The Establishment Clause is a particular example of the Constitution’s separation of powers. The concept of separation of powers is often ascribed solely to the question of the proper relationship between the federal branches, but the entire Constitution is governed by the overarching principle that society is best served when centers of power are kept separate. Indeed, the most important contribution the American experiment has made to liberty may well be its extension of the concept of separation to the church-state relationship.
The very natures of church and state have made their separation frustratingly difficult. Each is authoritative, each is capable of exercising power in the public sphere with great alacrity, and each is more than willing to use the other to accomplish its own ends. The Supreme Court has had its hands full, to say the least.
In the midst of a chorus of criticism trained on supposed inconsistencies in the Supreme Court’s Establishment Clause cases, the Court has been admirably forthright about the fact that the establishment cases pose formidable challenges to judgment. The decisions have “reflected the need to rely on careful judgment–not simple categories. . . .” 2 The Justices have not flinched from their difficult duty under the Establishment Clause and have displayed relatively little angst over the conclusions they have reached. While Justice Scalia has been an intense critic of the Court’s apparent inconsistencies, 3 the Court has simply acknowledged the inevitable difficulty of the issues and forged ahead. This simultaneous admission of the difficult judgments posed by establishment questions and the Court’s relative comfort with its conclusions was captured in Chief Justice Burger’s opinion for the Court in Walz v. Tax Commission, 4 when  he stated: “This is a ‘tight rope’ and one we have successfully traversed.” 5 Underneath the tight rope lie crocodiles hungry for power–church and state among them.
There seems to be an almost irresistible impulse to find a particular “test” that would lead to clear answers for all the difficult questions posed by the Establishment Clause. The curious history of the so-called “Lemon test” illustrates my point. In Lemon v. Kurtzman, 6 the Court articulated several factors that had been used in previous establishment cases. 7 The Court did not claim to be crafting a bright-line test that would make it possible to leave judgment behind in all future establishment cases but rather characterized its mission with some modesty. These were the factors that had been used in previous cases, the Court told us, and they were now being gathered together. 8 Within a matter of years, the Lemon factors were transformed into the “Lemon test,” and this so-called test was subjected to withering criticism for its failure to predict results in close cases. 9 Yet, the Lemon Court’s summary of previous standards was not intended to foreclose further analysis under the Clause. This criticism of the Lemon factors is misplaced. The fact that it did not forestall the need for the further application of judgment in establishment cases is not remarkable. The hard cases are inevitable. In a particularly difficult case in which to articulate definitive establishment standards, Board of Education of Kiryas Joel Village School District v. Grumet, 10 Justice O’Connor declared what the case law had already revealed: the Establishment Clause lends itself to no “Grand Unified Theory.” 11
 The task of balancing church-state power is unavoidably difficult. Power does not exist in static form. Rather, it is plastic in its permutations and infinitely creative in its drive to realize itself. Thus, the Establishment Clause is charged with the task of policing a constantly changing boundary between two dynamic social entities. It is hardly surprising that the “cases arising under these Clauses have presented some of the most perplexing questions to come before the Court.”12 The search for a unified theory is a diversion from the hard task assigned by the Establishment Clause.
There can be no grand unified doctrine under the Clause because of a congeries of factors, including the ontological status of religion, the nature of power, and the larger constitutional milieu within which church-state relations are set. On the basis of these circumstances, this Article rejects the search for a unified doctrine. Instead, it focuses attention on constitutional history as well as political and theological realities to explain and clarify the contingent nature of the Court’s doctrine. In the final section, this Article applies these insights to school voucher schemes. Like other establishment issues, school vouchers require the courts to apply careful and nuanced judgment regarding a particular scheme against a backdrop of church-state relations.
I. POWER IN THE CONSTITUTIONAL DOMAIN
To say that there is no bright-line rule that will generate reliable answers in every establishment case is not to say that there are no principles capable of guiding the courts. The most important constitutional guiding principle is that all concentrations of power should be kept separate to the extent possible. The Establishment Clause rests on three constitutional fundamentals intended to distribute power throughout society. First, all concentrations of power can be turned toward tyrannical ends. 13 Second, liberty is best achieved by promoting a balance of power between social entities within society. 14 Third, religion will operate in the public, political  sphere.15 Through the Free Exercise and Establishment Clauses, the Constitution attempts to achieve a rough balance of power between church and state, simultaneously strengthening and cabining each. 16
As a historical matter, it is best to begin any analysis of the Bill of Rights, including the Establishment Clause, with the debates at the Constitutional Convention. The Federalists, and particularly James Madison, did not view the First Amendment as a new substantive limitation on the main text of the Constitution but rather as an iteration of the fact that the enumerated powers had not granted Congress power over religion. In other words, the First Amendment was thought to be a consistent, though wordier, statement of principles already inherent in the original draft. Thus, it makes sense to examine the Convention to divine the Framers’ views on religion.
At the Constitutional Convention, religion was treated like any other social entity capable of holding power. 17 Wherever the Framers looked, they accepted as a fact that men could and would use their power to accomplish evil, rather than good, ends: “From the nature of man we may be sure, that those who have power in their hands will not give it up while they can retain it. On the contrary we know they will always when they can rather increase it.” 18 In James Madison’s words, “The truth was that all men having power ought to be distrusted to a certain degree.” 19
There was broad consensus on the end to be avoided–tyranny from any social center of power. The Framers typically focused on the choice of the best means to avoid tyranny. All those with power were assumed to be interested in exceeding their authority. 20 Thus, this new national government  the Framers were crafting was constructed on the presupposition that its power was likely to be misused. 21
To avoid undue concentrations of power, the Framers sought to achieve a balance of power between identified social entities. 22 They envisioned society as a solar system or a watch. 23Each part had an assigned and independent role. But they were roles that were alsointerdependent. One force was intended to counterbalance every other force. Society was to be governed by the concept of separation of powers, a phrase unfortunately relegated in contemporary constitutional lore solely to the three branches of the federal government.
For the Framers, the appropriate exercise of power fell between two extremes. The one holding power could exercise it ineffectually or overbearingly. Either extreme was unacceptable. 24 Thus, James Wilson catalogued the types of bad, or tyrannical, government as follows: “Bad governments are of two sorts. First, that which does too little. Second, that which does too much: that which fails through weakness; and that which destroys through oppression.” 25 The Articles of Confederation had produced a government that was too weak. The Framers desired to construct a government that was strong, but not too strong.
