Human Service Organizations and Moral Work

There is another sense in which working on people is inherently moral. Fundamental to such work are decisions about the allocation of resources to the clients. These may include money, time, and expertise. Inevitably, the demand for these resources outstrips their supply, resulting in a system of rationing (e.g., first come first served, clients with greatest perceived need, younger over older patients). Rationing resources to clients involves a moral categorization of deservingness. Whatever may be the rationale and merit of the allocation rule, fundamentally it conveys an evaluation of social worth, since some clients become more deserving than others (Roth 1972; Prottas 1979; Lipsky 1980). As a result, evaluations of social worth locate and reaffirm the place of clients in a moral stratification system that determines their rights and claims to scarce resources. As I show later, these systems of moral evaluations determine and rationalize the activities of the workers.

The New Institutional Perspective

By shifting the emphasis from technical rationality to institutionalized rules as the engine that drives organizational forms, the new institutional perspective has provided a powerful framework for understanding human services organizations. It has sensitized us to the importance of organizational structure as a manifestation of institutional rules rather than technology. C. R. Hinings and C. Greenwood (1988, p. 8) express the relationship between institutional values and structure in the concept of "archetype," which they define as "a particular composition of ideas, beliefs and values connected to structures and systems." Moreover, the emphasis on conformity with the institutional environment points to processes of structural isomorphism; organizations in the same sector acquire similar structures that the institutional environment imposes on them through coercive, mimetic, and normative processes (DiMaggio and Powell 1983).

While we recognize that human service organizations are embedded in institutionalized moral systems, it is important to emphasize that these systems may lack consensus or internal consistency (D'Aunno, Sutton, and Price 1991). Moreover, as I will point out, moral work is highly contextualized, reflecting the particular cultural, political, and economic exigencies of the local community in which such work takes place. As a result, these organizations have to make choices among contending moral systems or conflicting rules within them that, in turn, will be reflected in the elements that constitute their organizational forms. The moral choices are not only reflected in the structure but also in the service technologies themselves. S. R. Barley and P. S. Tolbert (1997) use the concept of "scripts" to denote the mechanism by which institutional rules are enacted in the technology. They define scripts as "observable, recurrent activities and patterns of interactions characteristic of a particular setting" and propose that institutions are enacted through them (Barley and Tolbert 1997, p. 5). In his study of two hospital radiology departments, Barley (1986) showed how the scripts that CT scanner technicians used embodied the institution of medical dominance. Finally, the new institutional perspective has also shown how both the organization and its workers can be active agents in deciding which moral rules to enact or ignore. C. Oliver (1991), for example, proposes that organizations can engage in a variety of tactics in response to institutional pressures, ranging from avoidance to manipulation, especially when there is a lack of institutional consensus. Thus, the organizations themselves can undertake moral entrepreneurship-mobilizing constituencies and developing network relations that reinforce and institutionalize their own moral beliefs (Zucker 1988). This is exemplified by the feminist health centers C. Hyde has studied (1992). Moreover, workers in organizations doing moral work are active interpreters and promoters of moral rules, not the least of which are their own. For example, welfare workers often evoke their personal moral standards in deciding to sanction recipients for noncompliance (Hasenfeld and Weaver 1996).

Moral Work and Organizational Forms

When a client is accorded high social worth, the staff are motivated to mobilize all the necessary organizational resources to affirm such a status. In contrast, when a client is viewed as morally deficient she becomes "undeserving" and is subject to a moral test before gaining access to organizational resources. Mothers who become single parents because of the death of a working spouse are morally deserving of universal benefits (i.e., Survivor's Benefits). Mothers who become single parents because their spouse deserted them are morally undeserving and can only apply for means-tested public assistance (Fraser 1989). Attribution of responsibility signifies whether the clients themselves are morally responsible for their predicament or whether they are victims of circumstances beyond their control. This assumption, in turn, affects the degree to which the organization puts the onus on the clients to justify their claim for services. In the first instance, clients must often undergo "repentance" or publicly profess their moral deficiencies to qualify for services. For example, applicants for general assistance are assumed to be responsible for their predicaments because of lack of a work ethic and must undergo a work test (i.e., participate in work activities) to obtain relief In contrast, persons eligible for unemployment compensation are assumed to be victims of economic circumstances. Assumptions about amenability to change influence the degree to which the organization commits itself to bringing about change in the client's circumstances. Students tracked into vocational versus academic tracks are assumed to lack the capacity to excel intellectually, and the school is less likely to invest in them because they are socially devalued (Oakes 1995). Similarly, assumptions about the desired end results influence the service goals and objectives of the organization. Schools that truly believe that developmentally disabled children can be educated to function in the normal society" commit themselves to finding effective educational technologies that can integrate the children into regular classrooms (Handler 1986). Other schools that only give lip services to the idea of mainstreaming are more likely to find reasons to segregate these children (Weatherley and Lipsky 1977). Finally, whether the organization treats its clients as objects or subjects determines the extent to which clients will have a voice in what is done to them. Organizations that treat their clients as subjects encourage them to become active participants and to have a voice in the decisions about their course of service. In contrast, when clients are treated as objects, they are worked on rather than worked with.

