The Minnesota Integrated Services Project: Final Report on an Initiative to Improve Outcomes for Hard-to-Employ Welfare Recipients Karin Martinson

Transcription

1 The Minnesota Integrated Services Project: Final Report on an Initiative to Improve Outcomes for Hard-to-Employ Welfare Recipients Karin Martinson KARIN MARTINSON CAROLINE RATCLIFFE KATIE VINOPAL JOANNA PARNES The Urban Institute January 2009 Center on Labor, Human Services, and Population

2 The Minnesota Integrated Services Project: Final Report on an Initiative to Improve Outcomes for Hard-to-Employ Welfare Recipients Karin Martinson Karin Martinson Caroline Ratcliffe Katie Vinopal Joanna Parnes The Urban Institute January 2009 The Urban Institute 2100 M Street, NW Washington, DC This report was prepared for the Minnesota Department of Human Services under Contract Number A The views expressed are those of the authors and should not be attributed to Minnesota DHS or to the Urban Institute, its trustees, or its funders.

3 Acknowledgements The evaluation of the Minnesota Integrated Services Project (ISP) was conducted by the Urban Institute under contract to the Minnesota Department of Human Services (DHS). The evaluation is funded by the McKnight Foundation and DHS. The authors are especially grateful to the many people involved with the Minnesota ISP who took time to meet with us and help us learn about their programs. We also would like thank Leslie Crichton, Jane Delage, Marti Fischbach, Mark Kleczewski, and Vania Meyer from the Minnesota Department of Human Services for reviewing earlier drafts and providing comments.

4 Table of Contents Acknowledgements Executive Summary... ES-1 Chapter I: Introduction...1 Policy and Program Context... 1 Defining Service Integration... 3 Project Goals and Sites... 4 The ISP Evaluation and Data Sources... 9 Chapter II: The ISP Programs: Service Strategies and Participant Characteristics Overview of ISP Program Models: What Type of Service Integration Was Achieved? Description of the ISP Programs Anoka County: Partners for Family Success Chisago: Chisago County Integrated Services Project...23 Crow Wing County: Crow Wing County Integrated Services Project Hennepin County: Gateway to Success Ramsey County: Integrated Services Project Red Lake: Mino Aanokii St. Louis County: Project Hope Washington County: Integrated Services Project Chapter III: Change in ISP Participants Employment, Earnings, Government Benefits Receipt, and Employability Measure Scores Employment and Earnings MFIP Receipt and Benefit Levels Supplemental Security Income Receipt Summary of Patterns of Employment, Earnings, MFIP, and SSI ISP Participant Scores on the Employability Measure Chapter IV: Summary of ISP Participants Outcomes for Each Site Chapter V: Discussion of Key Findings and Policy Implications References Appendix A: Employability Measure... A-1 Appendix B: Supplemental Tables... B-1 Appendix C: Regression Approach...C-1

5 Exhibits and Tables Executive Summary Exhibit ES-1: Minnesota Integrated Services Projects... ES-6 Chapter I Exhibit 1-1: Location of the Minnesota Integrated Services Projects... 6 Table 1-1: Economic and Demographic Profile of the Minnesota Counties in which the Integrated Services Project Sites are Located... 7 Chapter II Exhibit 2-1: Minnesota Integrated Services Projects Table 2-1: Characteristics of ISP Participants at Enrollment, by Site Table 2-2: Types of Services and Assistance Received by ISP Participants Who Completed Initial Assessment within a Six-Month Follow-Up Period, by Site Chapter III Exhibit 3-1: Average Quarterly Employment Rate and Earnings Amount in the Two Years Before and After Enrollment Table 3-1: Changes in Employment, Earnings, and MFIP Receipt Over Time Exhibit 3-2: Average Monthly MFIP Benefit Receipt and Benefit Amount in the Two Years Before and After Enrollment Table 3-2: ISP Participants' Supplemental Security Income Table 3-3: Changes in Employability Measure Scores in the 6 and 12 Month Follow-Up Periods Appendix B Table B-1: Demographic Characteristics of ISP Participants at Enrollment, by Site... B-2 Table B-2: Educational Attainment and Prevalence of Learning Disabilities for ISP Participants at Enrollment, by Site... B-3 Table B-3: Prevalence of Physical and Mental Health, Substance Abuse, and Domestic Violence Barriers for ISP Participants at Enrollment, by Site... B-4 Table B-4: Criminal History, Housing Situations, Modes of Transportation, and Prevalence of Multiple Barriers for ISP Participants at Enrollment, by Site... B-5 Table B-5: Changes in Employment, Earnings, and MFIP Receipt Over Time in Anoka... B-6 Table B-6: ISP Participants' Supplemental Security Income in Anoka... B-7 Appendix C Table C-1: Logit Regression of ISP Participants' Employment-Regression Coefficients... C-5 Table C-2: Logit Regression of ISP Participants Employment-Odds Ratios.. C-7

