The Charitable Choice provisions of the “Personal Responsibility and Work Opportunity Act of 1996,”[1] encourage states to contract with Faith Based Organizations (FBOs) for delivery of social services to welfare recipients. The relevant language of the Act allows such FBOs to retain their pervasively sectarian nature. They may require that employees share their faith, but may not use tax dollars for overtly religious purposes, or  condition receipt of benefits upon religious participation. Charitable Choice legislation was premised upon the assumption that FBOs do a better job at lower cost than traditional social service providers (an assumption that is neither supported nor contradicted by available data, because no comprehensive research on the issue is available.)

Charitable Choice thus extends to a large number of religious organizations an invitation to participate in a trend that has been reshaping government—particularly at the state and local level—for at least the past twenty years.[2] That trend is often referred to as “privatization” or “reinvention;” however, those terms are more general and encompass a large number of programs and initiatives, including outright sales of government operations.[3]  Charitable Choice legislation is the most recent example of the most prominent of these methods to “reinvent” government: the vastly increased use of private for-profit and nonprofit providers to deliver government services pursuant to contractual agreements.  While government has always purchased goods and services in the market, the practice has grown significantly[4] and has extended into social services where contractors are more likely to be nonprofit agencies with social missions than business organizations seeking profits, and where it presents issues quite different from agreements to purchase computers or to pave city streets.[5]

As contracting out becomes a more and more common method of delivering government services, the nature of “the state” and hence of “state action” as that term is used in constitutional analysis becomes an evolving concern.[6] There is significant evidence that the growth of contracting, coupled with an unrealistic and narrow understanding of state action, has created a jurisprudence that is, as one scholar has put it, “significantly underprotective of constitutional rights.”[7]

In traditional political discourse, we consider the nature of government or debate the proper role of the state. Rarely do we turn to discussion of what the state is.   One can refer, of course, to Augustine’s “Earthly City,” Aristotle’s polis, or Kant’s “system of public right,” but I think D.D. Raphael accurately summarizes the contemporary idea of the state by defining it as “an association having universal compulsory jurisdiction within territorial boundaries.”[8]  The two elements of that definition—territoriality and a monopoly on the right to use certain types of force or power—are arguably integral to popular understanding of the concept of statehood. Both elements, however, are undergoing redefinition.

In industrialized nations, and perhaps elsewhere, the growth of the global economy and the worldwide penetration of the Internet are increasingly challenging traditional notions of territorial jurisdiction. In America, the steady expansion of government since the New Deal has already required us to rethink the relationship between government power and fundamental rights. While rights were traditionally defined as limitations on the coercive power of the state, lawyers and political philosophers now speak of both negative and positive liberties and debate the propriety of recognizing affirmative “entitlements.”  With the advent of widespread contracting, where a growing number of services are provided by and paid for by government but delivered by contractors, we must ask new questions as well. Some are technical: should contractors have the same First Amendment free speech rights as government employees[9]?  Should the definition of government power include the power of the purse? Other questions are political: are partnerships with businesses and nonprofit organizations creating a new definition of government?  Is privatization extending, rather than shrinking, the state? Does the substitution of an independent contractor for an employee equate to a reduction in the scope of government, as proponents believe?  Or, as Donald Kettl has suggested,[10] does the substitution operate instead to shift the locus but not the scope of government activity, and thereby blur the boundaries between public and private, making it ever more difficult to decide where “public” stops and “private” begins?[11] If we are altering traditional definitions of public and private by virtue of these new relationships, what is the effect of that alteration on a constitutional system that depends upon the distinction as a fundamental safeguard of private rights?

However we understand government, a central tenet of democratic regimes is that the state must be accountable to its citizens. Contracting out complicates accountability in a number of ways.[12]  In an important book published in 1993, Nonprofits for Hire, Steven Rathgeb Smith and Michael Lipsky explored a variety of issues raised for government and the nonprofit sector by virtue of the increasing reliance upon government contracts.

                        “American social policy is in the midst of a dramatic restructuring of the

                        way public social services are provided. Although government funding of

nonprofit service organizations dates to the colonial period, only in the last 25 years did this government-nonprofit strategy emerge as a widespread and favored tool of public service delivery. But entrusting the most vulnerable citizens and the most delicate service tasks to private agencies is not simply a matter of choice between “making” or “buying” services. This might be the case when one considers contracting out for pencils, computer services, or strategic weapons. But when it comes to purchasing the care and control of drug addicts, the safety and nurturing of children, the relief of hunger and the regulation of family life (through child protective activities) from private agencies, other values than efficiency are at stake. We contend that the impact of this transformation on the future of the American welfare state has not received adequate attention.”[13]

Among the issues Smith and Lipsky explore is the transfer of state power to private providers.