The Framers’ general attitude of distrust was not limited to the government or politicians. Even religion was worthy of distrust. While the Framers acknowledged religion’s power and potential goodness, 26 they were more than a little concerned that religion could exceed its appropriate bounds in the political sphere. 27 When the Framers did refer to religion during the Debates, it was not for the purpose of singing its praises but  rather for the purpose of pointing to the fact that it, like all other social entities, is capable of overstepping its bounds. In particular, Madison pointed out that religious parties had lobbied the British Parliament to abuse its power to regulate “the qualifications both of the electors, and the elected . . . .” 28 This view was consistent with Madison’s earlier views expressed in his famous Memorial and Remonstrance in which he criticized religious assessments as a “dangerous abuse of power.” 29 As the Constitution worked itself out in practice, James Madison continued to voice such fears. In the years following his presidency, Madison warned of the “danger of silent accumulations & encroachments by Ecclesiastical Bodies,” saying that they “have not sufficiently engaged attention in the U.S.” 30
This distrustful attitude toward religion is hardly surprising; many in the states had fled Europe to escape religious oppression at the hands of powerful churches. 31 Moreover, one must take into account the predominantly Protestant mind set of many of the Framers. The Reformation occurred a brief two centuries before the Framing. Europe witnessed a series of events in that era that will continue to reverberate as long as Protestant sects remain distinct from the Catholic Church. The eighteenth century Protestant could not help but know and believe that religion had abused the people’s trust and undermined the public good at some time or another. This distrust of religion is evidenced in James Madison’s criticism of the Anglican Church in Virginia, when he complained that “pride, ignorance and knavery” prevailed among the priesthood while “Vice and Wickedness” afflicted the laity. 32 Indeed, inter-sect disputes at the time of the Framing were commonplace. 33
For the Framers, religion manifested itself in the political culture as religious sects or factions capable of wielding significant power. 34 The Framers valued religious liberty and therefore believed in protecting religion  from a potentially tyrannical state, but they equally regarded religion itself as potentially tyrannical. 35 In their view, religion was capable of applying political pressure in ways that were unacceptable in a republican democracy. 36 This complex attitude was basically Calvinist and in some senses paradoxical: they had faith in the good that human institutions can accomplish but also an acceptance of the fact that all human institutions are inevitably fallible. 37
In sum, a balance of societal powers is the Constitution’s graspable but elusive goal. Church and state not only fit into this model, but also demand such treatment if the maximum liberty is to be grasped.
II. THE POWER OF RELIGION
The constitutional matrix with its focus on the balance of power invites courts to assess–either overtly or covertly–the relative power of religion and state. Without doubt, the power of the government has increased since the time of the Framing. Equally, the power of religion has increased. At this stage in history, neither entity should give solace to those–like the Framers–who fear abuses of power.
A. Religion’s Power in Contemporary Society
There is a tendency in modern discourse to assume that religion is marginalized or trivialized in American society. 38 The implicit argument has been that as the state and culture have increased in power and scope, religion has remained relatively the same or lost ground and therefore has been proportionally disabled in the political sphere. 39 Another branch of this theory is that religion has been relegated to a smaller portion of people’s lives. The argument is that religion must compete against pop culture, the government, and the global marketplace for a spot in the believer’s life. The problem with the marginalization thesis is that it conflicts with sociological data and political facts.
 Since the 1993 publication of Stephen Carter’s tome to trivialization, The Culture of Disbelief, 40 the President and Congress have been striving mightily to prove that they take religion seriously. The Religious Freedom Restoration Act (“RFRA”), 41 which accorded religion more power against government than it has ever had in this country’s history, and the Federal Workplace Guidelines speak to the power of religion in politics today. 42 The huge margin by which RFRA was passed evidences the power of religion in Congress. 43 Moreover, RFRA’s legislative history, which is barren of critical analysis of its likely untoward effects on many elements of society including children, schools, prison administration, the environment, and historical and cultural preservation, demonstrates an almost mindless deference to religion that the Framers would have eschewed. The comments by members of Congress and the President on RFRA’s passage and the President’s remarks on the day the Guidelines were released make absolutely clear that the federal political branches are acutely aware of religion’s power in the marketplace and eager to serve it. 44 More recent hearings on the Religious Liberty Protection Act evidence a continuing zeal to serve religion. 45 Similar deference to religious interests can be found in the states. 46 The cases and statutes discussed below reinforce this political reality. 47
Father Andrew Greeley has presented statistical evidence that strongly indicates religion has not been marginalized in people’s lives. Indeed, he asserts that religion is just as much, if not more, a force in this era than it was in the supposedly deeply religious Middle Ages. 48
The case simply has not been made that religion’s power is weaker against the state than it was in the era of the framing. The Framers’ discussion of religion at the Constitutional Convention is just as insightful  today as it was then. They were not concerned about religion’s marginalization or the secularization of society. Nor were they infected by the Marxist or enlightenment dogma that holds that religion is a social fiction that will give way to a truer reality. The Framers treated religion not as a step in human social evolution, but rather as an unavoidable fact of human existence.
B. The Power of Religion to Obtain State Aid
Religion in the United States has tended to fare well in obtaining state aid. It has succeeded at the legislative level frequently and also fared increasingly well under Establishment Clause review of such legislation. Its successes bring to mind James Madison’s warning that “no man or class of men ought, on account of religion to be invested with peculiar emoluments or privileges.” 49
1. The Legislature: The Willingness of the State to Satisfy Religion’s Demands in the Political Arena
Just as Madison predicted, religion has not been reticent in the public realm or unwilling to approach the public trough. 50 In recent history, politically savvy groups, from the Christian Coalition to the Christian Science Church to the Coalition for the Free Exercise of Religion, have been active players in the political process eager for certain governmental favors. The very existence of the Establishment Clause cases discussed in the next section proves that religion has not been shy in the public realm. It would strain credulity to argue that the state benefits disputed in the establishment cases were simply the largesse and the initiative of state legislators. Common sense informs us that religious individuals and institutions have been pushing for such benefits for decades and will continue to do so. Indeed, these cases were before the Supreme Court and were therefore deemed worthy of lengthy and costly litigation only because the religious individual or institution desired to take advantage of particular public assistance. The Supreme Court’s cases only touch the tip of the political iceberg.
In each era, particular religious entities have taken center stage. Today, we hear most about two coalitions, the Christian Coalition and the Coalition for the Free Exercise of Religion. The question of their power is an empirical one. The following is an introduction to religious involvement  and successes in the legislative process.
The political muscle of the Christian right was forcefully expressed by Pat Robertson last year when he declared with some confidence that his organization would tell Congress what to do on a variety of issues. 51 The Christian Science Church also has been extremely active in the state legislatures lobbying for exemptions from mandatory childhood immunization for communicable diseases, medical care requirements, and child neglect and abuse laws. Forty-five states, including the District of Columbia, permit parents to forego immunization for their children for religious reasons. 52 Those believing in spiritual treatment also have sought to avoid criminal prosecution or civil penalties when a child is permanently disabled or dies as a result of the failure to obtain medical treatment. 53 They have succeeded on this score in a large number of states.54 For example, in Oregon, a faith-healing sect was responsible for the death of three children. The local prosecutor sought to indict members of the sect but could not because Oregon state law exempted from the child neglect laws those who failed to obtain medical care for their children because of religious beliefs. 55
Faith-healing sects also have had significant successes at the federal level. 56 Under recent Medicare regulations that were amended so as not to be so obviously sect-specific after they became the subject of a lawsuit and media attention, faith-healing “nurses” who had trained one week to  assist those near death received federal Medicare funding. 57 In other words, individuals who did not believe in medical care were receiving federal funds intended for medical care. Similarly, the most popular HMO bill in the 105th Congress contained a provision that would have permitted faith-healing nurses to receive HMO insurance funds, and to do so without satisfying the rigorous screening requirements attached to any medical provider’s request for payment. 58
The most impressive accumulation of power by a religious entity in recent years resides in the Coalition for the Free Exercise of Religion (the “Coalition”). This organization, which now consists of approximately eighty organized religions and some civil liberties groups, was formed to correct the Supreme Court’s free exercise decision in Employment Division v. Smith. 59 The Court inSmith held that the Free Exercise Clause did not require the state of Oregon to pay unemployment compensation to fired state-paid drug counselors who used a narcotic, peyote, even if it was used in a religious ceremony. 60 The Coalition’s first tack was to attempt to overrule Smith through congressional enactment, the Religious Freedom Restoration Act (“RFRA”).61 RFRA attempted to alter dramatically the existing balance of power between church and state by forcing every government that substantially burdened religious conduct to prove that its law was enacted for a “compelling interest” and was the “least restrictive means” of achieving that interest. 62 This new benefit for religion was granted without serious scrutiny by Congress. 63The Supreme Court, however, held RFRA unconstitutional as beyond Congress’s authority. 64
The Coalition’s defeat with RFRA did not halt its political momentum  or tarnish its political influence. Within weeks of the Court’s decision invalidating RFRA, the House held hearings in which members of the Coalition were permitted to vent about the decision. 65 The Coalition then turned to two separate tasks: introduction of state-level rfras in all fifty states and a new federalrfra. 66 They have paired up with the Rutherford Institute in the states and succeeded in obtaining such legislation in Florida and in Illinois as well as a state constitutional amendment in Alabama. 67 They also succeeded in guiding such legislation through the California legislature, though it was vetoed in its entirety by then-Governor Peter Wilson. 68
At the federal level, the Coalition attempted to reinstate the RFRA regime through other congressional powers, including the Commerce Clause, the Spending Clause, and Section 5 of the Fourteenth Amendment. The bill was entitled the Religious Liberty Protection Act of 1998. 69 The hearings for the bill were heavily stacked by members of the Coalition, and members of Congress appeared receptive to their message at that time. 70
The list of religiously motivated political lobbying is longer than this Article can possibly document, but in addition to the above, three other examples are worthy of mention. First, the Catholic Church has been quite active in lobbying for school vouchers, both from the pulpit and in the  public arena. 71 Various religious entities also have pushed for the charitable choice proposals that would funnel welfare funding from state or federal agencies into the hands of religious institutions. 72 Finally, the Native American Church has been extremely successful in asking for exemptions from generally applicable narcotics laws for the use of peyote in religious ceremonies. 73
 Whatever one thinks about any of these examples as a policy matter, and I would support some and not others, their very existence proves that religion is an energetic actor in the political realm that will push for its own ends. Religion is not a passive participant in the political process but rather a potent presence with the capacity to overreach. 74 It continues to deserve the mantle of distrust Madison placed upon it when the courts approach establishment questions.
