Armin Grosz, Sarah Grosz and Naftali Grosz v. The City of Miami Beach, Florida

GOLDBERG, Senior Circuit Judge: This case calls for the accommodation of two important values, both embodied in the spirit and letter of the Constitution: free exercise of religion and the effective use by *731 a state of its police powers. That accommodation consists of a balancing process, leading to a scheme of compromise between the two values that best accords with constitutional mandates. The accommodation we make today, between Appellees’ interest in holding religious services in their home and the zoning powers of the City of Miami Beach (“the City”), arises out of the following facts. I. FACTS A joint stipulation of facts filed in the court below reveals the background of this case. 1. In 1977 plaintiffs, United States citizens, purchased property at 3401 Prairie Drive, Miami Beach, Florida. Naftali and Sarah Grosz use the house as their principal residence, and Armin Grosz lives there during parts of the year. In addition to a house, the property includes a separate building which was previously used as a garage and a recreational room. 2. For many decades prior to this purchase, the property has been zoned RS-4 by the City, for single-family residential use, and plaintiffs bought the house subject to a standard deed restriction expressly citing this single-family residential restriction. Plaintiffs have not requested rezoning of the property. 3. Some years after the purchase, plaintiffs applied for permits for internal modifications to the accessory structure, stating that the building would be remodeled for “playroom use.” They were specifically informed by the City that the structure could not be remodeled as a religious institution. The accessory structure was not externally modified, but plaintiffs specifically stocked the inside of the building for group religious services, including benches to seat over 30 persons, Torahs, Arks, a Menorah, skull caps, an eternal light, numerous prayer books, shawls, and other items of religious significance. Internally the structure has much of the physical indicia of a small synagogue or “shul,” and is not operated as a playroom. Externally the building was not modified. 4. Prior to the remodeling, plaintiffs were aware that Miami Beach Zoning Ordinance No. 1891 had been construed by the City to prohibit churches, synagogues and similarly organized religious congregations in single-family residential zones, and that this construction had been upheld by the State and Federal trial and appellate courts. 5. Naftali Grosz, a Rabbi and the leader of an orthodox Jewish sect, is aged and somewhat infirm but is not immobile. It is a requirement of his religion to conduct religious services twice daily in a congregation of at least ten adult males, and it is convenient for Naftali Grosz to use this structure for that purpose. For many years the Grosz family has conducted group religious services in whichever home they have occupied. Plaintiffs could, however, conduct such services in many other areas within the City of Miami Beach, including an area within four blocks of their home. 6. Most of the congregation at the Grosz home are friends, family members or neighbors. The typical congregation is between ten and twenty males who regularly assemble at the building for religious services. However, occasionally (generally in the winter months and on Saturdays) the congregation contains as many as fifty persons, some of whom are neither friends, family members or neighbors. Plaintiffs do not exclude members of the public from attending these services and on rare occasions, in order to ensure the presence of ten adult males, have solicited persons to attend the services. Because Naftali Grosz is a leader of a particular orthodox Jewish sect, nonresidents of Miami Beach who belong to the sect use the accessory building as their principal or sole place of worship visiting South Florida. Naftali Grosz has referred to the congregation as a shul, and a witness, who is a neighbor of plaintiffs, has testified that persons have come to her house asking for directions to the “Grosz shul.” Plaintiffs would not refuse *732 contributions for the services, but do not solicit such contributions. 7. The religious services last one-half hour in the morning and one-half hour in the late afternoon every day. On Saturdays and religious holidays each service may take two to three hours. Daily services usually cause no substantial disturbance to the neighborhood, but well-attended services have disturbed neighbors as a result of persons seeking directions to the Grosz shul, as a result of chanting and singing during the services, and as a result of the occasionally large congregations of worshippers at the property. 8. On February 6, 1981, the plaintiffs were given a “notice of violation” by the City threatening prosecution which would result in conviction of a misdemeanor charge. This notice of violation of the zoning ordinance was prompted by citizen complaints to the City. 9. The City did not and would not prosecute plaintiffs for praying in their home with ten friends, neighbors, and relatives, even on a regular basis, but rather because of the specific conduct described above. The City of Miami Beach permits churches, synagogues, and other religious institutions to operate freely in every zoning district of the City except the RS-4 single-family district and has enforced the single-family residential limitation equally against Christian, Jewish and “non-traditional” religious groups. Numerous other residential zones of the City, constituting at least 50% of the City’s territory, specifically and expressly list churches and synagogues as permitted users. 