Introduction

Welfare reform provided the opportunity to do more than expand or contract services; it created the chance for states to rethink the whole approach—philosophy, scope, and delivery—of provision to the poor. In most urban areas, a variety of non-profit, private and public organizations work to serve the poor. Frequently, these organizations have different but overlapping jurisdictions and special but similar missions. Proposals for and attempts at coordination are frequent. However, initiatives rarely achieve the benefits meaningful collaboration could provide. While non-governmental organizations (NGOs) and public agencies alike speak about avoiding duplicate services, often their resulting specialization necessitates a fragmentary approach to services for poor families. Recognizing this, some states and localities have promoted one-stop centers where a variety of service providers are housed under a single roof. This mechanistic solution has met with only partial success.

Meaningful collaboration that combines resources to address family needs holistically is often difficult since NGOs are frequently in competition with one another for contracts or grants and public welfare agencies are bound by strict rules and regulations. Historical tensions between private and public service providers too often serve to make collaboration undesirable for one or more parties. In their handbook for the "new public management," Governing By Network, Goldsmith and Eggers note that such partnerships are often plagued by "goal incongruence," "fragmented coordination," and "communication meltdown," since traditional bureaucracies lack the internal capacity for such complex relationships.¹

In the best environment and with the right leadership, however, the increased freedom from rules and regulations allowed under the Personal Responsibility Work Opportunity Reconciliation Act of 1996 (PRWORA) created a unique opportunity to change the dynamic that historically made such collaborations difficult. The state of Oregon took that opportunity seriously and pursued a strategy of welfare reform that enhanced rather than frustrated community partnerships and that attempted real coordination between normally estranged, or at least separate, public agencies and local groups. Between 1995 and 2001, Oregon provided a model for comprehensive welfare reform based upon: (1) a coherent strategy of community involvement; (2) flexibility from regulations in implementing welfare reform; and (3) a commitment to excellent leadership.

The Oregon model was unique because it did not employ the same kind of punitive policy tools that other states used to reduce caseloads under the Temporary Assistance for Needy Families (TANF) program. Without the use of time limits, family caps, new eligibility restrictions, or benefit reductions, Oregon reduced its caseload nearly 61 percent compared to the average national decline of 56 percent during the same time period.² Monthly benefit levels for a family of three remained at a modest $503, but this stayed well above the national average of $414 and Oregon achieved impressively high work participation rates during this time.³ For example, in fiscal year 2001, Oregon ranked sixth in the nation for percentage of welfare clients meeting TANF work participation standards.⁴ Early assessments of the work program found that eighteen months after welfare, 93 percent of those who left the program were still employed.⁵ Later assessments, under different methodology, found a more modest number of 58 percent employed after two years but these are relatively strong measures for work retention.⁶ In comparison, Wisconsin—a state often cited as a national model—admittedly achieved higher rates of TANF caseload reduction and work participation rates but created a work system with "uncommon severity," an extensive diversion program to keep the poor from welfare in the first place, and two-year time limits.⁷

The research presented here was conducted from 1998 to 2001 and represents a best-case illustration of what was possible with ample resources under welfare reform.⁸ The system I describe below reflects my observations of four welfare offices across the state where I was able to view caseworker–client interaction, attend classes designed for recipients, and extensively interview caseworkers, managers, and staff from partnering agencies. I had access to a number of intradepartmental publications as well as central office staff. Numerous people inside and outside the agency felt free to offer critiques of the system and to point out areas for improvement. Partnering staff, in particular, were quite candid about the shortcomings of caseworkers and the welfare agency. However, their intimate knowledge of personnel and procedures reinforced the picture of a highly collaborative system where staff worked closely together toward common goals.

I argue that such a community approach to welfare provision produces favorable outcomes for engagement with work without the more punitive elements favored by some states and politicians. Welfare reform "worked" in Oregon in the narrow (but important) sense of reducing caseloads and moving people into jobs but also in the broader sense of focusing public and private resources in a coordinated fashion to meet the needs of welfare recipients. Of course, moving poor and often poorly educated women with dependent children into jobs does not necessarily improve their incomes enough to escape poverty or make their lives demonstrably easier, especially in the short run.⁹ There are serious limitations to the work-first strategy specifically and to means-tested programs in general no matter how they are managed. The community approach outlined here is not without its concerns since it requires fairly high levels of bureaucratic discretion for street-level workers and assumes the existence of community partners engaged in poverty work. Nevertheless, absent political demand for broader and universal income and work supports, enhancing community collaboration and local social capital offers the best promise for making welfare laws work.

A Community-Approach to Welfare Provision

Finding inadequate the empirical and normative implications of viewing public policy with a market model, Deborah Stone writes: "Because politics and policy can happen only in communities, community must be the starting point of our polis. Public policy is about communities trying to achieve something as communities."¹⁰ As a corrective to rational choice analysis that understands political decisionmaking merely as a series of attempts to maximize benefits and reduce costs, Stone's argument about the centrality of communities is insightful. Empirically, people are motivated by a variety of forces, including loyalty, the public interest, caring relationships, and power.¹¹ Normatively, as Stone so convincingly demonstrates, the exchange relationship and market mechanisms are limited conceptions about political community, ones that do not adequately reflect either the dominant understanding of social life or its aspirations.

As a normative matter, the attempt to narrowly "rationalize" policies through expert cost–benefit analysis is misguided to the extent that it is antagonistic to democratic politics. The danger on the other extreme is to assume that the mere aggregation of interests is the sum total of self-government. Meaningful self-government demands that policy makers operate not on communities but with them, that they break down the walls between public and private to think inclusively as a political community. As a practical matter, few policies are approached with that degree of broad coherent vision. Incrementalism and parochialism are more generally characteristic of policy change, as new programs are often layered on top of existing programs.¹²

Supporters of welfare reform argued that block grants constituted a break with the past that would genuinely increase state flexibility to experiment and to adjust programs to local conditions. The assumption of federalist structures is that subnational communities are meaningful political units; enriching the control of local units will lead to better outcomes for that area and ultimately for the nation as a whole. As Philip Selznick argues, "if the federal principle is honored, the integrity of the parts must be nurtured and protected."¹³ Conversely, critics of welfare reform argue that U.S. welfare policy suffers from its state and local administration; universalism is real reform and the desired goal. PRWORA, testing the very idea of federalism, asked policymakers to determine exactly what are the benefits (if any) of devolution to state and local governmental entities. In welfare reform, we find the promise and the pitfalls of our federalist structure and ideal.