Following the new institutional perspective, I propose that these moral assumptions find expression in organizational forms and practices. By organizational forms and practices I mean especially (a) the organizational output goals, (b) the interorganizational network or task environment, (c) the service technology, (d) the organization of work, and (e) client-worker relations (see also Hannan and Freeman 1989). Specifically, the output goals will reflect the assumptions the organization makes about the social worth of the clients and the desired end results. The interorganizational network will manifest the commitment of the organization to mobilize the necessary resources to achieve the desired end results and to legitimate its moral assumptions. The service technology is also influenced by assumptions about the clients' social worth, attribution of responsibility, and amenability to change. The organization of work reflects the value placed on the needs of the client in contrast to the needs of the workers. Client-worker relations reflect, in addition, assumptions about the clients as objects or as subjects.

Organizations that ascribe high social worth to their clients-seeing them as victims of circumstances beyond their control and viewing them as amenable to change and entitled to an active voice in the decisions about their service trajectories-are more likely to be structured as "client-centered" organizations. By this I mean that organizational ideologies will place a high degree of commitment on the clients; the inter-organizational network will focus on mobilizing needed resources and services for them; the service technology will be highly individualized and tailored to the specific attributes and needs of the clients; the formal structure will be debureaucratized; and staff-client relations will be extensive and characterized by a high degree of mutual trust. Examples of such organizations are the feminist health centers Hyde (1992) studied, especially during their formative years, or the response of the Madison (Wis.) School District to the needs of developmentally disabled children (Handler 1986). In contrast, organizations that ascribe to their clients low social worth-attributing responsibility to innate deficiencies in the clients themselves, not seeing them as amenable to change, and treating them as objects-are more likely to develop organizational forms that will demean them. In such organizations ideologies toward the clients tend to be punitive, interorganizational relations are formed mostly with other social control agencies to affirm the moral inferiority of the clients, the service technology is highly routinized and bureaucratized, and client-staff relations are limited and based on suspicion and mistrust. Contemporary welfare departments exemplify such forms (Prottas 1979; Bane and Ellwood 1994).

Organizations that encounter multiple and conflicting moral systems are likely to give them expression within the organization through service units that are decoupled from each other. It is not uncommon for such organizations to have, for example, multiple service technologies that are guided by different if not conflicting moral assumptions (Strauss et al. 1985). Examples of such organizations are mental health agencies that have added drug abuse treatment units, or what I. D'Aunno and R. H. Price call "hybrid organizations" (1985). By decoupling the different service units and their respective technologies from each other, conflict is avoided and legitimacy is maintained.

Organizations cannot always institutionalize all of their moral assumptions, that is, give them an expression in their organizational forms. The institutionalization process requires adequate resources, knowledge and expertise, and cooperative external network relations. Yet, the organization may experience obstacles in attaining them. This is particularly the case when interest groups controlling key resources that the organization needs do not accept its moral assumptions. As I show later, the failure of welfare departments to implement a "rehabilitation" ideology was due, in part, to a lack of adequate external and internal resources. Moreover, organizations may change their moral assumptions in order to rationalize organizational forms that arise from the need to adapt to changing environmental exigencies. In other words, when the needs of the organization to survive and adapt no longer support its original moral assumptions, these assumptions, in turn, are likely to be changed to fit the new reality (Scott 1969). Thus, the moral assumptions both constitute and are constituted by organizational forms and practices.