1 The “notice of violation,” mentioned in the stipulations was issued because of the City’s view that Appellees’ twice daily performance of religious ceremonies on their property conflicts with use restrictions embodied in City Ordinance No. 1891 § 6-1 (“the Ordinance”). 2 This conclusion stems from the City’s position that religious ceremonies conducted on the Grosz property occasionally constitute organized, publicly attended religious services. 3 II. PROCEDURE BELOW AND ISSUE ON APPEAL Appellees sued the City in the United States District Court for the Southern District of Florida, seeking a determination that the Ordinance is unconstitutional on its face because of overbreadth and vagueness, and unconstitutional as it was sought to be applied to them. Appellees prayed for declaratory and injunctive relief and damages. Following an evidentiary hearing, the district court granted a preliminary injunction which restrained the City from bringing criminal prosecutions against Ap-pellees based on “conduct or circumstances which were the subject of the notice of violation.” The court denied the parties’ cross motions for summary judgment, finding that genuine issues of material fact still existed. Because both parties were deter *733 mined to resolve the case in a summary procedure, they prepared and filed the joint stipulation of facts reproduced above. Based on these stipulations, the trial court granted summary judgment as to the merits of plaintiffs’ claims. The court found the ordinance neither vague nor overbroad, and thus not facially unconstitutional. However, the City’s application of the zoning code was found to impose a burden on Appellees’ free exercise of religion. Concluding that the City’s interest in enforcing its zoning laws did not rise to the level of a compelling state interest, the court held the Ordinance to be unconstitutional as applied. This appeal follows. The issue facing this Court is whether the trial court properly balanced the competing governmental and religious interests. 4 Because we conclude that the court below erred in its balancing, we reverse the finding of unconstitutionality. III. AN EXPEDITION INTO FREE EXERCISE DOCTRINE Interpretation and application of the free exercise clause has created, within that field of constitutional doctrine, areas dotted by unanswered questions. Fortunately, those unanswered questions do not hinder resolution of the issues in this case. A tour through the doctrine including the areas of uncertainty, will serve, however, to highlight the reasons behind our confidence of result in this case. In the course of this tour, we first traverse the surface of two threshold principles of free exercise doctrine. We then turn to the doctrinal centerpiece, balancing itself, and discuss the weighing of government interests, the weighing of religious interests, and the reaching of a final equilibrium. A. Thresholds Before a court balances competing governmental and religious interest, the challenged government action must pass two threshold tests. The first test distinguishes government regulation of religious beliefs and opinions from restrictions affecting religious conduct. The government may never regulate religious beliefs; but, the Constitution does not prohibit absolutely government regulation of religious conduct. Given a regulation’s focus on conduct, government action passes this first threshold. Braunfeld v. Brown, 366 U.S. 599, 603, 81 S.Ct. 1144, 1146, 6 L.Ed.2d 563 (1961); Cantwell v. Connecticut, 310 U.S. 296, 303-304, 60 S.Ct. 900, 84 L.Ed. 1213 (1940). 5 The second threshold principle requires that a law have both a secular purpose and a secular effect to pass constitutional muster. First, a law may not have a sectarian purpose — governmental action violates the Constitution if it is based upon disagreement with religious tenets or practices, or if it is aimed at impeding religion. Braunfeld v. Brown, supra, 366 U.S. at 607, 81 S.Ct. at 1148. Second, a law violates the free exercise clause if the “essential effect” of the government action is to influence negatively the pursuit of religious activity or the expression of religious belief. Id. *734 This is not to say that any government actions significantly affecting religion fail this threshold test. Rather, any nonsecular effect, regardless of its significance, must be only an incident of the secular effect. 6 Past these two thresholds, we now begin our journey into balancing. B. Balancing If a government action challenged under the free exercise clause survives passage through the belief/conduct and secular purpose and effect thresholds, the court then faces the difficult task of balancing government interests against the impunged religious interest. This constitutional balancing is not a simple process. We gain some sense of direction, however, from one basic principle: the balance depends upon the cost to the government of altering its activity to allow the religious practice to continue unimpeded versus the cost to the religious interest imposed by the government activity. This principle marks the path of least impairment of constitutional values. And, although previous court balancings provide us more detailed guidance in assigning weights to the competing values, the balancing map is far from complete. While we turn to examine previous precedent for the guidance that it can supply, we must keep in mind that our journey is into an area of the free exercise landscape where solid ground occasionally gives way to crevices of uncertainty. 