6 Table C-3: OLS Regression of ISP Participants' Earnings... C-9 Table C-4: Logit Regression of ISP Participants' MFIP Receipt-Regression Coefficients C-11 Table C-5: Logit Regression of ISP Participants MFIP Receipt-Odds Ratios C-15 Table C-6: OLS Regression of ISP Participants' MFIP Benefit Amount. C-19

7 EXECUTIVE SUMMARY This report presents the final results for the evaluation of the Minnesota Integrated Services Projects (ISP) operating in eight sites across the state. Conceived by the Minnesota Department of Human Services (DHS) and enacted in 2005, the ISP seeks to address the needs of long-term cash assistance recipients in the Minnesota Family Investment Program (MFIP), many of whom are in danger of reaching their time limit on cash assistance benefits. The project was designed to address the serious and complicated barriers these families face by requiring a more coordinated response from the human service system. To this end, DHS provided grants to eight sites to address the multiple needs of long-term MFIP recipients. ISP aims to improve both economic and family-related outcomes for this population by increasing access to more comprehensive services that address multiple needs, coordinating services provided by multiple service systems, and focusing on the needs of both adults and children in the household. The primary goals of this report are to (1) describe the changes in the economic outcomes and family-related outcomes of ISP participants in each site over a two-year follow-up period, (2) provide estimates of the relationship between ISP participation and participants employment and MFIP outcomes, controlling for county economic conditions and participant characteristics, and (3) provide conclusions and policy recommendations based on the experiences of the ISPs. This report also summarizes findings from the implementation study on the design and operation of the ISP models from previous evaluation reports. The Evaluation of the Minnesota Integrated Services Projects MFIP is Minnesota s Temporary Assistance to Needy Families (TANF) program and, like most TANF programs across the country, it requires all cash assistance recipients to work or participate in employment-related services or risk financial penalties (known as sanctions). MFIP also establishes a lifetime limit on the receipt of cash benefits of 60 months, with some extensions allowed, and permits recipients to keep some of their benefits when they go to work, until their adjusted income reaches 115 percent of the federal poverty level. While the MFIP program has experienced considerable success in moving some individuals off welfare and into work, concerns have grown over how to meet the needs of those who remain on cash assistance for long periods of time, many of whom face significant barriers to employment. Relatively few strategies targeted at hard-to-employ welfare recipients have been evaluated, with families continuing to fare poorly over time even in the most successful efforts. Policy and Program Context The Minnesota Integrated Services Project is designed to address the multi-faceted needs of long-term MFIP recipients. In developing the project, state administrators recognized that a significant portion of this population receive diagnosis or services through other public systems, including mental health, chemical dependency, disability, child protection, domestic violence, and services for children with special needs. This reflects that the current delivery systems providing support to those with significant challenges are generally organized around singleissue expertise, with limited communication or coordination across systems. Moreover, many of ES-1

8 these barriers were not diagnosed or identified until individuals were close to meeting their time limits on cash assistance and little time was available to provide needed assistance. To encourage a more coordinated response from the human services system to address the serious and complicated barriers, DHS provided grants to address the multiple needs of longterm MFIP recipients. DHS established four primary goals of the ISPs: (1) identify employment barriers earlier in the family s time on cash assistance; (2) work with both adults and children in each family; (3) fundamentally change the way services are delivered so they are provided in an accessible, integrated, and cost-effective manner; and (4) identify policy and system issues that interfere with the delivery of services to the adults and children in these families. Eight sites representing diverse locations across the state were selected for the ISP: Anoka County, Chisago County, Crow Wing County, Hennepin County, Ramsey County, the Red Lake Band of Chippewa Indians, St. Louis County, and Washington County. The Chisago and St. Louis projects are regional in nature and include several surrounding counties. Each site received funding to operate their program for three years, although in the last year of the project resources were provided to extend the ISPs an additional year. Defining Service Integration To understand the nature of the service integration that occurred in the Minnesota ISPs, it is useful to discuss the meaning of service integration. The desire to simplify and streamline client processes through service integration is often cited as a solution to the wide range of uncoordinated programs that exist at the local level. Over the years, the terms integration and coordination as well as collaboration and linkages have often been used interchangeably and with varying connotations and meanings. It is generally recognized that there is no single definition of service integration. While the coordination of service delivery systems usually takes place at the local level, studies have shown that a initiative to coordinate may either be locally developed ( bottom-up coordination) or may be encouraged or imposed by federal or state officials ( top-down coordination). Studies also recognize a distinction between administrative service integration strategies with more ambitious goals focused on reforming the delivery system and more modest operational strategies focused on linking clients to existing services, without altering the program funding process, service agency responsibility, or organizational structures. When developing the ISPs, DHS did not provide a specific definition of service integration. While certain partners were mandated (county human services agency, a managed health care plan, and a community-based health clinic), ISP sites were given significant discretion in determining how to structure and operate their service integration models. This approach builds on the county administered welfare system in Minnesota, where counties are given significant latitude in designing a range of programs. Methodology The ISP evaluation, funded by the McKnight Foundation and DHS, is a multicomponent study employing a range of research strategies and data sources. The evaluation includes an implementation study; a study of participants employment, welfare, and family-related ES-2

