.

The idea that some persons are entitled to ignore the laws that others are required obey, and that this privilege depends upon the actors' system of beliefs, is extraordinary and transparently inconsistent with our constitutional values. In the debate over RFRA, the degree to which this idea was alien to our constitutional tradition was obscured by a misreading of the Supreme Court's religious liberty jurisprudence in the three decades preceding the Court's decision in Department of Employment Services v. Smith, 474 U.S. 872 (1990). During that period, the Court gave lip-service to the proposition that government behavior that penalized persons for doing that which was essential to their religious commitments should be measured against the rigors of the compelling state interest test.

Two crucially important facts went largely unobserved during the RFRA debate. First, while the Court spoke broadly, it acted extremely narrowly. Only one isolated group was ever permitted to defy a general legal rule on the basis of the compelling interest test. That was the Amish, who were permitted to direct the development of their teenage children outside the framework of what the State of Wisconsin recognized as a school. One other group prevailed in the Court's many pre-Smith exemptions cases. The Court protected people who were presumptively entitled to claim unemployment insurance benefits; who had deep religious reasons for refusing an available job; and who faced a serious danger that those reasons might be treated with hostility by state bureaucrats. Outside of these two small groups, every other attempt by any religious person or group to invoke the compelling state interest test failed. In every other branch of constitutional jurisprudence, the compelling state interest test was strict in theory, but fatal in fact; here it was strict in theory but notoriously feeble in fact. The Smith Court did not cause or even precipitate the test's demise. The Court merely announced what had long been true.

The second thing that went largely unobserved in the RFRA debate was the fact that RFRA --- and now, even more, RLPA -- proposed a much more sweeping form of the compelling state interest test than had ever been even the nominal rule in the Supreme Court. As the Court observed in Flores, RFRA imposed "a least restrictive means requirement ... that was not used in the pre-Smith jurisprudence RFRA purported to codify." 117 S. Ct. at 2171. Sections 2(b)(2) and 3(b)(1)(A) of RLPA repeat this innovation. As constitutional commentators widely recognize, the least restrictive means requirement is the element that gives the compelling state interest test its special rigor in other contexts. More significantly still, through its extraordinarily capacious definition of the exercise of religion RLPA extends the potential coverage of the compelling state interest test to a far wider range of cases than was ever contemplated by the Supreme Court's most sweeping statements. We explore the implications of this last observation in the section that follows.

I.2. RLPA's Novel and Unprecedented Definition of the Exercise of Religion. RLPA exacerbates RFRA's Establishment Clause problems. Section 6(1) of RLPA defines "religious exercise" to mean "an act or refusal to act that is substantially motivated by a religious belief, whether or not the religious exercise is compulsory or central to a larger system of religious belief." RLPA also amends RFRA to incorporate this new language. Section 7(a)(3). This definition is new. It appeared neither in RFRA nor in the Supreme Court's pre-Smith jurisprudence. Under RFRA, few courts had insisted that religious exercise be "compulsory" in order to trigger the statute's provisions, but most courts had held, in effect, that RFRA applied only to "substantial burdens" upon beliefs which were in some way and to some degree "important" to religious believers.⁽²⁾

RLPA's definition of religious exercise threatens to increase the extent to which RFRA favored religion over non-religion. Under RFRA, it was possible to argue that a burden upon religious exercise was not "substantial" if it affected only optional practices for which adequate substitutes were available. For example, under RFRA, several churches running soup-kitchens in residential neighborhoods sought zoning exemptions which, they conceded, were unavailable to comparably situated secular charities. In these cases, it was possible to argue that no "substantial burden" upon religious practice existed: the churches were free to run soup-kitchens in other locations, and they were free to engage in other charitable practices which, as a matter of their own religious doctrine, were equally worthy. See, e.g., Daytona Rescue Mission, Inc. v. City of Daytona Beach, 885 F. Supp. 1554, 1560 (MD Fla. 1995). When successful, arguments of this kind mitigated the RFRA's favoritism for religion.

It is not clear that these arguments would remain available under RLPA. To be sure, Sections 6(1) and 7(a)(3) define "religious exercise," not "substantial burden." Courts might find burdens upon religious exercise insubstantial if they affected only unimportant practices or if they left religious believers other, equally acceptable means by which to pursue their religious convictions. That construction of the "substantial burden" test, however, might render Section 7(a)(3) nugatory; if so, courts would be loathe to accept it. For that reason, RLPA exacerbates RFRA's already troubling disparity between the treatment of religious and non-religious interests. RLPA might fail to survive scrutiny under the Establishment Clause even if RFRA (without RLPA's amendments) could have done so.