1. The Burden on the Government. —In general, the burden on the governmental interest depends upon the importance of the underlying policy interests and the degree of impairment of those interests if the regulation were changed to impose no burden on religious conduct. One principle that has emerged in free exercise doctrine, the “least restrictive means test,” reflects the logic of our calculus. That is simply, if the government can effectuate its policy through a nonburdening technique the degree of impairment equals zero. In Cantwell v. Connecticut, 310 U.S. 296, 60 S.Ct. 900, 84 L.Ed. 1213 (1940), the Supreme Court reversed the convictions of three Jehovah’s Witnesses for breach of the peace and for soliciting contributions without a state certificate. The defendants’ conduct consisted of going house to house and playing an anti-Catholic record to those who would listen. In reversing the convictions, the Supreme Court reasoned that the government’s actions went beyond the least drastic, most narrowly constructed means of accomplishing state goals. Prevention of fraud and maintenance of public safety were held to require neither a discretionary licensing system for religious solicitation, nor arrest of nonthreatening, nonabusive individuals for breach of the peace. The government did not need to burden religious practice to accomplish its secular goals. In cases like Cantwell, where the government suffers no burden in changing its actions to accommodate religious conduct, any sectarian interest regardless of how lightly it weighs wins the balancing. Properly analyzed, therefore, the least restrictive means test constitutes just a special application of the general balancing test. Free exercise doctrine suggests another step in implementing the government burden formula outlined above: we must consider the impact of a specific, religion-based exception upon government’s policy objectives. In contrast to the clarity of the least restrictive means test, however, ambiguities pervade the standard for determining whether a religion-based exception is constitutionally mandated. In Braunfeld v. Brown, supra, 366 U.S. 599, 81 S.Ct. 1144, 6 L.Ed.2d 563, the Supreme Court upheld Pennsylvania’s Sunday retail sales prohibition against a free exercise challenge. A merchant and member of the Orthodox Jewish Faith claimed that the mandated Sunday closing prevented him from compensating for the business he lost by following his faith’s command that he not work on Saturday. In rejecting his contention that the state must grant a religion-based exception to the Sunday closing laws, the *735 Court was deferential to government assertions that any exception would be unworkable. The court relied on “reason and experience” in concluding that potential problems in administration and enforcement necessitated the law’s blanket application. In contrast, Sherbert v. Verner, 374 U.S. 398, 83 S.Ct. 1790, 10 L.Ed.2d 965 (1963), while declining to overrule Braunfeld in any respect, applied a religious exception standard much more exacting of the government. Sherbert struck down a state’s application of its law where the effect was to pressure a worker to abandon her religious convictions regarding the day of rest. The worker had been discharged from her job because of her refusal to work on Saturday, the Sabbath day of her faith. Application of the state’s unemployment compensation eligibility rules resulted in denial of her claim for unemployment compensation benefits. In discussing the government’s asserted interests in not granting an exception to the eligibility regulations the Court rejected, as unsupported by any proof in the record, the contention that an exception would induce fraudulent compensation claims based on feigned religious objections. Government arguments that administrative problems would result from having to determine the legitimacy of exemption claims and that granting exceptions would undermine the soundness of the unemployment fund also failed to impress the Court. But then, in United States v. Lee, 455 U.S. 252, 102 S.Ct. 1051, 71 L.Ed.2d 127 (1982), government arguments similar to the ones that failed in Sherbert convinced the Court that a religion-based exception to a challenged law would be too costly. The free exercise claim in Lee stemmed from the federal government’s collection of social security taxes from all employers. Although the payment of social security taxes violated the religious beliefs of the Amish, the Court held that the government’s interest in the integrity of the social security system outweighed the burden on religion. The rejection of a religion-based exception rested explicitly upon concerns about the potential impact of such an exception on the program’s fiscal soundness. Conflicting indications from Sherbert on one hand and Braunfeld and Lee on the other leave open important questions as to when a religious exemption is mandated: How much weight should be given the potential for feigned religious claims for exemption? To what extent can a court consider administrative costs and inconvenience as opposed to necessary impairment of the primary policy goal? Generally, what degree of impact upon the government’s broad policy goal should be tolerated before rejecting exceptions? 