welfare reform in illinois - northwestern university · secretary of the illinois department of...

Welfare Reform in Illinois: Is the moderate approach working? Second annual report from the Illinois Families Study University Consortium on Welfare Reform May 2002

Dan A. Lewis, PhD Amy Bush Stevens, MSW, MPH

Northwestern University

Kristen Shook Slack, PhD, MSW University of Wisconsin at Madison

Bong Joo Lee, PhD

University of Chicago

Paul Kleppner, PhD Northern Illinois University

James Lewis, PhD

Roosevelt University

Stephanie Riger, PhD University of Illinois at Chicago

Robert Goerge, PhD

University of Chicago

Illinois Families Study Institute for Policy Research

Northwestern University 2040 Sheridan Road Evanston, IL 60208

May 1, 2002 Linda Reneé Baker Secretary of the Illinois Department of Human Services 100 South Grand Avenue East Springfield, IL 62762 Dear Secretary Baker, On behalf of the University Consortium on Welfare Reform and in accordance with the requirements of the Welfare Reform Research and Accountability Act (P.A. 90-74), it is my great pleasure to forward to you, for transmittal to the Illinois General Assembly, our report on the second year of the Illinois Families Study. The report, titled Welfare Reform in Illinois: Is the moderate approach working?, has both an executive summary and a complete technical report. The transmission of the report to you completes our responsibilities in the second year of the project as required by law. I want to make it clear to you and other members of the Governor’s cabinet that my colleagues and I are available to assist in interpreting the results and facilitating further analysis. I also want to compliment the Illinois Department of Human Services for the cooperation we have experienced in preparing this document. Your staff provided us with invaluable assistance over the last three years as the Consortium and the Department developed a very productive working arrangement. I believe that this partnership will be a model for other states that seek an objective understanding of the impact of welfare reform, an understanding that will assist policymakers in improving service delivery. This report focuses on the extent to which Illinois has achieved the goals of welfare reform and how our state’s experience may be instructive nationally as the US Congress grapples with TANF reauthorization. We hope this approach informs the discussion both here at home and at the federal level. Next year and the years to come will bring new reports and further analysis of Illinois’ efforts to improve the lives of poor families. We stand ready to assist the General Assembly and the executive branch of the Illinois government in the challenges ahead. Sincerely, Dan A. Lewis, Professor of Education and Social Policy, and Principal Investigator, University Consortium on Welfare Reform

Second annual report from the Illinois Families Study i

Acknowledgements We are particularly grateful to the several foundations and government agencies that have supported the Illinois Families Study (IFS). The first three years of the project were made possible by the Joyce Foundation, the John D. and Catherine T. MacArthur Foundation, the Woods Fund of Chicago, the Annie E. Casey Foundation, the Polk Bros. Foundation, the Searle Fund, the Illinois Board of Higher Education, the Administration for Children and Families, and the National Institute of Child Health and Human Development. More recently, we have received additional support from the US Department of Education, the National Institute of Justice, and the Illinois Department of Human Services. We thank the 1,183 respondents who completed interviews two years in a row for this study. This report would not be possible without them. The Metro Chicago Information Center (MCIC) conducted data collection, field work, and data entry for both rounds of interviews. Woody Carter, Research Director, manages the IFS data collection at MCIC. A special thanks is extended to MCIC’s interviewing teams in Chicago, Peoria, and East St. Louis. They are an extremely professional group of interviewers who have tirelessly followed and successfully interviewed the very mobile group of families participating in this study. One of the most satisfying aspects of this work has been the partnership we have formed with the Illinois Department of Human Services (IDHS). The department has gone out of its way to assist with the project. Dave Gruenenfelder and Alan Whittaker, in particular, have provided valuable feedback and assistance. Many other IDHS staff have also provided feedback on drafts of this report. The Illinois Department of Public Assistance (IDPA) also provided assistance. The Legislative Advisory Committee, comprised of several members of the Illinois State Legislature, provides ongoing consultation to the IFS staff on policy relevance and dissemination. The members of this committee include Senators Steven Rauschenberger and Miguel del Valle, and Representatives Barbara Flynn Currie and Rosemary Mulligan. Numerous graduate students from Northwestern University and the University of Wisconsin at Madison contributed significantly to this report. Lisa Altenbernd and Marla McDaniel provided feedback on report revisions, in addition to conducting data cleaning and data analysis. Amber Stitziel Pareja, Irene Carvalho, Morgan Doran, Matt Smith, Joan Yoo, Lynette Renner, and Alan Puckett also conducted data cleaning and data analysis. Natalie Phillips-Hamblett, IFS Project Assistant, also contributed to the report through research and administrative support. Jane Holl, Principal Investigator of the Illinois Families Study: Child Well-Being Supplement, provides valuable support and expertise to research design and data collection oversight for the overall study.

ii Welfare Reform in Illinois: Is the moderate approach working?

Several staff members at the Institute for Policy Research at Northwestern University provided administrative assistance to the project. These individuals include Ellen Feldman, Ruth McCullough, Patti Huiras, and Fay Lomax Cook. Audrey Chambers edited this report and Alice Murray assisted with formatting. The Principal Investigators of the study are Dan Lewis (Northwestern University), Paul Kleppner (Northern Illinois University), James Lewis (Roosevelt University), Stephanie Riger (University of Illinois at Chicago), Bong Joo Lee (University of Chicago), and Robert Goerge (University of Chicago). This report does not necessarily reflect the opinions of those who provided assistance to the investigators and project staff in its development. All correspondence related to this report should be directed to Amy Bush Stevens, IFS Project Coordinator, at 847-491-5889 ([email protected]), or Dan Lewis, Principal Investigator, at 847-491-8722 ([email protected]).

Second annual report from the Illinois Families Study iii

Table of contents

Executive summary ...........................................................................................xi Introduction.........................................................................................................1

Purpose of this report ...................................................................................................... 1 Goals of welfare reform.................................................................................................. 2 Key features of Illinois TANF and related programs ..................................................... 2 Why is Illinois important to the national reauthorization debate?.................................. 3 Welfare in Illinois during the 1990s: Historical and policy context............................... 5 A moderate approach to reform: Illinois and its regional context .................................. 6 AFDC/TANF caseload dynamics in Illinois during the 1990s....................................... 7 Other demographic characteristics of Illinois ................................................................. 9

Objectives and research design......................................................................10 Objectives ..................................................................................................................... 10 Sample and sample selection ........................................................................................ 10 Data sources .................................................................................................................. 11 Response rates............................................................................................................... 13 Analysis strategy........................................................................................................... 13

Respondent characteristics.............................................................................15 Non-respondents ........................................................................................................... 16

Have the main goals of PRWORA been realized in Illinois? .........................18 Work and welfare.......................................................................................................... 18 Family formation .......................................................................................................... 26 Summary ....................................................................................................................... 28

Life after welfare reform: How are families faring?........................................30 Income and poverty....................................................................................................... 30 Material hardship .......................................................................................................... 32 Job characteristics ......................................................................................................... 37 Physical and mental health............................................................................................ 42 Health insurance............................................................................................................ 43 Child care ...................................................................................................................... 45 Satisfaction with welfare............................................................................................... 47

Making work pay in Illinois: Use of work supports and other services ......50 Earnings disregards and stopping the clock.................................................................. 50 Food stamps and Medicaid ........................................................................................... 54 Other work supports...................................................................................................... 54 Knowledge of work supports and other policies........................................................... 58 Most valued benefit....................................................................................................... 59 Does it pay to move from welfare to work? ................................................................. 60

iv Welfare Reform in Illinois: Is the moderate approach working?

Time limits and grant closures (“sanctions”).................................................63 Who is most likely to reach the lifetime limit?............................................................. 63 How are families affected by losing TANF? ................................................................ 64

Putting it all together: What helps families attain self-sufficiency?.............68 Additional description of methods................................................................................ 71

Discussion.........................................................................................................73 Have the federal goals of welfare reform been achieved in Illinois? ........................... 73 How are families doing? ............................................................................................... 73 What about the more vulnerable families? ................................................................... 74 Do welfare reform policies and work supports make work pay in Illinois?................. 75 What is the overall impact of the moderate Illinois approach to welfare reform? ....... 76

Conclusions and policy implications..............................................................77 Work supports are key to the success of welfare reform.............................................. 77 Family well-being looking better, but still needs improvement ................................... 79 Family formation .......................................................................................................... 80

Recommendations............................................................................................81 Recommendations to federal policymakers for TANF reauthorization........................ 81 Recommendations to Illinois policymakers.................................................................. 82

References ........................................................................................................83

Appendices .......................................................................................................84 Appendix A: The Illinois TANF Program.................................................................... 85 Appendix B: Welfare policies and caseload trends in Illinois and other Midwestern states.............................................................................................................................. 88 Appendix C: IFS Study Regions................................................................................... 93 Appendix D: IFS Research Team, Advisory Groups, and Collaborative Partners....... 97 Appendix E: IFS supplemental studies ....................................................................... 100 Appendix F: Three families tell their stories .............................................................. 102