The Origins and Transformation of Institutional Moral Rules

Yet, these broad moral edicts get their own particular spin in the local community, the second source for institutional moral rules. When moral work is conflictual and ambiguous, especially regarding the control and management of deviance, upper-level politicians delegate considerable discretion to the local level. In doing so, they buffer themselves from the controversies that surround the symbols they espouse, and they need not be concerned with the difficult issues of implementing the programs that must do the moral work. Local officials, in turn, design the programs in response to the local political economies and the moral assumptions that justify them. Welfare programs are a prime example. Each local welfare office is distinctive and reflective of the dominant moral assumptions in the local community in which it is embedded. Indeed, when local conditions change so do the moral assumptions guiding the program. When Massachusetts experienced an economic boom and a consequent decline in its welfare rolls, it instituted a welfare-to-work program, the Employment and Training Choices Program, that viewed welfare recipients in favorable moral terms. The program was in effect voluntary, it provided welfare recipients with many services as incentives to participate, and it aggressively developed job opportunities for the recipients. However, with a declining economy, rising welfare rolls, and a shift to a politically conservative government, welfare recipients were redefined as the "enemy." The program was transformed to become mandatory and punitive, setting strict time limits on receipt of aid and putting the onus on the recipients to find jobs (Handler and Hasenfeld 1997).

The organization itself is a third source of moral rules through its own entrepreneurship. The feminist health centers Hyde (1992) studied pursued and instituted new moral rules in their services and internal structure based on a feminist ideology. The ideology was expressed in service goals that aimed to give women control over their own health through self-help and participation in social action, and in an internal structure that was based on collective governance. The Madison School District, in contrast to other schools, assumed that disabled children were morally worthy, amenable to change, and could be effectively mainstreamed It developed an organizational ideology that saw parents as part of the solution and sought actively to involve them in curricular decisions about their children. It gave parents knowledge, promoted their active participation, and assigned them advocates who negotiated on their behalf with school officials (Handler 1986).

When organizations can form coalitions and join other interest groups that share their moral rules, they may be able to influence social legislation in order to institutionalize these rules. When they are successful, their organizational forms and practices also gain prominence through normative and mimetic processes. For example, the National Alliance for the Mentally Ill, through an effective coalition of parent groups, has been very successful in changing dominant conceptions about schizophrenia, resulting in significant changes in treatment practices (Hatfield 1991).

Organizations also change their own moral rules when they need to rationalize their adaptive strategies in the face of a changing environment. R. Scott's (1969) classic study of agencies for the blind shows how sheltered workshops changed their moral assumptions from the position that the blind ought to be integrated into the regular labor market to a position that they ought to be protected from such a market. The change was a result of pressures on the workshops to maintain their fiscal viability and, thus, to keep the more productive and able blind persons. In other words, moral assumptions not only underlie new organizational forms and practices but are also used to justify forms and practices that arise out of the need to adapt to external political and economic forces.

Organizational forms and practices produce moral consequences by the way in which clients are treated and services are delivered. Although these forms might be justified based on technical rationality or efficiency, they implicitly generate and reinforce moral conceptions about their clients. These conceptions may, of course, be incompatible with publicly espoused belief systems. Yet, these implicit conceptions have greater currency in guiding the behavior of the staff and are, therefore, institutionalized in the organizational form and organizational practice. Being mutually reinforcing, forms and assumptions reproduce themselves over time.

Finally, at the microlevel, workers are a fourth source of moral rules. Workers engage in moral entrepreneurship through their own actions. Inevitably, in organizations that do moral work, staff members exercise considerable discretion. The organization is dependent on them to interpret the rules and apply them to specific cases. No matter how many rules the organization promulgates, it is left to the line staff to gather and interpret the information about their clients (Lipsky 1980). They can always manipulate the information and find rules or organizationally sanctioned rationales to justify their actions. Clients, especially when they lack power, become dependent on the workers to construct their cases in moral terms, and they have little recourse to redress such constructions (Handler 1986). Therefore, the personal belief systems of the workers play a significant role in operationalizing the service technologies and, particularly, in shaping client-staff relations (Hasenfeld and Weaver 1996). Moreover, workers develop practices that enable them to cope with and manage the particular exigencies they encounter, such as the amount of time and resources available to them and the size of their caseloads. They typify their clients and make service decisions that take into account these factors. Furthermore, they rationalize their actions by morally constructing their clients. That is, they engage in moral entrepreneurship. These personal belief systems and moral rationalizations are shared among groups of staff members because they have similar backgrounds, training, and experiences; they face the same work exigencies; and they communicate with each other about their work situation. It is through this process of sharing that personal moral entrepreneurship becomes institutionalized in organizational practices. In effect, they represent social policies enacted from below (Lipsky 1984; Sandfort 1999).

The "Iron Cage" Revisited?