2. The Burden on Religion. — Having examined existing guidelines for measuring the burden on government, our journey changes direction and we turn to the religion side of the calculus. The formula for calculating the burden on religion essentially parallels the formula used to figure the government’s burden. The importance of the burdened practice within the particular religion’s doctrines and the degree of interference caused by the government both figure into the calculus. Symmetry between the government burden formula and the religion burden formula is not complete, however. Although a focus on the importance of a religious belief occasionally proves instructive, 7 the determination of how important the practice is can prove more difficult than evaluating the importance of a government policy. Religious doctrine may exist, to a large extent, as a reflection of individual adherents’ interpretations. Reliable indicia of the importance of particular religious conduct may be hard to find. Courts, therefore, often restrict themselves to determining whether the challenged conduct is rooted in religious belief or involves only secular, philosophical or personal choices. See, e.g., Wisconsin v. Yoder, 406 U.S. 205, 215-216, 92 S.Ct. 1526, *736 32 L.Ed.2d 15 (1972); EEOC v. Mississippi College, 626 F.2d 477, 488 (1980). Only conduct flowing from religious belief merits free exercise protection; no weight measures on the side of religion unless the government action ultimately affects a religious practice. This latter determination is also difficult, but it is one that must be made in all free exercise challenges. Wisconsin v. Yoder, supra, 406 U.S. at 215-216, 92 S.Ct. at 1533. Id. Finer distinctions, as to the weight of the burden, must usually be based upon the degree of interference element in the formula. Focusing on the degree of interference factor, it is clear that when the government totally precludes religious conduct by imposing criminal sanctions, the burden weighs at its heaviest. See Wisconsin v. Yoder, supra, 406 U.S. at 215, 92 S.Ct. at 1533; Cantwell v. Connecticut, supra, 310 U.S. at 307-308, 60 S.Ct. at 904-905. Less clear is how courts are to estimate the burden on religion when the government does not criminally prohibit religious conduct but infringes upon free exercise in some other way. In Braunfeld v. Brown, supra, 366 U.S. 599, 81 S.Ct. 1144, 6 L.Ed.2d 563, the Supreme Court appeared to indicate that burdens on religion, other than affirmative impositions of criminal penalties, weigh lightly against government interests. Creating a direct/indirect dichotomy with which to classify burdens on religion, the Court said: In [cases where religious practices conflict with the public interest], to make accommodation between the religious action and an exercise of state authority is a particularly delicate task, [citations omitted] because resolution in favor of the State results in the choice to the individual of either abandoning his religious principle or facing criminal prosecution. But ... this is not the case before us because the statute at bar does not make unlawful any religious practices of appellants; the Sunday law simply regulates a secular activity and, as applied to appellants, operates so as to make the practice of their religious beliefs more expensive. Braunfeld v. Brown, supra, 366 U.S. at 605, 81 S.Ct. at 1147. The Court further noted that to strike down legislation which imposes only indirect burdens would, absent the most critical scrutiny, “radically restrict the operating latitude of the legislature.” Id. This distinction led to a conclusion that, if the State regulates conduct by enacting a general law within its power, the purpose and effect which is to advance the State’s secular goals, the statute is valid despite its indirect burden on religious observance unless the State may accomplish its purpose by means which do not impose such a burden. Id. at 1148. Under such an analysis the real-world impact of a burden — in Braun-feld, the merchant’s choice between his religion and his business — becomes a second level inquiry. By creating a presumption of validity for laws meeting the secular purpose and effect standard and imposing only indirect burdens, Braunfeld produced a wide fissure through which certain government actions can escape thorough-going balancing. The Supreme Court moved to narrow the width of that fissure two years later in Sherbert v. Verner, supra, 374 U.S. 398, 83 S.Ct. 1790, 10 L.Ed.2d 965. The state’s defense of its unemployment compensation eligibility rules centered on the fact that the burden on religious conduct did not result from criminal liability. Rather, the state maintained that the burden on religious conduct flowed “indirectly” from state welfare legislation. The Court rejected this argument in broad terms. Here not only is it apparent that appellant’s declared ineligibility for benefits derived