Second annual report from the Illinois Families Study v

Figures Figure 1: Child poverty, AFDC/TANF caseloads, and unemployment in

Illinois, 1990-2000 ...........................................................................5 Figure 2: Illinois AFDC/TANF caseload, entries, and exits (1990-2001).........8 Figure 3: Trends in work, 1998-2001 (percent of IFS sample members with

reported earnings; n=1,899) ..........................................................19 Figure 4: Trends in reported earnings, 1998-2001 (median quarterly earnings

for those with earnings in each quarter) ........................................20 Figure 5: Trends in TANF and food stamp use and Medicaid enrollment

among adults, 1998-2001 (number of IFS sample members active for each program, by month; n=1,899) ..........................................21

Figure 6a: Transition from work-only status in 1999-00 (2001 status among

those in the work-only group in 1999-00) ......................................24 Figure 6b: Transition from combined work and TANF status in 1999-00 (2001

status among those in the work-and-TANF group in 1999-00) ......24 Figure 6c: Transition from TANF-only status in 1999-00 (2001 status among

those in the TANF-only group in 1999-00).....................................25 Figure 6d: Transition from neither work nor TANF in 1999-00 (2001 status

among those in the no-work/no-TANF group in 1999-00)..............25 Figure 7: Percent of children in poverty by study county, 1995 and 1997 ....94 Figure 8: Percent change in TANF grantees by study county, 1996-1999 ...95

vi Welfare Reform in Illinois: Is the moderate approach working?

Tables Table 1: IFS data sources (included in this report) ......................................11 Table 2: IFS survey response rates.............................................................13 Table 3: Residence of respondents in 1999-00 (n=1,183)...........................15 Table 4: Demographic characteristics in 1999-00 (n=1,183) .......................16 Table 5: Characteristics of sample members by survey response status....17 Table 6: Number or quarters worked, 1998-2001 (percent of IFS sample

members with reported earnings; n=1,899) ...................................19 Table 7: Changes in benefit packages for adults, 1999-2001 (Percent of IFS

sample members active for each program, by month; n=1,899)....22 Table 8: Work status and TANF use ...........................................................22 Table 9: Work and TANF use patterns ........................................................23 Table 10: Marriage and cohabitation .............................................................26 Table 11: Births .............................................................................................26 Table 12: Births, by 1999 characteristics.......................................................28 Table 13: Changes in mean and median annual income ..............................31 Table 14: Income ratio and poverty ...............................................................31 Table 15: Material hardships experienced since last interview/ in past year .33 Table 16: Food insecurity since last interview/ in past year...........................34 Table 17: Homelessness and problems with housing conditions since last

interview/ in past year....................................................................35 Table 18: Perceived financial situation ..........................................................36 Table 19: Use of charities and crisis assistance since last interview/ in past

year ...............................................................................................36 Table 20: Median and mean hourly wage at current job................................37

Second annual report from the Illinois Families Study vii

Table 21: Changes in hourly wage for individuals between 1999-00 and 2001 (among those employed in 1999-00 and 2001) .............................37

Table 22: Employer benefits received at current job (among workers)..........38 Table 23: Industry (of current job) .................................................................38 Table 24: Full and part-time work ..................................................................39 Table 25: Changes in work status for individuals between 1999-00 and 2001

(among those employed in 1999-00 and 2001) .............................40 Table 26: Length of time it takes to get to work, including dropping off children

for child care..................................................................................40 Table 27: “How satisfied are you with your current main job?” ......................41 Table 28: Received a promotion or pay raise from current employer in past

year ...............................................................................................41 Table 29: Job retention..................................................................................42 Table 30: Overall health status of respondents and their children.................42 Table 31: Depression ....................................................................................43 Table 32: Respondents’ current health insurance coverage .........................43 Table 33: Children’s current health insurance coverage ...............................44 Table 34: Gaps in health coverage for respondents and their children, during

past year........................................................................................45 Table 35: Main type of child care arrangement used last week.....................46 Table 36: “How many different child care arrangements did child have during

past 12 months?”...........................................................................46 Table 37: Child care problems experienced since last interview/ in past year

(among parents of at least one child under age 12) ......................47 Table 38: Satisfaction with current welfare worker: percent who “somewhat”

or “strongly agree” .........................................................................48 Table 39: Attitudes about welfare reform: percent who “somewhat” or

“strongly agree” .............................................................................49

viii Welfare Reform in Illinois: Is the moderate approach working?

Table 40: Respondents who had their clocks stopped (while receiving TANF

and working)..................................................................................51 Table 41: Characteristics of respondents who had their clocks stopped (while

receiving TANF and working) ........................................................52 Table 42: “Stopped clock” use and material hardship ...................................53 Table 43: Child Care Subsidy Receipt...........................................................54 Table 44: Self-reported use of and need for the child care subsidy...............55 Table 45: Current receipt of housing assistance ...........................................55 Table 46: Current receipt of child support .....................................................56 Table 47: Earned Income Tax Credit (EITC) receipt .....................................57 Table 48: Participation in job training, work experience, and education

programs since last interview/ in past year....................................58 Table 49: Knowledge of welfare policies .......................................................59 Table 50: “If you could pick one thing, what would you say is most important

to your family’s well-being?” ..........................................................60 Table 51: Material hardship by work and TANF status..................................62 Table 52: Moving towards the 60-month time limit, 1997-2001 .....................63 Table 53: Characteristics of respondents by risk for reaching the time limit ..64 Table 54: Percent of respondents who experienced a full-grant loss due to

“non-cooperation” ..........................................................................65 Table 55: Characteristics of respondents who experienced a full-grant loss for

“non-cooperation” ..........................................................................66 Table 56: Full-grant losses and material hardship.........................................67 Table 57: What helps families to meet the goals of welfare reform and attain

self-sufficiency?.............................................................................70 Table 58: Characteristics of the Illinois TANF program .................................85

Second annual report from the Illinois Families Study ix

Table 59: Comparison of six Midwestern state TANF programs ...................89 Table 60: Maximum monthly benefits and earnings levels in six Midwestern

states (for a family of three; as of October 2000) ..........................92 Table 61: Standard of need and AFDC benefits for six Midwestern states,

January 1994.................................................................................92 Table 62: AFDC/TANF caseload declines in six Midwestern states, 1994-

2000 ..............................................................................................92 Table 63: Number of TANF grantees (monthly average) by study county,

1996, 1999, and 2000....................................................................95 Table 64: Race by study county, 2000 ..........................................................96 Table 65: Hispanic origin by study county, 2000 ...........................................96

x Welfare Reform in Illinois: Is the moderate approach working?

Second annual report from the Illinois Families Study xi

Executive summary Purpose of this report Illinois, like most other states in the nation, is now more than four years into the major welfare reforms that were passed in 1996 and implemented in 1997. This report assesses the success of the Illinois approach in achieving the formal goals of the reforms, and its impact on the well-being of the families affected by the reforms. This report has two primary audiences. First, the Illinois Families Study is being conducted in response to a mandate by the Illinois legislature. The goal of the study is to inform state policymakers about how Illinois families have been faring since the implementation of Temporary Assistance for Needy Families (TANF) and other welfare reform policies. Second, this report contributes to the national discussion about the reauthorization of TANF by providing evidence on the impact of the reforms in a large state with a moderate TANF program. Because Illinois welfare reform policies are considered to be moderate, or even generous, in comparison to other states, and because Illinois enjoyed low unemployment rates in the late 1990s, the state provides an excellent example of the effects of the “middle of the road” approach to reform under good economic conditions. Methods The Illinois Families Study is longitudinal and will follow the same group of families for six years. The core of the study is an annual in-person survey of a random sample of adults who were primary TANF grantees in the fall of 1998, a little more than a year after TANF was implemented. Participants were selected from nine Illinois counties that were stratified by two regions: Cook County (including the city of Chicago) Eight Downstate counties (including the cities of East St. Louis and Peoria, and rural

counties surrounding Peoria) Together, these nine counties represent approximately 75% of the state TANF caseload. They also represent cities and towns of varying sizes and demographic makeup. This report draws upon two sources of data: Survey data: results of two annual in-person interviews, one conducted between

November 1999 and September 2000 and a follow-up interview conducted between February 2001 and September 2001

Administrative records: information about the use of TANF, Medicaid, food stamps, and child care subsidies, and employment and wage records (linked to the same families who were interviewed)

A total of 1,183 respondents were interviewed both in 1999-00 and in 2001. The response rate in 1999-00 was 72% (1,363 respondents) and was 87% in 2001 (1,183

xii Welfare Reform in Illinois: Is the moderate approach working?