I argue that the flexibility provided by federal welfare reform legislation offered an opportunity for states to develop an approach to welfare provision centered in communities and utilizing community partners. While it is not new for state agencies to contract for services provided outside the agency, simple contracting should not be equated with a community approach to public policy issues.¹⁴ A truly collaborative approach has three important elements:

1. commitment to consensus building with multiple stakeholders;
2. resource sharing; and
3. staff integration.

Meaningful partnerships with local community groups greatly enhance the functioning of the welfare agency and its delivery of services to clients. But just as importantly, the partnerships provide a foundation for innovation and experimentation that can help fulfill the promise of federalism.

The use of the word "community" over the more popular "network" is purposeful. The literature on policy networks is informative and an important corrective to bureaucratic studies that assume a hierarchical approach to policy formation and implementation. However, much of the aim of the policy network analysis is against traditional government bureaucracy both as an empirical understanding of how government operates and a normative claim about how government should operate. It often extols the benefits of introducing market mechanisms or private partnerships for better service delivery to consumers. Public administration scholar Lester Salamon notes that the current popularity of "reinvented, downsized, privatized, devolved" government has its source in "a newfound faith in liberal economic theories."¹⁵ Networks contract with private entities for services on behalf of consumers; the community approach partners with local organizations on behalf of fellow citizens.

In his in-depth study of the politics and implementation of welfare reform in Wisconsin, Lawrence Mead documents the "radical privatization of welfare" that resulted from the state's effort to "reinvent" government.¹⁶ Mead admiringly describes the process of administrative competition unleashed by welfare reform where "competitive and market pressures largely replaced accountability through the political process, as in traditional bureaucracy."¹⁷ The competitive system in Wisconsin mirrored the program that the welfare agency implemented. Mead writes: "W-2 [the Wisconsin welfare plan] is individualist in spirit... . The plan mentions supporting the family and community, but these goals are secondary."¹⁸ While welfare reform implementation in Oregon resembled networks and reinvention, the style was much more cooperative and collaborative and focused on consensus building, not competition.

A commitment to consensus building within the policy community creates a set of local stakeholders who are invested in making welfare reform succeed. Consensus among historically antagonistic organizations or even among those with different missions and cultures is not easily achieved. It requires meetings, face-to-face dialogue, and compromise. There are often deep divisions between local service providers who believe their role is one of advocacy for welfare clients and career bureaucrats trying to implement new policy rules and regulations. Culture clashes are also common between public agencies housed in different departments, under different leadership, or working with a different clientele. Without first the commitment to reach some kind of consensus among disparate organizations, the other two elements of partnership—resource sharing and staff integration—will produce only limited benefits.

Resource sharing helps organizational relationships in obvious ways. Community service providers who receive funding from state agencies are less likely to be publicly critical of those agencies, more willing to engage in dialogue with those agencies, and more likely to feel they have a vested interest in helping those agencies succeed in their goals. Funding relationships, however, can also be a source of contention since recipients of contracts will always want more money with less oversight. If the entire relationship is only about the terms of contract obligations, then the benefits of such partnerships are minimal. When resource sharing occurs in an environment of consensus building, the focus can shift to other issues: How well is the partnership meeting the needs of recipients? What will it take to do so?

Creating a consensus-driven environment among multiple participants who share resources is an important start for developing a collaborative approach to welfare provision. When state agencies expand the understanding of sharing resources to include personnel, the potential exists for a much-strengthened partnership. Staff integration provides expertise and opportunities for information sharing that can work to enhance both organizations, but the benefits go further, particularly among organizations that have a history of distrust, as do local service providers and welfare agencies. When staff from one organization are placed under the same roof and integrated into the functioning of another, they have sustained and direct interaction with individual workers and processes. Therefore, there are fewer occasions for suspicion and misinformation. When a question arises about a case or situation, staff can meet face-to-face to resolve the issue quickly. Alternatively, the management of each organization feels confident that "their people" are protecting their agency's interests out "on the front line."

Elinor Ostrom's work is instructive here. Although Ostrom writes about solving tragedy of the commons situations in communities, cooperation between multiple service providers, both public and private, resembles the dilemmas faced by fishermen who by over-fishing jeopardize their futures.¹⁹ Both public agencies and charitable groups have the same goal: moving people out of poverty and improving their lives. They also realize that given the nature of the problems of poverty and the limitations of their own organizations, working together is in the best interest for furtherance of that goal. But as Ostrom explains, "participants may simply have no capacity to communicate with one another, no way to develop trust, and no sense that they must share a common future."²⁰ Changing the institutional arrangements can have important effects, not just for the flow of information, but also for the social capital that is created. As Robert Putnam demonstrates, where social capital exists, where personnel from differing organizations, with varied backgrounds and philosophies, develop norms and networks of trust and reciprocity, "everyday business and social transactions are less costly."²¹ In sum, a community-based approach to welfare reform creates an environment conducive to enhancing social capital through consensus building, resource sharing, and staff integration.

The Community Approach in Practice

The Oregon Department of Adult and Family Services (AFS) was one of the few state agencies to take seriously the proposition that programs should not only be tailored for local communities; local communities should be partners in the goal of improving the well-being of poor families. The strategy of pursuing a collaborative approach to welfare services was evident before passage of PRWORA, began at the state level, and continued through the district and branch levels of AFS.