solely from the practice of her religion, but the pressure upon her to forego that practice is unmistakable. The ruling forces her to choose between following the precepts of her religion and forfeiting benefits, on the one hand, and abandoning one of the precepts of her religion in order to accept work, on the other hand. Governmental imposition of such a choice puts the same kind of burden upon the free exercise of religion as *737 would a fine imposed against appellant for her Saturday worship. Sherbert v. Verner, supra, 374 U.S. at 404, 83 S.Ct. at 1794. Recently, the Court employed identical reasoning to strike down a state’s denial of unemployment compensation benefits when a worker quit because of religious objections to working in his employer’s weapons production department. Thomas v. Review Board of the Ind. Employment Security Div., 450 U.S. 707, 101 S.Ct. 1425, 67 L.Ed.2d 624 (1981). Although Sherbert specifically avoids overruling Braunfeld, combined with Thomas it significantly transforms Braunfeld's direct/indirect distinction. Affirmative imposition of criminal penalties no longer distinguishes direct from indirect governmental burdens. Absent the illumination that the penalty versus benefit distinction provided, the contours and consequences of the direct/indirect distinction are less obvious. It is especially unclear whether Braunfeld still creates presumptive validity for governmental actions that impose only indirect burdens. 8 No Supreme Court case since Braunfeld has relied on the direct/indirect rhetoric to uphold governmental action. One unifying principle for calculating the burden on religion remains intact, however. Governmental actions that burden religious conduct by focusing on the conduct itself, regardless of whether the action constitutes an imposition of a penalty or a denial of a benefit, impose heavy burdens on religion. 3. The Final Balance.—In our examination of the methods by which burdens on government and religion are calculated two principles surfaced that merit repetition here. The first flows from the least restrictive means test. When the government can as easily achieve its policy objective without burdening religious conduct, the burden on the government is zero and the scale necessarily reads against upholding the government action. Similarly, the second principle operates when the burden on religion is zero. When the burden imposed by the government rests on conduct rooted only in secular philosophy or personal preference, the scale always reads in favor of upholding the government action. Another principle appears if we reexamine the Supreme Court’s balancings in Sherbert v. Verner, supra, 374 U.S. 398, 83 S.Ct. 1790, 10 L.Ed.2d 965, Thomas v. Review Board of the Ind. Employment Security Div., 450 U.S. 707, 101 S.Ct. 1425, 67 L.Ed.2d 624, and Wisconsin v. Yoder, supra, 406 U.S. 205, 92 S.Ct. 1526, 32 L.Ed.2d 15. The Supreme Court suggests in Sherbert and Thomas that once a government action passes the threshold tests of conduct focus and secular purpose and effect, a showing of “compelling state interest” on the government side will justify inroads on religious liberty. Yoder articulates the standard differently, referring to “interests of the highest order.” Substantively, however, the same message emanates from all three cases. If avoiding the burden on government rises to the very upper ranges of government interest, a free exercise challenge will fail. Beyond this basic principle, though, we must make our way without the guiding beacons of sure and certain legal principles. When the government interest in continuing an action that burdens religion does not rise to those upper ranges, the standards are vague as to what government interests will counterbalance free exercise interests. In Gillette v. United States, 401 U.S. 437, 91 S.Ct. 828, 28 L.Ed.2d 168 (1971), individuals, who opposed the Vietnam War on the grounds of conscience and religion and desired exemption from military service, asserted that the government’s restriction of conscientious objector status to those who objected to all war rather than a particular war violated their free exercise rights. The Court held the government’s “substantial” interest in maintaining a fair and even *738 handed system of exempting individuals from military service sufficient to override the asserted burden on religion. Johnson v. Robison, 415 U.S. 361, 94 S.Ct. 1160, 39 L.Ed.2d 389 (1974) indicated that a whole range of government interests can justify burdening religion, the requisite level depending upon the significance of that burden. There, the free exercise challenge involving the government’s denial of veteran’s educational benefits to conscientious objectors who performed alternative civilian service failed. The government’s interest proved to be of “a kind and weight” to justify the very minor burden on religion. This language suggests that an ad hoc balancing is appropriate when existing, broad principles do not command the result. Viewed collectively, relevant Supreme Court cases provide only the most general hints for reaching a proper, final balance of free exercise and government interests. Little background exists as to what constitutes a “substantial” government interest. And, courts called upon to make an ad hoc balance receive precedent guidance only in the discrete factual