respondents). Ninety-three percent of respondents consented to allow IFS researchers to access their individual administrative records. All analyses of survey data presented here are weighted to adjust for regional stratification and non-response. Have the main goals of PRWORA been realized in Illinois? The primary goals of PRWORA were to promote work and marriage and to decrease welfare dependence and births to unmarried women. In Illinois, the goal to decrease welfare dependence was met, efforts to increase work were moderately successful, and “family formation” goals were less successful. Between 1998 and 2001, work and earnings increased slightly while TANF use dropped sharply. About half of the IFS sample was working in 2001, while about one-quarter was still receiving TANF. There was a significant increase in the proportion of respondents who were relying on work only (with no TANF), and a corresponding decrease in the proportion relying only on welfare. The proportion of respondents who were neither receiving TANF nor working increased 10 percentage points between the 1999-00 and 2001 interviews. Overall, large declines in TANF use were not matched by comparable increases in work, indicating that some families were left with neither work nor welfare as a source of support, although they may have had other sources of support. Efforts to decrease welfare dependence, therefore, appear to have been very successful, while efforts to increase work were moderately successful. Between the 1999-00 and 2001 interviews, there was no change in marriage rates and a slight decrease in childbearing. Overall, only 10% of respondents were married in 2001. There were fewer births to unmarried women in 2000 than in 1999, although it appears that this decline was caused more by personal characteristics such as aging and family size, than by policies related to welfare. PRWORA’s “family formation” goals, particularly related to marriage, do not appear to have been met in Illinois during this time period. How are families doing? Looking across a broad range of indicators of well-being, families appeared to be doing slightly better in 2001 than in 1999-00. Although many families saw significant improvements, some still struggled with poverty, hardship, and meeting other needs. Overall, income for these families has been increasing, although incomes continued to be very low. Combining earnings and all benefits (e.g., TANF and food stamps), the average family income in 1999-00 was only $11,812, up from $7,475 in 1998. The median annual income rose from $6,250 in 1998 to $8,750 in 1999-00. Seventy-six percent of the families were living below the poverty line in 1999-00.

Second annual report from the Illinois Families Study xiii

Between 1999-00 and 2001, there was an overall decrease in the number of families who experienced material hardship, although the majority (56%) still reported some hardship. Most housing affordability problems increased, with more respondents saying they had trouble paying their rent or mortgage or were evicted. Homelessness and food insecurity, however, saw modest declines. Homelessness declined from 7% in 1999-00 to 5% in 2001. This includes families who stayed in a shelter, car, abandoned building, “on the streets,” or temporarily (less than two weeks) doubled up with a friend or relative. Nine percent of respondents in 2001 said they “sometimes” or “often” didn’t have enough food. Overall, most respondents perceived their financial situation to be improving or staying the same. Over two-thirds (67%) said they “somewhat” or “strongly” agreed that they can generally afford to buy the things they need. Eighty-three percent of respondents at both surveys, however, said they “worry a lot about having enough money in the future.” Between 1999-00 and 2001, respondents made significant gains in their hourly wages and in employer-sponsored benefits, although wages still remained fairly low and most working respondents did not receive benefits from their employers. The median wage in 2001 was $8.00 per hour, and 23% of working respondents said they received health insurance from their employer for themselves. In 2001, 41% of workers said they had received a pay raise or promotion in the past year. In 2001, employed respondents worked an average of 32 hours per week, down slightly from 33 hours in 1999-00. Seventy-one percent of all employed respondents were working full time (30+ hours per week) in 2001. There were overall declines in job satisfaction among workers between 1999-00 and 2001, although 73% of respondents in 2001 still reported they were “very” or “somewhat” satisfied with their job. Overall, health status appeared to be improving or staying the same by 2001. There was no change in self-reported health status for respondents and a slight improvement in the parent-reported health status of children. Fewer adults reported depressive symptoms. Between 1999-00 and 2001, health insurance coverage improved for children and worsened for parents. Significant increases in employer-sponsored health insurance were not enough to offset large declines in Medicaid receipt for adults, resulting in one-quarter of respondents being uninsured in 2001. Changes in health insurance status for children were more positive and less dramatic than for adults. Increases in KidCare (S-CHIP) and employer-sponsored coverage helped to offset declines in Medicaid, resulting in a decrease in the number of uninsured children in 2001. In 2001, 9% of respondents had at least one child who was uninsured. Use of formal child care arrangements, such as child care centers and Head Start, was fairly rare. In 2001, almost half of the children under age 12 in the sample were being cared for by a relative or some other family member. Stability of child care arrangements

xiv Welfare Reform in Illinois: Is the moderate approach working?

improved slightly from 1999-00, with 90% of children having only 0-1 child care arrangements within the 12 months prior to the 2001 interview. Quality and cost were the most common child care concerns, although most parents appeared to be satisfied with their child care arrangements. Most respondents expressed positive feelings about their welfare worker and specified welfare policies (time limits and work requirements). Satisfaction with workers and policies increased further from 1999-00 to 2001. Fully 92% said they “somewhat” or “strongly” agreed that “it is a good idea to require people on welfare to work.” Making work pay in Illinois: Use of work supports and other services Work does indeed “pay” in Illinois. Those who were working in 1999-00 experienced less material hardship in 2001 than those who were not working in 1999-00. Despite fairly high rates of awareness of work support policies, many respondents were not receiving the work support benefits provided in Illinois. With the exception of the child care subsidy, use of work supports and other services appeared to decrease over time. The majority of respondents (84%) had their TANF clocks stopped at some point between July 1997 and February 2001. This means they had a “stopped clock” due to employment while they were receiving TANF. Decreased material hardship among longer-term “stopped clock” users indicates that the policy can have a positive impact if used for an extensive period of time. Food stamp and Medicaid use has dropped considerably, with a little more than half of the sample still receiving each of these benefits in 2001. Use of child care subsidies increased slightly between 1999 and 2000, although only 55% of workers with a child under age 12 were receiving this benefit in September 2000. About one-third of respondents reported receiving some kind of housing subsidy in 2001, down slightly from 1999-00. Sixteen percent of respondents received formal child support payments in the year prior to their 2001 interview. More than half, however, said they received informal child support from their child(ren)’s other parent, usually the father. Participation in job readiness, job training, work experience, basic education, and secondary education programs declined between 1999-00 and 2001, while participation in postsecondary education saw no change but remained very low (3%). Most respondents did not participate in these types of programs.

Second annual report from the Illinois Families Study xv

The majority of respondents were aware of earnings disregards, transitional Medicaid, and continuing food stamps, while slightly less than half said they knew that the months when they work at least 30 hours per week do not count toward the time limit (“stopped clock” policy). Knowledge of these benefits increased significantly between 1999-00 and 2001. Looking at the relationship between work and welfare status in 1999-00 and material hardship in 2001, we found that work was protective against material hardship. Those who were relying on TANF only, or had neither work nor TANF in 1999-00 consistently reported more hardship in 2001 than did those who were working. These findings indicate that work does indeed “pay” and that staying off welfare alone does not seem to protect families from hardship. It is also important to note that although work seems to prevent some material hardship, levels of hardship were still high even among those who were working and off welfare; 46% of the work-only group experienced some hardship in 2001. Time limits and grant closures (“sanctions”) Although most families are not at risk of reaching the time limit within the next year, certain “vulnerable” groups can be identified that are at risk of “using up” their TANF months. Most families we studied are not at risk of reaching the time limit in July 2002. As of February 2001, about one-quarter had “used up” more than half of their 60-month life-time limit. Those closer to reaching the time limit were more likely to have chronic health problems, more children, or a child with a limiting health condition. Those who had “used up” less than half of their TANF months were more likely to be working and to have a high school degree. We were not able to measure the impact of sanctions, per se. Instead, looking at a similar group of families—those who lost their full TANF grant due to “non-cooperation”—we found that families who lost their benefits involuntarily were more likely to experience material hardship and to earn lower hourly wages compared to other families. Putting it all together: What helps families attain self-sufficiency? Work and work supports seem to promote positive outcomes, while full-grant losses and health problems are associated with negative outcomes. Employer-sponsored health insurance, child care subsidies, and the “stopped clock” option appear to help families to get and keep jobs. Those who work, in turn, were less likely to experience material hardship, whereas those who experienced a full-grant loss for non-cooperation had, on average, lower earnings and more material hardship. Having a high school degree or GED helped respondents to keep working and earn higher wages.

xvi Welfare Reform in Illinois: Is the moderate approach working?

Respondents with poor health were less likely to be working and more likely to experience material hardship. Even after controlling for work status in 1999-00, respondents with the following characteristics prior to 2001 were significantly more likely to be working in 2001: Had a job with employer-sponsored health insurance Received a child care subsidy Used the “stopped clock” (even after controlling for number of months on welfare) Had a high school degree or GED

Respondents who reported a health problem in 1999-00 were less likely to be working in 2001. Respondents with the following characteristics were significantly more likely to be earning higher hourly wages relative to others who were working in 2001: Had a job with employer-sponsored health insurance Had a high school degree or GED

Respondents who experienced a full-grant loss for “non-cooperation” in 1999-00 earned lower hourly wages. Respondents who were working in 1999-00 in a job with employer-sponsored health insurance were much less likely to be on welfare in 2001. Respondents with the following characteristics were significantly more likely to experience material hardship in 2001: Had a health problem Experienced a full-grant loss for “non-cooperation”

Respondents who were working in 1999-00 were less likely to experience material hardship in 2001. Conclusions Work supports are key to the success of welfare reform.