2. The Cases: The Willingness of the Courts to Uphold the State’s Beneficence
While religion has not been shy about requesting or accepting public assistance, the Court has tended to permit an increasing amount of aid to sectarian institutions. Beginning with the Court’s first establishment decision upholding free transportation for children attending nonpublic–primarily sectarian–schools, 75 the Court has had to address a large variety of schemes that benefit sectarian institutions. Every imaginable aspect of a sectarian school’s budget has been the object of state legislation at some point: transportation; 76 books; 77 maintenance and repair;78 tuition assistance through tax deductions, benefits, or direct grants; 79 teachers’ salaries; 80field trip transportation; 81 test and scoring services; 82 diagnostic and therapeutic services; 83on-site special education teachers; 84 specially drawn school districts; 85 vocational rehabilitation; 86 interpreters for the deaf; 87 and  property tax exemptions. 88
As the previous section shows, though, the political reality of religion’s power extends well beyond the scenarios addressed in the Court’s establishment cases. 89 Too often, the Court treats the benefits disputed in the establishment cases as isolated, and therefore likely benign, events. The apparently benign provision of a tax deduction for sectarian school tuition or textbook loans does not appear as benign, however, when one understands that they are integral pieces of a larger puzzle of political pressure. The political realities prove the fundamental soundness of the Court’s admonition in Committee for Public Education and Religious Liberty v. Nyquist 90 that a lax establishment doctrine will lead states “openly [to] subsidize” sectarian interests. 91
III. THE ESTABLISHMENT CLAUSE
The history of the Establishment Clause argues in favor of retaining a distrustful attitude of those with power, whether they operate under the guise of church or state. The Court’s contemporary doctrine needs to be realigned to serve this fundamental constitutional insight.
A. The Inadequacy of Nonpreferentialism to Meet the Evils Intended to Be Redressed by the Establishment Clause
There is precious little regarding the framing of the Establishment Clause on which we can hang our hats. Some have examined selective bits of the Clause’s history to conclude, along with Chief Justice Rehnquist, that the Clause creates a nonpreferentialist rule. Such a rule precludes the establishment of a single church by the state but permits the government to favor religion on an evenhanded basis. The nonpreferentialist rule is seductively simple, seemingly offering a path away from the current thicket of establishment cases. As a historical matter, however, it is an oversimplification of the milieu against which the Establishment Clause was drafted and adopted.
The nonpreferentialist school is mistaken. The Establishment Clause was drafted in an era when most states did not have single establishments. Rather, they had multiple establishments. 92There were a number of states that established not the Episcopal Church but rather Protestantism. In  each community, the people would vote on who would be their minister from a range of denominations. That denominational choice would then preach in the church constructed with citizen’s tax dollars.
This concept of multiple establishment was an innovation in the states. In contrast, in Europe, establishments had been sect-specific–for example, the Church of England or the Roman Catholic Church. 93 Because the era surrounding the Establishment Clause featured multiple establishments rather prominently, the prohibition on establishment in the Establishment Clause should not be taken to mean only a prohibition on the Europeanstyled single-church establishment. Rather, the Clause was intended to prevent the establishment of a class of religion or even of religion generically. Thus, it does not mean only that a single church cannot be joined to the state, but it also means that no set of religions can be joined to the state.
Nonpreferentialism forbids the evil of a single-church state, but it fails to redress the inappropriate exercise and accumulation of power when religious entities band together. Indeed, it invites the political union of religious entities seeking similar ends as a way of avoiding establishment strictures. In other words, it creates an incentive for the religious sects that Madison believed would not join together (and therefore would pose little risk for the balance of power between church and state) to join together. One might argue that the nonpreferentialism “rule” is accurate historically on the ground that the Framers did not believe disparate religions would come together and therefore the only evil the Establishment Clause needed to address was single-sect overreaching. Such an empirical basis for nonpreferentialism has been withdrawn in this era in which religious entities have learned that working together can serve their mutual ends in the political sphere. 94 The balance of power cannot be served therefore through simple protection against single-sect domination but rather through the more complicated task of finding an appropriate balance of power between state and church.
This reading is consistent with the many statements at the Constitutional Convention that the Constitution is intended to demarcate boundaries of power between social entities and to keep the various powers discrete. 95 In Madison’s words, “religion and Government will exist in  greater purity, without [rather] than with the aid of Government.” 96 Church and state are both better off on this account if they remain significantly distant. The hard question is how distant they must be and how that distance is to be achieved. This is a matter of line drawing, obviously not a simple task in this arena. 97
Given the ever-changing state of play between church and state, any landmark in establishment jurisprudence has been wrung for meaning. Most notably, Thomas Jefferson’s famous “wall of separation” between church and state has been one of the pivotal metaphors in the Establishment Clause debate. 98 Unfortunately, the metaphor can be misleading because it seems to offer an either-or proposition: there is a wall or there is not. But that is an unhelpful tack because the either-or approach focuses solely on the wall between church and state, rather than on what lies on either side. An overly constricted focus on the wall itself forces the inquiry into its potential characteristics: from brick and mortar to porous. 99 Yet the wall is not what matters most under the Establishment Clause. What matters more is the shape and strength of religion and state on either side of that wall. Depending on those characteristics, the wall needs to be either brick and mortar, porous, or somewhere in between. It is the Court’s job to custom design a series of walls over time, taking into account the relative power of church and state at that time in that scenario. Thus, “it has never been thought either possible or desirable to enforce a regime of total separation.” 100 Equally, it has never been thought prudent to remove the wall altogether.