                        “Like teachers, police officers and welfare workers, service providers

                        in the nonprofit sector manage scarce resources by coping with their jobs

                        in such a way as to render them not simply implementers of public policy,

                        but ‘makers’ of public policy.”[14]

Smith and Lipsky contend that workers in nonprofit agencies executing government contracts must be considered agents of government,[15]and express concern that political accountability has been compromised by the lack of transparency which is an inevitable component of such arrangements. As they note, from the point of view of welfare state clients, advocacy groups, and the general public, the state has disappeared, making it extremely difficult to assess political responsibility[16] or ensure accountability.[17] That difficulty will inevitably be compounded when government services are provided by pervasively sectarian, faith-based organizations pursuant to Charitable Choice.[18]

One traditional way to enforce government accountability is through the courts. Just as a lack of transparency in contracting relationships can impede political accountability, I will argue in Section III that the failure of the state action doctrine to keep pace with the political reality of government contracting significantly undermines our ability to achieve constitutional accountability. That said, however, an appropriate solution must be carefully crafted, as I discuss in Section IV.  If the state action doctrine does not change to accommodate new realities, we are in danger of losing an important constitutional check on the exercise of administrative power. But if contractors become state actors for all or most purposes, such an extension of the doctrine would further blur our ability to define government in a way that allows us to hold it politically accountable.

We rely upon our understanding of the state action doctrine in order to know when we may ask the courts to restrain government agencies. If we do not have comprehensible rules defining those actions we may legally attribute to the state, the efficacy of constitutional litigation is undermined. If  we are unable to convey to citizens the boundaries of government’s legal responsibilities, our ability to fashion appropriate political remedies will also be compromised.

III           

                                                Current State Action Doctrine  

As Robert Gilmour and Laura Jensen have noted,

“When the relationship between government and citizen becomes more complex than that between a mere commodity or service provider and its customers, more than marketplace efficiency is required to hold the government and its proxies and surrogates accountable for their exercise of authority on behalf of the state.” [29]

Acknowledging the need for such rules and actually fashioning them have proved to be very different matters. As one commentator has wryly noted, the Court’s ‘sifting’ and ‘weighing’ in state action cases “differs from Justice Stewart’s famous ‘I know it when I see it’ standard for judging obscenity mainly in the comparative precision of the latter.”[30]  On one hand, the mere fact that a regulatory agency exercises oversight of a licensee and has thus implicitly approved the licensee conduct at issue has been held insufficient to attribute an action to the state.[31] On the other hand, where government intentionally funds an unconstitutional program conducted by private actors, the Courts have generally found state action.[32]

In Jackson v. Metropolitan Edison Co.[33]Justice Rhenquist summarized what has been called the “nexus theory” of state action jurisprudence.[34]

“While the principle that private action is immune from the restrictions of the Fourteenth Amendment is well established and easily stated, the question whether particular conduct is “private” on the one hand, or “state action” on the other, frequently admits of no easy answer…[The] inquiry must be whether there is a sufficiently close nexus between the State and the challenged action of the regulated entity so that the action of the latter may be fairly treated as that of the State itself.”[35]

The nature and degree of interrelationship sufficient to establish such a nexus remains uncertain. As commentators have noted,[36] there are three general theories pursuant to which the courts have found a sufficient nexus to support state action: the public function test; the government ‘entanglement’ theory, and cases where there has been specific authorization or encouragement of the challenged activity. The public function test was established in Marsh v. Alabama.[37] Marsh involved the distribution of religious literature on the streets of a so-called “company town,” where all property, including the streets, was privately owned. In language that arguably foreshadows the instant inquiry, the Court found state action in “permitting a corporation to govern a community of citizens so as to restrict their fundamental liberties.”[38] 

