Make a Donation

Skip to featured navigation for this section, if applicable.

Skip to general site navigation

Human Services at Risk

by Dennis W. Felty, Keystone Human Services, Inc. and Marshall B. Jones, Pennsylvania State University


Publicly supported social services in America are undergoing a major revolution. After the inception and growth of the welfare state came a pervasive switch to the delivery of services by private, mainly nonprofit organizations. Now comes what promises to be an equally pervasive switch from fee-for-service to managed care. This time, moreover, the dominance of nonprofit organizations in public social services is very much in question. The possibility that managed care might work out to the advantage of recipients is discussed, as is the possibility that it might not. In the latter connection, a formula is advanced for containing "inappropriate" incentives.

The advent of managed care in publicly supported social services has raised basic questions about the role of nonprofit and for-profit organizations, incentives for management and ownership, and quality of service. We do not argue that for-profit organizations necessarily degrade quality of care to make money, nor that nonprofit organizations may not pursue other goals, for example, company size, market dominance, or executive salaries, more assiduously than recipient welfare. We do argue that managed care in the public sector will not produce quality care at reduced cost unless tangential incentives are constrained. Specifically, we argue for a government imposed cap of approximately 11 percent of contract value on all centralized costs, including profits or retained earnings, executive salaries, risk reserve fees, and directors' fees. If the case for such a cap can be successfully made, the way will be open to describing how managed care might work in the public sector, that is, how it might work to the advantage of social service recipients. It does not follow, of course, that managed care will work well just because it might. Some account of what might go wrong, of threats to the managed care revolution, is also in order.

Public Welfare in America

Publicly supported social services have undergone three revolutions in this century. The first was the inception and growth of the American welfare state. With the Social Security Act of 1935, the federal government for the first time insured against old age and unemployment and provided aid for dependent children, child welfare, the blind and otherwise disabled, and maternal and child health. With implementation of Medicare and Medicaid in 1965, the federal government began financing health care. In addition, President Lyndon Johnson's War on Poverty initiated many other social services, including Head Start, food stamps, and community action projects. Today, the welfare state provides a long list of social services, including programs for runaway youth, child abuse prevention, residential treatment, mental health, early intervention, shelters for battered women and the homeless, hospices for AIDS patients, day care, drug and alcohol rehabilitation, family crisis intervention, instruction in English as a second language, employment and training programs, and many more. Also in the postwar years, many human services became "entitlements," that is, services provided unconditionally at government expense to persons who meet specific eligibility requirements.

The second revolution was a quiet one and went largely unnoticed by social scientists until it was well advanced. Many people thought that the advent and rapid growth of the welfare state would signal an end to any role for the private sector in public welfare. Private philanthropy would be superfluous in a society where social and charitable programs were funded by government. The event turned out quite otherwise. Between 1940 and 1989, the number of nonprofit organizations not affiliated with any religious denomination (secular nonprofits) increased from 12,500 to just under a million—an 80-fold increase in 50 years. In the same span of years the number of business corporations increased less than seven-fold.1

This parallel between the growth curves of secular nonprofits and the welfare state was no accident. The nonprofit sector grew and diversified as it did largely because it was funded by government. The 1967 amendments to the Social Security Act specifically encouraged government agencies to purchase social services from nonprofit agencies rather than provide those services directly, and underscored this encouragement by providing matching grants that tripled the value of contributed goods and services. Beginning in the early 1970s, government policy moved hundreds of thousands of people with mental illness or mental retardation from state institutions into community settings. The vast majority were cared for by government-supported nonprofit agencies, many of which were founded at the beginning of the deinstitutionalization process and grew rapidly as funding increased regularly over the next 3 decades.

By 1980, $6.5 billion of federal funds were channeled annually to nonprofit providers of social services. Including other kinds of services, for example, health care, education, and arts or culture, the amount of money channeled by the federal government to nonprofit organizations rose to $40.4 billion.2 A 1982 study conducted by the Urban Institute at 16 representative sites around the country found that 56 percent of all government-funded social services were delivered by nonprofit agencies, 4 percent by for-profit, and 40 percent by government agencies.3 In some fields, government-funded services were delivered almost entirely by nonprofit agencies. Prior to the 1960s, for example, public care of the mentally ill was delivered almost exclusively in large public institutions. By 1985, of 21,948 separate facilities and programs funded by state mental health agencies, 20,909 were private, predominantly nonprofit agencies. In New York City, many municipally funded social services, including community development, mental health, mental retardation, alcoholism services, day care, and Head Start, were delivered entirely through purchase-of-service agreements.4

The prevailing perception of the relationship between government and private organizations was that they complemented each other. What government should not do, private organizations might, and what private organizations could not do, government might. With such an understanding, government and the nonprofit sector would be expected to operate in largely different spheres, to concern themselves with different kinds of programs. What in fact happened was a separation of roles, a clear division of labor in the same spheres. Government's role was to fund social services and the nonprofits' role was to deliver them.5

This second revolution is still continuing. It is even deepening. Many state and local governments now seek to privatize not only the delivery of social services but the management of large statewide and county systems. Service systems that may be affected include child protection, community care of the mentally ill and mentally retarded, and treatment of substance abuse.