contexts involved in prior balancings. IV. A REPEAT TRIP: THE BALANCE IN THIS CASE We know now at least the broad contours of the path we must follow in making an accommodation between the City’s interest in zoning enforcement and Appellees’ free exercise interest. We must first apply the conduct/belief and secular effect and purpose tests. Should the government action pass these tests we- then must balance the cost to the government of altering its activity to allow the religious practice to continue unimpeded against the cost to the religious interest imposed by the government action. A. Thresholds In this case, the thresholds pass quickly beneath our feet. The City’s zoning law affects prayer and religious services, and so involves conduct. Therefore, balancing not absolutism is appropriate. That the law has both secular purpose and effect is noncontroversial. No one contends that zoning laws are based upon disagreement with religious tenets or are aimed at impeding religion. Similarly, given zoning’s historical function in protecting public health and welfare, see Village of Belle Terre v. Boraas, 416 U.S. 1, 94 S.Ct. 1536, 39 L.Ed.2d 797 (1974), and the incidental nature of the asserted burden on religion, the essential effect of zoning laws is clearly secular. B. The Burden on Government The City of Miami Beach asserts a governmental interest in enforcing its zoning laws so as to preserve the residential quality of its RS-4 zones. By so doing the City protects the zones’ inhabitants from problems of traffic, noise and litter, avoids spot zoning, and preserves a coherent land use zoning plan. The Supreme Court has acknowledged the importance of zoning objectives, stating that segregation of residential from nonresidential neighborhoods “will increase the safety and security of home life, greatly tend to prevent street accidents, especially to children by reducing traffic and resulting confusion, ... decrease noise ... [and] preserve a more favorable environment in which to raise children.” Village of Euclid v. Ambler Realty Co., 272 U.S. 365, 394, 47 S.Ct. 114, 120, 71 L.Ed. 303 (1926). See also, Village of Belle Terre v. Boraas, supra, 416 U.S. at 9, 94 S.Ct. at 1541. The City asserts a significant governmental objective in the case at bar. Gatherings for organized religious services produce, as do other substantial gatherings of people, crowds, noise and disturbance. In fact, the parties’ stipulations reveal that the City was acting pursuant to neighbors’ complaints to end the disturbance caused by Appellees’ conduct. Given this total inconsistency between the accomplishment of the City’s policy objectives and the continuance of Appellees’ conduct, the government action in this case easily passes the least restrictive means test. Doctrine also requires that we consider the impact of a religious based exemption to zoning enforcement. In that regard we *739 find that granting an exception would defeat City zoning policy in all neighborhoods where that exception was asserted. Maintenance of the residential quality of a neighborhood requires zoning law enforcement whenever that quality is threatened. Moreover, no principled way exists to limit an exception’s costs just to the harm it would create in this case. Crowds of 500 would be as permissible as crowds of 50. Problems of administering the exception such as distinguishing valid religious claims from feigned ones, therefore, need not even be considered. A religion based exception would clearly and substantially impair the City’s policy objectives. Together, the important objectives underlying zoning and the degree of infringement of those objectives caused by allowing the religious conduct to continue place a heavy weight on the government’s side of the balancing scale. C. The Burden on Religion In calculating the burden on religion, we first determine whether the conduct interfered with constitutes religious practice. The religion of Appellee, Naftali Grosz, requires him to conduct religious services twice daily in the company of at least ten adult males. Solicitation of neighborhood residents to attend and the participation of congregations larger than ten, the conduct on which the City based its notice of violation, are not integral to Appellees’ faith. However, the trial judge made no findings, and the record is not clear regarding the extent to which, if any, these objectionable but nonessential practices aid Appellees in gathering ten men and conducting the required services. We assume therefore, that the nonessential practices further the religious conduct. We must also assume, then, that Appellees suffer some degree of burden on their free exercise rights. Turning to the significance of that burden, we note that Miami Beach does not prohibit religious conduct per se. Rather, the City prohibits acts in furtherance of this conduct in certain geographical areas. The relevant question is to what degree does the City’s exclusion of Appellees’ activities from RS-4 zoned areas burden religious conduct. The City’s zoning regulations permit organized, publicly attended religious activities in all zoning districts except the RS-4 single family districts. The zones that allow religious institutions to operate constitute one half of the City’s territory. Appellees’ home lies within four blocks of such a district. Appellees do not confront the limited choice of ceasing their conduct or incurring criminal liability. Alternatively, they may conduct the required services in suitably zoned areas, either by securing another site away from their current house or by making their home elsewhere in the city. We cannot know the exact impact upon Appellees, in terms of convenience, dollars or aesthetics, that a location change would entail. The burden imposed, though, plainly does not rise to the level of criminal liability, loss of livelihood, or denial of a basic income sustaining public welfare benefit. 