9 outcomes based on administrative data; and a regression analysis examining the relationship between the interventions and participants employment, earnings, and welfare receipt. The report studies individuals who enrolled in the program April 2005 through June 2006, a total of 987 participants across all sites. For each of the eight sites, with administrative data provided by DHS, we analyze individual s monthly MFIP benefits and their quarterly earnings in jobs covered by the Minnesota unemployment insurance (UI) program over a two-year preenrollment and a two-year post-enrollment period. The study also examines data from the Employability Measure, an instrument developed by DHS for program staff to assess and track participant status in 11 areas related to economic success: child behavior, dependent care, education, financial, health, housing, legal, personal skills, safe living environment, social support, and transportation. This Employability Measure instrument was designed to be administered by program staff at ISP enrollment as well as every six months following enrollment, so that changes over time could be measured. However, the follow-up Employability Measures were not consistently administered across the sites. Other data sources include participants self-reported responses at ISP enrollment to a number of questions regarding specific barriers, including mental health, chemical dependency, learning disabilities, and criminal history. The implementation study is based on site visits and phone interviews by Urban Institute staff with program staff, a review of a sample of ISP case files in each site to document the types and levels of services received by ISP participants within a 6 to 18 month follow-up period, and focus groups with ISP participants in four sites (Anoka, Hennepin, Ramsey, and Washington). Implementation of the Integrated Services Projects: Target Group, Program Design, and Services This section summarizes key findings from the ISP implementation study describing each of the ISPs, target group and participant characteristics, and program design and services. See Martinson et al. (2007) for more detail on participant characteristics and program design and services. Target Group and Participant Characteristics Most of the ISPs established a broad target group of individuals on MFIP who have significant barriers to employment without specifically defining the type, number, or severity of barriers. A few sites target a more narrow population. Because of the nature of the eligibility criteria, there is significant discretion in determining who is eligible for and referred to the ISP. The most common method the sites use to identify appropriate families for ISP is direct referrals from MFIP, with MFIP staff making referrals within relatively broad parameters. Some ISPs also receive referrals from other agencies, although this is less common. The program in Ramsey is unique in that it specifically targets individuals with mental illness. In addition to serving a broader range of families, Anoka targets ES-3

10 individuals who are likely to be eligible for the Supplemental Security Income (SSI) program, a federal income supplement program for the disabled. ISP participants have a wide range of employment-related barriers, with mental health barriers the most prominent in most sites. Anoka and Ramsey counties serve populations with higher incidence of barriers, particularly mental and physical health problems. At the time of ISP enrollment, more than one-third of ISP participants in most sites reported that they had a mental health condition that made it hard to work. A smaller but still substantial segment of ISP participants reported barriers related to physical health, substance abuse, learning disabilities, and domestic violence issues. The ISP population also exhibited a high prevalence of multiple barriers to employment, with over 50 percent reporting two or more barriers in most sites. There are differences across the sites in terms of the types of barriers faced by ISP participants, with those in Anoka and Ramsey facing the most significant barriers. Over 60 percent of ISP participants in Anoka, and over 80 percent in Ramsey, reported a mental health condition that made it difficult to work and an even greater proportion said they had ever been diagnosed with depression. In these two sites, approximately 40 percent of the ISP participants had a physical condition that made it difficult to work and over 70 percent had two or more barriers to employment. Program Design The ISPs are relatively small and in most sites served those who volunteered to participate in the program. Programs in the ISP are small by design, with target enrollments of 100 to 200 participants, and a small staff, ranging from 5 to 13 individuals. Once individuals are referred to the program, they are contacted by ISP staff who inform them of services available through the ISP. Referred individuals do not become participants in the program until they agree to enroll and complete an initial assessment documenting the circumstances and barriers the individual and their family face. With the exception of Crow Wing, the ISPs complement but are separate from the standard MFIP program in each county. For the most part, individuals volunteered to participate and were not sanctioned for non-compliance with ISP, except in Crow Wing where ISP was not separate from the MFIP program and in Washington, where individuals were mandated to participate once they volunteered. All programs established partnerships with other organizations to serve ISP participants, although they varied in terms of the closeness of the collaboration. Most do not have connections with a wide number of partners from other service delivery systems. Linkages with experts in mental health, child protection, substance abuse, and public health are most common. A brief description of each ISP is provided in Exhibit ES-1. County social service agencies play an important role in several programs and serve as the lead operational agency in two sites. However, ES-4