The Organization of Welfare Departments as Moral Practices

The "Suitable Home" Ideology (1935-50)¹

The early years of the federal Aid to Dependent Children were heavily influenced by the Mothers' Pension philosophy and, specifically, by the ambiguous moral concept of a "suitable home." It was driven by the domestic code of the mother as a homemaker responsible for maintaining a good home life and providing for the proper moral, physical, and mental development of her children. Single mothers who did not conform to the moral code were viewed as having low social worth and as a danger to the community. Because of their moral flaw, they were not perceived as amenable to change. Thus, to qualify for aid the mother had to prove that she was morally "fit." Not surprisingly, in many states coverage was limited to what became known as the "gilt-edged widows." While Aid to Dependent Children legislation did not require that children live in "suitable homes," the concept nonetheless was widely practiced and endorsed by the professional community, including the American Public Welfare Association and officials of the Children's Bureau (Bell 1965, p. 29).

Local welfare departments were left to interpret the concept of the "suitable home" according to prevailing community standards. Thus, local offices and individual workers engaged in their own moral entrepreneurship in defining what was meant by suitable homes, particularly since the workers exercised enormous discretion. As W. Bell (1965, p. 41) put it, "Emphasis was placed according to the importance attached to certain subjective standards by the community, the agency and the individual worker." Typically, it meant that mothers with illegitimate children and women of color were excluded from aid. Southern states were particularly blatant in their moral degradation of African Americans. Louisiana was the first to adopt the "employable mother" rule "requiring all AFDC families with children seven years old or older to be refused assistance as long as the mother was presumed to be employable in the fields" (Piven and Cloward 1971, p. 134).

There were also considerable local variations in the organization of welfare offices and the aid eligibility practices that they instituted. Political appointees with strong local ties and values staffed local offices (quoted in Bell 1965, p. 37). Only in 1939 did federal legislation require that a merit system be used to select workers. Still, most workers were not trained professionals in social work doctrine and practice (Bell 1965, p. 134).

Over time, a service technology evolved that placed considerable emphasis on surveillance and the imposition of coercive standards of conduct. During the 1950s and early 1960s, with the increased migration of African Americans to the large urban centers and the advent of the civil rights movement, many local communities felt morally threatened by the rise of families of color (including those with illegitimate children) seeking aid and by the increasing costs of aid. Many communities reacted by instituting more punitive practices. In the infamous case of Newburgh, New York, which instituted highly restrictive welfare policies, one of the councilmen stated, "this is not a racial issue, but there's hardly an incentive to a naturally lazy people to work if they can exist without working" (reference illegible). Workers made sure that the mothers did not maintain a liaison with any man ("man-in-the-house" rule) and made eligibility for aid contingent on establishing paternity. Midnight raids in search of the elusive man were not uncommon. Fear of fraud resulted in extensive use of "collateral contacts" with relatives, neighbors, friends, landlords, merchants, employers, schools, police departments, health agencies, public agencies administering unemployment compensation and old age and survivors' insurance, banks, and credit bureaus to uncover hidden family resources (Bell 1965, p. 87). Welfare offices developed specialized investigation and fraud detection units. Workers used their authority and considerable discretion to invade homes at will and to demand that their norms of conduct regarding child rearing and child care, money management, and work be followed. The relationship between the welfare workers and their clients was based on mistrust, suspicion, and the presumption of client immorality. Welfare applicants and recipients had few if any rights.

The "rediscovery" of poverty in America (i.e., Michael Harrington's 1962 book) and the ascendancy to power of liberal advocacy groups, including professional social workers, coupled with the continued rise in the welfare caseloads and costs prompted a reexamination of welfare strategies at the federal level. A moral shift took place that viewed welfare recipients less as outcasts and more as persons suffering from social and personal pathologies who needed rehabilitation. The desired outcome was independence and self-sufficiency, and it was assumed that most recipients were amenable to change through the provision of casework. These beliefs were expressed in the 1956 amendment to the Social Security Act that provided for a federal match of 50 percent of the administrative costs for social services and in the 1962 amendment that increased federal matching to 75 percent of the administrative costs (Derthick 1970, pp. 129-38).

However, the triumph of the social casework ideology proved elusive to institutionalize in organizational practices. First, the goal of rehabilitation had to compete with the goal of eligibility determination and redetermination. Second, both at the federal and state levels, insufficient resources were allocated for rehabilitation. In states such as Massachusetts, caseworkers were overwhelmed by large caseloads, and most had no professional social work training.² Moreover, there were no resources to forge effective interorganizational relations with other service organizations in order to provide needed services such as job training, day care, and rehabilitative services. Fourth, the service technology-social casework-was highly indeterminate. It was expressed in the professional rhetoric of "producing change in the lives of welfare recipients through the techniques of counseling, advice and guidance" (Handler and Hollingsworth 1971, p. 55). Or, as M. Derthick put it, "Casework, in short, is what the caseworker does" (1970, p. 136). The new mandate did not increase the actual services that welfare recipients obtained but greatly expanded the paperwork for the workers, who had to document how each contact with the client constituted a "unit of service" so that the department would qualify for the federal match.

