Ancillary Services to Support Welfare to Work. A Summary of Overarching Issues


The preceding ten chapters review the need for and provision of services to address barriers to employment. Several issues common to each are important to highlight and discuss more explicitly: (1) interagency coordination, (2) infrastructure, and (3) additional research needs. While much of this report has focused on what local welfare agencies should be and are doing to help clients overcome barriers to employment, these three crucial areas need the attention of a community much broader than local welfare agency staff. These issues suggest that stakeholders at the state and federal levels have additional roles to play in addressing the nation's goal of helping those on welfare become productive, self-sufficient members of society.

Interagency Coordination

The descriptions of current welfare agency programs in the preceding ten chapters make clear the extent to which service provision for welfare clients is increasingly dependent upon a coordinated system of service delivery. This dependency has resulted from agencies' need to coordinate with other local and specialized providers for services that address barriers to employment faced by clients previously exempt from participation requirements. Current efforts to coordinate services reveal several areas of concern, however, that should be addressed at the federal, state, or local level: (1) overlap or duplication of effort, (2) incompatible service systems, (3) minimal employer involvement, and (4) limited funding coordination.

There is likely to be duplication of effort and overlap of services provided by welfare, health, labor, housing, criminal justice, and educational organizations in the absence of a well-coordinated system. There are two reasons in particular to be concerned about this duplication or overlap: (1) with limited resources, social service providers cannot afford to waste funds; and (2) clients could be faced with a fragmented and confusing set of services characterized by gaps, inconsistencies, and incompatible procedures and expectations. Part of the challenge in developing a coordinated system within the social service community is addressing the fact that agencies often differ in their missions or objectives, levels of staff expertise, participant expectations, time frames for service provision, assessment methods, intake and monitoring procedures, and the language they use in discussing mutual concerns. Efforts that fail to acknowledge and address these crucial differences jeopardize successful service provision. Efforts that instead establish cross-agency committees, provide cross-agency staff training, or physically locate staff from different agencies together, for example, hold greater promise of fostering the kind of partnerships necessary to coordinate services effectively.

Because welfare agencies are working toward moving clients into the workforce, coordinated efforts need to extend to employers as well. However, there is little evidence of much coordination involving employers directly. Program models that we have categorized as "employment integration" are generally those for which the welfare agency has included a work-related component, for example, treatment in conjunction with skills training or job placement services. Other than in the case of employer-sponsored transportation services or employer accommodations for those with disabilities, programs tend not to include the direct involvement of employers. As a result, employers are too often missing from local area discussions and efforts aimed at designing a coordinated system of services that will move welfare clients into the workforce. Without greater employer involvement, welfare recipients are at risk of losing their jobs and returning to public assistance.

Finally, a coordinated system of service provision is constrained by the current fragmentation of funding. The system generates confusion about which agency is responsible for shouldering the costs of service provision and is hampered by legislative constraints on expenditures, including restrictions on the use of TANF funds for medical services, on the amount of TANF funds that can be transferred to the Social Services Block Grant, and on the use of Medicaid funding.

This fragmented and restrictive system also complicates decisions concerning the appropriate allocation of resources: whether to focus on the harder-to-serve or the more easily employed populations, on the welfare or the low-income populations, on those faced with one barrier to employment as opposed to another, or on those in need of expensive as opposed to low-cost services. A coordinated system that permits combining funding from a variety of sources would allow state and local agencies the flexibility they need to design services in response to client needs rather than funding constraints.

Additional federal or state responses that might assist with service coordination include (1) requiring that funding proposals be based on joint funding arrangements, (2) providing incentive payments to social service agencies that cooperate with other organizations, (3) supporting cross-agency training efforts that help staff develop common systems and assumptions for client services, (4) encouraging efforts designed specifically to support technical teams that help create partnerships between local service agencies, and (5) investing in computerized information systems that coordinate client information across a number of organizations and that track service provision.

To address the issues of service coordination, state-level staff from Ohio Family and Children First and the Ohio Department of Human Services have recently published a report entitled "State Resources Guidebook: 101 Ideas for Partnering at the Local Level."(1) This is a compilation of welfare agency efforts throughout the state to provide services to welfare recipients; these services are largely designed around coordinated systems and partnerships between social service agencies. Because the sharing between agencies of experiences and lessons learned provides some of the best direction for future efforts at service coordination, this step toward equipping staff with a critical tool toward this effort--the means to establish relationships and communicate with peers--holds promise for future efforts in the state.