11 in five programs (Chisago, Hennepin, Ramsey, St. Louis, and Washington), the lead organization is a nonprofit community-based organization. These organizations bring a range of expertise to the program, and some have experience as MFIP employment service providers. Reflecting the need to address the prevalence of mental health barriers, Anoka, Hennepin, Ramsey, and Washington established partnerships that provide expertise in this area. Anoka, Crow Wing, and Washington involve partners to assist with child protection services; Anoka, Crow Wing, and Hennepin include a public health component; and Crow Wing and Hennepin have partnerships to provide expertise in chemical dependency issues. While all the ISPs partner with a managed health care plan as required, in most programs, these organizations did not play a significant role. Overall, the ISPs generally limited the number of partners from other service delivery systems that they established formal connections with and focused on establishing linkages with a few key organizations. With some exceptions, most programs did not expand the number of partners involved in their programs as they matured. Hennepin is notable for making more significant changes to its initial program design by adding institutional partners with expertise in specific substantive areas. In the other sites, several partnerships did not work out as intended, with some partners less involved in the ISP than originally planned. Of all the sites, the program in Red Lake had the most trouble getting off the ground and implementing its model as intended. Primarily because of changes in leadership and shifting priorities, key partnerships were never established, and the level of services provided to participants is relatively low compared to the other sites. Three sites use a team approach that involves bringing staff with expertise in different areas to provide services to ISP participants, with all staff housed at the same physical location. Anoka, Crow Wing, and Hennepin use a team-based approach where participants may work with different staff or more than one staff person depending on their needs and the issues they are facing. However, within this single approach, there were different models. Anoka brought together a multidisciplinary team with staff from different divisions within the county human services division to provide expertise in a range of areas including juvenile and criminal justice, employment and vocational rehabilitation, public health, child protection, mental health, chemical dependency, and housing. The program also includes an emphasis on applying for SSI, with about half of the participants working exclusively on this issue with a dedicated staff person. Hennepin County contracts with other organizations and uses staff within the organization to provide a team of colocated staff with expertise in employment, mental health, substance abuse, domestic violence, public health, and youth issues. This program is also located at a community health clinic that provides many resources, including a food bank and an on-site medical, dental, and mental health clinic. Finally, the program in Crow Wing enhances the current MFIP program by incorporating joint supervision of program staff from the child protection division and the inclusion of a chemical dependency supervisor for case consultation, from within the Department of Human Services, and a public health nurse. ES-5

12 Exhibit ES-1 Minnesota Integrated Services Projects Team-based approach Anoka. This, project, housed in the Anoka County Human Services Division, features a multidisciplinary service team that provides intensive case management and service coordination for participants. Staff specialize in specific areas, including juvenile and criminal justice, employment and vocational rehabilitation, public health, child protection, mental health, chemical dependency, and housing, and participants are assigned to a case manager based on the barriers they are facing. A staff disability advocate assists participants with the SSI application process. Whenever possible, services for the family are provided in-house by the project team, though team members also connect with other professionals involved with the family. Crow Wing. Operated by Crow Wing County Social Services (CWCSS), this project builds on a segment of the county s existing MFIP program that targets hard-to-employ MFIP recipients, the Tier 3 program. MFIP recipients who have not found a job through the county s standard MFIP program are transferred to a MFIP outreach specialist at CWCSS who provides case management services and referrals to appropriate community resources. Supervisors from Child Protection Services (CPS) and Chemical Dependency divisions at CWCSS provide guidance and enhance coordination of services. An ISP specialist works with chemically dependent mothers, and a nurse from the Crow Wing Public Health Agency is also available to provide services as needed and participate in monthly staff meetings. The program also features an 8-week support group. Hennepin. Sponsored by NorthPoint Health and Wellness Center, Inc., a community-based health and human services agency, the core service of this program is one-on-one case management provided by Family Facilitators employed by NorthPoint as well as several MFIP employment service providers. Family Facilitators connect participants and their families with services in the community that address employment and other barriers. Staff from African American Family Services and Turning Point, two community-based social service agencies, assist participants with chemical dependency and domestic violence issues, and an onsite psychologist provides mental health assessments and counseling. Monthly support groups are a primary program activity. Service brokering approach Chisago. This program operates in a five-county region and is managed by Communities Investing in Families, a nonprofit organization with experience working with low-income families. Family advocates work one on one with participants to address barriers, refer them to additional resources in the community, and coordinate with MFIP employment counselors and other service providers who work with their clients. Red Lake. This project is operated by the Tribal Council of the Red Lake Band of Chippewa Indians. Through multidisciplinary case management, community workers link hard-to-employ MFIP recipients with appropriate services and programs on the reservation, such as GED courses. The program also provides transportation assistance and instruction in traditional work activities for clients, such as wreath-making, beading, and gardening. St. Louis. This project is operated in four counties by a set of community action agencies and the Minnesota Chippewa Tribe. Family employment advocates assess the needs of families and work with them one on one to help connect them with appropriate community resources to address a range of issues, including transportation, housing, substance abuse, child care, child support, probation, education, mental health, physical health, and domestic violence. Circles of Support, where participants are matched with community allies to provide support, is a key program component. Washington. Operated by HIRED, a nonprofit organization that provides MFIP employment services, this project aims to reduce the likelihood that residents will relocate and assist those who have relocated to reestablish services in Washington County. Its larger goal is to facilitate case coordination across systems for families involved with multiple service providers. Integrated Services coordinators complete an in-depth assessment, make individualized referrals to a wide range of services, and communicate with other professionals involved with the family to better coordinate services. The program has established a working relationship with the county s community mental health clinic to provide access to psychological evaluations and mental health services for clients. Single service approach Ramsey. This initiative integrates mental health rehabilitation expertise into the county MFIP employment services program, while accessing new funding outside the regular MFIP allocation. The ISP provides financial support to several providers to meet capacity and certification standards to provide services under Adult Rehabilitative Mental Health Services (ARMHS). ARMHS aim to help individuals with serious mental illness improve functionality, and services are billed directly to Medical Assistance (Minnesota s Medicaid program). Each agency has flexibility to provide ARMHS services or partner with an agency that provides ARMHS services. ES-6