Nonpreferentialism builds an insuperable wall between the state and any one religious entity but removes the wall altogether when the religious entity is an accumulation of sects. Its oversimplification of the evils to be redressed by the Clause thus ties the hands of the courts when they address the relevant balance of power between church and state.
B. Establishment Clause Doctrine
The Supreme Court’s doctrine in the Establishment Clause arena has  been treated to more internal and external criticism for its lack of consistency, perhaps, than any other constitutional doctrine. 101 Yet, this is one of those constitutional arenas where consistency is more a “hobgoblin” 102 than a helpful criterion.
The existential fact of changing power relations between these two most authoritative structures of human existence forces the doctrinal analysis into a fluid mode that must focus on the facts of each case. This political, theological, and ontological reality forecloses the possibility that any one case or controversy necessarily should predetermine whether an arguably similar law should be held valid in the future. 103 The relevant facts in any establishment decision include the particular issue at stake but also the social context against which the church-state relationship must be assessed. For example, the inscription of “In God We Trust” on our coins in contemporary culture has not caused the courts concern, but a ruling by the Securities and Exchange Commission that all stocks must bear the same inscription would indicate a shift in church-state relations that is likely to trigger serious concerns. The logical result is that particular arrangements between church and state that may achieve an appropriate balance of power in one era may not do so in another.
The Establishment Clause does not lend itself to a “Grand Unified Theory.” 104 Rather, it charges the courts with delineating the boundaries between church and state over time. This is an arena where lamentations over inconsistent doctrine are beside the point. 105 Because church and state ever will reach for an increase in power (either alone or together) and therefore their relative power can change in an infinitely creative number  of ways, mere predictability of the standards to be applied cannot and should not be the final goal of Establishment Clause doctrine. Rote application of bright-line rules to similar factual skeletons would hand church and state a too easily manipulable regime. 106
In fact, only a minority of the Court has trumpeted the necessity of consistency, while the majority has followed the more difficult path of divining whether particular church-state relations approximate the constitutional goal of a balance of power. The demands for consistency are but a distraction from the more important question of the appropriate allocation of power between state and religion.
From the beginning of the Court’s establishment jurisprudence, it has used accumulations of power as a touchstone for establishment violations, rather than any single bright-line test. In the establishment cases, the Court’s attention has been trained, appropriately, not on devising a bright-line test, but rather on the political reality of the balance of power between church and state presented in each case. The Court’s analysis is more akin to Goldilocks’ approach (too hot, too cold, just right) than any rigid logical formula. In this arena, the Court has internalized to a degree the Framers’ distrustful attitude toward any entity exercising power. As it has done so, establishment doctrine has evolved into a context-dependent and era-dependent balancing approach, which affords the Court maximum flexibility to identify inappropriate relationships of power. 107
The Court’s context-dependent and era-dependent doctrine has accreted so that the clause can redress not any one particular evil but rather a series of evils that have revealed themselves as time marches on. The underlying question posed throughout the establishment cases is whether the balance of power between church and state is tipped by the particular law under attack. The presumption standing behind this question is that the current status quo likely presents an acceptable balance of power. It is not the only acceptable balance that might be struck, but it is acceptable at this stage in history, because the earmark of an inappropriate balance–tyranny by either church or state–is not evident.
This explains the Court’s willingness to uphold practices that are embedded in society, even if they would appear suspect if they were first introduced  today. For example, in Walz v. Tax Commission, 108 the Court upheld by a vote of eight to one real property tax exemptions for religious entities. 109 Without question, such exemptions are a significant source of wealth redistribution to the churches. Indeed, many churches have reached a point where they could not operate without such exemptions. Yet, the Court did not invalidate the property tax exemption. The important question is why.
Although the Walz Court walked through its existing doctrinal approach, the Court’s trump card was the fact that such arrangements had been in place since the framing of the Constitution. 110After years in place, property tax exemptions did not appear to the Court to have resulted in either church or state becoming tyrannical. Its analysis was political and pragmatic. In response to Justice Douglas’s dissent, in which he argued that tax exemptions are the first step in establishing a state church, the majority replied: “If tax exemption can be seen as this first step toward ‘establishment’ of religion, as Mr. Justice Douglas fears, the second step has been long in coming. Any move that realistically ‘establishes’ a church or tends to do so can be dealt with ‘while this Court sits.'” 111 Thus, the Court refused to take a categorical approach to establishment analysis as it embraced a case-by-case, era-by-era analysis that turns on the practical political realities of the time.
In 1971, twenty-four years after its first foray into interpreting the Establishment Clause, 112 the Court reviewed its establishment cases to conclude that various factors are relevant to establishment analysis. The Court summarized the factors it had considered as follows: “First, the statute must have a secular legislative purpose; second, its principal or primary effect must be one that neither advances nor inhibits religion; finally, the statute must not foster ‘an excessive government entanglement with religion.'” 113 This collection of disparate factors was geared to meet abuses of power coming from any of a number of directions: from the state purposefully attempting to elevate a religion or religions above all other societal interests, from the churches achieving favorable legislative effects simply because they were churches, and from the inappropriate intermingling  of church and state power. 114 The Court was concerned not only with intentional state action that resulted in an inappropriate shift of power toward religion but also with religious windfalls in the legislatures and with the potential mischief created when the two entities share power. With its announcement of the Lemon factors, the Court consciously positioned itself as the guardian against the complicated play of power between church and state mutually or individually overreaching.
Thus, the Court’s doctrine has been crafted to address a multiplicity of evils rather than any single evil. In the Court’s words, “What our cases require is careful examination of any law challenged on establishment grounds with a view to ascertaining whether it furthers any of the evils against which that Clause protects.” 115
1. The Talisman “Neutrality” Has Torqued the Doctrine and Opened the Door Wider for Religion to Operate to Achieve Its Political Goals
Some of the most vexing cases under the Establishment Clause have involved state aid to religion. This is an arena where the Court has clearly applied a pragmatic balancing approach. It is also an arena where the concept of “neutrality” has shifted the analysis away from the Framer’s fundamental insight that religion is capable of acting against the public good. Like nonpreferentialism, neutrality does not offer a magical elixir that can lay to rest establishment quandaries in the future.
In a stream of cases, the Court has examined each aid situation as an individual set of unique facts. They have brought to these cases a variety of factors that have accreted to the establishment doctrine over time and attempted to determine if the balance of power in that particular scenario was tipped.
From its first establishment case, the Court adamantly has refused to read the Establishment Clause as an absolute ban on aid to religion. In Everson v. Board of Education, 116 the Court ruled that public school districts could bus parochial as well as public school students. 117 As a realistic matter, free bus transportation is a boon to both religious school students and the schools themselves. 118 The state, thus, is assisting religion. In dissent, Justice Jackson assessed the facts of the case to conclude that  such assistance was unconstitutional.
The rest of the Court, however, rejected the notion that free transportation crossed the establishment divide. The members of the Court reasoned from a determinedly pragmatic perspective. They worked from the proposition that children must attend school and that they had earlier ruled that students must be permitted to attend nonpublic schools. 119 Moreover, it was a fact of contemporary reality that churches received a vast range of “assistance” from the community in the form of police and fire protection, sewer systems, public highways, and sidewalks. 120 This was a “general program” that did not amount to a contribution of “money to the [sectarian] schools” or “support them.” 121 As a pragmatic matter, then, there was no fear that the state was supporting the churches or furthering their religious mission.