Meanwhile, the third revolution, managed care, has already begun. In 1993 Pennsylvania announced its intention to move all Medicaid funding into managed care systems by the year 2000. By early 1995, 23 percent of Medicaid recipients nationwide were enrolled in managed care plans.6 The expectation is that, in the effort to cap funding and limit financial liability for entitlement-based services, all state and local governments will transition to managed care delivery systems and reimbursement structures. Community care for the mentally ill and mentally retarded, drug and alcohol rehabilitation, family support, children and youth programs, early intervention, even perhaps juvenile justice will all be increasingly delivered by private organizations under managed care contracts.

Managed Care

A pervasive transition to managed care in the public sector has far-reaching implications for nonprofit provider organizations and for the services they currently provide. In regard to social services managed care has three critical features. First, the contracting company agrees unconditionally to provide stipulated services to a specified population of persons at a fixed contract cost. Second, the contracting agency must meet the terms of this contract without regard to costs. If the agency miscalculates the number of persons who will need services or the extent of those services, it must still provide them and cover the cost out of its own reserve resources. The contracting agency must, in short, assume risk.

The third key feature of managed care is a consequence of the first two. The at-risk contract is an insurance instrument and, like all insurance instruments, it depends on the law of large numbers. Even when the contracting agency calculates cost projections correctly, it courts disaster unless the population of enrolled persons is large enough to accommodate statistical variation within the defined population. In small populations, the risk is great that random variation in the needs of the enrolled population could result in a ruinous cost overrun. Hence, to be successful, managed care organizations (MCOs) must have large enrollments.

The term "behavioral health" refers to counseling or treatment of mental illness and psychological problems, especially substance abuse. Until recently, managed behavioral health care has been the province of often very large, for-profit companies operating in the commercial market (not in the public sector). Some of the largest of these companies have member enrollments in the tens of millions. Capital reserves are often very large. In the early stages of managed care, these organizations tended to limit themselves to financial and administrative operations. They had little involvement in actually providing care. Their function was to provide the capital reserves required (by law) to sign managed care contracts and to administer operations. Care tended to be delivered by provider organizations or individuals under contract with the MCOs. This is changing, however. Increasingly, for-profit managed care organizations are acquiring service-delivery capacity—through purchase, joint venture or partnerships. Of the 12 largest behavioral health managed care companies, over one third own or control a significant service delivery capacity.

The main strength of the provider organizations is that they generally have extensive experience not only in providing care but in contracting with the public sector. This experience, however, has been exclusively under fee-for-service contracts. Nonprofit providers are small or, at most, modestly sized organizations, with slender capital reserves. To compete in the emerging public-sector managed care market, nonprofit providers will have to become much larger and more diversified. The independent "boutique" provider is an endangered organizational species. In addition, and most important, to be successful, nonprofit providers will have to become MCOs themselves. There is no legal obstacle to doing so. MCOs may be either for-profit or nonprofit. There is a financial obstacle, however. To qualify as an MCO, a firm must show that it is able to accept and manage risk, and to do that it must have access to significant capital.

It can be argued that nonprofit provider organizations should simply accept the situation and let the MCOs organize behavioral health care and other human services in the public sector, while the provider organizations do what they know how to do, deliver care. The counter argument is that a clean split between the two kinds of organizations cannot be maintained and has, in fact, already broken down as MCOs acquire service-delivery capacity and nonprofit providers struggle to acquire MCO capacity. Even if a clean split could be maintained, perhaps it should not be. A division of labor between MCOs and provider organizations might conceivably be clean, but not even. Decisive authority lies with the MCOs. Their size and capital reserves and, most important, their ability to accept and manage risk give them a bargaining position that the provider organizations cannot match. If an MCO and subcontracting provider organizations clash, the MCO, as the risk bearing entity, will generally prevail. There is, moreover, good reason to believe that they will clash because the cultures of the two kinds of organizations are markedly different.

Chief executive officers (CEOs) of major for-profit MCOs and Health Maintenance Organizations command salaries comparable to those of other large publicly traded companies while the salaries of CEOs in the larger nonprofit organizations average approximately $50,000 and exceed $100,000 in only the largest nonprofit organizations.7 Publicly traded, for-profit MCOs have a primary obligation to shareholders to maximize return on investment. Profit margins in government-funded behavioral health are frequently reported to range from 10 percent to 20 percent of gross revenue, on contracts that can often be in the hundreds of millions of dollars. The retained earnings of nonprofit providers are limited by state and federal regulation and in most cases will not exceed 1 percent to 3 percent of gross revenue. In addition, these earnings are not distributed to ownership (there is none) but used to buttress operations or develop new services. Historically, MCO executives have backgrounds in business, while the personnel of community nonprofit providers, even at executive levels, usually have social service backgrounds.