Marsh has been so strictly limited as to suggest a very narrow scope indeed for the public function test. In Flagg Bros. v Brooks[39] the Court considered whether action taken by a New York warehouseman, explicitly permitted by the New York version of the Uniform Commercial Credit Code, was attributable to the state by virtue of that statutory authorization. Plaintiffs, not unreasonably, had argued that when a state legislature passes a law expressly authorizing a private party to take an action previously reserved to the state, that authorization should be held to be state action. The “primary contention was that New York had delegated to Flagg Brothers a power traditionally exclusively reserved to the State.”[40] Writing for the Court, Justice Rhenquist began his analysis by stating  “While many functions have been traditionally performed by governments, very few have been ‘exclusively reserved to the State.’”  The “mere acquiescence” with a private action by government was insufficient to make that action attributable to the State, and the state action doctrine would not apply to “a number of state and municipal functions,” among them education, fire and police protection, and tax collection.[41] Rhenquist did add a disclaimer of sorts, noting that “we express no view as to the extent, if any, to which a city or State might be free to delegate to private parties the performance of such functions and thereby avoid the strictures of the Fourteenth Amendment.”[42]  That question has gained urgency with the proliferation of contracting; it remains essentially unanswered.

In language that is prescient for purposes of this inquiry, the dissent in Flagg protested that

            “[T]he distinctions between “permission” and “compulsion” on the one hand and

            “exclusive” and “nonexclusive” on the other, cannot be determinative factors in

state action analysis. There is no great chasm between “permission” and “compulsion” requiring particular state action to fall within one or the other definitional camp.”[43]

And in a footnote, the dissent worried that under the majority’s theory, “[T]he State can shield its legislation affecting property interests from due process scrutiny by delegating authority to private partners.”[44]

The Court has found state action when cooperation between the public and private sectors is so close as to render them substantially inseparable. This is inevitably a fact-sensitive inquiry, and relatively minor factual distinctions have produced very different results. In Burton v. Wilmington Parking Authority,[45] for example, the Court noted the “public aspects” of a restaurant charged with racial discrimination, primarily attributable to the fact that it was a lessee in a publicly owned building. However, the ruling made it clear that not every lease of public property would be considered a sufficient entanglement to justify a finding of state action. Licensing and regulation have sometimes been held sufficient, but only when these activities are deemed “intensive.”[46] The mere approval of a liquor license has been held insufficient[47] as has regulation of and payment for nursing home care[48] and the grant by government of a monopoly to represent the United States in the Olympics.[49]

The 1983 case of Blum v. Yaretsky[50] is an excellent example of the inadequacies of current state action doctrine. The case involved an alleged due process violation arising out of involuntary discharges and transfers of Medicaid patients in a nursing home. Rhenquist, writing for the Court, declined to find state action. He articulated the standard to be met in the following language:

                        “…a state normally can be held responsible for a private decision

                        only when it has exercised coercive power or has provided such

                        significant encouragement, either overt or covert, that the choice must in

                        law be deemed to be that of the State.”

Acknowledging that over 90% (and perhaps as many as 99%) of the patients in the facility were being paid for by the government, and that the nursing home was subject to pervasive governmental regulation, the Rhenquist majority nevertheless held

                        “That programs undertaken by the State result in substantial funding of the

activities of a private entity is no more persuasive than the fact of regulation of such an entity in demonstrating that the State is responsible for decisions made by the entity in the course of its business.”

In an acerbic dissent joined by Marshall, Justice Brennan underscored the facile nature of this analysis. Noting that a determination whether state action is present will depend upon “a realistic and delicate appraisal of the State’s involvement in the total context of the action taken,” Brennan wrote

                        “The Court’s analysis in this case [proceeds] upon a premise that is

                        factually unfounded…A doctor who prescribes drugs for a patient on the

                        basis of his independent medical judgment is not rendered a state actor

                        merely because the State may reimburse the patient in different amounts

                        depending upon which drug is prescribed. But the level of care decisions

                        in this case, even when characterized as the ‘independent’ decision of the

                        nursing home, have far less to do with the exercise of independent

professional judgment than they do with the State’s desire to save money…On the contrary, the two levels of long-term institutionalized

care enshrined in the Medicaid scheme are legislative constructs, designed to serve governmental cost-containment policies.