The Court has taken this pragmatic approach through its aid cases to determine that a state could not fund maintenance and repair costs for sectarian schools, provide tuition reimbursement, grant income tax benefits, 122 provide instructional materials and auxiliary services programs, 123 free transportation for field trips, 124 or on-site special education teachers. 125 The state has been permitted, however, to provide free transportation, 126 property tax exemptions, 127 textbook loans, 128 secular textbook loans, standardized tests and scoring services, diagnostic and therapeutic services, 129 vocational rehabilitation assistance for the blind, 130 tax deductions for tuition, textbooks, and transportation, 131 an interpreter for a deaf student, 132 and on-site teachers for special education. 133 There are more cases permitting a particular type of aid than not. In other words, the Court has not been parsimonious about permitting the state to aid religious schools. Through piecemeal efforts–a book here, a bus there–state legislatures have been permitted to provide a great deal of assistance to students who  are in sectarian schools. State allocations to discrete student needs when examined on a case-by-case basis simply have not appeared to tip the balance of power between church and state to any significant degree.
Yet, not all aid arrangements have passed muster. In Committee for Public Education and Religious Liberty v. Nyquist, 134 the Court held that repair and maintenance grants and tuition reimbursement grants were unconstitutional. 135 The Court also invalidated the grant program that permitted poor parents to receive reimbursement of “$ 50 for each grade school child and $ 100 for each high school child . . . [provided] the amount of state reimbursement [does] not exceed 50%” of actual tuition paid. 136 In a 6-3 decision, the Court rejected these programs on the ground that both programs permitted a stream of income to benefit sectarian schools without any assurance that the expenditures were used for the secular teaching of the schools rather than the religious. In other words, the programs lacked any safeguards that would prevent the grants from turning into direct state support for religious schools.
The repair and maintenance grants failed because the state made no effort to restrict the monetary payments for repair and maintenance to solely secular purposes. Moreover, the Court implied that such monitoring would run afoul of the Constitution as well. 137 Thus, the payment to the schools was insufficiently policed and incapable of being adequately policed.
The tuition reimbursements, which were limited to 50% of nonpublic school tuition, also failed on the same reasoning, with the Court saying that they lacked “a guarantee that state funds will not be used to finance religious education.” 138 The State “must be certain” that its money is not being used to “inculcate religion.” 139
The Nyquist Court then stepped back to its pragmatic perch and assumed that whatever line they drew in the establishment cases would be subject to constant pressure from both sides. The 50% limitation in the tuition reimbursement scheme was inadequate to ensure that public monies would not support religious inculcation. It recalled that in Earley v. DiCenso, a companion case toLemon v. Kurtzman, 140 a salary supplement to  teachers of secular subjects in all schools was invalidated even though it was limited to 15% of the teacher’s annual salary. The Court had reasoned that the mere assumption that a low percentage–15%–would subsidize secular rather than religious subjects was insufficient. The state in Earley could not rely “on the assumption that, whatever a secular teacher’s inabilities to refrain from mixing the religious with the secular, he would surely devote at least 15% of his efforts to purely secular education, thus exhausting the state grant. It takes little imagination to perceive the extent to which States might openly subsidize parochial schools under such a loose standard of scrutiny.” 141
In sum, the primary concern in Nyquist was that the financial assistance could find its way into religious purposes. The underlying assumption, which rings of the Framers’ presuppositions regarding the exercise of power, was that there would be a constant political push to aid parochial schools and a willingness in those institutions to convert public monies to religious ends. Neither state nor church could be utterly trusted, and therefore schemes without mechanisms of limitation could not pass muster.
In a sea change, a 5-4 Court in Mueller v. Allen 142 did not find the same constitutional errors in a Minnesota law permitting taxpayers to deduct expenses incurred in providing tuition, textbooks, and transportation for their children. The Court’s focus shifted in this case. No longer was the Court worried either overtly or implicitly about abuses of power by either church or state. Rather, the Court operated at a farther level of abstraction. Indeed, the burden shifted. In Nyquist, the Court held that it was necessary to be “certain” that the funds did not further religious purposes.143 In Mueller, however, the Court stated that to invalidate an aid program benefitting religious institutions, there must be “certainty” that there is an empirical benefit to religious institutions.144
The Court concluded that the tax deductions did not advance religion because they applied to both private and public school students. This was an “attenuated financial benefit,” 145 rather than a direct payment into the schools, and therefore looked more like the cases in which aid was approved than the invalidated scheme in Nyquist. The Court explained that the scheme in Nyquistonly benefitted parents of children in nonpublic schools but that the Mueller scheme benefitted children in all schools.  Thus, “neutrality” saved the day.
From the Framers’ perspective, this is odd reasoning at best. The Court looks more like an ostrich than any other creature. The distrust of all social entities inherent in the Constitution is mediated in favor of a presumption that “neutral” state aid does not effect a real benefit for religion. This is, undoubtedly, a twist on establishment jurisprudence, but it requires careful parsing to determine exactly how the Court has torqued the doctrine.
The Court does not, in fact, take issue with the Framers’ warnings about abusive exercises of power. The Court continues to operate within a Calvinist framework: “‘What is at stake as a matter of policy [in Establishment Clause cases] is preventing that kind and degree of government involvement in religious life that, as history teaches us, is apt to lead to strife and frequently strain a political system to the breaking point.'” 146 Having acknowledged the structural role of the Establishment Clause, however, the Court made an empirical determination that there need be less worry in this century about the evils to be addressed by the Establishment Clause than in the Framers’ time:
At this point in the 20th century we are quite far removed from the dangers that prompted the Framers to include the Establishment Clause in the Bill of Rights. The risk of significant religious or denominational control over our democratic processes–or even of deep political division along religious lines–is remote, and when viewed against the positive contributions of sectarian schools, any such risk seems entirely tolerable in light of the continuing oversight of this Court. 147
Thus, the Court decided Mueller under the same legal analysis that was applied in Nyquist but against a differently understood empirical reality. The Framers’ pragmatic distrust was assuaged by an apparently tolerable status quo. The result was allegiance to neutrality combined with a sanguine assumption that religions would not exercise its power in the public sphere inappropriately or in opposition to the public good.
In sum, establishment jurisprudence has turned on two considerations: (1) a doctrinal foundation that distrusts inappropriate accumulations of power in either church or state, and (2) the finding of a constitutional fact, an empirical finding that measures the threat to liberty engendered by certain church-state relations in a particular era. At some level, the cases  have held to the Framers’ views under the first prong, but have widely departed from the second.
2. From Aguilar to Agostini: A Case Study in Context-Dependent Balancing
One of the most explicit examples of the Court’s pragmatic, power-balancing approach in establishment cases exists in the Aguilar/Agostini pair of decisions. In Aguilar v. Felton, 148 the Court held that state-paid special education teachers were prohibited from going on the premises of sectarian schools. 149 The Court explained as follows: “Though a comprehensive system of supervision might conceivably prevent teachers from having the primary effect of advancing religion, such a system would inevitably lead to an unconstitutional administrative entanglement between church and state.” 150
The Court’s decision resulted in some interesting, and even comical, arrangements in the New York City school system, where millions of dollars were spent to locate trailers on the streets outside sectarian schools where special education teachers could assist the sectarian school students. 151 This strained arrangement continued for over a decade.
The same case then returned to the Supreme Court, with a new name, Agostini v. Felton. 152The Court this time held that public school teachers could go on the premises of the sectarian schools where their tasks were plainly not in furtherance of the religious mission of the school.153 Only ten years apart, the Court reached opposite conclusions under the Establishment Clause in precisely the same case.