II. FEDERALISM ISSUES.

II.1. Spending Power Issues. Section 2(a)(1) of RLPA attempts to regulate the ability of state and local governments to "substantially burden ... religious exercise ... in a program or activity ... that receives federal financial assistance." That Section is an effort to draw upon Congress' spending power. The Supreme Court has held that Congress has broad discretion to impose conditions upon the use of federal money by state and local governments. The leading case is South Dakota v. Dole, 483 U.S. 203 (1987). In Dole, the Court upheld a statute which provided that states would lose federal highway funds if they did not raise the drinking age to 21. South Dakota objected to the statute on the ground that, under the Twenty-First Amendment, liquor laws were a matter of state rather than national control. The Supreme Court rejected this argument, reasoning that states could retain control over their drinking ages if they were willing to reject the offer of federal funds.

The Court's construction of the spending power in Dole was generous, but it was not unlimited. The Court emphasized that "our cases have suggested (without significant elaboration) that conditions on federal grants might be illegitimate if they are unrelated 'to the federal interest in particular national projects or programs.'" In Dole, the Court reasoned that "the condition imposed by Congress is directly related to one of the main purposes for which highway funds are expended--safe interstate travel." By raising the drinking age, the Court suggested, states would further the purposes of federal transportation law. Yet, unless Dole's nexus requirement is entirely meaningless, RLPA cannot possibly satisfy it. RLPA applies to all religious conduct and it applies to all federal spending programs. It defies belief to think that accommodating religious conduct, regardless of its nature, supports the goals of every federal expenditure, regardless of its purpose. Indeed, RLPA's compelling state interest test is blatantly inconsistent with that idea: it would require states to accommodate religious conduct even at the expense of the core goals of any given program unless those goals rose to the level of a "compelling state interest."

In effect, RLPA assumes that once federal dollars touch some activity or program, the activity or program is federalized top-to-bottom: it then becomes fair game for congressional regulation regardless of whether the regulation has anything to do with the federal government's initial spending program. That is not what the Supreme Court said in Dole, and it is not a sensible reading of the Constitution.

These considerations are sufficient to scuttle Section 2(a)(1) of RLPA, but it suffers from an additional constitutional defect. In Dole, states remained free to legislate whatever drinking age they preferred. If they departed from the federal standard, the penalty was forfeiture of federal funding. RLPA is not written that way. It does not provide that states will forfeit federal funds unless they enact state-law versions of RFRA or RLPA; instead, it subjects the states directly to private rights of action under federal law. This objection is somewhat technical in character, and there are ways around it. For example, the Court might construe RFRA as imposing conditions on every offer of funding which the national government makes to the states; on this theory, RLPA's regulation would effectively result from a "contract" between the states and the federal government, rather than from direct regulation by the federal government. It is not obvious, however, that this theory would or should succeed.⁽³⁾

II.2. Commerce Clause Issues. Section 2(a)(2) of RLPA attempts to regulate the ability of state and local governments to "substantially burden religious exercise in or affecting commerce." That Section is an effort to draw upon Congress' commerce power. The Court has construed the commerce power generously including, of course, in connection with congressional efforts to prohibit discrimination. The case most often cited in this connection is Katzenbach v. McClung, 379 U.S. 294 (1964). In McClung, the Court upheld application of Title II of the Civil Rights Act of 1964 to Ollie's Barbecue, a restaurant in Birmingham, Alabama. The Court said Congress had power to prohibit race discrimination by Ollie's Barbecue on the following theory: by refusing to serve African-Americans, Ollie's Barbecue diminished the volume of business it did, and it thereby diminished demand for food products that moved in interstate commerce. The effect of one restaurant's actions might be small, but Congress was entitled to consider the aggregate effects of all restaurants similarly situated.

McClung grants Congress expansive authority, but that authority is not unlimited. Even in McClung, the Court insisted that Congress must identify some "connection between discrimination and the movement of interstate commerce." The Court upheld Title II only because the legislative record included "ample basis for the conclusion that ... restaurants ... sold less interstate goods because of ... discrimination." It is impossible to imagine, much less substantiate, any such basis for RLPA. Religious conduct varies tremendously and unpredictably. From the standpoint of interstate commerce, religious activity is a random vector. There is no reason to believe that it promotes, diminishes, obstructs, or facilitates interstate commerce. Nor is there any reason to think that requiring government to accommodate religion would have any predictable effect whatsoever upon interstate commerce.