13 The other ISPs primarily rely on the efforts of case managers in bringing services together for participants. Known as a service brokering approach, under this model, program staff are responsible for referring participants to other agencies in the community and coordinating these services based on the individual needs of participants. While used to some extent by sites using the team approach, this was the primary method for coordinating services in Chisago, Red Lake, St. Louis, and Washington. Under this approach, primarily through the efforts of line staff, the ISPs coordinate a wide range of services that address the multiple needs of individuals on their caseload including physical or mental health, substance abuse, housing, special needs of children and domestic violence. ISP staff may draw on the expertise of organizational partners, but participants are referred to these organizations to receive assistance rather than having on-site staff person provide the services. A single service approach, unique to Ramsey, focuses on providing in-depth assistance in one service area mental health. This program systemically brings rehabilitation expertise in mental health, with a focus on improving daily life skills and functionality, into the county MFIP but by design does not generally address the other service needs of participants. ISP Program Services ISP staff maintain low caseloads, which allow them to maintain frequent contact ISP participants. An important aspect of the ISP model in all sites is that staff have very low caseloads, ranging between about 10 to 40 cases, with most caseloads around 20. Although frequency varies across sites and among individual staff, staff consistently report very frequent contact with participants, with many maintaining weekly contact with clients over long periods of time. In addition to working with participants, many staff strive to make connections with staff from other service delivery systems that may be providing services to participants. Results from the focus groups conducted in four sites indicate that participants were generally positive and most had developed a very strong bond with their ISP worker. Focus group participants reported frequent contact with their ISP worker, either by phone or in face-to-face meetings, and generally described feeling that their worker was highly accessible and willing to help. Each ISP provides a high level of service in multiple areas, with assistance with mental health, employment, and child-related issues most common. A case-file review for a sample of cases examined the type of assistance received by ISP participants in specific areas in each site. This analysis shows high rates of assistance in multiple areas in most sites, reflecting both the voluntary nature of the program, where ES-7

14 those who are enrolled have an interest in receiving services, and the intensive nature of the services provided by program staff. Providing insights into the breadth of services provided, we found that across sites participants received services in six assistance areas on average within a 6-to-18 month period. Our review also reveals substantial differences across sites in the types of services provided, demonstrating the differing program goals and design of each of the sites ISPs as well as differences in the needs of their participants. Important patterns of service receipt are: Assistance with mental health issues is widely provided across all sites, reflecting the high incidence of mental health barriers among ISP participants. Services typically include assistance in diagnosing mental health problems and referrals to professionals who provide counseling, therapy, or drug treatment. Assistance in applying for SSI is a major focus of the program in Anoka County, where 90 percent of ISP participants received assistance in this area, and Ramsey, where 50 percent did so. This focus reflects the very high level of mental and physical health barriers among the ISP participants in these two sites. Employment is emphasized, although to varying degrees, in the ISPs. This emphasis was particularly strong in Crow Wing, St. Louis, and Washington, where over 80 percent of participants received employment-related services (primarily individualized job search), and weakest in Anoka and Ramsey where about one-third of the participants received assistance in this area. Another key goal of the ISP is to more effectively address the needs of the entire family, and most programs show high levels of assistance in addressing the needs of participants children. These results are consistent with findings from the focus groups, where ISP participants described receiving intensive services in multiple areas of assistance and, in general, felt they would not have received these services were it not for ISP. Compared to the service brokering approach, sites using the team approach generally provide participants easier access to specialized services and have more opportunities to bring together staff with different expertise to discuss specific cases. Under the team-based approach, staff with expertise in a range of areas are easily accessible by ISP participants, with no additional referrals or scheduling needed to receive assistance. Because team members are colocated and under the same management structure, staff discuss how to best coordinate service needs in different substantive areas for individual cases through regular, formal meetings, as well as informal discussions. Sites relying on a service brokering approach are generally not as successful in bringing staff from different service systems together to provide more coordinated services, although they sometimes have phone contact with other individual workers involved with the family, typically on an as-needed basis. In addition, ISP staff in programs with this model sometimes faced capacity or budgetary constraints at ES-8