Thanks in part to several strong work supports available in Illinois, we found that work does indeed “pay” for many families. Material hardship and poverty, however, continue to be prevalent in this population and those leaving welfare often enter jobs with low wages and few benefits. Policymakers therefore have an obligation to follow through on their end of the new “social contract” embodied in the 1996 reforms by ensuring that working and leaving welfare are sustainable and translate into concrete gains for parents and their children. A comprehensive network of work supports for low-income families is the key to meeting this state obligation.

Second annual report from the Illinois Families Study xvii

In Illinois, three work supports stand out as being particularly critical for promoting work and self-sufficiency: The “stopped clock” for employed TANF recipients Child care subsidies Medicaid and KidCare (S-CHIP)

These supports appear to be very helpful to families, although it is important to note that not all families who need Medicaid and child care subsidies are receiving them. Family well-being is looking better, but still needs improvement.

Looking across a broad range of indicators of well-being, families appeared to be doing slightly better in 2001 than in 1999-00. Although many families experienced some instability or hardship, many of the most severe hardships, such as homelessness and food insecurity, decreased slightly over the two-year study period. Three problems, however, stand out as obstacles to work, self-sufficiency, and well-being: Poor health and lack of health insurance Low levels of education Difficulty paying for housing

Recommendations to federal policymakers for TANF reauthorization Time limits The “stopped clock” policy for employed TANF recipients has been successfully

implemented in Illinois and should be adopted nationally. The “stopped clock” policy for TANF recipients who are participating in

postsecondary education, providing in-home care for a medically dependent child, or caring for a sick child or spouse should also be encouraged by federal legislation.

Medicaid and S-CHIP Public programs are the most viable source of health insurance coverage for most

current and former welfare recipients. The federal government should provide incentives to states to make Medicaid more

accessible to parents by: Raising income eligibility cutoffs Expanding Transitional Medicaid Assistance (TMA) beyond 6-12 months after

TANF exit The federal government should facilitate and encourage S-CHIP waivers that allow

states to expand their S-CHIP benefits to parents. The federal government should reward states for high enrollment in Medicaid and S-

CHIP. Outreach and marketing efforts should be encouraged to increase awareness of these programs.

xviii Welfare Reform in Illinois: Is the moderate approach working?

Child care Child care assistance should receive the highest possible levels of financial support

from the federal government. States should be rewarded for high take-up rates for their child care subsidies. The Illinois approach to providing child care subsidies with eligibility tied to income

(instead of TANF status) should be encouraged in other states that have more fragmented child care programs.

Education and training The federal government should encourage states to: Include secondary and postsecondary education as a legitimate work activity Allow TANF recipients to combine work and school activities to meet the weekly

number of hours specified by the work requirement Make sure recipients are aware of available educational opportunities

Work requirements Given the demands of single parenting and recent declines in the economy, requiring

a 40-hour work week is not recommended. The current requirement of 30-35 hours per week appears to be working well and should be maintained. Requiring a full 40-hour week may discourage some recipients and could be counterproductive.

Recommendations to Illinois policymakers Restore child care funding in fiscal year 2003. Maintain child care subsidies as a high

spending priority. Increase take-up of the child care subsidy. Expand income eligibility cutoffs for Medicaid coverage for adults and extend the

provision of Transitional Medicaid Assistance to at least 12 months, if not longer, for TANF leavers. (Eliminate the 6-month re-certification.)

Extend KidCare to parents through the FamilyCare program. Raise the TANF monthly cash grant. This may help to decrease the high levels of

material hardship among TANF recipients. (Concurrent maintenance or expansion of existing work supports will ensure that work still “pays more” than TANF receipt.)

Ensure that TANF and non-TANF working poor families have access to emergency assistance funds. This will help to alleviate material hardship and may help parents to stay in the work force (e.g., through emergency transportation assistance).

Enact policies that promote affordable housing. Increase the earnings eligibility cutoff for exiting TANF. Ensure that families who

leave TANF due to earnings have an income above the Federal Poverty Line.

Second annual report from the Illinois Families Study 1

Introduction Purpose of this report Illinois, like most other states in the nation, is now four years into the major welfare reforms that were passed in 1996 and implemented in 1997. This report assesses both the success of the Illinois approach in achieving the formal goals of the reforms (Did the law do what it was meant to do?), and its impact on the welfare recipients whose behavior it was intended to change (Did it help or hurt the people it affected?). We do this with two audiences in mind. First, the Illinois Families Study was designed to meet the requirements of the Welfare Reform Research and Accountability Act (P.A. 90-74), a measure passed by the Illinois General Assembly in 1997 that required the Illinois Department of Human Services (IDHS) to appoint outside researchers to implement a six-year panel study of welfare reform in Illinois. The primary goal of the study, therefore, is to inform state legislators, IDHS administrators, and other state and local policymakers about the experiences of families and children in the Illinois welfare program. Second, in addition to informing local policymakers, this report contributes to the national discussion about whether to modify, expand, or maintain the current reforms. Congress must reauthorize the welfare reform legislation by October 1, 2002 and there will be much discussion and debate about where the nation should go from here. This report will describe the Illinois experience and suggest that this state’s approach could be something that other states might learn from and, with some modifications, adopt. In the pages that follow, we will describe the provisions of the Illinois TANF program and demonstrate why Illinois is an important state to examine. We will also set the context in which Illinois operates by describing the state’s previous reform efforts and economic and demographic trends, and by comparing the Illinois welfare reform experience to that of other Midwestern states. This context will help to frame the study’s findings and also to caution the reader to keep economic conditions and other policy changes in mind when trying to identify the effects of the 1996 welfare reform law. We then go on to look at five key questions: 1. Have the federal goals of welfare reform been achieved in Illinois?

a. Has welfare dependence decreased? b. Have work, marriage, and two-parent families been successfully promoted?

2. How are families doing? a. Are families better off when they work and/or leave welfare? b. Are families able to meet their children’s needs as they move from welfare to

work? c. How do current and former recipients feel about the welfare system and

policies? 3. How are the most vulnerable families faring?

a. Who is most likely to hit the time limit? b. What is the impact of sanctions? c. What kinds of obstacles to self-sufficiency do these families face?

4. Do welfare reform policies and work supports make work pay in Illinois?

2 Welfare Reform in Illinois: Is the moderate approach working?

5. What is the overall impact of Illinois’ moderate approach to welfare reform? Goals of welfare reform When Congress passed the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA), it replaced the federal entitlement program for low-income families with children known as Aid to Families with Dependent Children (AFDC) with Temporary Assistance for Needy Families (TANF) block grants. According to the PRWORA legislation, the overall goals of TANF are to: Provide assistance to needy families so that children may be cared for in their own

homes or in the homes of relatives End the dependence of needy parents on government benefits by promoting job

preparation, work, and marriage Prevent and reduce the incidence of out-of-wedlock pregnancies and establish annual

numerical goals for preventing and reducing the incidence of these pregnancies Encourage the formation and maintenance of two-parent families1

Although each of these goals addresses family formation issues, most of the content of the legislation actually focused on promoting work and decreasing welfare dependence. Illinois, like many other states, chose increased work and decreased welfare dependence as its top priorities for welfare reform, placing less emphasis on the family formation goals. It is important to remember that the national and Illinois approaches to welfare reform focused, from their inception, on reducing welfare dependency, and only secondarily on reducing poverty. The underlying assumptions of the reforms were that staying on welfare under the old rules and regulations did grave damage to recipients and violated the central tenets of the American value system. Work was not only to be encouraged, but required, and help was to be temporary for the overwhelming majority of recipients. Proponents of the new “work-first” system argued that if these goals were achieved, the lives of the poor would be improved and public support for social programs would be strengthened. Arguing that poverty is often caused by welfare dependency, they also predicted that poverty rates would be reduced by the new reforms. Key features of Illinois TANF and related programs The Illinois General Assembly crafted the state’s TANF plan during its 1997 session and the Illinois Department of Human Services (IDHS) began implementing the program in July 1997. The Illinois plan incorporated major features of the state’s earlier experiments with welfare reform, including Earnfare, Get a Job, WorkFirst employment programs, a “family cap,” and the Work Pays income disregard program. Key features of the Illinois TANF program and related services include:

1 Personal Responsibility and Work Opportunity Reconciliation Act of 1996, H.R. 3734, Section 401.


Time limits: 60-month lifetime limit on TANF cash assistance “Stopped clock” provision: the time limit clock stops for recipients who: Work at least 30 hours per week and still qualify for assistance (30 hours for

single-parent families and 35 hours for two-parent families) Are full-time students in a postsecondary degree program with a minimum

GPA of 2.5 (36-month limit; must be a single parent)2 Provide constant in-home care for a medically dependent child3 Provide care for a disabled child or spouse2

Income disregards (Work Pays program): two-thirds of earned income is

disregarded when determining benefit levels (For example, a parent earning $300 monthly receives a $100 decrease in her monthly TANF grant.)