In my view, an accurate and realistic appraisal of the procedures actually employed in the State of New York leaves no doubt that not only has the state established the treatment levels and utilization review in order to further its own fiscal goals, but that the State [has] set forth precisely the standards upon which the level-of-care decisions are to be made, and has delegated administration of the program to the nursing home operators, rather than assume the burden of administering the program itself.”[51] (emphasis in original)

While the Court’s decision in Richardson was limited to the immunity issue, and did not explicitly address the question of state action, lower courts have not hesitated to find state action in private prison and institutional detention cases,[56] often noting that the power to deprive an individual of liberty is a quintessentially governmental power.[57]  This line of reasoning is persuasive, but it is difficult to reconcile with cases like Wade v. Byles,[58] where a private company providing security to a public housing project was held not to be a public actor despite the fact that the guards had authority to carry guns, arrest people, and use deadly force.[59]

Complicating state action jurisprudence even further is the tendency of reviewing courts to apply different standards of analysis depending upon the nature of the Constitutional right involved, without, however, articulating the basis for those differences. Commentators have noted that, in cases involving racial discrimination or implicating  First Amendment religious liberties, the Court has been much more willing to find—or assume--state action.[60]  

A recent controversy in Kentucky[61] illustrates this double standard in the public/private distinction, and highlights some of the issues that are likely to become more urgent as states implement Charitable Choice legislation. The Kentucky Baptist home for Children serves 3,600 orphaned, dependent, abused and neglected children on an annual budget of $15.6 million. Approximately $12 million of that amount comes from the State of Kentucky, which is also responsible for placing with the Home the vast majority of the children served. The home is the largest state-contracted provider of such services in the state.

  The Home’s mission statement includes the following language:

                        “We are a Christian ministry that, through God’s direction

                        and leadership, reaches out to children and families with Christ’s

                        love and compassion. We are committed to presenting a clear

                        message of Christian values.”

A female social worker with the home was terminated in October, 1998.  The letter of termination explicitly stated that the termination was because her “admitted homosexual lifestyle is contrary to Kentucky Baptist Homes for Children’s core values.” (There was no dispute about the reason for the termination, no suggestion that the employee was unqualified or had failed to perform adequately.  Indeed, the home had actively recruited her from her prior position.) The day she was terminated, the following memorandum was circulated to the entire staff, announcing a new Board employment policy. That policy read, in pertinent part,

“Homosexuality is a lifestyle that would prohibit employment with Kentucky Baptist Homes for Children. The Board does not encourage or intend for staff to

seek out people within the organization who may live an alternate lifestyle; we will, however, act according to Board policy if a situation is brought to our attention.”

                                                           

 
Ironically, in litigation involving Charitable Choice, the First Amendment may provide a remedy unavailable in other privatization contexts.  In a sense, the Establishment Clause  allows the courts to evade the state action issue, by recasting the issues in terms of government support of religion. But finding a violation of the Establishment Clause also requires the presence of state action, whether that requirement is articulated or not. A recent Sixth Circuit case is illustrative of the point: Hartman v. Stone[70] involved a free exercise challenge to a U.S. Army regulation prohibiting authorized child care providers on military bases from engaging in any type of religious activity in the home while providing child care. The Army conceded that the prohibition burdened the free exercise rights of those using the child care services, since only “authorized” child care providers could operate on the base. The Army justified its regulation on the grounds that the providers were acting as a proxy for government, and therefore religious activity was inappropriate under the Establishment Clause. The Sixth Circuit rejected the Army’s argument not because of any disagreement with the theoretical construct, but because it found that the providers were not, in fact, proxies for the government. It is difficult to argue that such analyses are somehow distinct and distinguishable from the state action cases; indeed, asking whether a private party is a “proxy” for government is asking whether the state has acted through that party.  It does not further clarity or consistency to approach state involvement for purposes of Establishment Clause or Equal Protection[71] jurisprudence as if it were conceptually divorced from conventional state action analysis.
If certain liberties protected by the Establishment Clause and the Equal Protection doctrine are to be accorded a greater importance than, say, due process guarantees, the Courts should say so explicitly, and should just as explicitly justify the distinction.
                                               
            IV.
                                                    Discussion
 
How should the concept of state action be understood? What would a coherent jurisprudence look like?  The Warren Court seemed to be groping toward a holistic approach that was promising[72]; however, the Berger and Rhenquist courts have retreated from that analytical approach, restricted the scope of the doctrine and arguably further confused the issues.[73]
 
Any workable state action doctrine will require flexible application, especially during a period in which we are “reinventing” government. Flexibility need not trump consistency and predictability, however. Certain characteristics of the relationship between government and private entities will always be relevant to the inquiry whether an action can be fairly attributed to the state.  Among the dispositive elements will be the existence, nature and extent of government funding; the nature and extent of government control of the activity in question; the extent to which government has authorized a contractor to exercise government powers; and a functional (holistic) analysis. All of these tests have been used to determine the presence or absence of state action, and none would require a paradigm shift in our approach to the law in this area.
 