Commitment to Consensus-Building with Multiple Stakeholders

As early as 1989, under the direction of Democratic Governor Neil Goldschmidt, Oregon began a strategic planning initiative called "Oregon Benchmarks." Citizen meetings were held across the state to determine what people and communities thought were important societal goals.²² The state revised and expanded the original economic benchmarks every two years to include goals in the areas of the economy, human development, quality of life, and the environment. The benchmarks are specific (decrease the high school drop-out rate; increase the percentage of Oregonians who volunteer at least 50 hours per year to civic, community, or nonprofit activities; decrease the percentage of families living below the poverty line; increase the percentage of seniors who are able to live independently) and are assigned numerical targets.²³ With goals instead of specific programs in mind, the benchmarks encouraged thinking and initiatives across multiple public agencies and communities.

The benchmarks initiative grew and expanded from its original mostly visionary policy. The Oregon Progress Board (OPB), a bipartisan citizen group, formally reports to the legislature every two years on the status of achievement (or non-achievement) of the benchmarks. But the OPB also took an active role in data collection, realizing that Oregon needed, first, a set of baseline data and, second, more and better information about the current well-being of Oregonians. OPB commissioned a study of adult literacy and with help from the Annie E. Casey Foundation collected data on young children, which was used to help formulate and prioritize indicators.²⁴ Data collection is an ongoing and evolving process, as is benchmark creation. The development of OPB and Oregon Benchmarks benefited from consecutive Democratic governors committed to the process in addition to Republican legislative support for the idea of results-driven government. Still, Executive Director of the OPB in 1997, Jeffrey Tryens, warned that benchmarks must be understood as a bipartisan initiative, one that takes time: "the value of the process is that it does not have a prescribed end," which does not always please those on fiscal and/or electoral cycles.²⁵ The political alliance that nurtured the benchmark process was always under the threat of unraveling.

In 1995, when the Welfare Reform Work Group proposed a number of goals designed to reduce child poverty and decrease the number of families receiving public assistance, the resulting report explicitly called for the creation of a "community-based delivery system."²⁶ AFS then convened district meetings around the state and told community leaders that if they worked with state agencies to move people off of public assistance and into jobs, then the savings from the reduced welfare rolls would be re-invested in the community. This re-investment strategy demonstrated to local community groups that the state was not seeking to reduce its investment in the poor, that the self-sufficiency goal for families in welfare reform was genuine and not part of a budget- cutting initiative. The focus on goals, not new program rules or regulations, served to unite otherwise disparate interests and disparate organizational imperatives. Coming as it did within the context of the benchmarking process, the AFS initiative enjoyed more legitimacy, coherence, and support than it might have on its own.

Next, in each Oregon district, the state welfare agency formally partnered with either a community college or a local Job Training Partnership Association (JTPA). The district and the Prime Contractor (PC) set the terms for their partnership together, made arrangements unique for each district, and incorporated other service providers and community organizations as necessary. One AFS office boasted fourteen partners in 2000, ranging from day care providers to teen services to mental health counseling. The list of partners provided by AFS contained both nonprofit service providers and other public agencies, namely, the Employment Department and Services to Children and Families (SCF). Often collaborative relationships with other state agencies require just as much careful development as do partnerships with outside service providers. In fact, many caseworkers and managers expressed more frustrations with other public agencies than with the NGOs. According to a number of AFS staff, the Employment Department had a more rigid bureaucratic culture that made it hard for them to collaborate. SCF, on the other hand, was too reticent to act in cases that AFS staffers felt were egregious or desperately in need of preventive attention. In order to facilitate these relationships, each district office assigned a staff person exclusively to exploring and maintaining community partnerships, but others throughout the organization were encouraged to participate in community events and forge their own relationships with local schools and civic groups. One example of this kind of commitment was the AFS staff in Lebanon, Oregon, that in May 1998 volunteered to make fundraising calls for the Center Against Rape and Domestic Violence, raising $1,500 for the organization.²⁷

Sandie Hoback, the head of AFS between 1996 and 2001, called the collaborative approach an "empowerment strategy."²⁸ Hoback's empowerment strategy focused on communities by giving local agencies the power to make strategic decisions, forge alliances, and work on good outcomes instead of merely administering programs. She met frequently with community groups and critics to talk about what the agency was doing and should be doing better. At first, a number of people thought AFS's new approach was disingenuous, and many organizations dealing with the problem of domestic abuse were very skeptical. But over time trust developed between the local shelters and branch offices; AFS was viewed as part of the solution to domestic violence. When asked how this was accomplished, Hoback explained that AFS just kept "bringing them to the table," and responded to a number of the criticisms. The results were extremely positive. In 1998, four separate battered women's organizations gave awards to local welfare offices for their work with domestic violence victims.

More conservative critics of welfare policy—notably from the business community—were equally supportive of welfare reform due to the piloting and eventual adoption of a small subsidized work program, JOBS Plus. Instead of cash assistance and Foodstamps, JOBS Plus recipients worked in private sector jobs subsidized with state and federal money. AFS engaged local business leaders to recruit other employers and to take an active role in the implementation of JOBS Plus. Although JOBS Plus remained small and voluntary, it effectively neutralized conservative opposition to the broader welfare reform experiment.

The creation of these partnerships—in domestic violence work, for private sector employment and beyond—was part of a coherent strategy that started at the state level, was pursued by both the Department of Human Resources (DHR) and AFS, and was incorporated into the mission of every district. A field services representative reported that not every district has had the same success at forging these relationships, a variation he suggested was attributable to leadership. Some partnerships were more conflictual than others. The best ones utilized a high level of resource sharing and staff integration.

Resource Sharing and Staff Integration

As part of Hoback's "empowerment strategy," local or district offices made their own decisions about the exact process or arrangement for handling clients, although they shared many of the same components. In general, a new client attended an orientation by AFS staff before meeting with a caseworker. The orientation explained the focus on work and the supports available for working families. Usually at this time any immediate needs were addressed, including day care. The welfare agency expected recipients to participate in job-search or job-readiness activities. The agency provided day care vouchers for these activities and workers discussed how to find, interview, and assess appropriate childcare. What counted for participation in the JOBS program was fairly extensive. Table 1 is a breakdown of activities for two different time periods. Both of these data points are snapshots of time; program participants were often enrolled in one or more of the programs over the course of the time or even concurrently.