Sanctions: three-step sanction process ending in full-grant sanction after three occurrences of non-cooperation or after three months in a first or second-step sanction

Family cap: no additional cash benefits for children born ten months or more after initial enrollment

Work requirements: 30-35 hours per week; broad definition of work-related activities, including community service, substance abuse treatment, domestic violence counseling, foster parenting, and postsecondary education

Cash benefits: maximum monthly cash benefit of $377 for a family of three in urban areas (unchanged from AFDC)

Child care subsidies: for all families at or below 50% of 1997 state median income, regardless of TANF status; parent co-pay required

Transitional Medicaid Assistance: for 6-12 months after leaving TANF (depending on income)

KidCare (S-CHIP): subsidized health insurance for children in families whose income is below at least 185% of the Federal Poverty Line (FPL)4

See Appendix A for more information about the Illinois TANF Program. Why is Illinois important to the national reauthorization debate? Signaling a profound shift from federal to state responsibility for welfare programs, PRWORA established TANF block grants, which allowed states a great deal of control over the design and implementation of their welfare programs. Each state is therefore a unique “laboratory” for examining different approaches to welfare reform. Illinois is an important state to examine for several reasons. First, in August 1996 the state’s AFDC caseload was the fourth largest in the nation,5 making it home to a large proportion of the nation’s welfare recipients. Second, Illinois welfare reform policies are considered to be moderate, or even generous, in comparison to other states, providing an

2 This rule was not implemented until January 1999. 3 This rule was not implemented until July 2001. 4 There are several different KidCare programs with differing levels of eligibility. 5 U.S. Department of Health and Human Services (www.acf.dhhs.gov). By 2001, the Illinois TANF caseload was the 9th largest in the nation.


excellent example of the “middle of the road” approach to reform with a mix of work incentives and penalties for non-cooperation. Illinois was ranked tenth highest nationwide by the Tufts Scale, a tool designed to indicate which state welfare policies were most likely to strengthen the economic well-being of recipient families.6 Significant investments in child care subsidies, generous income disregards, and a provision that stops the time limit clock for TANF recipients who are working at least 30-35 hours per week helped Illinois to achieve this high ranking. Illinois’ low benefit levels, a “family cap,” and a full-grant sanction for non-cooperation, however, indicate that the Illinois TANF program also includes many of the deterrents and penalties common to other states. Third, both TANF caseloads and child poverty rates fell faster in Illinois than in the nation overall in the mid-1990s, while unemployment rates have been similar to the national average. Between 1996 and 2000 the Illinois AFDC/TANF caseload declined by 61%, compared to 49% nationwide.7 Between 1993 and 1998, the child poverty rate fell by 34% in Illinois, compared to a decline of only 17% nationally.8 (Since 1998, however, the Illinois child poverty rate has begun to increase. See Figure 1.) Like the rest of the nation, Illinois enjoyed relatively low unemployment rates in the late 1990s. Indeed, the Illinois unemployment rate was slightly below the national rate from 1994 to 1997. Illinois unemployment rates have ranged from a high of 7.6% in 1992 to a low of 4.4% in 2000.9 The findings of this study therefore reflect the effects of welfare reform in a large state with moderate policies that has experienced positive trends in caseload and poverty reduction, along with a strong economy. We infer, therefore, that the outcomes reported here reflect a “better case scenario” for welfare reform that is somewhat generalizable to the nation as a whole, but more so to states with moderate or incentive-based welfare policies and favorable economic climates.

6 Center on Hunger and Poverty, Tufts University, 1998. 7 U.S. Department of Health and Human Services (www.acf.dhhs.gov/news/stats/caseload.htm) 8 National Center for Children in Poverty, 2000. 9 Illinois Department of Employment Security. (http://lmi.ides.state.il.us)


Figure 1: Child poverty, AFDC/TANF caseloads, and unemployment in Illinois, 1990-2000

23.1%

17.4%

4.4%

7.6%

53,280 cases

208,646 cases

0.0

5.0

10.0

15.0

20.0

25.0

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Chi

ld P

over

ty a

nd U

nem

ploy

men

t Rat

e

-

50,000

100,000

150,000

200,000

250,000

AFD

C/T

AN

F C

asel

oad

Child Poverty Rate Unemployment Rate AFDC/TANF Caseload (as of September)

Sources: U.S. Census Bureau (child poverty rate); Illinois Department of Human Services, calculations by Chapin Hall (AFDC/TANF caseload); Illinois Department of Employment Security (unemployment rate) Welfare in Illinois during the 1990s: Historical and policy context Historically, Illinois has taken a somewhat conservative approach to welfare programs, with taxpayers and policymakers reluctant to spend state funds on programs for the poor.10 Compared to neighboring Midwestern states, Illinois’ AFDC/TANF benefit levels and spending commitments have been low, indicating that these programs generally have been a low spending priority for the state. During the 1980s and the early 1990s, the value of monthly public assistance payments in Illinois declined greatly. By 1995, monthly AFDC payments in Illinois accounted for just 40% of the Standard of Need11. In 1994 the Illinois AFDC benefit amount was below other Midwestern states (IA, MI, MN, WI), and the size of the subsidy was lower than all but Ohio and Missouri in terms of its proportion of the Standard of Need (see Appendix B). The absolute increase in Illinois AFDC benefits between 1970 and 1994 (58%) was the lowest

10 Lewis, George, and Puntenney, 1995. 11 Each state sets its own Standard of Need for determining AFDC eligibility. The standard referred to the amount of money a family needs to live, by family size, for that particular state. (Administration for Children and Families, www.acf.dhhs.gov/programs/afdc/afdc.txt)


percentage increase among Midwestern states, as well as being far below the national average increase of 100%12. Like most other states during the pre-PRWORA period, Illinois instituted a number of reforms to its AFDC program and implemented several demonstration projects designed to promote employment among welfare recipients. Examples of these efforts include Project Chance, the state’s Job Opportunities and Basic Skills Training (JOBS) program; Earnfare, Get a Job, and Work First employment programs; and a “family cap” policy. Work Pays, an income disregard program designed to ease the transition from welfare to work, was first implemented in 1993 and later became a core component of the state’s TANF program. Unlike previous AFDC policies in which most earnings were deducted from the monthly grant, Work Pays reduces assistance by $1 for every $3 earned. A moderate approach to reform: Illinois and its regional context When compared to policies in the other five upper Midwest states (Indiana, Michigan, Minnesota, Ohio, and Wisconsin), the Illinois TANF program appears to be somewhat moderate. Like Minnesota and Wisconsin, Illinois opted for the 60-month time limit, while Michigan imposed no time limit at all and Indiana (24 months) and Ohio (36 months) implemented shorter limits. A comparatively generous Illinois TANF policy is the “stopped clock” provision for recipients who meet specific work requirements; no other upper Midwest state currently provides this benefit. Illinois joins Indiana and Wisconsin as the only “family cap” states in the region. As mentioned above, Illinois has low benefit rates. After combining TANF cash payments and food stamps, only Indiana and Ohio have lower overall benefit levels than Illinois. The combined benefit leaves a family of three at 59% of the Federal Poverty Line (FPL) in Illinois, compared to a low of 52% in Indiana and a high of 76% in Wisconsin (see Table 60, page 92). The generous Illinois earnings disregards help to mitigate this disparity somewhat for working families, although the earnings level at which families are no longer eligible for welfare is still below the poverty line in Illinois (and in all other states in the region, with the exception of Minnesota). An Illinois family of three, therefore, must exit TANF once its earnings reach 93% of the FPL, compared to 115% in Minnesota, 97% in Indiana, and 55% in Wisconsin (see Table 60, page 92). Illinois is also comparatively moderate in its provision of other work supports. Along with all the other upper Midwest states, Illinois provides up to 12 months of Transitional Medicaid Assistance (TMA), although several other states in the nation have extended this benefit for up to 24 or even 36 months.13 For children in most age groups, income eligibility standards for Medicaid and the State Children’s Health Insurance Program (S-CHIP, or KidCare) are comparatively low in Illinois. The family income cut off for Medicaid for children ages 1-18 years in Illinois is 133% FPL, for example, compared to 275% in Minnesota and 200% in Ohio (see Table 59, page 89). Illinois has heavily funded child care subsidies in recent years. Although the income eligibility level for this subsidy is comparatively low in Illinois (160% FPL, or 50% of the 1997 state median 12 Lewis, George, and Puntenney, 1995. 13 Welfare Information Network (www.welfareinfo.org)