The theory of Section 2(a)(2) of RLPA is largely parallel to the theory of Section 2(a)(1): it presupposes that once the congressional commerce power touches some activity or practice, that activity or practice becomes federalized top-to-bottom: it becomes fair game for congressional regulation regardless of whether the regulation has anything to do with promoting interstate commerce. That is not what the Supreme Court said in McClung. It is flatly inconsistent with the Supreme Court's recent decision in United States v. Lopez, 115 S.Ct. 1624 (1995), which held, inter alia, that Congress cannot regulate guns simply because they at one time entered the stream of interstate commerce.

II.3. Issues Pertaining to Section Five of the Fourteenth Amendment. In Section 3(b), RLPA purports to limit the zoning authority of state and local governments. This Section of RLPA appears under the heading, "Enforcement of the Free Exercise Clause." It is meant to apply to all land use cases, not just to those where the legislation's dubious invocations of the spending and commerce clause are apt. Apparently, this Section, like RFRA before it, depends for its validity on Congress' power to enforce the Fourteenth Amendment. That power was, of course, the focus of the Supreme Court's decision in Flores. There, the Court emphasized that Section Five does not permit Congress to displace the Court's judgments about the content of constitutional rights. Exercises of power under Section Five are valid only so long as they serve to put in place a scheme of remedies for rights which the Court itself is willing to recognize. Flores, 117 S. Ct. at 2163-64, 2171-72.

In Flores, the Court emphasized that "Congress must have wide latitude in determining" what measures are well-suited to remedy constitutional violations. Id., at 2164. Nevertheless, Section 3(b) of RLPA unquestionably repeats the vices that proved fatal to RFRA. Section 3(b) involves a sweeping and unwarranted federalization of local decision-making. It is no exaggeration to say that, under RLPA, any encounter between a religious organization and a local zoning authority would become a matter for federal adjudication. This remarkable preemption of local authority cannot be defended as a reasonable mechanism to remedy or prevent discrimination against religious interests. No doubt zoning administrators sometimes abuse their authority to harm unpopular churches. But that problem is not reasonably attacked by extending all churches--no matter how rich, how powerful, or how favored in law--a blanket writ to challenge the zoning ordinances which every other citizen and institution must respect. What the Court said about RFRA is equally true of Section 3(b) of RLPA: "The stringent test [it] demands of state law reflects a lack of proportionality or congruence between the means adopted and the legitimate end to be achieved." 117 S. Ct. at 2171. Section 3(b) of RLPA is therefore starkly unconstitutional under Flores.

III. SEPARATION OF POWERS ISSUES.

Section 3(a) contains a remarkable assault on the judiciary's authority to make independent judgments about the meaning of the Constitution. It presumes, under the guise of enforcing the Fourteenth Amendment, to articulate "presumptions" which courts must respect when applying its First Amendment jurisprudence. In particular, the Section purports to increase the government's burden of persuasion in Free Exercise Clause cases. Because Section 3(a) attempts to deprive the courts of the authority to interpret the Constitution, it is patently unconstitutional. There are two doctrinal paths to that conclusion. The simplest runs through Flores. The Court said clearly in Flores that Congress may not use its Fourteenth Amendment powers to alter the substance of the Court's interpretations of the Fourteenth Amendment. Section 3(a) of RLPA offends this conclusion more blatantly than RFRA did, and the Court would undoubtedly find it unconstitutional.

There is, however, an even more fundamental doctrinal objection to Section 3(a). In United States v. Klein, 80 U.S. (3 Wall.) 128 (1871), the Supreme Court held that Congress may not specify a "rule of decision" for courts. Courts must be able to decide for themselves how to apply statutes or the Constitution. In the realm of statutory interpretation, Klein is difficult to apply: in some sense, of course, Congress specifies a "rule of decision" for courts every time it writes a statute. Christopher L. Eisgruber and Lawrence G. Sager, Why the Religious Freedom Restoration Act is Unconstitutional, 69 N.Y.U. L. Rev. 437, 470 (1994). RLPA, however, is a text-book violation of Klein. It attempts to compel judges to respect Congress' judgment, rather than their own, when interpreting the Constitution. And it forces judges to act as though they and adopted Congress' constitutional judgment as their own. Congress has the power and responsibility to arrive at its own view of constitutional substance, of course. But Congress is obliged to permit the Court this same independence of judgment.

CONCLUSION

RLPA's constitutional defects are not technicalities. On the contrary, they all reflect strong claims on the policy judgment of the members of Congress who wish to act on behalf of religious liberty. Congress may well want to assure that religiously-motivated persons are treated fairly and that their interests are reasonably accommodated. But Congress surely does not want to sweepingly favor religiously-motivated persons over the vast majority of citizens conscientiously leading their lives, and to do so at the expense of the democratically-shaped rule of law. Likewise, Congress surely does not want to generate what Justice Kennedy in Flores correctly characterized as "... a considerable intrusion into the States' traditional prerogatives and general authority to regulate for the health and welfare of their citizens." And finally, Congress should want to act as the Supreme Court's partner in the pursuit of political justice for American citizens, not as its adversary. That is the admirable tradition into which, for example, Title VII and the Voting Rights Act fall. RFRA was a false start, and Congress need not and should not perpetuate RFRA's mistakes.