These differences are worth noting because they relate to motivation. We will refer to motivation as "inappropriate" according as people are motivated toward other ends than the provision of quality care, for example, toward profit margins, company size, market dominance, or executive salaries. For most goods it doesn't matter whether motivation is appropriate or not. It matters little whether the makers of mousetraps are trying to make a better mousetrap or trying to make money. What matters is the trap itself and whether it serves the needs of its purchasers. For behavioral health, however, and other services intended to benefit poor and vulnerable people, appropriateness of motivation may indeed matter.

Motivation and Incentives

Social Services Are Not Commodities

Most products are commodities and, if publicly supported social services were adequately regarded as commodities, providers' motivations would be no more relevant than the motivations of mousetrap makers. If it were complete, commodification would erase any distinction between social services and ordinary business products and pave the way to a purely business understanding of service provision. There would be no case for limiting profits and salaries, no more, at least, than there would be for capping salaries in, say, the automobile industry. Implicit, therefore, in the argument we are making is the contention that social services are not adequately regarded as commodities.

The quality of a commodity is defined by its customers. What matters are the customers' needs and preferences as they define them. If potential purchasers do not like the product or are unwilling to pay the price, they do not buy it. They shop around for other products that might serve the same purpose, talk to friends; they may buy a product and then return it. For services it is much the same, except that sometimes customers want the service provider to be motivated in specific ways. They may, for example, want a doctor to care. Again, however, the customer may pick and choose.

Publicly supported social services, however, do not meet this description of a commodified product or service, in the first instance because the recipients of those services are not the customer, that is, the person who actually purchases the service.8 They are not, therefore, in position to exercise the controls that customers usually exercise over what they buy. They cannot shop around or switch. They frequently have only the choice to accept or reject what is provided and rejection is not usually a realistic option. In addition, people with mental retardation or mental illness or children in the child welfare system may not know or be able to consult with knowledgeable others. Vulnerable and easily misled, they lack experience and power as informed consumers and are easily imposed upon.

If publicly provided social services are to be regarded as commodities, it falls to government to play the role of informed consumer. A consumer ordinarily knows (has personal contact with) the product or service that he or she is buying. The best the government could do in this respect would be to make on-site inspections of the services it provides. Unfortunately, even monthly visits would not be adequate, especially if announced in advance. Cleaning up and putting on a good face for the inspector is an old routine. In practice, on-site inspections are much less frequent than monthly and often do not get beyond administrative offices. It could hardly be otherwise. Frequent on-site visiting would be extremely expensive.

An alternative approach is for government to require reports or to favor providers who make reports on the services they provide and outcomes achieved. These reports might involve sophisticated, state-of-the-art management information systems. In fact, most MCOs today tout precisely such systems as the main reason, second only to their capital reserves, why governments and provider organizations should partner with them.

The question is: do these systems suffice to commodify social services? Do they allow us to ignore motivation and incentives as considerations in company and system performance? The answer to this question is, no, they do not. First, human services are often delivered in a context that is unavoidably messy and subjective.

Second, the providers often must accommodate multiple constituencies, with different ideas about what constitutes a positive outcome. In the case of a child in serious difficulty, these constituencies might include the natural parents, foster parents, the courts, the school system, protection and advocacy organizations, the community, and taxpayers, not to mention the child. In such circumstances, the notion of a highly commodified product may be very unrealistic.

Third, an information system is no better than the information that goes into it. So what kinds of information go into these systems? The best systems include many things: schedules, health status, remediation of symptoms, employment, participation in the community, involvement with family, presence of friends, adequacy and stability of housing, absence of incidents and changes in quality of life indicators. Probably most relevant to the issue of commodification are recipient satisfaction surveys.

One asks the recipients to what extent they are satisfied with the services they are receiving. Such surveys have at least the merit of involving the recipient. Unfortunately, satisfaction surveys are notoriously misleading. The root of the difficulty is that human beings are adaptive organisms, at least most of the time. Satisfaction has two components: first the objective facts of the matter and, second, the evaluation one places on those facts. When the evaluation is negative, one may change the facts or change the evaluation. Facts, however, are often difficult to change. The upshot is that satisfaction with health, for example, remains roughly the same regardless of age. Octogenarians are as satisfied with their health as people 40 years younger. Objectively, their health is not as good as the health of younger people, but as people age their standards change. When octogenarians say that their health is okay, they mean "relative to other people our age." People judge the facts of their lives against frames of reference that change with their circumstances. Hence, satisfaction surveys have limited usefulness in assessing the quality of service recipients receive.9

None of this implies that management information systems are not helpful and should not be used. It is important to have reliable information about recipients and programs, especially for appropriately motivated management. Nevertheless, this helpfulness is not so great that it allows one to treat social services as if they were a manufactured product, every one exactly the same as every other of the same description.