15 organizations where they made referrals, which affected the timeliness and nature of services received by ISP participants. Overall, sites using a team-based approach were able to provide more integrated services than those using a service brokering approach. While the Minnesota ISPs provide a comprehensive set of services to their participants, for the most part, they did not successfully reduce the number of systems individuals are involved in or coordinate the actions of larger service delivery systems. The ISPs focus on identifying a wide range of services that are appropriate for individuals and assist individuals in accessing these services, sometimes through working with individual staff in other service delivery systems. To achieve a more systemic type of service integration, stronger mandates or guidance may be needed from state or county officials. The ISPs found it difficult to achieve systemic change without the leverage provided by this type of top-down approach. The ISPs were also very small and designed to operate for a limited period, which made it difficult to affect broader, long-run changes in the service delivery system. In addition, at initial implementation, focusing on a single geographic area may also be more productive than developing pilots covering a wider region as was done in the Chisago and St. Louis sites. Changes in Employment, Earnings, MFIP and SSI Receipt, and Employability Measure Scores by ISP Participants This analysis examines the trends in ISP participants employment, earnings, and benefit receipt two years before they enrolled in the ISP program and two years after enrollment. We also calculate differences in these outcomes before and after individuals enroll in the ISP. This analysis also examines changes in participants SSI receipt and Employability Measure scores, which provide information on barriers related to family stability. To ensure our comparison of pre- and post-enrollment earnings is based on participants more permanent and steady level of earnings, we define the pre- and post-enrollment periods to exclude the period immediately before and after enrollment. For our analysis of employment and earnings which is based on quarterly data, we follow the recent literature and define the pre-enrollment period to include the fifth through eighth quarter before enrollment and the post-enrollment period to include the fifth through eighth quarter after enrollment. For our analysis of MFIP receipt and MFIP benefit levels, which is based on monthly data, we define the pre-enrollment period to include the 13 th through 24 th month before enrollment and the post-enrollment period to include the 13 th through 24 th month after enrollment. Our examination of participants' employment, earnings, and MFIP receipt uses both descriptive and multivariate regression analyses. The descriptive analyses show patterns and changes in the outcomes over time. The multivariate regression analysis provides estimates of the relationship between ISP participation and participants employment and MFIP outcomes, controlling for county economic conditions (unemployment rate) and participant characteristics (gender, age, race, and educational attainment measured at ES-9

16 enrollment). A key benefit of the regression analysis over the descriptive analysis is that it takes account of the effects of the economy, as measured by the unemployment rate, on outcomes. It is important to note that the regressions cannot rule out that unmeasured factors aside from the program services may be responsible for some of the estimated relationships. These possible factors could include variables that have been shown to be related to MFIP outcomes, such as having a second parent in the household, mobility, serious mental health and chemical dependency diagnoses, marital status, number of children, whether the person is a U.S. citizen or speaks English, and county poverty rate. Also, small sample sizes make it more difficult to detect statistically significant relationships, particularly gains that are small in magnitude. Key findings from the analysis are as follows: Employment rates and average earnings tend to fall in the quarters leading up to ISP enrollment and increase in the quarters following enrollment. Conversely, MFIP receipt rates and average MFIP benefits tend to rise prior to enrollment and decline after enrollment. An analysis of ISP participants employment and earnings in the two years before and after ISP enrollment shows that employment and earnings generally decline prior to ISP enrollment and increase after enrollment. In Anoka, for example, about 40 percent of ISP participants were employed two years before enrollment, but only 15 percent were employed in the quarter of enrollment. During the two subsequent years, the employment rate fluctuated between 20 percent and 30 percent. Average earnings show a similar pattern. Patterns of MFIP receipt and MFIP benefit levels are inversely related to trends in employment and earnings. Across the sites, MFIP receipt tends to increase in the two years before ISP enrollment and fall in the two years after enrollment. The descriptive analysis shows that ISP participants in two sites Hennepin and St. Louis experienced statistically significant increases in employment and earnings between the pre- and post-enrollment periods. However, the regression-adjusted differences show statistically significant increases in employment and earnings for ISP participants in Hennepin only. The descriptive results show an increase in the employment and earnings of ISP participants in Hennepin and St. Louis between the pre- and post-enrollment periods. Hennepin s employment rate increased by 20.4 percentage points (from 31.7 percent to 52.1 percent) and average quarterly earnings increased by $907 (from $974 to $1,881). St. Louis s increases are more modest than Hennepin s, but still substantial. The employment rate increased by 9.7 percentage points between the pre- and post-enrollment periods (from 41.5 percent to 51.2 percent) and average quarterly earnings increased by $327 (from $911 to $1,237). Results from the regression analysis, which control for the county unemployment rate and selected participant characteristics, suggest that only Hennepin s program may be associated with increases in both employment and earnings. Hennepin s regressionadjusted employment rate increased by 11.3 percentage points and average quarterly ES-10