Money: Where government money passes to a presumptively private entity, a threshold inquiry should be whether the transaction is a purchase of goods or services, which should not constitute state action, and other types of funding, which may.  Purchase involves a product or service that is generally available and relatively standardized, where production of the good or performance of the service is substantially, if not entirely, controlled by the vendor. When the state goes into the market looking for computer support or engineering services, for accountants to perform an audit, for asphalt to use in paving projects, or similar widely traded goods and services, it is relatively clear that government is simply making a purchase.  Even where the transaction is apparently a purchase of services, however, a significant long-term relationship between a contractor and government, where the government’s business constitutes a majority of the contractor’s income, should be held to raise a rebuttable presumption of state action. When a single “customer” accounts for most of a contractor’s income, that customer clearly has the power to dictate behavior. The existence of that leverage justifies raising the presumption, which can be rebutted by evidence that no active participation by government in fact occurred.
 
Control: Where government controls the manner in which work is done, there should be a finding of state action, without requiring a finding of explicit authorization of the disputed act.[74] This rule is consistent with the Court’s current articulation of the law, if not its practice. Mere regulation should continue to be insufficient, but something less than direct or total control should implicate the state. Where the regulatory scheme substantially limits the options available to the private entity, or prescribes goals and delimits acceptable methods for achieving them, the state should bear responsibility for the results. Indicators relevant to the issue of control—besides money—are ownership, relationship, degree of regulatory supervision and authority, and the actual course of dealing between the parties. Courts might helpfully analogize with existing tests to determine whether a worker is an employee or an independent contractor. Many of the same considerations will be relevant.[75] 
 
Agency. When the state authorizes a private entity to act on its behalf, it creates an agency relationship. When the agent is authorized to exercise powers that are essentially, even if not exclusively, governmental, we are justified in finding that government has acted through that agent.[76] Here, the law of agency and partnership could be particularly helpful and relevant, and it is curious that those principles have not been applied, even by analogy.  When government cloaks a contractor with real or apparent authority to act on its behalf, the ensuing actions should be deemed governmental.[77]
 
Function:  Each of these inquiries is an attempt to answer the question whether the private entity is acting as a proxy for government under the facts of the case. If the contractual relationship has replaced government employees who were previously providing the service, common sense suggests the contractor is a government proxy. Where government is responsible for delivery of the service, there should be at least a rebuttable presumption of state action.  The “exclusive function” test is not good public policy, because the issue is not whether the activity is one that only government does or has ever done, or should do. The issue is whether it is government that is actually “doing” the activity in question.[78]  Where government undertakes an activity, funds it, authorizes a contractor to act on its behalf, and effectively dictates the manner in which it is done, that activity should fairly be attributed to government.
 
Courts have been reluctant to burden governmental units—and ultimately taxpayers--with  liability for the actions of private contractors, but the concern is arguably misplaced. Government can protect tax dollars by contracting for hold harmless or other indemnification provisions. Liability is a recognized cost of doing business, and the allocation of costs is a proper subject for contractual negotiation. Indeed, it could be argued that a refusal to allow government to evade its constitutional responsibilities through contracting will force an explicit recognition and accommodation of potential liability costs. Such a result would benefit everyone: private contractors, government units, and most of all citizens who have a right to demand fiscal, political and legal accountability from those they elect to office.
 
When government acts, government should be accountable. The instrument government chooses should not alter that result. Whether government delivers drug counseling or job placement or any other service through a state agency or a faith-based organization, the program is state action.  Courts should recognize that reality and require adherence to constitutional standards.
 
                                                                        V.
                                                               Conclusion
 Sometimes the most obvious rules produce the most tortured applications. It is a truism of every high school and college government class that the Bill of Rights applies only to the government, that there must be state action in order to find a constitutional infringement.  But by “reinventing” government, we have created mutants and hybrids, neither public nor private, but some admixture of the two. The courts have encountered those mutations much as the blind men encountered the elephant: one finding a snake, one a wall, one a tree trunk. Unless the courts can come to grips with the whole animal, protean and evolving as it is, and fashion a coherent jurisprudence that will safeguard the distinction between public and private and thus protect constitutional liberties without engulfing truly private enterprises, we will find ourselves dealing with 19^th century animals in a 21^st century legal zoo.