Table 1 - Participation in JOBS activities.

Full table

Of these six program categories, AFS staff was responsible for just Program Entry, while partnering agencies or businesses conducted the rest of the activities. Although not all partners received AFS funding, the agency devoted a significant amount of its budget to outside organizations. AFS administrative costs, including case management, accounted for only 10 percent of the AFS budget in 1998.²⁹ Meanwhile, between fiscal years 1998 and 2001, AFS devoted 58 percent of its total JOBS budget to contracting agencies. In many cases, a client did not know when she was talking to an AFS person or someone hired by the community college, counseling center, or day care referral agency. Even when partnered staff introduced themselves as being from another agency, it was clear that they were part of the ordinary functioning of the agency and not a separate organizational actor.

AFS caseworkers met regularly with clients and assessed what activity or sets of activities were suitable for the individual case. Partnering staff frequently offered recommendations and helped caseworkers adjust plans and sometimes expectations. Generally, a client started with some form of structured job search, run by a Community College or JTPA. For clients who were clearly struggling or unable to meet the expectations of the program, the caseworker called team meetings with job placement specialists, mental health experts, AFS supervisors and others who were deemed appropriate. This kind of team management proved most useful when staff from other agencies worked closely together and when multiple service providers had contact with the client. Team meetings designed to generate ideas also provided oversight of the caseworker–client relationship.

AFS supported its team staffing philosophy with Employee Development Team (EDT) courses that were free for all DHR staff and partnering agencies. A survey of the training schedule for fifteen months between December 1997 and June 1999 revealed that nearly one-third of the EDT classes explicitly tried to develop team skills. In a letter to all DHR staff communicating the department's philosophy of service provision, DHR director Gary Weeks wrote, "Collaboration is not a mere buzzword. It's essential if we are going to use scarce resources wisely, if we are to view clients holistically, and if we are to truly connect with communities."³⁰ Clearly, team-based decisionmaking was a priority for the agency.

Caseworkers and partner staff out-sourced to the branch offices reported improved communication and therefore service to clients. Much interaction occurred on an informal basis outside of team meetings. Caseworkers concerned about alcohol abuse, for example, could alert a counselor about to speak to a group of job seekers. Job placement specialists became aware of literacy problems that were not admitted to caseworkers during the initial assessment. These pieces of information were transmitted more quickly and easily in environments where staff interacted on a daily basis. The ability to solve conflicts quickly and share information face-to-face helped to build trust between staff of separate organizations.

Out-sourced staff continued to be supervised by the agency that hired them. Therefore, they frequently acted as liaisons between AFS and the management of the community organization, improving communication not just among front-line workers but also managers. The partnering agencies were not uncritical of AFS procedures or particular AFS caseworkers. The staffs of the partnering agencies were, on the whole, better educated and trained in a specialized area. But it was because of those reasons that partnering agencies liked to be placed at AFS offices; they felt they could have an important impact on how clients were treated. Although AFS workers held the authority to make determinations about placement, in practice they often consulted with and respected the viewpoints of partner staff.

Furthermore, it was not just caseworkers who changed their attitudes and behaviors after working closely with partnered agency staff. One community college employee outsourced at a branch office stated that sixteen years ago he thought AFS was the "Evil Empire." Years later, he willingly worked at the agency and agreed with the emphasis at AFS on thinking every welfare recipient is capable of work; the challenge, he argued, was just to figure out the clients' needs. As one caseworker reported, both local community colleges and AFS saw themselves as advocates for the poor but not each other. Integrating staff changed that dynamic. Sharing resources and staff also meant that welfare clients received more services under one roof. Even when services could not be provided at the same location, having the face-to-face contact with someone from the service agency was important and more effective than merely handing out a referral with a phone number.

Although adding outside staff to AFS offices was the most common form of integration in Oregon, staff sharing did not occur in one direction. For example, AFS personnel were assigned to an abstinence-based prevention initiative for teens that was developed with the Department of Education. Other personnel were assigned to local community college programs, the Employment Department, SCF, battered women's shelters, and job placement centers. In each case, AFS sought to strengthen relationships and to work in concert over similar goals: reducing teen pregnancy; helping domestic abuse victims; moving people into jobs; supporting work efforts. In many cases, these relationships were in flux, changing and adjusting as the need arose.

Much of the benefit of such initiatives is difficult to measure. The collaborative emphasis appeared to encourage a more holistic approach to the needs of families. Caseworkers reported that team meetings were an important method for seeing "outside of the box," that they often found insight into an issue from the interaction with specialists or community leaders. On the other side, service providers enjoyed the access to caseworkers, resources, and clients. One mental health counselor out-sourced to a local AFS branch explained that he received some of his clients from direct referrals from caseworkers, but many clients came to him after he spoke at an orientation meeting or talked about domestic abuse in a life skills class.

Although integrated into the functioning of AFS branch offices, many of these out-sourced staff felt sufficiently separate from the agency to conduct activities and forge relationships on their own terms. On a day I visited a "job readiness" class in North Salem, a drug/alcohol treatment counselor presented a talk about drugs that was amazingly detailed and factual. The counselor did not preach or deliver an anti-drug message but instead outlined the symptoms of bipolar disorder, post-traumatic stress disorder, and the medications used to treat both. Further, he explained in great detail how serotonin re-uptake inhibitors work and how individual metabolic rates affect reactions to various drugs. Moving on to street drugs, the counselor described how some people self-medicate in response to post traumatic stress disorder. During the course of the presentation, a number of clients spoke openly about medications that they were currently taking or experiences of drug abuse by family members.

Institutionalized Flexibility

The commitment on the part of AFS to partner with local community organizations and other important actors in addressing the needs of families would have meant little if the state (or federal government) had not been open to such arrangements. The state of Oregon and its legislature granted AFS considerable room for experimentation. In adopting the strategy of Oregon Benchmarks—focusing on outcomes and performance measures—the state had essentially institutionalized this flexibility.