income), the program has several advantages. The child care subsidy is universally available; eligibility is tied to income, not to participation in TANF. There are no waiting lists for the subsidy, and it is not time limited (see Table 59, page 89). Unlike Michigan and Ohio, Illinois has a state Earned Income Tax Credit (EITC), although the credit is non refundable and relative limited (5%), making it less progressive than the refundable credits available in Minnesota (33%) and Wisconsin (4%-43%; see Table 59, page 89).14 (These percentages represent the percent of the federal EITC refund.) (See Appendix B.) Compared to neighboring states, Illinois provides TANF recipients with unprecedented access to postsecondary education. Single parents can attend school full-time for up to 36 months, while working toward a 2-year or 4-year degree. All other upper Midwest states place much tighter restrictions on post-secondary education (Indiana, Michigan, Minnesota, and Ohio), or do not allow it at all (Wisconsin). (See Appendix B.) In sum, when compared to neighboring states, Illinois welfare and work support policies range from being highly supportive or “generous” (“stopped clock” provision, generous earnings disregards, and broad access to postsecondary education) to being fairly restrictive (“family cap,” low benefit rates, and relatively low Medicaid and S-CHIP income eligibility cut offs). The overall effect is a “middle of the road” approach to welfare reform that provides valuable supports and incentives for some parents, but still leaves many families below the poverty line when they exit welfare. It is important to note that there have been dramatic shifts in the Illinois Department of Human Services’ TANF expenditures over the past few years. Funds have been shifted away from cash benefits and towards support services, such as the child care subsidy. In 1996, for example, cash assistance made up 66% of all AFDC spending in Illinois. By 2000, only 27% of TANF spending went directly to cash benefits. In comparison, child care represented 38% of TANF spending in 2000, up from 13% of AFDC spending in 199615. AFDC/TANF caseload dynamics in Illinois during the 1990s In an analysis of the Illinois AFDC/TANF caseload from 1990 to 1999, researchers from the Chapin Hall Center for Children at the University of Chicago examined the trends in total caseload, entries, exits, duration of time on welfare, recidivism, and earnings for recipients during the pre-Work Pays (1990-93), post-Work Pays/ pre-TANF (1993-97), and post-TANF (1997-99) periods.16 The average time spent on AFDC in the pre-Work Pays period was only 9 to 10 months, although recidivism was fairly high, with about one-quarter of those who left returning to the rolls within 4 to 5 months. During the 1990s, duration of time spent on welfare declined from a median of 11 months in 1994 to 7 months in 1999. Recidivism also declined during this period, indicating that the implementation of Work Pays and TANF, along with lower unemployment rates, were associated with decreases in welfare dependency in Illinois. Although marked by 14 Center on Budget and Policy Priorities, 2001. 15 WELPAN, January 2002. 16 Lee, Goerge, and Dilts, 2000


occasional dips, exits from AFDC/TANF were relatively stable while entries to the program declined significantly since late 1997. As a result of decreased duration on welfare, lower recidivism, and fewer recipients entering the program, the overall AFDC/TANF caseload saw an unprecedented decline of 75% from about 200,000 in 1994 to under 50,000 in 2001 (see Figure 2). Figure 2: Illinois AFDC/TANF caseload, entries, and exits (1990-2001)

0

50000

100000

150000

200000

250000

Jan-9

0

Aug-90

Mar-91

Oct-91

May-92

Dec-92

Jul-9

3

Feb-94

Sep-94

Apr-95

Nov-95

Jun-9

6

Jan-9

7

Aug-97

Mar-98

Oct-98

May-99

Dec-99

Jul-0

0

Feb-01

Cas

eloa

d

0

5000

10000

15000

20000

25000

Entr

ies

and

Exits

Caseload Exits (6-month moving average) Entries (6-month moving average)

Source: Illinois Department of Human Services, calculations by Chapin Hall The same study also found that after the implementation of Work Pays in 1993, earnings among those on welfare steadily increased. The average monthly earnings for the 1991 AFDC entry cohort were $47, compared to $9617 for the 1999 TANF cohort. A companion study of welfare leavers (exit cohorts) in Illinois from 1995 to 1999 concluded that recipients were increasingly more likely to have earnings when they left assistance in the late 1990s, but that earnings remained fairly low and have not led to an improvement in overall economic well-being for these families.18 Overall, the Illinois AFDC/TANF caseload declines have been somewhat more dramatic than other upper Midwest states. Focusing on the period between 1994 and 2000, Wisconsin led the way with a 78% caseload decline. Illinois and Michigan followed,

17 Reported earnings are in 1999 dollars and are adjusted for inflation using the Consumer Price Index (CPI). 18 Lee, Goerge, and Dilts, 2001


with each state experiencing a 69% decline in the AFDC/TANF caseload during that time period. Declines in Minnesota (38%), Indiana (49%), and Ohio (63%) were slower. In addition to rapid caseload declines, Illinois has been recognized for its performance in other areas. The state received performance bonuses from the federal government for reductions in out-of-wedlock births and for employment retention among TANF recipients. Other demographic characteristics of Illinois It is important to note that the racial makeup of the Illinois TANF caseload is markedly different from the nation as a whole. While the average state TANF caseload was 38.5% African American in 1999, 72.3% of Illinois TANF recipients were African American.19 This difference is reflected in the sample for this study. Additionally, Cook County (Chicago) is home to a very high proportion of the state’s TANF caseload. In 1999, Cook County residents made up about 75% of the Illinois TANF caseload.20 The state welfare population is therefore heavily urban, although this study sample also includes rural and smaller urban counties. Illinois tends to score average to below-average on indicators of child well-being. Illinois ranked 30th in the nation in the 1997 Kids Count composite score, which includes a broad range of child indicators such as child poverty, infant mortality, and the teen birth rate.21 Child health insurance rates, however, are slightly higher in Illinois. In 1998, 15% of low-income children living in the United States did not have health insurance, compared to 13% of low-income children in Illinois.22 For additional background information about Illinois, see Appendix C.

19 U.S. Department of Health and Human Services, 1999. (www.acf.dhhs.gov) 20 Voices for Illinois Children, 1999. 21 Annie E. Casey Foundation, 2000. 22 Annie E. Casey Foundation (www.aecf.org)


Objectives and research design Objectives The primary goal of the Illinois Families Study (IFS) is to inform legislators, state program administrators, social service providers, advocates, and other policymakers about the experiences of families and children in the state’s TANF program. To do so, the study will follow a representative sample of families from nine Illinois counties over six years to document how those families respond to the changes brought about by TANF. More specifically, the Illinois Families Study aims to inform policymakers by fulfilling three major objectives:

1) Broaden the scope of state welfare reform research to include key indicators of well-being (e.g., health, life events, hardships), in addition to employment outcomes and welfare-use patterns

2) Describe the impact of and need for government-funded support services 3) Describe changes in workforce attachment, economic status, and family well-

being over an extended period of time A more detailed description of the study’s background and purpose is available in the first-year report, Work, Welfare, and Well-Being: An independent look at welfare reform in Illinois (available at www.northwestern.edu/ipr/research/IFS.html). Sample and sample selection A sample of 1,899 TANF grantees was selected from the 1998 welfare caseloads in nine Illinois counties: Cook, Knox, Stark, Marshall, Fulton, Peoria, Woodford, Tazewell, and St. Clair. The nine counties were selected to ensure variability in regional representation, given that the majority of Illinois welfare recipients reside in the metropolitan Chicago area (Cook County). The eight “non-Cook” or “Downstate” counties included in the study represent mid-size urban, small urban, and rural areas. The combined welfare population of all nine counties represents approximately three-quarters of the statewide welfare caseload. The sample is stratified by region: half of the sample members were randomly selected from Cook County, and half were randomly selected from the combined Downstate counties. This ensured that a sufficient number of sample members were included from the Downstate regions (which enhances the reliability of the statistics generated for these regions). However, this design also makes the sample less representative of the actual distribution of welfare recipients in the combined nine counties. Therefore, base weights were constructed to correct for the over-representation of Downstate sample members. In this way, the statistics generated for this report can be interpreted as representative of the population of welfare recipients from the nine study counties. A more detailed description of the IFS study design and sampling strategy is available in Work, Welfare, and Well-Being: An independent look at welfare reform in Illinois (available at www.northwestern.edu/ipr/research/IFS.html).


Data sources The IFS draws upon annual survey data, a wide range of administrative records, and qualitative research. This report focuses primarily on findings from two waves of interviews and selected administrative records. Table 1 displays the data sources that are drawn upon in this report. Table 1: IFS data sources (included in this report) Survey data Time period available Wave 1 IFS interview (“1999-00 interview”)

Interviews conducted: November 1999-September 2000

Respondents were asked about activities “within last 12 months”

Wave 2 IFS interview (“2001 interview”)

Interviews conducted: February 2001-September 2001

Respondents were asked about activities “since your last interview”

Administrative records TANF, Medicaid enrollment, and food stamp use (IDHS)

September 1998- February 2001

Child care subsidies (IDHS)

September 1999 and September 2000

Employment and wages (IDES, Unemployment Insurance records)