Capping Centralized Costs

Of course, motivational appropriateness cannot be measured either. There is no way positively to ensure that the people who administer and deliver social services are motivated to optimize quality. The best one can do is to make sure that unnecessary incentives for other motivations not be present. Board directors, for example, in many of the larger behavioral health MCOs are often paid large fees. It is unusual for the directors of a nonprofit provider to be paid fees. The directors serve because it suits their idea of themselves to help vulnerable and disadvantaged people. If it is possible, however, to obtain the services of experienced and capable people as board directors without paying fees, there seems no good reason to pay them. It is an unnecessary and potentially trouble-making expense.

Profits and executive salaries are also, of course, cases in point. The idea of appropriate motivation is unquestionably strained by large salaries and wide profit margins. This "strain," however, is not necessarily best viewed as an issue in ethics. What really matters is how much of total government expenditures directly benefits recipients. If government appropriates $100 million dollars for behavioral health, how much should be spent on direct care and how much for "overhead"? The particulars of any such split, just what they should be, would require a lengthy discussion. Fortunately, the state of Pennsylvania affords a good, current example of what seems to us a reasonable approach.

Pennsylvania will not fund behavioral health, managed care under Medicaid unless "centralized costs" are less than 11 percent of total contract value. Administrative costs of the MCO, profits, executive salaries, risk fees, and directors' fees are all included under the 11% cap. This formula seems sensible to us, not only for Medicaid in Pennsylvania but, as a generic model, for social services throughout the country. It applies equally to for-profit and nonprofit organizations. A nonprofit's retained earnings are treated by the formula in the same way as a for-profit's return to ownership. And, by lumping all costs other than direct service delivery together, it clearly recognizes that profit is not the only form that inappropriate motivation may take.10

It doesn't include every such form, of course. Conflicts of interest, for example, can also generate inappropriate motivation. Many nonprofit companies exclude from board membership anyone who has financial interests that could create a conflict of interest. While not part of the Pennsylvania formula, this practice is entirely in harmony with it and could be added as a supplement.

Some sources of inappropriate motivation are almost impossible to exclude. The life of a health care executive is almost glamorous these days, what with mergers and acquisitions, high-level business meetings, trips, and fast-breaking news. It is quite possible for a person's head to get so full of deals and dollars that he or she forgets what the purpose of the business is. The Pennsylvania formula does not exclude this possibility, and it is difficult to think of a supplement that could exclude it. When things settle down, however, it may not matter.11

The problem with inappropriate motivation is not that executives cannot earn large salaries and at the same time perform effectively as managers. The problem is the opportunity cost of all motivations other than to deliver quality care. Money spent on profits and large salaries or pursuing market dominance is money not spent on service delivery. Executive salaries should be competitive, of course, and there is no objection to profits provided that, together with other centralized costs, they do not exceed a small part (» 11%) of total contract value. Absent such a cap, however, the only bar to ever-expanding centralized costs and ever-diminishing support for service delivery is the principled restraint of social service executives, directors, and owners. Inappropriate incentives need to be contained because leaders in social services cannot be expected voluntarily and with few exceptions to forego goals which their counterparts in unequivocally commercial enterprise usually pursue.

A Positive Prospect

Integrated Service Delivery

The managed care revolution is not driven primarily by considerations of service quality but by costs and the possibility of containing them. Nevertheless, managed care could work well in the public sector. A managed care contract does not, for example, encourage dependency. The fewer the services a recipient needs and the sooner unneeded services are withdrawn the better for the MCO; and the recipients are no worse and possibly better off for not receiving unneeded services. Fee-for-service contracts, on the other hand, reward a service provider for retaining a recipient in care as long as possible. The longer and more variously the recipient is served, the more richly the provider is rewarded. This difference is certainly worth noting. The main hope in managed care, however, is the possibility it holds out of integrated service delivery.

Under traditional fee-for-service, the social services were extremely fragmented. Each agency typically provided a single service, and an individual family might have to deal with a dozen or more agencies (for example, family support, early intervention, child abuse, alcohol, recreation, etc.). Furthermore, the agencies were rarely coordinated and frequently rivalrous. They often duplicated one another's work or, worse, worked at cross-purposes.