The re-investment strategy described in the previous section would not have been possible without the consent and cooperation of the federal government and the state of Oregon. The state use of benchmarks as a strategic planning tool set the stage for the Oregon Option, an intergovernmental plan for creating a new type of service delivery to meet the state's benchmarks. The idea was to jettison the traditional bureaucratic approach, with its stress on specific programs and its burdensome administrative rules, in favor of one focused on outcomes. Capitalizing on the Clinton Administration focus on re-inventing government, the Oregon Progress Board argued that "the intergovernmental system for delivering assistance and services through federal grants and mandates to state and local governments has broken down in a tangle of good intentions gone awry."³¹ The Oregon Option proposed to change funding streams and federal regulations in exchange for results-driven accountability. State and federal leaders met both in Portland and Washington DC to discuss the Oregon Option. In 1994, state officials signed a Memorandum of Understanding with Vice President Al Gore and top Clinton administration officials including Donna Shalala, Robert Reich, Janet Reno, and Alice Rivlin.³² The Memorandum set the goal of improving intergovernmental service delivery by identifying benchmarks and removing barriers to achieving good program results.³³

Under the Oregon Option, the Department of Human Resources sought federal waivers designed to both reduce the reporting requirements and allow the state to keep a portion of program savings from reductions in the welfare caseload.³⁴ The agreement stipulated that Oregon must reduce the caseload by more than the national average in order to keep the savings. The federal waivers were granted in March 1996, just months before PRWORA and the transition to block grants. Still, the early action on the part of Oregon allowed the state to recoup $7–10 million for reinvestment. Block grants under PRWORA essentially continued this relationship since the federal grant does not decrease as the caseload decreases.

By 1999, the benchmarks initiative, which emphasized strategic planning based on outcome goals and collaborative partnerships for meeting those goals, was used by over thirty state agencies.³⁵ The focus on community goals served as a way to bring disparate groups together. The Workforce Option, one of the five program clusters sponsored by the Department of Human Resources, billed itself as "an initiative involving state, local and federal partners committed to transforming our workforce programs into a more integrated system that is customer-focused, collaborative, and oriented toward outcomes."³⁶ The message was consistent and supported by all levels of state administration and the legislature: the administration was interested in goals and outcomes, not in specific program requirements or regulations.

Administrator Sandie Hoback took that message to her staff at AFS. District managers and branch directors reported having a great deal of freedom in tailoring their programs, organizing their staff, and trying new ideas. Successful ideas proliferated to other districts through email and regional meetings to be replicated where applicable. Over and over again, caseworkers explained that they were encouraged to be creative and, even more importantly, to be "allowed to fail" if something they tried did not work well.

The managers and caseworkers in the branch offices expressed pride in their innovations and communicated a sense of ownership over programs. A healthy level of competition appeared to exist across the districts both because senior management praised and encouraged innovative programs and because other branches often adopted successful ideas. One branch bragged about pioneering the JOBS Plus program, being involved at the ground level to create policy for the entire state. DHR profiled one district in Eastern Oregon for its partnership with an airplane manufacturer and local community college to provide composites training for jobs in the aerospace, boating, and motorcycle industries.³⁷ That partnership then served as a model for a computer guided machine tools training program in Portland. Sometimes the innovative program ideas required tremendous resources and coordination with multiple entities, but not always. The manager of a rural branch in Drain, Oregon, started an experimental support group using quilting. Billed as a "life-skills" class, the all-female group of recipients and former recipients "talk[ed], counsel[ed] and sew[ed]" twice a week with donated materials and returned the finished product back to the community.³⁸

Beginning in 1999, AFS focused increasingly on job retention and wage growth as caseloads declined dramatically. Giving local districts and offices the opportunity to experiment was particularly helpful as the tasks of the bureaucracy changed and the agency moved into unknown territory. The Dalles branch pioneered a computer class for working former recipients after they told AFS that they needed computer skills to advance in their jobs (and make enough money to not need additional supports like day care and Foodstamps). Together with the local community college, the branch office launched a course entitled "Introduction to Technology" that took recipients through the basics of assembling a computer out of its box, installing software, using Windows, word processing, and spreadsheet programs as well as emailing and researching on the Web. Upon completion of the 10-week, 60-hour class, participants were allowed to take home the computer that they had used in class. This course exemplified the AFS approach to welfare provision: demand high levels of participation and cooperation but create the conditions necessary for people to succeed and make the effort worthwhile to participants.

Branch directors complained that there were still some areas of welfare provision where, due to continuing federal regulations, they could not employ the same level of creative energy and flexibility. In June 1999, Oregon's error rate for Foodstamps exceeded acceptable federal levels, which meant that management had to start re-emphasizing accuracy and eligibility rules and requirements. The main drawback for "creative" bureaucratic management is that so much discretion can produce errors or bad judgment. And since AFS could set the terms for client "participation," caseworkers exercised much more control over clients' lives than under the old system. Discretion can be abused.

A full discussion of the issue of bureaucratic discretion cannot be accomplished here; however, it is important to note that the issue was not only abuse of discretion for which some safeguards can be instituted, but variance itself. Case management focused on identifying client strengths and needs, but both are highly subject to interpretation. Is a client not cooperating because of an unvoiced problem of domestic abuse or does she exhibit what some caseworkers called an "entitlement" attitude that needs to be reshaped with strict oversight? How a caseworker responds to the same kind of information varies depending upon one's individual style and particular set of experiences. Some of the caseworkers interviewed thought of themselves as mothers or coaches or consultants. One caseworker, a former counselor for people with mental disabilities, stated that she tends to see clients as supremely able, even in some of the most difficult circumstances, compared to her previous mental health patients. A former police officer turned caseworker was much quicker to resort to formal procedures and sanctions to ensure cooperation by recipients, whereas another caseworker in the same branch expressed a much more lenient and liberal orientation to client behaviors and attitudes. A mothering type might have spent considerable time talking about self-esteem issues or engaging in "hand-holding" discussions, while former welfare recipients hired by the agency tended to be quite strict about compliance and individual responsibility. Even within the bounds of the rules and appropriate (non-abusive) action by caseworkers, the variance across caseworkers' personal styles was quite pronounced.