The easy critique of the Aguilar to Agostini story is that the Court’s Establishment Clause jurisprudence is in disarray, that the Justices do not know what they are doing in this arena, and, therefore, some quick-fix is needed. 154 I would explain the decisions more charitably. The Court’s  willingness to overturn Aguilar does not necessarily signal a lack of direction. Rather, it indicates that the Court’s compass is gauging political realities, rather than abstract principles. 155 The decision in Agostini reversing Aguilar is evidence that the Court is monitoring its establishment decisions to determine that the proper balance between church and state has been reached. Had Aguilar led to an efficient and successful delievery of services to private school systems, it is highly unlikely the Court would have been so eager to overrule it. When the result of its decisions is nonsensical, expensive, and the likelihood of tyranny is low, were the Court to reverse itself, it does so. Not because it does not understand how and why to apply the Establishment Clause’s prohibitions, but precisely because it does.
The Court’s establishment cases, thus, turn on constitutional factfinding regarding the relative power of church and state. To date, this factfinding has been stated as a set of presuppositions and has been subject to precious little hard analysis. The empirical relationship between church and state is crucial to determining whether there is a risk of establishment. 156 This empirical foundation to establishment analysis has gone unnoticed as scholars and Justices have debated the doctrine without acknowledging the factual presuppositions underlying their analysis. Thus, the cases turn on unexamined presuppositions about the sociological status of religion in society.
In this game of power, which is played on a field of unexamined presuppositions, characterization may be everything. The more unthreatening religion is made to appear, the more likely aid to religion will appear to be unthreatening. One can expect religion to don the garb of ineffectualness as it improves its political tactics.
Whether religion will become a threat to liberty can be examined on two axes. The first question is how much power religion holds. The second question is what kind of power religion is being permitted to exercise. The establishment cases have tended to answer both questions in ways that privilege religion. First, they have assumed too often that religion is a relatively powerless player in the political sphere. Thus, they have ruled, a little bit of discrete aid, 157 an opening legislative prayer, 158 and a Christmas  display balanced by other religious and secular symbols, 159 can do no harm to the church-state balance.
Second, they have assumed that religion exercises benign power. As recently as 1970, the Court has treated religion to be protected under the Religion Clauses as a rather homogeneous entity that has a “harmonious relationship to the community.” 160 The Court justified real property tax exemptions for all churches on the ground that “certain entities . . . that foster [the community’s] ‘moral or mental improvement,’ should not be inhibited in their activities by property taxation or the hazard of loss of those properties for nonpayment of taxes.” 161 The Court seems to believe that religion is a milquetoast set of beliefs that will necessarily enhance the public good. The modern Court has generally failed to acknowledge the wide range of religions that may pose a threat to the general welfare such as the Aryan Brotherhood or the Satanists or even the multitude of mainstream religions, each of which is capable of acting in ways that undermine the public good.
Yet, as the Court has resorted to oversimplified and whitewashed versions of religion in deciding the establishment aid cases, it has held to a more realistic view of religion in the free exercise cases. In a strangely schizophrenic twist, at the same time that the Court’s establishment doctrine has treated religion as a benign and nonthreatening force, its free exercise doctrine has treated it as a force capable of undermining the public good. This attitude spans its first free exercise case,Reynolds v. United States, 162 to its more recent jurisprudence. In Employment Division v. Smith, 163 the Court held that religion does not have a constitutional right to trump generally applicable, neutral laws. 164 In Smith in particular, the Native American religion’s practice of smoking peyote, an illegal narcotic, could not trump the state’s laws against illegal drug use. 165One of the major premises of this conclusion is that religion is capable of doing that which is inconsistent with the general welfare, which is protected by neutral laws of general applicability. Just as it had said in its first free exercise case, the Court reaffirmed in Smith that the Religion Clauses do not  permit an individual “to become a law unto himself.” 166 Religion is not a necessarily benign force permissibly fostered by the state but rather a forceful entity capable of undermining the public good. It is also an entity capable of protecting itself through the political process where its demands can be assessed according to the greater public good. 167 This relatively less flattering, though more accurate portrayal of religion as a capable political actor, brings to the foreground the clashes between law and religion that have been swept aside in the Court’s establishment discourse. The examples of such clashes are numerous: religions have claimed the right to override a wide variety of laws, including laws protecting children, 168 local land use laws, 169 and school safety rules. 170 The inevitable conflict between laws for the general welfare and religion is unlikely to abate any time soon. Likewise, religion’s propensity to ask for state aid is unlikely to vanish. These political realities should inform the factual presuppositions undergirding the courts’ aid cases, including those addressing school vouchers.
IV. SCHOOL VOUCHERS AND POWER
Despite the tight budgets most states and local governments labor under, there is a serious push to institute voucher systems. The policy justifications for voucher schemes tend to turn on the claims that there is a crisis in public education and that public schools deserve to be subjected to competition. 171 Without disagreeing with this policy perspective, I doubt very much that voucher schemes that funnel public monies to sectarian schools are sufficiently constitutional to justify the interest currently focused upon them. The constitutionality of voucher schemes turns on two elements: constitutional fact-finding regarding the current state of play between church and state,172 and application of the doctrine to that  constitutional fact-finding.
All voucher schemes are means by which local governments provide tuition assistance for students to attend nonpublic schools. Obviously, public schools receive public assistance. Aid to nonpublic schools comes in at least three flavors. Assistance can be provided to (1) sectarian school students only, (2) nonsectarian school students only, or to (3) all nonpublic school students, sectarian and nonsectarian.
The first flavor–aid to sectarian students only–is undoubtedly unconstitutional. Assistance directedonly to religious schools violates even the most lax establishment doctrine. 173
The second flavor–aid to nonsectarian students only–is likely constitutional on the grounds that the exclusion of religious schools is necessary to avoid serious establishment difficulties. This is precisely the reasoning employed in Corporation of Presiding Bishops v. Amos 174 to uphold Title VII’s exemption of religious employers from the proscription of discrimination on the basis of religion. 175 Such a scheme could be subject to a free exercise challenge, however, if the excluded religious schools could show that they were excluded on the basis of animus. 176 In the absence of animus, such a program should pass free exercise muster. 177
It is the third flavor on which this Article will focus: the voucher scheme that is facially neutral, i.e., that provides tuition assistance to all nonpublic students, whether they attend a sectarian or nonsectarian school. 178 Nonpreferentialism that fails to investigate the empirical relationship between church and state would argue in favor of such a scheme. It would also pass a facial “neutrality” test. Once one starts exploring beneath the surface, however, potential establishment evils appear.
First, a facially neutral scheme may be a pretext to privileging a particular religion. If all of the nonpublic schools in a particular district are  from one sect, then the balance of power between the state and religious sects has tipped too far in favor of a single religion.
Second, a facially neutral scheme may also, as a factual matter, favor religion over irreligion. 179A system in which all of the nonpublic schools are pervasively sectarian, though not all of a single sect, permits a significant union of church and state power that should be troubling. In this era, in which the vast majority of nonpublic schools are pervasively sectarian, 180 facial neutrality is likely not substantive neutrality. 181 While programs that “neutrally provide state assistance to a broad spectrum of citizens [may not be] readily subject to challenge under the Establishment Clause” 182 under the Court’s reasoning in Mueller and Bowen v. Kendrick, 183 the political reality is that voucher schemes at the end of the twentieth century in the vast majority of jurisdictions in the United States are unlikely to provide assistance to a “broad spectrum of citizens.” 184
Third, a facially neutral scheme may distribute public moneys to sectarian institutions in a manner roughly equivalent to the money distributed to nonpublic nonsectarian institutions. The question is whether this evenhanded distribution forestalls establishment concerns. This is where “judgment–not simple categories” 185 must come into play in each case.