9 In comparison to the religious infringements analyzed in previous free exercise cases the burden here stands towards the lower end of the spectrum. D. The Final Balance In discussing the process for reaching a final balance we noted earlier that courts are frequently forced to undertake an ad hoc balancing when existing free exercise doctrine does not command a specific result. But, such ad hoc balancings need not always be based only on appeals to a court’s basic intuitive sense. Fortunately, the instant case arises in a factual context in which substantial, relevant case precedent exists to guide our balancing. This case is not the first to involve balancing government’s interest in restricting the location of *740 religious conduct. Prince v. Massachusetts, supra, 321 U.S. at 169, 64 S.Ct. at 443, upheld the convictions of a Jehovah’s Witness for violating child labor laws. State law prohibited children from selling anything in the streets. The defendant, the custodian of a nine-year-old girl, asserted the child’s free exercise rights to sell religious literature on public streets. The defendant claimed that the child’s distribution of literature accorded with religious tenets of the Jehovah’s Witness faith. The Supreme Court held the state’s interest in children’s welfare permitted the state to wholly prohibit children from selling religious literature on the streets. With respect to adults, the Court observed that although the literature distribution could not be altogether banned, it could be “regulated within reasonable limits in accommodation to the primary and other incidental uses [of streets].” Id. at 169, 64 S.Ct. at 443. International Society for Krishna Consciousness v. Eaves, 601 F.2d 809 (5th Cir.1979) involved the Society’s First Amendment challenge to a municipally owned airport’s restrictions on distribution of literature and solicitation of funds. Accepting the Society’s claim that religious tenets required followers to solicit funds, this Court nevertheless upheld the airport restrictions. Government may regulate place and manner of religious expression as long as there is no content classification and so long as the regulation is reasonable. Id. at 827. Admittedly, restriction of religious conduct on public streets and in airports is less burdensome than restriction of such conduct on an individual’s property. The Prince and the Krishna cases, however, establish the principle that government can exercise its police powers to limit the place and manner of religious conduct, despite a significant burden on free exercise. A decision from the Sixth Circuit, Lakewood Congregation of Jehovah’s Witnesses v. City of Lakewood, 699 F.2d 303 (1983), gave that principle specific application in a zoning context. In Lakewood, a church congregation challenged a zoning ordinance that prohibited their building a church on a lot they had previously purchased. The Sixth Circuit characterized the infringement on religious freedom as an “inconvenient economic burden” and a “subjective aesthetic burden.” The City of Lakewood’s interest in creating residential districts to promote its citizens’ health and well being was held to outweigh the congregation’s religious interests. We think Lakewood’s balancing process reached the correct result in a ease very similar to this one. The Sixth Circuit faced, if anything, a closer balance than the one called for today — as opposed to the one half of Miami Beach territory where Appellees may conduct their religious services, the City of Lakewood permits church buildings on only around ten percent of its land. We glean a final bit of support for our holding from a Supreme Court dismissal for want of a substantial federal question in Corporation of the Presiding Bishop v. City of Porterville, 338 U.S. 805, 70 S.Ct. 78, (1949). The state court had held that churches may be excluded from residential areas consistently with the free exercise clause. Justice Vinson, writing in a later case, American Communications Ass’n v. Douds, 339 U.S. 382, 397, 70 S.Ct. 674, 94 L.Ed. 925 (1950), explained the Supreme Court’s dismissal in the California case. When the effect of a statute or ordinance upon the exercise of First Amendment freedoms is relatively small and the public interest to be protected is substantial, it is obvious that a rigid test requiring a showing of imminent danger to the security of the nation is an absurdity. We recently dismissed for want of substan-tiality an appeal in which a church group contended that its First Amendment rights were violated by a municipal zoning ordinance preventing the building of churches in certain residential areas. V. CONCLUSION Our journey at an end, we now examine the scale and determine how the two conflicting constitutional values, free exercise rights and the