September 1998- March 2001

More detailed descriptions of each data source are provided below. Survey data The core of the IFS is a longitudinal panel study that will track the same group of families for six years. These surveys include questions about a number of topics, ranging from the characteristics of the current or most recent job, other types of work activities (e.g., job training, education), income, and welfare experiences to housing characteristics, family structure, life events, physical health, and other indicators of well-being (e.g., depression, self-efficacy, etc.). Most interviews are conducted in person, usually in the respondent’s home. All survey data collection is managed and conducted by Metro Chicago Information Center (MCIC). The first of the annual surveys was administered between November 1999 and September 2000. The second annual survey was administered between February 2001 and September 2001. The average length of time between surveys was 13 months, with a range of 6 to 26 months. This is important to understand when analyses representing “changes” between survey waves are presented, since not every respondent has an equal time interval between interviews. In the first survey (1999-00), questions were asked


about the current month or the “previous 12 months,” while in the second survey (2001), the same questions were asked about the current month or the time period “since the last interview.” This means that the “reference” time periods are not perfectly consistent across surveys, although most respondents were interviewed reasonably close to one year after their initial survey; three-quarters received their second interview within 8 to 16 months after their first interview. Differences in interval lengths translate into shorter versus longer “risk” or “exposure” periods. In other words, any given respondent may have a longer (or shorter) time period between surveys for certain events to occur (e.g., getting a job, losing a job, getting married, giving birth, etc.) compared to any other respondent. Administrative data The IFS survey data is also linked to administrative data from the Illinois Department of Human Services (IDHS), the Illinois Department of Employment Security (IDES), and the Illinois Department of Children and Family Services (IDCFS) using a probabilistic linking process.23 The primary administrative data sources for this report are the Illinois Department of Human Services (welfare benefits and support services) and the Illinois Department of Employment Security (quarterly information on employment status and wages, as reported by employers). Specific administrative information used for selected analyses in this report include: Welfare benefit data: Information on the receipt of cash TANF benefits and food stamps and Medicaid enrollment for September 1998 through January 2001. Work support data: Data on child care subsidies from the Illinois Department of Human Services for September 1999 and 2000. Employment and wage data: Wages are reported quarterly to IDES by employers who pay unemployment insurance on behalf of their employees. Quarterly information on employment and earnings is available from September 1998 through March 2001. Ninety-three percent of the 1999-00 survey respondents (N=1,261) provided consent to access administrative data related to themselves and their children. In this report, analyses incorporating information from administrative data for the post-1999-00 survey time period relies on this subsample of respondents. Difference of means tests comparing consenters and nonconsenters, using administrative data provided as part of the original sampling frame, do not suggest that there is any systematic bias between these two groups in terms of employment or welfare use. Qualitative data The purpose of the IFS qualitative component is to gain a deeper understanding of the process families undergo as they transition from welfare to work. Researchers interviewed 69 Chicago-area participants in the summer of 2000, and completed follow- 23 A matching process that utilizes personal identifying information (e.g., Social Security numbers, names, birth dates, etc.) to link across data sources.


up interviews with 58 of those participants during the summer of 2001. Although not the focus of this report, three selected profiles from the qualitative component are included in Appendix F. Response rates Of the original 1,899 sample members selected for the IFS in 1998, 1,363 were interviewed for the 1999-00 survey, for a response rate of 72% (see Table 2). This initial response rate is consistent with and, in many cases, greater than the response rates of other large-scale surveys of welfare-involved or welfare-eligible individuals. Studies that have achieved higher response rates tend to be conducted within regions that are less urban than Cook County, and within only one region as opposed to several sites. In 2001, 1,183 of the 1999-00 respondents completed a second survey interview (see Table 2). This represents an 87% retention rate across survey waves. Throughout this report, survey data is generally presented for this group of 1,183 respondents who completed interviews at both waves. Table 2: IFS survey response rates Number of completed interviews Response rate 1999-00 survey 1,363 72% 2001 survey 1,183 87% Overall, 63% of the original 1,899 sample members were interviewed at both time points (1999-00 and 2001). To adjust for attrition between survey waves, the same method that was used to adjust for nonresponse in 1999-00 was applied to adjust for nonresponse in 2001 (using administrative data from the original sampling frame, as well as a wide range of survey data from 1999-00). These two nonresponse adjustment weights, in conjunction with the base weights that correct for the stratification design, are applied to the final 2001 sample. One advantage of the IFS is the ability to construct nonresponse weights using sampling frame data on a number of key factors, including employment, welfare use, food stamp receipt, Medicaid enrollment, and a range of demographic factors. In constructing non-response weights, an algorithm was applied, which produced a set of mutually exclusive groups that together accounted for the greatest amount of variation in response probabilities (Panel Study of Income Dynamics, 1992). This strategy minimizes the effects of bias in the sample related to differences in respondents and nonrespondents. Analysis strategy All of the analyses utilizing only administrative data refer to the original sample of 1,899 individuals. These analyses are unweighted, and therefore over-represent the Downstate sample members.


Most of the analyses presented in this report (particularly those analyses incorporating survey data), are conducted using the final 2001 sample (n=1,183). This affords an assessment of changes at the individual level, since comparisons of the full 1999-00 sample with the final 2001 sample do not refer to the exact same set of individuals. With the adjustment and base weights applied, however, the statistics generated should reasonably (although not perfectly) approximate the characteristics and circumstances of the original representative sample (n=1,899). Analyses that combine both survey and post-1999-00 information from administrative data are restricted to those sample members who consented to allow IFS researchers to access administrative information on themselves and their family members (n=1,123). These analyses are weighted to adjust for the stratification design, and nonresponse in both survey waves. The analyses assessing changes in characteristics and circumstances across survey waves are presented as point-in-time statistics for the 2001 full sample, and for the 2001 sample by region (Cook versus Downstate). Statistics representing the percentage point changes from 1999-00 to 2001 are also provided for the full sample and for each region. Unless a different statistical test is noted, difference of means (or difference of proportions) tests were conducted to identify statistically significant changes in the analysis tables.


Respondent characteristics This section describes the characteristics of the 1,183 IFS respondents who completed both the 1999-00 and 2001 surveys. These characteristics are reported as they were measured at the baseline survey (1999-00). Like all other survey results presented in this report, the data presented here are weighted to adjust for regional stratification and non-response. As shown in Table 3, after weighting is applied, the majority of the IFS sample is from Cook County (90%), reflecting the state’s large proportion of TANF recipients in that county. Ten percent of the sample is from Downstate, with about half from St. Clair County (East St. Louis) and half from the Peoria region (Peoria County and the surrounding rural counties). Table 3: Residence of respondents in 1999-00 (n=1,183)

Weighted Unweighted Number Percent Number Percent

Cook County 1069 90% 548 46% Chicago 935 79% 484 41% Suburban Cook County 134 11% 64 5%

Downstate 114 10% 635 54% St. Clair County 65 5% 363 31% Peoria County 32 3% 184 16% Rural Counties (Peoria region) 17 1% 88 7%

Table 4 displays basic demographic characteristics of the IFS respondents. On average, an IFS respondent was female (97%), African American (85%), and 31.6 years of age when she was first interviewed in 1999-00. Respondents had on average 2.5 children and the average age of children was 7.6 years. More than half (63%) gave birth as a teenager and most have never been married (63%). Fifty-nine percent were high school graduates or had received their GED. Thirteen percent indicated that they were of Hispanic, Latino, or Chicano origin. With the exception of marriage rates and race/ethnicity, there were very few demographic differences across the two study regions. A much larger proportion of Cook County respondents (66%) had never been married, compared to Downstate respondents (42%). Nearly one-third of Downstate respondents were White (32%), compared to a much smaller proportion of the Cook County sample (11%). Only 1% of the Downstate respondents were of Hispanic, Latino, or Chicano origin, compared to 14% of those in Cook County. The majority of respondents in both regions were African American, although African Americans make up a larger proportion of the Cook County group (87%) than the Downstate group (67%).


Table 4: Demographic characteristics in 1999-00

1999-00 (n=1,183)

All Cook County

Downstate

Mean age 31.6 31.7 31.0 Female 97% 97% 97% Mean number of children 2.5 2.5 2.5 Mean age of children: Youngest child Oldest child All children (N=3,029)

5.2 9.6 7.6

5.2 9.6 7.6

5.7 9.8 7.6

Gave birth as a teenager 63% 63% 61% Marital status: Never married Divorced Separated Married Widowed

63% 12% 16% 7% 1%

66% 12% 15% 6% 1%

42% 16% 28% 13% 1%

High school graduate (including GED completion)

59% 58% 63%

Race: African American White Other

85% 13% 2%

87% 11% 2%

67% 32% 1%

Hispanic/Latino/Chicano 13% 14% 1% Non-respondents Table 5 presents selected characteristics of three groups of original IFS sample members (n=1,899): respondents who did not participate in any IFS interview; those who participated in only the 1999-00 interview; and those who participated in both the 1999-00 and 2001 interviews. There are some significant differences related to participation and/or continuation in the surveys, including gender, race, county of residence, history of TANF receipt, and current TANF, Medicaid, food stamps, and child care subsidy receipt. Work experience prior to the first survey was not associated with participation in the IFS. The above variables and others were used to construct the non-response weights for the 1999-00 survey data; when constructing non-response weights for the 2001 survey data, additional information from the survey was used, as well. In making adjustments for non-response, statistics generated in this report more accurately reflect the original 1,899 IFS sample members, who in turn are a representative group of 1998 welfare recipients from the nine study counties.