With the level of discretion enjoyed by caseworkers, both informal and formal checks were critical. The Oregon welfare agency prided itself on its record of employment placement. When bragging about its status as the state with the second largest caseload drop since 1994, AFS literature adds: "And unlike most other states, Oregon achieved its caseload declines through positive means (placing people in employment) rather than the punitive methods...used in some places."³⁹ Where some states restricted eligibility or utilized time limits to reduce their caseloads, Oregon did neither. There were sanctions (benefit reduction or benefit loss) for non-cooperating clients but no level of sanction could be instituted without a team meeting of service providers and AFS staff. For those who did lose benefits through the sanction process, they could reapply for benefits immediately. The philosophical approach to welfare reform communicated through training materials and by managers, together with the procedural protections, contributed to the formation of a bureaucratic culture that was fundamentally not punitive, but was uneven.

Leadership

Reorganizing a bureaucracy and maintaining an organizational culture that is both more responsive and innovative do not happen by mandate alone. The Oregon welfare agency benefited from excellent leadership from the beginning of its welfare reform initiatives. Absent that leadership, the community approach would have failed. To achieve public purposes, leadership counts. What influences the quality of the education in an elementary school? Leadership by the principal. What determines the quality of life in a prison? Leadership by the warden. What affects the performance of a welfare, training, and employment program? Leadership by the department's top managers.⁴⁰

Sandie Hoback, AFS Administrator from 1996 through 2001, received high praise from all levels of the agency. Her leadership, first as a Welfare Reform Coordinator and then as head Administrator, clearly was key to the success of the agency. Hoback performed well in meeting a number of management challenges identified by researcher Robert Behn in a study of a welfare-training program.⁴¹ The agency achieved an impressive level of coordination under her guidance. Staff felt that she supported them in their work and generally reported positive feelings about the direction of the agency and changes since welfare reform. The technical problems of service delivery were delegated to district and local offices, but the Central Office monitored performance measures for each branch. Thanks to dropping caseloads and block grant arrangements, AFS did not experience a budget shortfall during this period. Moreover, in the spring of 1999, Hoback was already planning for reductions in the block grant after 2002. Part of this strategy was to involve the local business community in supporting welfare reform initiatives and job training programs.

Hoback was clear about the mission of the agency as "helping Oregonians become and remain self-supporting." This focus meant that even as clients moved off of cash assistance or TANF, the welfare agency continued to track their job performance. As the caseloads declined, the emphasis shifted from moving people into jobs to moving people up the economic ladder and from reactive service delivery to preventive services. The central office consistently and effectively communicated this mission, even as programs evolved.

The most important skills Hoback brought to the job of Administrator involved her adept relations with constituent groups and community organizations. First, she came to the job with a background in Legal Aid and as a community organizer herself: she started the first battered women's shelter in Salem, the state capital. This experience gave her a tremendous amount of credibility and connections in the advocacy community. Hoback expressed a high regard for grass roots organizing and saw her task as creating communities of interest among a variety of organizations around families and family protection. So she was not hesitant to invite critics in and she stated that she wanted to hear what they had to say because she believed they were often right. Partnering organizations and community leaders reported a high level of accessibility to Hoback even if they did not get everything they wanted from the agency. Although it is rare for a welfare administrator to come to the position with Hoback's background, her strategy—focusing on a meaningful vision, holding people accountable for outcomes, adapting structures as necessary, supporting her staff, and bringing together a diverse group of participants—could be duplicated by any talented leader.

One of the reasons the staff at AFS expressed so much support of her and the changes the agency implemented was because they were involved in the process. One caseworker and union steward described the movement toward team management as a positive one for morale. Previously, the union issues revolved around worker safety in doing home visits, pay and benefits, and issues dealing with individual managers. During this period, caseworkers and managers made joint decisions about work assignments and caseworkers participated at various levels of management decisions. Hiring decisions required agreement from a committee of managers and line staff. The effort to include line workers in management decisions helped to boost morale and achieve better performance. As one manager explained, "You get a better program when people who deliver it are involved in the process."

The agency also professed a commitment to training staff and developing leaders from the local offices. In 1998, employees spent an average of fifty hours in training activities, much of it designed to develop skills, not merely learn program changes.⁴² "Mission Values Principles" training introduced the concept of principle-based decisionmaking to caseworkers who had significant discretion under welfare reform. The Central Office also created Agency Leadership Academy seminars that introduced new managers to the culture and language of partnership, performance-based management, and the centrality of workforce development.⁴³ AFS staff could temporarily "try out" new roles before they committed to changing jobs. New opportunities for advancement or for challenge frequently presented themselves as the agency adapted to a changing client base, launched new initiatives, or formed new partnerships.

The consistency with which caseworkers explained the mission of the agency and the ideals behind principle-based management and outcome-based performance reviews suggested deep penetration of the message and tremendous coherence despite the decentralized nature of welfare administration in Oregon. Not only did caseworkers understand the philosophical and programmatic thrusts of welfare reform, they believed in them. The fact that AFS was able to develop this high level of caseworker adoption of the program and its goals, along with community support, speaks well of both the community approach to welfare and the leadership in Oregon.

Wither/Whether the Community Approach?

The Oregon welfare system offered a model for an effective approach to welfare reform that changed the dynamics of community interaction between welfare agencies and local service providers. The social capital built within these local policy communities around the common goals of helping poor families become economically self-sufficient, reducing domestic violence and teen pregnancy, and increasing the welfare of all children aided in these endeavors. The key here is that these relationships were not built overnight; meaningful collaboration takes time to develop. Breaking down barriers between organizations is a process that happens slowly and in incremental steps.

Progress may be slow but it is not impossible. Other states could replicate Oregon's success if they engaged in a coherent strategy of community involvement that strives for consensus, provides relief from undue regulations, encourages innovation, and makes a commitment to attracting and developing good leadership.