To be true to the Establishment Clause’s fundamental political presuppositions, the courts would do best to adopt the Court’s approach in Nyquist, which is aligned with the Framers’ general attitude of distrust even if benefits are packaged with benefits for nonsectarian institutions. The Establishment Clause argues against the Court’s more trusting presuppositions in Mueller. 186Religion may have chosen to go behind the scenes in this republican democracy throughout the course of over two hundred years, but that does not make its drive to power any less a matter of concern.  We have not reached a point “far removed from the dangers that prompted the Framers to include the Establishment Clause in the Bill of Rights,” 187 but rather live in an era when we need be certain that religion is not overreaching. On the state of the existing political record, “it takes little imagination to perceive the extent to which States might openly subsidize parochial schools.” 188 Thus, the “channel [for direct aid to sectarian schools should be] a narrow one.” 189
How narrow the channel will be will depend on the questions posed in each establishment case by the courts. Where the benchmark is the relevant power between church (whether an individual sect or a collection of sects) and state, an appropriately distrustful court should ask the broaderLemon questions as well as the following: (1) how much money is being passed from public hands to sectarian hands, in raw dollars, in proportion to the sectarian institution’s budget, and in proportion to the government’s school budget; (2) who decides how much money is given to the sectarian institutions; (3) how is the money distributed to the sectarian institution; (4) which students are receiving the benefit of the money, for example, is it limited to a small class of the poor or does it provide money to those who would have been able to pay the tuition regardless; (5) how does the government know how its money is being spent by the sectarian institution; (6) may the sectarian institution raise its tuition once the government starts paying and then use the difference for purely sectarian purposes; (7) how will increases in funding be enacted in the future; and (8) is the same amount of money being handed to each sectarian institution, either by pro rata share or in raw dollars.
This is not an eight-part test to be amended to a three-part Lemon test. Rather, it reflects the beginning of an open-ended, challenge-specific inquiry into the relevant power between church and state before the scheme is instituted and after the scheme is in place. Admittedly, it provides precious little succor for those hoping for a safe haven in establishment cases that could protect one from the burden of exercising judgment. But it is the best that can be done.
Facially neutral voucher schemes posit a potential mingling of church-state power that should make anyone–including voucher supporters–step back for a minute. Voucher schemes open the statehouse doors to church lobbying for increases in voucher rates, lobbying regarding when payments are made and how they are made. It takes little imagination to extrapolate from these inevitable topics of lobbying in a voucher district to  a scenario where sectarian administrators are in as much consultation with public officials on a wide variety of school issues as are public school administrators.
The entanglement between church and state is the least of it. Competition between sects for aid likely will increase. We can take a page from history here. When New York lived under a multiple establishment in the seventeenth and eighteenth centuries, a system in which a variety of religions were preferred, “there was constant argument [between those supporting the Church of England and those supporting other Protestant denominations] concerning the disposition of tax funds for the support of religion.” 190 The reality is that money breeds political activity and that contemporary churches are very comfortable in the lobbying mode. In short, a voucher system opens the door to excessive and escalating entanglement even if the voucher system on its face looks neutral.
The result of this mingling of church-state power is unlikely to be good for the public good, and it may not even be good for the churches. It is instructive to examine the German system of tithing to get some perspective on the question of state aid to churches in the United States. In Germany, if one declares membership in a particular church, the state taxes one’s income by a flat rate. 191 If one does not pay the tithing tax, tax evasion charges follow. German tennis star Stefi Graf can attest to this. 192 I raise this point because the state’s involvement as the throughway mechanism for church income may have hurt church giving in Germany. 193
This rupture may have been caused in part by the fact that churches are now a step removed from their givers. The theory among givers apparently is that what the state takes is enough, even too much. Therefore, why give spontaneously, why give for particular causes? You have done your duty by paying into the state’s church fund.
This experience can be analogized to the voucher scenario. At this time, private schools are largely self-funded entities and the relationship between parent and school is cemented by tuition payments, i.e., money. The parents give a great deal in volunteer efforts to raise money and to  oversee the school’s actions. Will this continue to be just as true when the government operates as the funds gatherer and dispenser? Government involvement sunders the previously self-referential system and, therefore, may hinder fund-raising as it reduces the accountability of the sectarian school officials for the funds received.
When the money parents currently pay to sectarian schools is given to the parents by the state, the bonds of accountability have been stretched if not breached. If the school does not have enough money to fulfill its desired program, the school may well argue in favor of a tax increase rather than going to the parents. In a voucher regime, it is likely to do both. Or, when the school complains of inadequate funds, the sectarian school parents may well accuse the school of ineffective lobbying. State funding of sectarian schools therefore transforms the parents’ concerns into issues that are “not their problems” because the money does not originate in their pockets and travels directly to the school. Far from the benevolent contributors to a free market in education their supporters invoke, 194 voucher systems are an assault on the free market in educational systems because they rupture this otherwise close relationship between consumer and producer.
Current political reality thus counsels caution in the face of facially neutral voucher systems. Neither nonpreferentialism nor neutrality should be permitted to short-circuit the important, though difficult task of discerning the appropriate balance of power between church and state. There may be an “education emergency,” but tipping the balance of power between church and state by further opening the states’ coffers to religion trades in a serious policy dilemma for one that could shake the foundations of the republic. Fortunately, “the Constitution protects us from our own best intentions” by prohibiting us from “concentrating power in one location as an expedient solution to the crisis of the day.” 195
In this era, many voucher systems that permit governments to subsidize sectarian education cross the constitutional line. Along with other religious liberty legislation now being promoted in Congress and the  states, they are classic examples of overreaching. 196 True to the Framers’ fears, religion has not limited itself to an ineffective political institution over which there need be no concern. It is as eager in the political arena now as it has ever been in history. The push for school vouchers is just one part of the picture. The Court’s free exercise doctrine takes this political reality into account; its establishment cases have yet to do so.
The pervasive hand-wringing over inconsistency in the establishment cases has deflected attention from the more fundamental inquiry into the unexamined presuppositions in the establishment cases regarding the power exercised by religion. The charge of inconsistency is attributable to a misdirected focus on tangential aspects of the Court’s cases:
In attempting to articulate the scope of the two Religion Clauses, the Court’s opinions reflect the limitations inherent in formulating general principles on a case-by-case basis. The considerable internal inconsistency in the opinions of the Court derives from what, in retrospect, may have been too sweeping utterances on aspects of these clauses that seemed clear in relation to the particular cases but have limited meaning as general principles. 197
The courts properly stand as a brooding omnipresence behind the complex play of political forces between church and state. This complex task requires “room for play in the joints” of the doctrine. 198 Under the Establishment Clause, the courts need to be a substantial check on present abuses of power, not just legislative-like bodies issuing reliable edicts for the future. 199Thus, the inevitably sly permutations of power make ad hoc, case-by-case decision-making in this arena not only inevitable but a positive good.
James Madison issued a prescient warning that the “danger of silent accumulations and encroachments by Ecclesiastical Bodies has not sufficiently engaged attention in the U.S.” 200Establishment Clause doctrine needs to be re-innoculated by this healthy, albeit distrustful attitude.