Still, one should not lose sight of the fact that the changes in the moral assumptions had important effects on organizational practices. They transformed the welfare office to what may be best described as a "benign bureaucracy." J. F. Handler and E.J. Hollingsworth's (1971, p. 127) study of six welfare offices in Wisconsin showed that "for the vast majority of AIFDC families, social service means a caseworker's visit a little more than once every three months for a little less than forty minutes per visit, with an occasional client's call to her caseworker." The visit itself could be best described as a "friendly chat." Clients reported that workers mostly discussed issues about children, health, and general plans for future education and employment. The caseworkers avoided topics that could generate complaints or requests or make it difficult to deliver services. The clients themselves reported a high degree of trust in their caseworkers and were quite satisfied with the advice and counsel they received. Handler and Hollingsworth (1971) argued that because the workers did so little, especially in providing concrete services, they did not bother their clients, and the clients approved of the arrangement. That is, the clients were not morally degraded, but, equally, no expectations for change were demanded of them. The desired outcome of independence and self-sufficiency was a myth.

Again, it is important to recognize that considerable local variations existed in organizational practices. Bell (1965, p. 161) reported that departments that had more restrictive eligibility requirements were less likely to provide social services (i.e., casework). Moreover, each state and county provided its own operative definition of what "social services" meant.

Over time, the recognition that eligibility determination and social casework expressed incompatible moral assumptions and service technologies resulted in the two functions being decoupled. Indeed, the social work profession strongly advocated for the separation because it did not want to be associated with the morally problematic function of aid determination. In her famous editorial in Social Work, Gordon Hamilton strongly advocated for the separation, writing that "the money function disables or overwhelms the social services" (1962, p. 128). According to M.J. Bane and D. T. Ellwood (1994, p. 15), "Social Workers argued that the dual role of counselor and investigator was impossible to achieve. Such perceptions of coercion, accurate or not, poisoned the 'therapeutic' value of the counselor/client relationship." The federal government urged the separation on welfare departments in 1967 and mandated it in 1972 (Simon 1983, p. 1215).

The continued explosion in the welfare rolls by 36 percent between 1962 and 1967 served as proof that the rehabilitation model was not working (Bane and Ellwood 1994, p. 11). Many factors may have contributed to the continued growth (Handler and Hasenfeld 1991, pp. 117-19). Undoubtedly, a major cause was demographic-the rapid increase in single-parent families. However, it was also a period of expansion and extension of citizenship rights to previously disfranchised social groups, particularly ethnic minorities and women. It was a period of active political liberalism that resulted in the passage of the Civil Rights Act in 1964 and the Voting Rights Act in 1965. Several antipoverty measures were passed, some in response to the breakout of civil disorders in many urban centers (Piven and Cloward 1971). There were major changes in the legal culture. The rights of welfare recipients were greatly expanded by declaring many exclusionary practices illegal, including residency requirements, man-in-the-house rules, and employable-mother rules. Due process hearings were required to check arbitrary case closures, and highly discretionary grants for "special needs" were consolidated into uniform grant allocations. Liberal federal administrative regulations, combined with aggressive legal representation of the poor and the grassroots activities of the National Welfare Rights Organization, instituted a climate of entitlement.

Thus, the period from 1967 to 1972 witnessed a moral tilt toward defining welfare as an entitlement (Sosin 1986). Welfare departments came to accept the new moral assumptions that welfare recipients had a qualified entitlement to aid, that eligibility and the amount of aid could no longer be made contingent on the moral fitness or rehabilitative needs of the applicant, and that the poor were not necessarily in need of being "reformed." In turn, caseworkers eased eligibility determination and were less likely to use their discretion to withhold benefits (Sosin 1986, p. 271).

However, the continued rise in welfare costs and the number of recipients coupled with slow economic growth reinforced the sense that the program was out of control and that it was undermining dominant moral values. With implicit racial overtones, attention focused on "the large number of African-Americans, out-of-wedlock births, the moral consequences of marital disruption, single parenthood, and generational dependency" (Handler and Hasenfeld 1991, p. 120). Indeed, the passage of the 1967 Social Security Amendment reaffirmed these moral assumptions by making welfare receipt contingent on participation in a work incentive program (WIN). One of the administrative responses to the rising welfare costs was the reassertion of quality control to weed out "fraud, abuse and errors" (Brodkin and Lipsky 1983). In other words, welfare recipients were again recast as morally suspect and prone to abuse their entitlements. Beginning in 1972, sanctions were imposed on states that exceeded a certain error rate, and both supervisors and workers in local welfare offices were subject to penalties because of errors in determining eligibility and grant amounts. Welfare departments developed a set of highly complex and detailed administrative rules and regulations. The test of deservingness now meant being able to meet seemingly endless administrative requirements in a Kafkaesque bureaucracy.