The capacity of a state to sustain such a system, however, depends on continued resources and flexibility. After 2001, the Oregon welfare agency (now DHS) suffered significant budget cuts and retreated substantially from case management and the tracking of former recipients. With fewer resources, the agency greatly reduced the level of contracted services. In response, partnering agencies were forced to cut staff and services and to dramatically reduce outsourcing. Even more significantly, under the Bush administration, Oregon lost its waiver to operate without time limits. Tougher work participation rates (and financial penalties for states who do not comply) that go into effect October 2006 will significantly curb the flexibility enjoyed under PRWORA.

Even under the best circumstances, the Oregon welfare system raises concern. District variation meant that clients could have very different experiences and resources available to them depending on where they lived. Urban areas like Portland and Eugene tended to have a greater range of service providers to meet a variety of needs. More rural locations were forced to contend with serious transportation issues, limited economic opportunity and fewer services overall. A community-based approach is unlikely to be able to overcome these regional deficiencies.

Furthermore, Hoback's "empowerment strategy" did not necessarily empower clients. Individual caseworkers explained that they did ask clients what they needed to become self-sufficient. Some team meetings invited clients to participate. AFS staff who helped non-cash health plan and day care clients advance to better paying jobs regularly surveyed their clients for needed services. But client input was sought on an ad hoc basis, not institutionalized. Decisions could be and frequently were made against a client's preferences. In such a system, the problem of bureaucratic discretion must be monitored continuously. In interviews with Oregon welfare leavers, Acker and Morgen found a variety of responses to questions about caseworkers, both positive and negative:

• "I got a lot of extra special attention as far as her. She advocated for me. That's really basically how I got so far into the agency."
• "I always dreaded going into that office...Because it's like, I think they treat everybody who goes in there like they are taking advantage of the system."
• "I've had case-workers... that were awesome, and made you feel like family...[But a recent caseworker would not return phone calls after benefits were cut.] And she said, 'Well I have been busy.' And I said 'Well look now at where I am.' And that's when everything went downhill for me."⁴⁴

Clearly, some clients received poor treatment. Lipsky argues that the nature of social service provision means that street-level workers develop modes of "mass processing," which sometimes include "favoritism, stereotyping, and routinizing."⁴⁵ Nationally, welfare clients generally complain about all three, leading many of them to conclude that the welfare system is arbitrarily administered and unfair.⁴⁶

Still, Hoback's strategy of empowering caseworkers may have important benefits for clients. In a study of prison management, Maynard-Moody, Musheno, and Palumbo found that "the positive aspects of street-level influence can be maximized and the negative aspects minimized when service organizations are designed to engage, rather than mute, street-level worker perspectives on how policy should be implemented."⁴⁷

If street-level bureaucrats regularly make policy and enjoy broad discretion over clients, then ownership of the programs they are implementing through participatory structures (like team meetings and new program creation) and internalization of the agency's mission may be better antidotes than stricter application of the rules. The Oregon case suggests that the culture, mission, and organizational arrangements of an agency can serve as an important check on a street-level worker's judgment. Since, as Maynard-Moody and Musheno explain, "street-level cultural judgments are an irreducible element in governing the modern state,"⁴⁸ administrative arrangements and public–private partnerships should seek to enhance the community within which such bureaucrats operate. The social capital built between service providers in the community approach can result in enhanced and integrated service delivery to clients.

George Frederickson notes that the new public management is "the political science of making the fragmented and disarticulated state work."⁴⁹ The long-term trends in contracting, privatizing, and devolving are unlikely to see major reversals in the near future. Oregon's community approach to welfare provision outlined here—with its commitment to consensus building with multiple stakeholders, resource sharing, and staff integration—works against the current mode of greater fragmentation. Nevertheless, the fragmented structure of welfare policy enhanced by federalism and PRWORA creates particular problems of national equity that a community approach may be unable to address. Nevertheless, within the bounds of federalism, the Oregon case is an alternative to the competitive model of social provision and an illustration of another path offered by devolution that gives real flexibility to states.

Notes

¹ Stephen Goldsmith and William D. Eggers, Governing by Network: The New Shape of the Public Sector (Washington, DC: Brookings Institution Press, 2004), 40–47. This "new public management" relies heavily on public–private partnerships to bring about integrated and enhanced services.

² Author calculations based upon Oregon Department of Human Services data for TANF-BASIC Cases and TANF-UN Cases, 1995–2001.

³ Nisha Patel and Mark Greenberg, "Key State TANF Policies Affecting Microenterprise: Oregon," Center for Law and Social Policy, October 2002. http://www.clasp.org/publications/
welfare_policy.htm (May 24, 2006). Work participation rates for all families were 98.2 and 96.7 percent in 1998 and 1999, respectively. In 2000, the work participation rate dropped to 64 percent but that was still well above average for states.

⁴ Lawrence M. Mead, Government Matters: Welfare Reform in Wisconsin (Princeton, NJ: Princeton University Press, 2004), 6.

⁵ Adult and Family Services, Ways and Means Presentation (Salem: Oregon Department of Human Resources, 1999): I-2. Author's files.

⁶ Joan Acker, Sandra Morgen, and Lisa Gonzales, "Welfare Restructuring, Work and Poverty: Policy Implications from Oregon," Center for the Study of Women in Society, University of Oregon, 2002. Available online: http://aspe.hhs.gov/hsp/
leavers99/state-rpts/or/Fams-
who-left-TANF-FS.pdf (May 24, 2006).

⁷ Mead, Government Matters, 3, 134, 148.

⁸ Conditions in the state of Oregon and the nation changed considerably after 2001.

⁹ The likelihood of earnings growth and upward mobility is influenced by the quality and match of the initial job as well as job supports, but most studies find modest wage growth for more welfare recipients. See Julie Strawn, Mark Greenberg, and Steve Savner, "Improving Employment Outcomes under TANF," in The New World of Welfare, ed. Rebecca Blank and Ron Haskins (Washington, DC: Brookings Institution Press, 2001).

¹⁰ Deborah Stone, Policy Paradox and Political Reason (New York: Harper Collins, 1988), 14.