 APPENDIX A
LIST OF THE FORTY-FIVE STATES, AND THE DISTRICT OF COLUMBIA, THAT ALLOW IMMUNIZATION EXEMPTIONS FOR CHILDREN ATTENDING PUBLIC SCHOOL
1. Alaska: ALASKA STAT. § 14.30.125 (Michie 1998).
2. Arizona: ARIZ. REV. STAT. ANN. § 15-873 (West 1991).
3. Arkansas: ARK. CODE ANN. § 6-18-702 (Michie 1993 & Supp. 1997).
4. California: CAL. HEALTH & SAFETY CODE § 120365 (West 1996).
5. Colorado: COLO. REV. STAT. ANN. § 25-4-903 (West 1998).
6. Connecticut: CONN. GEN. STAT. ANN. § 10-204a (West 1996 & Supp. 1998).
7. Delaware: DEL. CODE ANN. tit. 14, § 131 (1997).
8. District of Columbia: D.C. CODE ANN. § 31-506 (1998).
9. Florida: FLA. STAT. ANN. § 232.032 (West 1998).
10. Georgia: GA. CODE ANN. § 20-2-771 (Harrison 1997).
11. Hawaii: HAW. REV. STAT. ANN. § 302A-1156 (Michie 1997).
12. Idaho: IDAHO CODE § 39-4802 (1997).
13. Illinois: 105 ILL. COMP. STAT. 5/27-8.1 (West 1993 & Supp. 1998).
14. Indiana: IND. CODE ANN. § 10-4-1-16 (West 1998).
15. Iowa: IOWA CODE ANN. § 140.13 (West 1998).
16. Kansas: KAN. STAT. ANN. § 72-5209 (1997).
17. Kentucky: KY. REV. STAT. ANN. § 214.034 (Michie 1997).
18. Louisiana: LA. REV. STAT. ANN. § 17.170 (West 1997).
19. Maine: ME. REV. STAT. ANN. tit. 20-A, § 6355 (West 1997).
20. Maryland: MD. CODE ANN., EDUC. § 7-403 (1997).
21. Massachusetts: MASS. GEN. LAWS. ANN. ch. 76, § 15 (West 1996).
22. Michigan: MICH. STAT. ANN. § 25.358(27) (Law. Co-op. 1998).
23. Minnesota: MINN. STAT. ANN. § 123.70 (West 1993).
24. Missouri: MO. ANN. STAT. § 167.181 (West 1991 & Supp. 1999).
25. Montana: MONT. CODE ANN. § 20-5-405 (1997).
26. Nevada: NEV. REV. STAT. § 392.437 (1996 & Supp. 1997).
27. New Hampshire: N.H. REV. STAT. ANN. § 141-C:20-c (1996).
28. New Jersey: N.J. STAT. ANN. § 26-1A-9.1 (West 1996).
29. New Mexico: N.M. STAT. ANN. § 24-5-3 (Michie 1998).
30. New York: N.Y. PUB. HEALTH LAW § 2164 (McKinney 1994 & Supp. 1999).
 31. North Carolina: N.C. GEN. STAT. § 130A-152 (1997).
32. North Dakota: N.D. CENT. CODE § 23-07-17.1 (1997).
33. Ohio: OHIO REV. CODE ANN. § 3313.671 (Anderson 1998).
34. Oregon: OR. REV. STAT. § 433.267 (1997).
35. Pennsylvania: PA. STAT. ANN. tit 24, § 13-1303a (West 1998).
36. Rhode Island: R.I. GEN. LAWS § 16-38-2 (1997).
37. South Carolina: S.C. CODE ANN. § 44-29-180 (Law. Co-op. 1997).
38. South Dakota: S.D. CODIFIED LAWS § 13-28-7.1 (Michie 1998).
39. Tennessee: TENN. CODE ANN. § 37-10-402 (1997).
40. Texas: TEX. EDUC. CODE ANN. § 38.001 (West 1997).
41. Utah: UTAH CODE ANN. § 53A-11-302.5 (1998).
42. Vermont: VT. STAT. ANN. tit. 18, § 1122 (1997).
43. Virginia: VA. CODE § 22.1-271.2 (Michie 1998).
44. Washington: WASH. REV. CODE ANN. § 28A210.090 (West 1997).
45. Wisconsin: WIS. STAT. ANN. § 252.04 (West 1997).
46. Wyoming: WYO. STAT. ANN. § 21-4-309 (Michie 1997).
 APPENDIX B
LIST OF THE THIRTY-FIVE STATES THAT ALLOW RELIGIOUS BELIEFS AS DEFENSE TO CHILD ABUSE, ENDANGERMENT, NEGLECT, MALTREATMENT
1. Alabama: ALA. CODE § 26-14-7.2 (1997).
2. California: CAL. PENAL CODE § 11165.2 (West 1998).
3. Colorado: COLO. REV. STAT. ANN. § 19-3-103 (West 1998).
4. Connecticut: CONN. GEN. STAT. ANN. § 17a-104 (West 1998).
5. Delaware: DEL. CODE ANN. tit. 31, § 403 (1997).
6. Florida: FLA. STAT. ANN. § 39.01 (West 1998).
7. Georgia: GA. REV. CODE ANN. § 19-7-5 (Harrison 1998).
8. Idaho: IDAHO CODE § 16-1602 (1997).
9. Iowa: IOWA CODE ANN. § 232.68 (West 1998).
10. Kansas: KAN. STAT. ANN. § 21-3608 (1997).
11. Kentucky: KY. REV. STAT. ANN. § 600.020 (Michie 1997).
12. Louisiana: LA. REV. STAT. ANN. § 2019(e)(4) (West 1997).
13. Maine: ME. REV. STAT. ANN. tit. 22, § 4010 (West 1997).
14. Michigan: MICH. STAT. ANN. § 722.634 (Lay. Co-op. 1998).
15. Mississippi: MISS. CODE ANN. § 43-21-105 (1998).
16. Missouri: MO. ANN. STAT. § 210.115.3 (West 1998).
17. Montana: MONT. CODE ANN. § 41-3-102 (1997).
18. New Hampshire: N.H. REV. STAT. ANN. § 169-C:3 (1997).
19. New Jersey: N.J. STAT. ANN. § 9:6-8.21 (West 1998).
20. New Mexico: N.M. STAT. ANN. § 32A-4-2 (Michie 1998).
21. North Dakota: N.D. CENT. CODE § 50-25.1-05.1 (1997).
22. Ohio: OHIO REV. CODE ANN. § 2919.22 (Anderson 1998).
23. Oklahoma: OKLA. STAT. tit. 10, § 7103 (1998).
24. Oregon: OR. REV. STAT. § 419B.OO5 (1997).
25. Pennsylvania: PA. STAT. ANN. tit. 23, § 6303 (West 1998).
26. Rhode Island: R.I. GEN. LAWS § 40-11-15 (1997).
27. South Carolina: S.C. CODE ANN. § 20-7-652 (Law. Co-op. 1998).
28. Tennessee: TENN. CODE ANN. § 39-15-402 (1998).
29. Utah: UTAH CODE ANN. § 76-5-109 (1998).
30. Vermont: VT. STAT. ANN. tit. 33, § 4912 (1997).
31. Virginia: VA. CODE § 63.1-248.2 (Michie 1998).
32. Washington: WASH. REV. CODE ANN. § 26.44.020 (West 1998).
33. West Virginia: W. VA. CODE § 61-8D-4 (1998).
34. Wisconsin: WIS. STAT. ANN. § 948.03 (West 1997).
35. Wyoming: WYO. STAT. ANN. § 14-3-202 (Michie 1998).