W. H. Simon (1983, p. 1199) suggested that welfare departments acquired a new organizational form characterized by formalization of entitlement, bureaucratization of administration, and proletarianization of the workforce. Not unlike the Internal Revenue Service, the goals of the welfare department were now to verify eligibility, write checks, and reduce errors (Bane and Ellwood 1994, p. 16). The interorganizational relations of the departments became highly circumscribed to those public agencies that could provide documentary verification of eligibility claims, such as family status, income and assets, unemployment and work history, health, birth of children, school enrollment, and living arrangements. In an attempt to purge discretion, rules and regulations were promulgated to govern every possible contingency, resulting in volumes of instructions being updated and changed, often before workers had a chance even to absorb the instructions that they had replaced. If prior norms allowed workers to exempt a car from an applicant's assets when it was judged to be "needed," now the car could be exempted only if its value was less than $1,500 (Simon 1983, p. 1202). The technology of eligibility determination became highly mechanistic and impersonal, focusing on the verification of every statement required for eligibility. Simon (1983, p. 1205) reported that "the Massachusetts welfare department gives applicants a list of thirty documents that they may be asked to supply ... [and] many of the documents must meet stringent technical requirements."

The organization of work was divided into several discrete work units. Applicants began at "intake," which consisted of staff specializing in eligibility determinations. Once eligibility was established, clients were typically assigned an "eligibility technician" responsible for periodic eligibility redetermination. These workers had large caseloads of about 200 cases. If the client was required to participate in a work program, she would be referred to a separate "human resources" unit (Simon 1983, p. 1216; Bane and Ellwood 1994, pp. 4-5).

To ensure reduction in errors, quality control units were established. They used statistical methods to select a sample of cases to review for possible overpayment errors. The review was very detailed, as states were subject to penalties if the error rates exceeded a certain acceptable level.³ There were two characteristics to these reviews, First, they were concerned with overpayments rather than underpayments (Brodkin and Lipsky 1983), reflecting the moral assumption that many welfare recipients were undeserving and prone to cheating. Second, most of the errors could be classified as "paper errors" unrelated to substantive eligibility determination (Simon 1983, p. 1211).

The new organization of work also led to further de-skilling of the workers to lower-level clerical positions. Indeed, in many welfare departments, the eligibility workers were not appreciably different from the applicants they processed and came to resent the recipients for getting an undeserved entitlement denied to them. The effects of the new organizational practices on worker-client relations were quite apparent. The division of labor and high staff turnover prevented the formation of any continuous relationships. The relations became far less trusting and more adversarial. Problem clients became a problem because they required more paperwork and more time to process the case, and they caused a greater probability of error. Discretion was exercised through the workers' control over information and when and how to invoke various regulations. The onus of responsibility fell on clients to prove and maintain their eligibility. Most case closures were due to administrative reasons-the failure of the clients to submit proper forms in a timely fashion. In short, paperwork replaced people work, resulting in what M. Lipsky (1984) termed "bureaucratic disentitlement." Benefits to recipients were curtailed and terminated under the guise of obscure and hidden bureaucratic rules.

The ascendancy of the neoconservatives to power signaled another shift in the moral assumptions about welfare recipients, culminating in a liberal and conservative "consensus" expressed in the Family Support Act of 1988 (Handler and Hasenfeld 1991, pp. 209-30) and more recently in the PRWORA. First, welfare benefits are no longer an entitlement. The social rights of welfare recipients are made contingent on meeting obligations to the state. As expressed by L. M. Mead (1986), accepting public aid signals a failure in citizenship that justifies the right of the state, in return for public assistance, to exercise paternal authority-requiring the recipients to work and demanding that they lead a moral life and raise their children to become law-abiding citizens. Second, welfare recipients are obligated to work for their relief. If in the past the model welfare recipient was the "gilt-edged widow," now it is the mother who works full-time while raising her children with proper care. Since the majority of mothers are now in the labor force, why not welfare mothers? Again, welfare is viewed as corroding the work ethic. Third, laying the specter of the 'underclass" squarely at their doorsteps has reaffirmed the moral condemnation of never-married mothers, especially African Americans and Latinos. Marriage, paternity, and child support have to be strengthened. Fourth, educational failure leads to welfare dependency. Therefore, poor teenaged mothers should be obligated to graduate from high school. Finally, local discretion is seen as the most effective way to respond to the problems of welfare dependency.