Of course, RFRA was motivated by a legitimate and important goal: the goal of assuring that religiously-motivated conduct is reasonably accommodated, that governmental actors are not insensitive or hostile to religious beliefs and commitments. Congress has an extremely important role to play in pursuing that goal. It can play that role in two different ways.

First, Congress can continue to police state and federal conduct for egregious failures of the duty of reasonable accommodation and correct those failures. This is a role that Congress has traditionally played to the great benefit of constitutional justice in the United States. Thus, for example, Congress directed the armed forces to make reasonable accommodation for the wearing of religiously mandated apparel (see 10 U.S.C. §774); and thus, Congress withdrew funding for a Forest Service road that would have harmed a sacred Native American site (see House Committee on Appropriations, Dept. of the Interior and Related Agencies Appropriations Bill, 1989, H.R. Rep. No. 713, 100th Congress, 2d Sess. 72 (1988)); and thus, Congress has provided church employers with exemptions from certain tax obligations that are inconsistent with their religious beliefs (see 26 U.S.C. §3121(w)(1)); and thus, Congress acted to specifically assure members of the Native American Church the ability to use Peyote as part of their sacrament of worship (see 42 U.S.C. §1996). This effort requires ongoing vigilance and nuance of legislative response, but Congress' performance in this context has been superb.

And second, Congress can enact more general legislation that offers broad protection to religiously-motivated persons against the possibility that their beliefs and commitments will be treated with insensitivity or hostility. This memorandum is not a good setting in which to explore the content of such legislation, but we would be glad to pursue the question with the Committee or any of its members.

What is critical to recognize for the moment is that RLPA is not such legislation. RLPA offers a distorted and untenable view of what religious liberty is, a view that Congress on reflection should not endorse; and RLPA streches notions of congressional authority to their breaking point, inviting the judicial articulation of constitutional limitations that Congress should not welcome. RLPA is unconstitutional, and if it were enacted, the Court would find it so to be. Congress has good reasons at the outset to choose a different vehicle to realize its altogether laudable concern for religious liberty.

Judiciary Homepage

Christopher L. Eisgruber

Professor of Law

New York University School of Law

Lawrence G. Sager

Robert B. McKay Professor of Law

New York University School of Law

1. ^1. Flores clearly invalidated RFRA with respect to the regulation of state and local government behavior. Courts have divided about whether Flores should be understood to invalidate RFRA with regard to regulation of federal behavior. Yet, regardless of whether RFRA's federal applications survived Flores, we expect that the federal courts should, and will, ultimately declare them to be unconstitutional. For reasons that are equally applicable to RLPA and so are discussed in this memorandum, we believe that RFRA is unconstitutional under the Supreme Court's Establishment Clause doctrine.

2. ^2. See, e.g., Mack v. O'Leary, 80 F.3d 1175, 1179 (7th Cir. 1996) ("a substantial burden on the free exercise of religion ... is one that forces adherents of a religion to refrain from religiously motivated conduct, inhibits or constrains conduct or expression that maintains a central tenet of a person's religious belief, or compels conduct or expression that is contrary to those beliefs"); Bryant v. Gomez, 46 F. 3d 948, 949 (9th Cir. 1995) (to meet the substantial burden standard, plaintiffs must point to a burden that is "'more than an inconvenience; the burden must be substantial and an interference with a tenet or belief that is central to religious doctrine.'" (quoting Graham v. C.I.R., 822 F.2d 844, 850-51 (9th Cir. 1987), aff'd sub nom. Hernandez v. Commissioner, 490 U.S. 680 (1988)); Thiry v. Carlson, 78 F.3d 1491, 1495 (10th Cir. 1996) ("To exceed the 'substantial burden' threshold, government regulation 'must significantly inhibit or constrain conduct or expression that manifests some central tenet of ... [an individual's] beliefs; must meaningfully curtail [an individual's] ability to express adherence to his or her faith; or must deny [an individual] reasonable opportunities to engage in those activities that are fundamental to [an individual's] religion'" (quoting Werner v. McCotter, 49 F. 3d 1476, 1480 (10th Cir. 1995) (brackets and ellisions added by the Thiry Court)); Cheffer v. Reno, 55 F.3d 1517, 1522 (11th Cir. 1995) (no substantial burden results if a government action "leaves ample avenues open for plaintiffs to express their deeply held belief[s]").