This situation cannot be corrected by calls for "interorganizational cooperation," what one theorist has called "a coordination model."12 The root of fragmentation in the traditional system lies in the method of payment. No agency is authorized or paid to address all of a family's problems. Nor is anyone paid for cooperating with other agencies; in fact, an agency may be penalized for not grinding its own axe, even if doing so works to the family's or other agencies' disadvantage. In managed care, however, the method of payment is radically altered, and this change works in several ways to promote integrated service delivery.

MCOs tend not only to be large but also diversified. The larger and more diverse its services, the less likely that an MCO's cost calculations will be upset by random variation. The costs of one service may be underestimated but those of another may be overestimated. Overall, they balance out.

For its part, government also benefits from large, diversified contracts; the fewer in number, the lower the cost of letting them. If a government could sign one contract for all of its social services, the savings would be very large indeed. Many fewer civil servants would be needed. Whole bureaus could be shut down. If one all-encompassing contract is not feasible, and there are many reasons why it might not be, then a few large contracts would be preferable to the present system of innumerable small contracts.

Pennsylvania, for example, plans a "carve-out" for behavioral health care. That is, the state proposes to contract separately for medical and behavioral health Medicaid services. Pennsylvania has given its counties a "right of first opportunity" to manage Medicaid behavioral health dollars, estimated to total one half billion dollars by the year 2000. To manage these funds, most counties will have to partner with a national MCO. At this point it seems probable that the half billion dollars will be split by a small number of companies. Funding for community care of persons with mental retardation, for drug and alcohol rehabilitation, family support, and children and youth may also move into managed care. A single contract for the entire state would be too big. Statewide and large regional contracts would sacrifice the local control and community connectedness of county systems. However, even integrated master contracts at the county level would be a great simplification over the present system and a great saving for the state.

Community and Functional Connectedness

That MCOs will provide multiple services is assured. It is both their intention and that of the governments which fund them. In addition, MCOs have the motivation or at least the incentive to integrate those services. A per capita contract puts the MCO at risk. If one home visit will suffice, it would be folly to make several. The MCOs can only lose by duplicating services or failing to coordinate them. However complicated an individual's or family's problems may be, it is in the MCO's interest to address them in a conceptually coherent way and for each service involved to know what the others are trying to do, when, and how. In addition, the MCO has an incentive to subcapitate risk to comprehensive, integrated provider systems. By doing so, the MCO passes risk on and does not incur the administrative cost of individual case management and service authorizations. By accepting risk, the delivery system gains greatly in flexibility and the ability to customize its services.

These services, however, are delivered at a local level, in communities, and focused on individuals and families. If they are to be successfully integrated, each service must know what other services are doing with a particular family. The delivery organization must be functionally connected at a local level. Such "functional connectedness" is possible only when the provider organization itself is, at most, regional in size. The multiple services provided by an integrated delivery system must be brought together managerially at a relatively low level. The lines of report must be reasonably short and for that to be the case the system itself must not be excessively large or geographically dispersed. The issue here is not ownership necessarily or even tables of organization, but the delivery system as an operational unit. Perhaps the system is owned as a subsidiary by a larger organization. At least in principle, that shouldn't be a bar to functional connectedness.

Other considerations work to the same conclusion. When a provider organization is regional or smaller, a community sees it as a local institution. Some of the organization's top leadership are local people. Some of the leadership may have family members or neighbors who are recipients. If the organization is also minimally bureaucratic, its procedures can be customized to local ways and expectations. The payoff for the provider organization is the positive regard, support, and cooperation of the community. Local people volunteer their services in many ways, from service on boards to chauffeuring clients from place to place. They contribute money and gifts in kind, for example, as discounts on goods or services. Not least important, the community experiences a sense of ownership and supports the organization politically, at both local and state levels.

Because behavioral health care is largely delivered in communities rather than in isolated settings, community relations affect the quality of care. In the future, managed behavioral health care in the public sector will be dominated by nonprofit providers which have acquired MCO capacity or by for-profit MCOs so positioned contractually with government that they can obligate provider organizations to subcontract with them. Which way it turns out will affect community relations and, through them, the probability that quality care is offered.

Involvement with the surrounding community is probably the best protection one can have against the abuse of recipients. The more people there are who regularly come into contact with recipients and the care they receive, the less likely abuse will occur. The probability of abuse decreases as the number of people who are observing or who know the recipient and his or her situation increases. It is in isolated places, empty stairwells and unlit streets, that abuse, like crime in general, is most likely to occur.13 People rarely volunteer their services, however, to for-profit companies, even if they are small; and volunteers are the most frequent points of contact with the community. In this respect, nonprofit MCO-providers have a clear advantage over for-profit companies.14

If the potential of managed care in the public sector is to be realized, a balance will have to be struck between large size and functional connectedness. If that can be done, however, the future would be promising. The nation would be covered by a mosaic of integrated delivery systems, each an incorporated entity under contract with a state or local government. These systems would seldom cover more than part of a state, except perhaps for a few small states, and would themselves be organized into smaller units, usually by counties. Some systems would be related to others by common ownership or use of a common information system; many more would be in competition with one another. In most places the mosaic would be more than one layer thick. That is, delivery systems would overlap, in such a way as to compete with one another directly on issues of quality, affordability, access, and choices offered to the people who use their services. Inappropriate incentives would be held within reasonable bounds and all systems would provide integrated services tied by voluntary participation and family and individual needs to the communities they serve. It is a pretty picture.