As welfare agencies assume more responsibility for meeting the diverse and demanding needs of their clients, they are faced increasingly with issues of infrastructure, both external and internal. Agencies are both reliant upon and obstructed in their efforts by the state of the external infrastructure, one of whose central features is the health of the local economy. Both the local unemployment rate and the location of job opportunities play an enormous role in whether clients are able to find productive employment and sever their reliance on public assistance. Other infrastructure issues external to the welfare agency itself include the adequacy of an area's local transportation system, which may or may not be able to address issues related to spatial mismatch, and the availability of local housing. In addition, agency efforts to expand the number of service providers--in the areas of child care, substance abuse, mental health, domestic violence, and vocational rehabilitation--are a direct response to an external infrastructure that is not able to respond to the needs of today's welfare clients.

Local welfare agencies are being forced to take on these issues of infrastructure in their efforts to serve clients successfully. This is, however, an enormous, unrealistic, and inappropriate responsibility for a local welfare agency to assume alone. For the most part, welfare agencies cannot do much, if anything, to address these broad issues of local infrastructure. An appropriate response should begin with state or federal recognition of the critical role that an area's infrastructure plays to the successful transition of clients from welfare to work. This should be followed by comprehensive and well-planned efforts to develop the local infrastructure. Such approaches will be more time-consuming than band-aid efforts that respond to immediate needs, however, and are therefore incompatible with the constraints that time limits place on service provision. A combined approach that facilitates the means to address clients' short-term needs but includes development of longer-term efforts to address these larger issues of local infrastructure will, then, be the most effective.

If forced to address the broad external infrastructure, welfare agencies risk devoting insufficient attention to the internal infrastructure--program staffing needs or internal systems for service provision. In each of the ten areas in this report, one or more of the central implementation issues pertained to these internal matters. What is undeniably clear from current agency efforts is that caseworker roles and responsibilities are changing rapidly and drastically and that the skills needed for these new roles are not always adequately considered or developed. Caseworkers are increasingly expected to assess clients' needs, be familiar with an extensive range of service options in numerous areas, operate effectively with external service providers, interview and interact with clients effectively, monitor client progress in several areas, and coordinate and prioritize services. Agencies are at risk of a significant "skills mismatch" as they assume the increased burdens for client needs that stem from welfare reform. If this skills mismatch is not adequately addressed, there is little reason to believe that any approach will prove effective. Efforts to address this skills mismatch, however, cost both time and money--resources that are generally in short supply within welfare agencies. Federal and state resources or statewide efforts, such as the one under way in Oregon to train the state's entire staff of caseworkers, are needed (see Program Models under Substance Abuse). The alternative of ignoring this issue could seriously impede the overarching objective of the recent legislation to help those on public assistance move from welfare to work.

Additional Research Needs

The implication, implementation, and effect of drastic policy changes such as those contained in the recent welfare reform legislation are understood only through adequate research. There is now sufficient experience with operational responses to welfare reform to indicate several areas in which we clearly need additional research to grasp its full impacts.

The first entails rethinking how clients should be appropriately and adequately assessed for services. Assessments can be made through the use of specific instruments (such as the SASSI, used for substance abuse) or through informal caseworker-client interactions that take place over a period of time. Choice of approach should be determined based on local preferences, but additional attention is needed in either case. A more formal method requires identification of the appropriate instrument, coupled with caseworker training to ensure that staff are comfortable with it and skilled in its use. Assessments conducted through informal client interactions require sensitivity to time limits and training of caseworkers so that an appropriate diagnosis does not allow client needs to be overlooked. By whatever method, the focus needs to be on identifying individual client problems and on determining the service or set of services that will help clients become or stay employed. Because the profile of the welfare population could change drastically and continuously under welfare reform, and because neither clients nor agencies can afford to let needs go unmet, this needs assessment must be done more efficiently, more regularly, and more thoroughly than at present.

Part of this effort will require greater consensus on how each of these ten barriers should be defined within a welfare context. Clearer barrier definitions and more regular methods of client assessment will, in turn, help to measure the percentage of recipients that face each barrier. Understanding how many clients are faced with which barriers to employment is the foundation for appropriate program design at the state and local levels. The wide ranges in estimates presented at the start of each chapter (and the full tables included in Appendix A) highlight the need for a clearer means of such determination. Current estimates are based on an ad hoc system that leaves large gaps in our knowledge concerning the magnitude of the problem. Though in some areas there is a convergence within these estimates that engenders some confidence about the magnitude of the problem, for many areas figures diverge substantially. We need additional research in several areas that contain few estimates (in particular, child care, housing, transportation, mental health, and multiple barriers), and we need current research that measures the TANF population rather than the AFDC population.