¹¹ Deborah Stone, Policy Paradox: The Art of Political Decision Making (New York: W.W. Norton, 1997), 33, 73–74.

¹² Laurence D. Brown, New Policies, New Politics: Government's Response to Government's Growth (Washington, DC: Brookings Institute, 1983).

¹³ Philip Selznick, "Afterward: Federalism and Community," in Dilemmas of Scale in America's Federal Democracy, ed. Martha Derthick (Washington, DC: Woodrow Wilson Center Press, and Cambridge: Cambridge University Press, 1999), 364.

¹⁴ For a comprehensive discussion of government contracting and public-private partnerships, see Donald Kettl, Sharing Power: Public Governance and Private Markets (Washington, DC: Brookings Institution Press, 1993) and Lester M. Salamon, ed. The Tools of Government: A Guide to the New Governance (New York: Oxford University Press, 2002).

¹⁵ Lester Salamon, "The New Governance and the Tools of Public Action: An Introduction," in The Tools of Government, ed. Salamon, 1.

¹⁶ Mead, Government Matters, 133.

¹⁷ Mead, Government Matters, 132.

¹⁸ Mead, Government Matters, 110.

¹⁹ Elinor Ostrom, Governing the Commons: The Evolution of Institutions for Collective Action (Cambridge: Cambridge University Press, 1990).

²⁰ Ostrom, Governing the Commons, 21.

²¹ Robert Putnam, Bowling Alone: The Collapse and Revival of American Community (New York: Random House, 2000), 288.

²² Jeffrey Tryens, Testimony before the U.S. House Committee on Government Reform and Oversight (Bethesda, MD: Congressional Information Service, October 31, 1997), Text from: Federal Document Clearing House Congressional Testimony. Available from: LexisNexis™ Congressional (Online Service).

²³ Oregon Progress Board, "Achieving the Oregon Shines Vision: The 1999 Benchmark Performance Report," (Salem, OR: Report to the Legislative Assembly, March 1999).

²⁴ Anne Lewis and Margaret Dunkle, The New Oregon Trail: Accountability for Results (Salem, OR: Institute for Education Leadership Policy Exchange, 1996).

²⁵ Tryens, "Testimony" (October 31, 1997).

²⁶ Oregon Progress Board, "A Brief History of the Oregon Progress Board," (Salem, OR: June 1999).

²⁷ Oregon Department of Human Services, Quarterly Performance Update for the Period Ending December 1998 (Salem: Adult and Family Services, 1998).

²⁸ Sandie Hoback, Interview with author (Salem, June, 1999).

²⁹ Oregon Department of Human Services, Quarterly Performance Update for the Period Ending December 1998 (Salem: Adult and Family Services, 1998), 12.

³⁰ Gary K. Weeks, "Confidentiality: a building block of partnerships," Currents (Oregon Department of Human Services, July 1998).

³¹ Oregon Progress Board, "The Oregon Option: A Proposed Model for Results-Driven Intergovernmental Service Delivery," (July 25, 1994), Oregon State Library: OR ED/P94.20p7c.3.

³² Anne Lewis and Margaret Dunkle, "The New Oregon Trail: Accountability for Results," 1996.

³³ Barbara Dyer, The Oregon Option: Early Lessons from a Performance Partnership on Building Results-Driven Accountibility (Washington, DC: The Alliance for Redesigning Government of the National Academy of Public Administration, July 1996).

³⁴ Oregon Progress Board, June 1999; Tryens, "Testimony" (October 31, 1997).

³⁵ Oregon Progress Board, March 1999.

³⁶ Oregon Department of Human Services, "Oregon Option Accomplishments to Date," (Salem: DHS, September 3, 1996).

³⁷ Dan Postrel, "Partners pull together to offer job training," Currents (Oregon DHS, May 1998).

³⁸ Patricia Feeny, "Quilting is healing for AFS clients: An experimental support group builds life skills and lasting relationships," Currents (Oregon DHS, September, 1998), 6.

³⁹ Oregon Department of Human Services, "How Well is AFS doing?" (Salem: DHS, May, 2000).

⁴⁰ Robert D. Behn, Leadership Counts: Lessons for Public Managers from the Massachusetts Welfare, Training and Employment Program (Cambridge: Harvard University Press, 1991), 216.

⁴¹ Behn, Leadership Counts, 217.

⁴² Oregon DHS, 1998.

⁴³ Oregon DHS, 1998.

⁴⁴ Joan Acker and Sandra Morgan, with Terri Heath, Kate Barry, Lisa Gonzales, and Jill Weigt, "Oregon Families Who Left Temporary Assistance to Needy Families (TANF) or Food Stamps: A Study of Economic and Family Well-Being from 1998–2000" (Center for the Study of Women in Society, University of Oregon, January 2001). The views and experiences of 78 Oregon welfare clients are presented in "Family Profiles."

⁴⁵ Michael Lipsky, Street-Level Bureaucracy: Dilemmas of the Individual in Public Services (New York: Russell Sage Foundation, 1980), xii.

⁴⁶ See also Ann Tickamyer, "Voices of Welfare Reform: Bureaucratic Rationality Versus the Perception of Welfare Participants," Affilia: Journal of Women and Social Work 15 (Summer 2000): 173–93.

⁴⁷ Steven Maynard-Moody, Michael Musheno, and Dennis Palumbo, "Street-Wise Social Policy: Resolving the Dilemma of Street-Level Influence and Successful Implementation," The Western Political Quarterly 43 (December 1990): 833–48.

⁴⁸ Steven Maynard-Moody and Michael Musheno, Cops, Teachers, Counselors: Stories from the Front Lines of Public Service (Ann Arbor: University of Michigan Press, 2003). See also Norma Riccucci, How Management Matters: Street-Level Bureaucrats and Welfare Reform (Washington, DC: Georgetown University Press, 2005), for the resilience of personal moral beliefs in caseworkers, charged with implementing reform.

⁴⁹ H. George Frederickson, "The Repositioning of American Public Administration" PS: Political Science and Politics, 32 (December 1999): 702.