Other Possible Outcomes

MCOs Versus Provider Organizations

Unfortunately, it is not the only possible outcome of the managed care revolution. The present situation is dominated by two starkly different arrays of organizations: on the one side, national, publicly traded for-profit MCOs with large capital reserves but little experience in service delivery and, on the other, nonprofit community providers with extensive experience caring for recipients in the public sector but slight, if any, capital reserves and minimal experience in creating or managing regional delivery systems.

The government-funded, behavioral health market is large, more than $50 billion in 1994,15 but nonprofit organizations active in it are relatively small, relative, that is, to for-profit corporations active in the nongovernmental market. Recently, for example, 100 human service agencies in Pennsylvania, almost all of them nonprofit and collectively providing more than $1 billion of service annually, formed a statewide "community behavioral health care cooperative." An average member of this cooperative provides $10 million of service annually. This is small compared to the giants of for-profit, behavioral health care.

This sharp division between nonprofit providers and MCOs in behavioral health is not expected to last and is, in fact, already changing. The nonprofits cannot survive unless they consolidate and acquire MCO capacity. As that happens, the commercial MCOs are motivated to protect their position in the market by acquiring direct service capacity. The provider organizations cannot be shuffled out of the picture. In one form or another they must remain because without them there is no service delivery. Once the provider organizations acquire MCO capacity, however, they no longer need the MCOs. They can contract with government themselves, and the MCOs will have no one with whom they can subcontract to deliver services. So the race is on.

But the runners are not staying in their lanes. At the same time that they are acquiring MCO capacity the provider organizations are partnering with MCOs. Community nonprofit providers have no experience with managed care and realize that they have much to learn from the commercial MCOs; and, of course, they need capital desperately. The MCOs, in their turn, are hiring experienced individuals from government and the nonprofit providers. They may not have any experience as organizations in the public sector, but they can hire people who do. The two kinds of organizations are not only changing but partnering with and interpenetrating one another.

Much Money To Be Made and Turbulence

To complicate matters, serious amounts of money are to be made. In the current system, administrative overhead at the state and county levels amounts to a large part of the social service dollar. Under fee-for-service, an original dollar must be processed through the state bureaucracy, then the county, and finally the administration of the provider organization. Because each service is contracted and monitored independently and there are very many such contracts to be let, the administrative and transactional costs to government are large. An MCO can offer a government purchaser an increase in quality and a major reduction in cost and still make a handsome profit. In one stroke, government substantially reduces its costs and at the same time transfers both the responsibility and the risk to the MCO. It is small wonder that, for the time being, many governments are not looking too closely at profit margins.

The largest amounts of money, however, are not made via profits or salaries but by buying, building, and selling companies. Once formed, nonprofit providers tend to continue in existence under the same auspices for long periods of time. For-profit MCOs, however, especially relatively small ones, have a short life expectancy. The industry is undergoing rapid consolidation; companies merge, acquire and are acquired by other organizations at a high rate. In 1994, for example, Green Spring merged with Magellan, Health Management Strategies was acquired by Charter Medical (a health maintenance organization, which was itself reorganized as a holding company), and Medco, with the help of a third company, separated from its original owner (Merck & Co.) and was renamed Merit Behavioral Care Corporation. It is by deals of this kind that the serious money is made.

Matters on the governmental side of the fence are not quite so turbulent. Nevertheless, there too things are changing. In particular, the regulatory environment in behavioral health care is likely to be greatly relaxed in the next few years. "Regulatory environment" refers to categorical funding, entitlement programs, and innumerable regulations regarding the physical environment in which care is provided, the credentials of care providers, their number and levels of training, and the clients themselves¾ who they may be, how housed or grouped. If these regulations are relaxed, the costs of behavioral health care will undoubtedly be reduced.

This relaxation in the regulatory environment is not necessarily permanent. At present, state governments are willing to allow more latitude to service providers—perhaps to encourage them to make the changes that operating in a managed care environment requires. In the future, however, the attitude of state governments may change. Relaxing regulations implies a greater delegation of authority to the private, provider organizations. Many people oppose any such delegation on principle because it attenuates accountability to public bodies and, therefore, to the voting public.16 When abuses occur, as inevitably they will, a cry will be raised to bring the provider organizations under tighter control, to regulate them more intensively. If that happens, the downside will be increased costs.