17 earnings increased by $465, or $1,860 annually. Although Hennepin ISP participants employment and earnings increased, their MFIP receipt and benefit levels did not significantly decline. This may be due to the income disregard which allows participants to keep some of their MFIP benefits when earnings are below a specified level. The descriptive analysis shows that ISP participants in four sites experienced statistically significant decreases in the MFIP receipt rate or in the average monthly MFIP benefit amount between the pre- and post-enrollment periods Anoka, Chisago, St. Louis, and Washington. Three of these sites Chisago, St. Louis, and Washington also had regression-adjusted declines. The descriptive results show that ISP participants in Chisago, St. Louis, and Washington experienced statistically significant declines in both their MFIP receipt rate and average MFIP benefit amount, while ISP participants in Anoka experienced statistically significant declines in only the MFIP benefit amount. The regression-adjusted differences show statistically significant declines in only three sites Chisago, St. Louis, and Washington. In St. Louis, ISP participants experience a decline in the amount of MFIP benefit received (but not a decline in the receipt rate). Specifically, average monthly MFIP benefits fell by $87. The regression-adjusted differences in MFIP receipt are similar in Chisago and Washington the MFIP receipt rate declined by 17.0 and 20.7 percentage points and the average monthly MFIP benefits among declined by $130 and $138, respectively. These MFIP declines occurred without a coinciding increase in employment or earnings. The reasons for these MFIP declines in the absence of employment and earnings gains are not clear. An analysis of ISP participants who reached their time limit during the study period shows that this was not a reason for the MFIP declines, with few individuals reaching their time limit and little variation across the sites. St. Louis ISP participants had increases in employment and earnings that are not statistically significant. 1 The magnitudes of these increases are not small, and the higher earnings could potentially account for the decline in MFIP benefits in this site. Washington is one of two sites where participation in ISP is mandatory and non-compliance could result in a MFIP sanction (or grant reduction). (A case-file review showed higher levels of sanctions in this site). This could potentially be a factor in explaining the MFIP reductions, although these did not occur in the other site where participation is mandatory and sanctions were incurred by participants (Crow Wing). The ISP programs in Anoka and Ramsey increased the proportion of participants who received SSI. Few ISP participants receive SSI prior to enrolling in the ISP program. Across the eight sites, only 15 individuals receive SSI benefits prior to ISP enrollment. By the end of the two-year follow-up period, an additional 122 ISP participants began receiving SSI. 1 Although findings from the descriptive analysis show increases in the employment and earnings of St. Louis ISP participants, the regression-adjusted differences are not statistically significant. ES-11

18 Anoka and Ramsey stand out as having the largest fraction of ISP participants that began receiving SSI after enrollment in ISP. In Ramsey, 30.1 percent of ISP participants began receiving SSI benefits after ISP enrollment, while 18.6 percent of Anoka s ISP participants began receiving SSI after ISP enrollment. In the other six sites, the percentage of ISP participants that begin receiving SSI after ISP enrollment is significantly lower and ranges between 2.1 percent and 8.1 percent. ISP participants in Anoka and Ramsey experienced unexpected statistically significant changes in participants outcomes. The analysis in Ramsey shows statistically significant changes in the unexpected direction. The regression-adjusted differences show statistically significant decrease in earnings and a statistically significant increase in the MFIP receipt rate. Similarly, Anoka s participants experience a statistically significant decline in their employment rate. These decreases in employment and earnings may be explained by the increases in SSI receipt in these sites. It is difficult for individuals to work once they receive SSI, both because individuals must have significant mental or physical health problems to be eligible for the program and because the SSI program includes many rules that discourage work. In addition, Ramsey s focus on rehabilitative services and improving everyday functioning may contribute to the increased MFIP receipt. Individuals may stay on MFIP longer than they would have in absence of the program in order to continue to these services. The Employability Measure scores show modest statistically significant gains on many elements in four sites Anoka, Chisago, St. Louis, and Washington. The results in Hennepin and Crow Wing are mixed and show follow-up Employability Measure scores that are both higher and lower than the initial scores. Ramsey makes no statistically significant gains, but experiences a statistically significant decline in one of the Employability Measure domains. Results from the Employability Measure scores should be interpreted cautiously because only a proportion of the ISP population completed a follow-up Employability Measure instrument so the results are not necessarily representative of the entire population. The analysis shows that Anoka has statistically significant gains at the six- and 12-month follow-ups in every domain except for legal and safe living environment. In Chisago, three domains education, health, and financial make statistically significant gains in both follow-up periods, while four domains child behavior, housing, social support, and transportation have a statistically significant gain in one of the follow-up periods. In St. Louis, nine average scores (all except legal and health) make statistically significant gains in at least one follow-up period. Washington ISP participants average scores make statistically significant gains during the first follow-up period in nine Employability Measure domains (all except legal and social support). Discussion of Key Findings and Policy Implications Many states and localities share an interest in finding strategies to more effectively assist individuals who receive cash assistance for long periods. Since this study uses non- ES-12