Once again welfare departments are being transformed in order to institutionalize these moral conceptions. Welfare departments have developed new organizational forms whose overall goal is to place recipients in the labor market and to ensure that teenaged parents remain in school. In most states, these new forms are add-ons to the current welfare departments and are decoupled from the income maintenance functions, although some have experimented in integrating both (Brock and Harknett 1997). The ambiguities inherent in these moral conceptions, combined with local discretion, have resulted in the proliferation of diverse organizational forms and practices, each echoing the moral imprint of the particular state and local community (Hagen and Lurie 1994). Communities may opt to emphasize one of two competing moral conceptions. The first views welfare recipients as morally deficient, especially in lacking a work ethic. The second views recipients as suffering from human capital and environmental deficits, such as lack of education and training. Departments that emphasize moral deficiency are more likely to blame the recipients for their predicament and to accord them low social worth. Therefore, they are more likely to institute mandatory participation, immediate job placement, and the threat of sanctions to elicit compliance. Departments that emphasize human capital deficits are more likely to view their clients as victims of circumstances and attribute to them higher social worth. They encourage voluntary participation, remedial education, and training, and they use persuasion to elicit compliance (Hasenfeld and Weaver 1996).

Still, the emerging dominant organizational form is modeled after Riverside, California, because it has been shown to be effective in moving a larger proportion of recipients into the labor market at low program costs (Riccio, Friedlander, and Freedman 1994). Led by a charismatic director, Riverside engaged in its own moral entrepreneurship and developed a program that was widely different from most other county programs in California. It did so by assuming that welfare recipients lacked in work ethic rather than in human capital and by adopting the philosophy that a low-paying, entry-level job was better than no job and could lead to a better job. So, it emphasized a strong employment message, inexpensive job search, and quick entry into the labor market. The message was addressed to the staff and the recipients. The program developed extensive relations with local employers by hiring job developers. The county was able to promise local employers job applicants "that afternoon." The staff were specially recruited for commitment to the mission of the agency and were tightly organized and monitored. Staff performance was rated, in large part, on job development and placement. The staff engaged in close monitoring of attendance and recipient job performance. The threat of sanctions was used frequently. The success of Riverside in reaching its goal of pushing recipients into the labor market at low service costs has led California and other states to institutionalize this service model in their recent welfare legislation (see, e.g., Holcomb et al. 1998; U.S. General Accounting Office 1998).

At the other end of the moral spectrum is Utah (Pavetti 1995; Pavetti and Duke 1995). Welfare officials there assume that recipients experience many employment barriers that can be removed through the provision of individually tailored services. Therefore, participation is broadly defined. For example, participation includes education, training, part-time employment, mental health counseling, parenting education, and substance abuse treatment. Eligibility workers have been replaced by "self-sufficiency" workers who develop a self-sufficiency plan for families applying for welfare for the first time. The staff is encouraged to develop strategies to engage recipients with multiple employment barriers. The assumption is that with appropriate help, multiple problem families can eventually become self-sufficient. The Utah program has a system of incentives and sanctions ranging from a bonus for participation to a loss of the grant for continued refusal to participate. If a family loses welfare for nonparticipation, it can requalify only after participating in a structured program "designed to overcome recipient's fears of change" (Pavetti 1995., p.3).

The service technology for working with long-term, multiple-barrier families reflects the moral assumptions and goals of the program. Specially trained workers have been hired and assigned small caseloads (30-35 cases). They provide home visits and one-on-one counseling and conduct regular reviews. Still, the staff felt that they were in "uncharted territory, often having to rely on trial and error to identify the best strategy for helping a family overcome their barriers to employment" (Pavetti 1995, p. 12). One can see why this organizational form may not last. The program is expensive, treatment is time-consuming, and results, measured by self-sufficiency, are not easily attained.