Competition May Fail

As managed care expands in the public sector and the turbulence so marked in its early phases subsides, something like a mosaic of integrated delivery systems may emerge. For a number of reasons, such a mosaic could also fail to emerge. A single, statewide contract, for example, could shut down competition. Provider organizations would have no choice but to contract with the MCO holding the contract and the state, once it had signed such a contract, would have difficulty switching to another MCO, at least not one with which it had had any previous dealings. The state would have to choose between a hypothetical delivery system and a known quantity. Barring scandal, the odds would strongly favor the entrenched MCO.

Competition could fail in other ways. The state might sign several regional contracts but only one in each region. In such a case the state would at least have some experience with multiple MCOs, but the MCOs themselves might be content with their local monopolies and have no interest in competing with one another. No matter how it develops, a failure of competition to develop would have several pernicious effects.

In the absence of competition the motivation to integrate service delivery is greatly weakened. In principle, an MCO's profit margin could be increased, but the margin might already be broad and for political reasons (not to disturb sleeping dogs) it might be preferable simply to leave matters where they were. If the geographic area covered was large, integrating service delivery would be difficult in any event and, in the absence of a clear self-serving motivation to do so, not worth the benefits that others for the most part would reap.

Containing inappropriate incentives depends on two engines, and competition is one of them. If competition fails, there will be no bar to excessive profits or salaries except the unwillingness of the state to fund them. The willingness or unwillingness of the state, however, is subject to political engineering. If the state is paying less for social services than it once paid (a circumstance that could easily coexist with inappropriate incentives), what does it matter if some people are making a great deal of money? After all, one expects people to make money in business. One forgets that publicly supported social services are not businesses in the usual sense of the word. Most businesses are not publicly supported and those that are, for example, some defense contractors, sell fully commodified products. Even then, profits are usually limited to 10 percent.

If inappropriate incentives (excessive profits and salaries, directors' fees) are not contained, if conflicts of interest are not excluded, no one should be surprised if the MCOs and provider organizations are not motivated to provide quality service. They have other things on their minds. And if the MCOs and provider organizations are not motivated to provide quality service, they probably won't. MCOs, after all, have something to gain by simply reducing services, not integrating them, and providing less care. They can make money that way and, if making money is the primary objective, the risk that costs will be reduced by allowing services to deteriorate cannot be ignored.

In the worst eventuality, the provision of social services degenerates into distributed warehousing. The recipients in such a warehouse are still cared for "in the community," but they might as well be encumbering the corridors of state institutions. Distributed in small numbers, stuffed into the nooks and crannies of an urban space, no one notices the emptiness of their lives or the absent-minded neglect of their caretakers. Meanwhile, the management information system assures everyone that all is more or less as it should be, with only the usual ups and downs to be expected when information about individuals or small groups is collected routinely.


Publicly supported human services in America are undergoing a massive transformation. In the next decade they will experience fundamental changes in financing (managed care), drastic consolidation, and entrance into the field by multibillion dollar national companies. These changes do not necessarily vacate the traditional values of local control, voluntary service, and quality care. A nationwide mosaic of modestly sized, comprehensive delivery systems, consistent with the impending changes, could retain those values and add a level of integration (case management) that the fee-for-service system lacks; and it could do so at reduced cost to government. The chances of such an outcome will be greatly improved if state governments: (1) adopt a formula for capping salaries, profits, and administrative costs similar to that in Pennsylvania and inclu ding an exclusion of conflicts of interest, (2) keep contracts small enough to assure community and functional connectedness, and (3) not only encourage competition in the abstract but make sure that they (the states) are never without an alternative organization to which they can turn in the event the present contractor provides less than quality service. By the same arguments reversed, managed care in publicly supported social services could work out badly. The outcome is especially critical because in either case a return to fee-for-service will almost certainly not be an option.