19 experimental regression analysis of small programs to examine the relationship between the ISP programs and employment, earnings, and MFIP receipt, the results should be interpreted cautiously. However, several implications can be drawn from the findings. The positive employment and earnings results in Hennepin County indicate the promise of this model which includes co-located staff with expertise in a number of areas including employment, mental health, substance abuse, domestic violence, and public health. The Hennepin ISP is notable because of the number of organizational partners as well as staff within the organization who provide on-site assistance in specific areas of expertise. In addition, case managers work closely with ISP participants and maintain low caseloads. The program is also located at a community health clinic that has many resources, including a food bank and medical, dental, and mental health clinic. The Hennepin ISP contracts with several other organizations to provide most of these specialized services to ISP participants and all staff are under the same management structure. Staff meet formally and informally to discuss the specific needs of individual cases. Finally, the Hennepin program includes a monthly support group that elicits high levels of participation by individuals in the ISP program. Two other sites (Anoka and Crow Wing) also use a team approach bringing together staff with expertise in different areas to provide services to ISP participants. However, while we cannot determine the specific reasons there are limited changes in economic outcomes over time in these sites, there are potential factors to consider. In Anoka, the ISP population has more mental and physical disabilities and the program has a strong focus on helping participants become eligible for SSI, which makes achieving increases in on employment and earnings more difficult. In Crow Wing, staff with expertise in child protection and substance abuse primarily provide supervision and case consultation rather than direct services. Crow Wing is also the only ISP site where the ISP services enhanced the existing MFIP program, rather than being a stand-alone program. Overall, while caution should be used given the limitations of the analysis of participant outcomes, the positive results for ISP participants in Hennepin warrant further consideration and evaluation of this approach. In particular, an evaluation using a random assignment research design would provide more definitive information on the effects of the program. Some aspects of the model may be difficult to replicate, given the unique array of services available at the community health clinic where the program is housed. This is particularly true in rural areas with more limited service environments. Increased employment and earnings may not be an appropriate goal for some longterm welfare recipients. Anoka and Ramsey counties both serve a population with significant mental and physical health issues, and the ISP program resulted in increases in SSI receipt and declines in employment and earnings among ISP participants. Reflecting the nature of their target population, assisting individuals in applying for SSI is a key service in both programs and the two programs have the lowest proportion of individuals receiving employment-related assistance among the ISPs. Given the significant barriers faced by the ISP populations in these sites, this emphasis on obtaining long-term income support may be appropriate for some individuals, particularly given the TANF time limits. However, these results also illustrate the challenge of identifying a target group ES-13

20 that can potentially benefit from services geared toward self-sufficiency rather than being more appropriate for income support programs for the disabled. A key issue is to identify appropriate screening and assessment tools to distinguish between those with temporary disabilities who have the potential to work, sometimes with accommodations and supports, and those with a significant level of permanent disability. A service brokering approach with low caseloads for staff may not be sufficient to address the persistent challenges of long-term cash assistance recipients. In the service brokering sites, program staff have very low caseloads and work with participants intensively to provide referrals to the services needed. However, while these services could have resulted in improvements in non-economic outcomes, the ISP participants in sites that used this approach did not experience increases in employment or earnings, and in some sites MFIP receipt declined. While caution should be used given small sample sizes on many of these sites, it may be that the brokered services, although coordinated by ISP staff, do not bring the full range of services needed to address participants needs. The service brokering approach can face limitations when other organizations have capacity or budgetary constraints. In addition, in this project, some of the sites using this approach are in rural areas and may face service shortages in particular areas. The non-economic benefits of the ISP programs remain difficult measure but are important to consider given the broad range of services provided. In addition to improving economic outcomes, the ISP initiative has relatively broad goals that include improving family-related outcomes and addressing the needs of children as well as adults in the family. Towards this end, the ISP programs generally provide a broad range of assistance in addition to employment-related services, particularly assistance with mental and physical health issues and issues related to their children. Indeed, over time, there are some increases in Employability Measure scores, an instrument designed to capture improvements in personal and family-related outcomes, although these results should be interpreted cautiously given data limitations. Nonetheless, given the hard-to-employ population served by the ISPs, it is important to recognize that these programs may achieve improved outcomes in areas other than employment gains and welfare reductions. There is a need to continue to develop measures of alternative measures of program accomplishments. Across all the sites, employment and earnings levels were very low both before and after enrollment in the program, indicating the significant challenge of designing effective program services for this population. In spite of these interventions, even in the most successful sites, the populations targeted by the ISPs continue to fare relatively poorly over time in terms of their overall employment and earnings and MFIP receipt. These findings are consistent with a range of other studies that show it is extremely difficult to improve the economic success of those who face significant barriers to employment. Overall, these results illustrate the persistent challenge of finding strategies to effectively assist individuals who receive cash benefits for long periods. ES-14