In her study of the work program in Chicago, E. Z. Brodkin (1997) showed how fiscal considerations led officials to modify their moral assumptions and the resulting treatment of welfare recipients. In trying to reduce costs while maximizing federal reimbursement, Illinois shifted the work program from a voluntary program emphasizing education and training to a mandatory program emphasizing job search. In doing so, the state simply redefined its moral assumptions about welfare recipients and their service needs. Other constraints added to the reduced social service approach to welfare recipients. Because the department deprofessionalized its casework staff, it was unable to staff the work program with trained workers. Moreover, because of union rules, most workers were recruited from the income eligibility and grant determination. units that emphasize highly bureaucratic routines. The pressure to met caseload quotas further eroded giving attention to the specific service needs of the recipients. As a result, the caseworkers undertook their own moral entrepreneurship to cope with a difficult situation: "Caseworker tended to define client needs to fit available slots, avoid eliciting service claims, and pressure clients to accept the bureaucratic construction of welfare rights and obligations" (Brodkin 1997, p. 15).

As noted above, PRWORA encouraged states to shift to a "work first ideology, resulting in the adoption of new organizational forms that emulate the Riverside model by trying to become employment placement agencies. Connecticut's Jobs First program is a good example(Bloom, Andes, and Nicholson 1998). The state's welfare-to-work program shifted from an emphasis on education and training to rapid job placement. The program limits cash assistance to 21 months, and recipients are required to participate in employment services such as job search and job club whose aims are quick job placement. Education and training are available only to recipients who fail to get a job after substantial and lengthy tries. In comparison to previous organizational practices the eligibility workers are more likely to emphasize the new employment regulations and time limit. The employment service workers are more likely to urge the participants to work and to take jobs that do not pay enough. The workers rely on sanctions to achieve compliance. Structurally, the employment service has become more bureaucratic (i.e., then is more paperwork), with a correspondent increase in caseloads (often exceeding 500 clients per worker).

In addition, PRWORA permits states to privatize the delivery of their welfare services, and several states and counties are taking up the option (see, e.g., Bernstein 1996; Hughes 1996; U.S. General Accounting Office 1997). Such efforts are rationalized in the name of efficiency. Still, the firms being hired to administer welfare do moral work. Whatever practices these firms institutionalize to optimize their profits inevitably produce moral consequences for their clients. First and foremost, recipients become commodified-their value to the firm is contingent on the revenues they generate. Hence, there is a potential that recipients with problems" who require more attention and resources are likely to be defined as "unprofitable." Second, if profits depend on reductions in the welfare rolls, the firms will have an incentive to terminate cases as expeditiously as possible. Therefore, greater emphasis will be placed on rapid job placements, and most likely sanctions will be the preferred mode to enforce compliance and to close cases. More ominously, as commodities welfare recipients are in danger of being stripped of whatever minimal social rights they have left. The firms will serve as buffers between the recipients and the public officials responsible for their welfare, and the officials will have greater incentives to side with the firms rather than with the recipients.

Conclusion

Within this perspective, I give particular attention to the multiplicity of moral rules these organizations encounter both externally and internally. As the analysis of welfare departments shows, organizational and personal moral entrepreneurship play a significant role in shaping organizational forms and practices. My analysis of the historical changes in the organizational forms and practices of welfare departments shows that while these changes reflect the transformations in moral assumptions about welfare recipients, they are also highly contextualized. There are wide variations in local practices in response to both locally defined moral ideologies and political economies. Therefore, contrary to the "iron cage" argument, human service organizations tend to display greater variations in organizational forms.

Future research on human service organizations needs to identify more explicitly the dynamic forces that produce variations in organizational practices and in the moral rules they enact. Recent attempts to integrate institutional, political economy, and structuration perspectives (e.g., Oliver 1992; Orlikowski 1992; Barley and Tolbert 1997; Sandfort 1999) may be a promising direction.

In one sense, the historical analysis also lends support to the emphasis of the new institutionalism on broad cultural influences that are powerful rationalizing agents of organizational practices. Despite the changes that welfare departments have undergone, certain central features have remained the same, echoing broad dominant and fairly stable moral conceptions about the able-bodied poor. These include viewing the poor as deviants and "others"; ensuring that the giving of aid will not corrode the work ethic; and upholding the moral code about family values, gender, and ethnicity by degrading the poor. No matter what specific organizational forms welfare departments institutionalize over time, these basic moral values have remained their guides.

References

Notes

I want to thank Mayer Zald for his insightful guidance and the three anonymous reviewers for their thoughtful comments. Throughout this article the term "welfare department" denotes the state or county administrative unit that administers AFDC or Temporary Aid for Needy Families.

1. Much of the following discussion is based on Bell (1965).

2. Welfare departments were also reluctant to hire professionally trained workers for fear that they would be left with an expensive workforce once federal funds dried up.

3. No state, however, was ever actually penalized despite findings of unacceptable error rates.

©2000 by The University of Chicago. All rights reserved