The second area in which we need additional research is program effectiveness, which is clear from the dearth of information available for the preceding ten chapters. Though a rigorous evaluation that involves random assignment of welfare recipients to different packages of services is generally not feasible or desirable for all agencies to conduct, a more modest approach that includes some measure of client performance and program costs would be informative. In addition to permitting some comparative analysis between programs, it would help staff at the local level measure program effectiveness and cost-effectiveness and gauge the need for program modifications. State and local agencies should require that programs be accompanied at least by small-scale efforts to collect this type of information.

A handful of current efforts that include rigorous evaluation or extensive data collection promise to provide some very helpful information concerning program effectiveness. These include evaluations of the Goodwill Employment and Training Welfare-to-Work Programs for those with disabilities; the Options/Opciones Program for victims of domestic violence in Chicago, Illinois, and the Minnesota Domestic Abuse Intervention Project in Duluth; the Maryland Department of Human Resources Welfare Assistance Grants program, which provides diversionary assistance to potential welfare recipients; the Work First New Jersey Housing Assistance Program, which addresses housing barriers for welfare clients in New Jersey; the New Hope Demonstration Project in Milwaukee, Wisconsin, which provides subsidized health insurance; the Work First New Jersey Substance Abuse Research Demonstration Project and the ten-state demonstration by the National Center on Addiction and Substance Abuse at Columbia University, which address substance abuse among welfare recipients; and the ACCESS programs in nine states, for those with multiple barriers to employment. It is critical that the level of support that has permitted the kinds of evaluations noted above be extended to evaluations of other efforts in other sites. Only with a broad base of research that permits an analysis of client outcomes, from data collected at the local level to these more extensive evaluation efforts, will we increase our understanding of the most effective and promising program designs.

Before efforts to evaluate programs' effectiveness go much further, however, agreement must be reached on what constitutes an appropriate measure of effectiveness. Measures of a program's effectiveness could range, for example, from employment-related outcomes (such as wages, hours worked, employer satisfaction, number of jobs held, length of employment) to parenting outcomes (such as abuse and neglect, measures of child health or development, use of good parenting skills, foster care placement rates), personal outcomes more directly related to treatment (such as abstinence from alcohol or drug use, separation from an abusive partner, reduction of a stress disorder), and agency outcomes related to program implementation (such as the number of clients served or caseworker satisfaction with new service approaches). Different service agencies often have differing performance objectives, making agreement on a measurement of effectiveness potentially difficult. It is not clear, for example, whether an agency's efforts to assist clients with serious substance addiction should be gauged by the employment earnings of those clients or on some measure of their abstinence from drug use. Development of consensus on appropriate performance measures may require a formal and coordinated effort at the national, state, or local level. However, without some process toward building this consensus, program evaluations are likely to measure a wide range of outcomes and to remain largely uninformative.

The third area in which we need additional research is understanding appropriate and effective service provision in the context of time limits. The pressure of time-limited welfare receipt has affected the design--chiefly the duration--of services to address the barriers discussed in this report. In some cases, this is resulting in both an appropriate and a welcome increase in efficiency. In areas where barriers can be addressed through tangible resources--areas such as specialized child care, transportation, housing, medical needs, and financial assistance--time limits are beginning to encourage welfare agencies to develop innovative and effective ways to bring clients and services together more quickly. In other cases, however, where barriers are addressed through some form of referral to "treatment"--for barriers such as disability, domestic violence, mental health, and substance abuse--the appropriate balance of the duration of services and a service response sensitive to time limits is much less clear. Are time limits allowing adequate opportunity for treatment, or are they restricting the length of treatment provision to the point of compromising its effectiveness? Are agencies and other service providers being forced to design programs that address client needs only in the short term and not in the long term? Additional research is needed to address these questions. Programs risk providing inadequate assistance if they give undue preference to the concern for time limits over the concern for addressing basic client needs.

TANF entails a fundamental shift to a work-oriented approach to assistance. This shift, in turn, requires a systemic response that demands more of the welfare system than previously. Provision of services to address the full range of barriers to employment will take creativity, a willingness to take risks, additional financial resources, and coordination between agencies that are not used to working together. Because our current knowledge regarding program practices is limited and the options for program design unlimited, progress toward a better understanding of what services to provide for which clients will depend upon a willingness to carefully examine programs. Equipped with better information, agencies can replicate programs that show the most promise of helping clients address barriers and make the transition from welfare to work.

1. For further information on this initiative, contact Marlene Preston-Rombach, Ohio Department of Human Services, at 513-852-3283.