state police power, are to be *741 accommodated. On the free exercise side of the balance weighs the burden that Ap-pellees bear of conducting their services in compliance with applicable zoning restrictions or relocating in a suitably zoned district. Countering on the government’s side is the substantial infringement of the City’s zoning policy that would occur were the conduct allowed to continue. The Prince and Krishna analysis regarding time, manner, and place restrictions on religion supports the view that the relative weights of the burdens favor the government. The Supreme Court’s pronouncement in Douds argues even more strongly for this conclusion. The judges who have precedentially performed balancings on the free exercise trapeze have encountered great difficulty with the weights and measures involved. Bal-ancings must avoid constitutionalizing secu-larly or sectarianizing the Constitution. In this area, where religious guarantees of the Constitution compete with the rights of government to perform its function in the modern era, certitude is difficult to attain. All should understand that we have not written today for every situation in which these issues might arise — only that we have done our best as amateur performers in solving this very, very delicate problem. We who perform on this flying trapeze may not always be daring and young, but we must avoid that slip that could take us into the doctrinal confusion below. We find that the burden upon government to allow Appellees’ conduct outweighs the burden upon the Appellees’ free exercise interest. Therefore, we reverse the trial court judgment as to the Ordinance’s unconstitutionality as applied. We remand to the district court with instructions to enter judgment in favor of the City. REVERSED and REMANDED. Footnotes [1] Record at 383-85. [2] City of Miami Beach Zoning Ordinance No. 1891 § 6-1, applicable to the Appellees’ property RS — 4 classification, provides in relevant part: B. Uses Permitted. No land, water or structure may be used in whole or in part, except for the following uses: 1. Single-family detached dwelling. 3. Accessory uses of above uses. “Single-family dwelling” is defined at § 3-1 of the ordinance as “a building designed for or occupied exclusively by one family.” “Accessory use” is defined at § 3-2 as “a subordinate use which is incidental to and customary in connection with the main building or use and which is located on the same lot with such main building or use.” The City reasoned that the omission of organized religious centers as permitted uses under § 6-1, while various other of the City’s zoning classifications specifically allow operation of houses of worship, made Appellees’ activities a clear use violation. [3] The joint stipulations of fact contain the following sentence: “The City did not and would not prosecute plaintiffs for praying in their home, with ten friends, neighbors and relatives, even on a regular basis, but rather because of the conduct described above.” Regardless of the City’s intent as to prosecutions for “other kinds” of conduct, we limit today’s ruling to the City’s actions regarding specific conduct documented in this case. [4] The Constitution’s protection of religious freedom derives most directly, of course, from the First Amendment: “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof ...” The free exercise clause of the First Amendment, relevant in this case, gains application to state and local governments through the Fourteenth Amendment’s due process clause. Cantwell v. Connecticut, 310 U.S. 296, 60 S.Ct. 900, 84 L.Ed. 1213 (1940). That states and localities should be able to effectively exercise police powers to benefit their citizens is implicit throughout the structure of the Constitution. In particular, the Tenth Amendment supports the exercise of such powers: “The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or the people.” [5] The belief/conduct distinction has survived, see, e.g., United States v. Middleton, 690 F.2d 820, 824 (11th Cir.1982), cert. denied, — U.S. —, 103 S.Ct. 1497, 75 L.Ed.2d 929; United States v. Holmes, 614 F.2d 985, 989 (5th Cir.1980), despite criticism by commentators, see L. Tribe, American Constitutional Law, § 14-9, at 837-38 (1978) and Giannella, Religious Liberty, Nonestablishment, and Doctrinal Development, Part I The Religious Liberty Guarantee, 80 Harv.L.Rev. 1381, 1387 (1967). [6] L. Tribe, supra note 5, § 14-9 at 838-40. [7] See Brown v. Dade Christian Schools, Inc., 556 F.2d 310, 321 (1977) (Goldberg, J., specially concurring). [8] At the very least the word “indirect” does not signal presumptive validity of a law as it did in Braunfeld. In striking down Indiana’s denial of unemployment benefits to an individual who quit his job because of religious convictions, the Court said, “While the compulsion may be indirect, the infringement upon free exercise is nonetheless substantial.” Thomas, supra, 450 U.S. at 718, 101 S.Ct. at 1432. [9] See Prince v. Massachusetts, 321 U.S. 158, 64 S.Ct. 438, 88 L.Ed. 645 (1944) (criminal liability); Gillette v. United States, 401 U.S. 437, 91 S.Ct. 828, 28 L.Ed.2d 168 (criminal liability); Braunfeld v. Brown, supra, 366 U.S. 599, 81 S.Ct. 1144, 6 L.Ed.2d 563 (loss of livelihood); Sherbert v. Verner, supra, 374 U.S. 398, 83 S.Ct. 1790, 10 L.Ed.2d 965 (loss of unemployment compensation).