  1. Peter Dobkin Hall, "Historical Perspectives on Nonprofit Organizations," in The Jossey-Bass Handbook of Nonprofit Leadership and Management, ed. Robert D. Herman and Associates (San Francisco: Jossey-Bass Publishers, 1994).
  2. Lester M. Salamon, America's Nonprofit Sector: A Primer (New York: The Foundation Center, 1992), p. 47.
  3. Lester M. Salamon, "Partners in Public Service: The Scope and Theory of Government-Nonprofit Relations," in The Nonprofit Sector: A Research Handbook, ed. Walter W. Powell (New Haven, Conn.: Yale University Press, 1987), p. 103.
  4. Steven Rathgeb Smith and Michael Lipsky, Nonprofits for Hire: The Welfare State in the Age of Contracting (Cambridge, Mass.: Harvard University Press, 1993), pp. 5,9.
  5. For further discussion of this division of labor, see Neil Gilbert, Capitalism and the Welfare State (New Haven, Conn.: Yale University Press, 1983); Jennifer R. Wolch, The Shadow State: Government and Voluntary Sector in Transition (New York The Foundation Center, 1990), p. 223; Smith and Lipsky (n. 4 above).
  6. Monica E. Oss, "System Evolution: Trends in Behavioral Health and Human Services Financing" (paper presented at the annual meeting of the Child Welfare League of America, Washington, D.C., March, 1996).
  7. "Great Extremes Exist in Executive Compensation," Open Minds: The Behavioral Health Industry Analyst (October 1995): 8.
  8. For more on the failure of social services to meet the usual definition of a market commodity, see Henry Hansmann, "Economic Theories of Nonprofit Organization," in Powell (n. 3 above); Lester M. Salamon, "Of Market Failure, Voluntary Failure, and Third-Party Government: Toward a Theory of Government-Nonprofit Relations in the Modern Welfare State," Journal of Voluntary Action Research 16 (1987):29-49.
  9. Gordon Bultena and Edward Powers, "Effects of Age-Grade Comparisons on Adjustment in Later Life," in Time, Roles, and Self in Old Age, ed. Jaber F. Gubrium (New York: Human Sciences Press, 1976).
  10. The Pennsylvania formula is cited here as illustrative only. There is nothing magical about the figure of 11 percent. Nor is it to be expected that the same figure would apply to all services or in all places. In fact, Pennsylvania itself is already negotiating different figures with different counties. There is, moreover, a giant "hole" in the Pennsylvania formula, namely, that it applies to MCOs only and not to subcontracting provider organizations. An MCO, however, may own a subcontracting provider organization, thus opening the way to what are called "second tier" profits. If centralized costs were 10 percent and 20 percent of contract value for the MCO and its subcontracting provider, respectively, only 70¢ on an appropriated dollar would directly benefit the recipients. The Pennsylvania formula says nothing about second tier-profits. A better formula would control them.
  11. Two further issues should be mentioned. The first is an issue in ownership. A nonprofit may own or control a for-profit company. In such an arrangement, an executive or board member of the parent company who holds equity in the subsidiary has a clear conflict of interest, because he or she is positioned to direct the parent company's policy, for example, by shifting costs, so as to favor the subsidiary and, hence, his or her own interest.
    The second issue concerns the distinction between "profits" and "return on capital investment." An MCO must have access to significant amounts of capital. It must either have or be able to arrange for many millions of dollars to be held in reserve against the possibility of a cost overrun. This money is not spent, however. If all goes well, the MCO will have all of its risk reserve when the contract ends. In fact, it may have more because the money held in reserve may also earn interest. In addition, the MCO or whoever puts up the risk reserve is entitled to a risk reserve fee, usually around 15 percent and paid for under the cap as a centralized cost. In contrast, the capital requirements of an MCO in terms of money it must spend in order to do business are light, often no more than a management information system and some computers. Now consider a company which invests $1 million in such software and equipment. The company then obtains a $100 million contract and meets its contractual obligations for $97 million, thereby making a $3 million profit. The profit is modest, 3 percent, but the return on investment is very large, 300 percent. It is not clear to us that return on investment either can or should be controlled, but its separateness from profit as a percentage of contract value should be recognized.
  12. Richard Bush, "Survival of the Nonprofit Spirit in a For-Profit World," Nonprofit and Voluntary Sector Quarterly 21 (1992): 391-410.
  13. Oscar Newman, Defensible Space: Crime Prevention through Urban Design (New York: Collier Books, 1972).
  14. This advantage could be mitigated by the use of volunteer inspectors. Social service motivation takes many forms, and one of them is a concern that people in care not be abused. A voluntary inspection system would have to include training (because there is something to know about both the process of inspection and what is being inspected) and oversight. The cost of such training and oversight would fall to government, but this cost would be incomparably less than the cost of hiring inspectors.
  15. "$54 Billion Spent on Behavioral Health Treatment," Open Minds, The Behavioral Health Industry Analyst (November 1994): 12. The total governmentally funded health market is, of course, much larger, approximately $300 billion in 1994.
  16. It is arguable that the attenuation is not very great. The lines of accountability that bind the heads of government agencies to elected officials and the voting public are often already so long and attenuated that they are only a little more binding than no lines at all. In addition, the directors of private organizations may be more exposed to legal action than the heads of governmental agencies. At the minimum, a suit against a private director is more likely to be settled quickly than a comparable suit against a government official.
For over forty years, Keystone Human Services has been steadfastly Advancing the Human Spirit. Learn how you can help. Join us.


  • Like Keystone Human Services on Facebook Like Keystone Human Services on Facebook
  • Follow Keystone Human Services on Twitter Follow Keystone Human Services on Twitter
Powered by Google