Table 5: Characteristics of sample members by survey response status Means and Proportions Characteristics:

Non-respondent (n=536)

Respondent in 1999-00 only

(n=180)

Respondent in 1999-00 and 2001

(n=1,183)

p-value

Age as of 9/98 30 31 30 Male gender 5% 6% 2% ***African American 70% 66% 74% *Hispanic 6% 9% 5% Cook County resident 51% 69% 46% ***Previous work experience as of 9/98

69% 66% 69%

Cumulative months on TANF as of 9/98

69 71 77 ***

Receiving TANF in 9/99

47% 62% 61% ***

Receiving Medicaid in 9/99

69% 82% 84% ***

Receiving food stamps in 9/99

55% 69% 71% ***

Received child care subsidies in 12/99

19% 31% 33% ***

***p<.001, **p<.01, *p< .05 (statistical significance for differences between time periods; Chi-square and ANOVA tests) Source: Illinois Department of Human Services (administrative records)


Have the main goals of PRWORA been realized in Illinois? Four of the primary goals of the federal welfare reform legislation, PRWORA, were to: Increase work Decrease welfare dependence Promote marriage Reduce “out-of-wedlock” births

In this section we draw upon administrative records and survey data to determine to what extent these four goals have been realized in Illinois. Work and welfare State administrative records for the original IFS sample (n=1,899) indicate that between 1998 and 2001, work and earnings increased slightly for the IFS sample, while TANF use dropped sharply. Figure 3 displays the percent of the sample who were working between the fourth quarter of 1998 and the first quarter of 2001. Work increased slightly in both Cook County and the Downstate region during this time period, although there was a slight decline in both regions after 2000. Overall, 46% of the sample was employed in the fourth quarter of 1998 and 49% were employed in the first quarter of 2001. The proportion of the sample that was working peaked at 53% in the fourth quarter of 1999 and again in the fourth quarter of 2000, and then began dropping in 2001, perhaps reflecting the downward trend in the economy or seasonal effects. Table 6 shows the duration of workforce participation. Eighteen percent of the sample had worked in all of the 10 quarters that were observed (1998 Q4-2001 Q1). Twenty percent of the sample had not worked in any of these quarters, indicating that 80% of the sample had worked at some point during this time period.


Figure 3: Trends in work, 1998-2001 (percent of IFS sample members with reported earnings; n=1,899)

52%50%

46%43%

49%46%

0%

10%

20%

30%

40%

50%

60%

1998Q4 1999Q1 1999Q2 1999Q3 1999Q4 2000Q1 2000Q2 2000Q3 2000Q4 2001Q1

Downstate

Cook County

All

Source: Unemployment Insurance wage data (administrative records) Table 6: Number or quarters worked, 1998-2001 (percent of IFS sample members with reported earnings; n=1,899) All

(n=1,899) Cook County

(n=947) Downstate

(n=952) No work (in any quarter, 1998Q4- 2001Q1)

20% 22% 18%

Worked in 1-3 quarters

18% 20% 17%


21% 20% 22%


23% 23% 22%

Worked all 10 quarters

18% 15% 21%

Source: Unemployment Insurance wage data (administrative records) Earnings rose between the fourth quarter of 1998 and the first quarter of 2001 for employed IFS sample members (see Figure 4). Overall, median quarterly earnings rose a total of $883, from $1,748 in the fourth quarter of 1998 to $2,631 in the first quarter of


2001. With the exception of the first quarter shown, Cook County sample members earned more than their downstate counterparts. Quarterly earnings for the overall group peaked at $2,715 in the fourth quarter of 2000, and then dropped to $2,631 in 2001, again reflecting the downturn in the economy or seasonal effects. Figure 4: Trends in reported earnings, 1998-2001 (median quarterly earnings for those with earnings in each quarter)

$1,748 (All)

$2,631

$1,785 (Downstate) $2,372

$1,696 (Cook)

$2,869

$-

$500

$1,000

$1,500

$2,000

$2,500

$3,000

$3,500

1998Q4 1999Q1 1999Q2 1999Q3 1999Q4 2000Q1 2000Q2 2000Q3 2000Q4 2001Q1

All

Downstate

Cook County

Source: Unemployment Insurance wage data (administrative records) Mirroring statewide trends, Figure 5 displays the sharp decline in TANF use among adults in the IFS sample. In September 1998, 96% of the sample was receiving TANF.24 By February 2001 only 24% of the sample was still on TANF. Declines in Medicaid enrollment and food stamp use were less pronounced, resulting in more than half of the sample members receiving these benefits in 2001. (Please note that the data presented in this section pertains only to adults. Medicaid for children and KidCare are discussed on page 44. Food stamps and Medicaid for adults are also discussed further on pages 43, 54, and 59-60.)

24 All IFS sample members were receiving TANF at some point between September and November 1998. The TANF rate for September 1998 is not 100% because sample members could have been receiving TANF in October and/or November 1998, but not in September 1998.


Figure 5: Trends in TANF and food stamp use and Medicaid enrollment among adults, 1998-2001 (number of IFS sample members active for each program, by month; n=1,899)

96% (TANF)

24%

61%

97% (Medicaid)

59%

90% (Food Stamps)

0%

20%

40%

60%

80%

100%

120%

TANF Grant

Medicaid

Food Stamps

Source: Illinois Department of Human Services (administrative records) Table 7 displays the changes in benefit packages, that is, the combination of cash (TANF), food stamp, and Medicaid benefits that sample members received. The proportion of sample members receiving none of these benefits rose steadily, from 18% in September 1999 to 30% in February 2001. Conversely, receipt of the entire package of benefits (TANF, food stamps, and Medicaid) dropped from 52% in September 1999 to 23% in February 2001. Other benefit packages remained relatively stable, although many more families were receiving Medicaid and food stamps (with no TANF) by September 2000. Overall, 27% of the sample members were receiving Medicaid and food stamps (with no TANF grant) in February 2001.


Table 7: Changes in benefit packages for adults, 1999-2001 (Percent of IFS sample members active for each program, by month; n=1,899) September 1999 September 2000 February 2001 No benefits 18% 26% 30% TANF, Medicaid, and Food Stamps

52% 30% 23%

Medicaid only 11% 12% 10% Food Stamps only 2% 7% 9% Medicaid and Food Stamps (no TANF)

14% 25% 27%

Other combinations of benefits

4% 1% 1%

Source: Illinois Department of Human Services (administrative records) The preceding analysis relied upon administrative data. The next set of tables (8-12) and figures (6a-6d) displays IFS survey data, and therefore relies upon self-reports and represents “points in time,” rather than trends over multiple months and years. The survey results are also weighted to adjust for regional stratification and non-response. As indicated in the administrative data above, responses from the 1999-00 and 2001 IFS surveys show a slight increase in the proportion of respondents who were working, and a much sharper decline in TANF use. The percent of survey respondents who were working at the time of the second interview increased 2 percentage points between 1999-00 and 2001, while the percent receiving TANF dropped 21 percentage points (see Table 8). Overall, in 2001, 53% of respondents were working and 31% were receiving TANF. Downstate respondents were more likely to be working (66% vs. 52% for Cook County), and to be off TANF (81% vs. 68% for Cook County). Table 8: Work status and TANF use

2001

Change from 1999-00 to 2001

All (n=1,182)

Cook County

(n=1,070)

Downstate (n=112-113)

All Cook County

Downstate

Working 53% 52% 66% +2% +1% +1% Not working 47% 48% 34% -2% -1% - 1% Receiving TANF 31% 32% 19% -21%*** -21%*** -29%***

Not receiving TANF 69% 68% 81% +21%*** +21%*** +29%***

***p<.001 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Table 9 displays the four combinations of work and welfare. Almost half of the entire sample were relying on work only (43%), while 20% were only receiving TANF. A small proportion (10%) was combining work and TANF, while almost one-quarter reported that they were neither working nor receiving TANF (27%). There was


statistically significant change in all four of the groups between 1999-00 and 2001. There were large increases in the work-only (+13 percentage points) and no-work/no-TANF (+10 percentage points) groups, and large decreases in the work-and-TANF (-11 percentage points) and TANF-only groups (-12 percentage points). While the increase in the work only group is somewhat encouraging, the similar increase in the no-work/no-TANF group is troubling. There were notable regional variations for these findings. In 2001, the two regions did not vary much in the work-and-TANF group or the no-work/no-TANF group. Downstate respondents (56%) were more likely than Cook County respondents (41%) to be relying on work only, while those in Cook County (22%) were more likely than those Downstate (9%) to be relying on TANF only (see Table 9). Downstate (25%) and Cook County (27%) had similar proportions of respondents in the no-work/no-TANF group. Table 9: Work and TANF use patterns

2001


All (n=1,182)

Cook County

(n=1,070)

Downstate (n=112)

All Cook County

Downstate

Work and TANF 10% 11% 10% -11%*** -10%*** -21%***

Work only 43% 41% 56% +13%*** +11%*** +22%***

TANF only 20% 22% 9% -12%*** -11%*** -8%** No work/ no TANF

27% 27% 25% +10%*** +10%*** +7%

***p<.001, **p<.01 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Figures 6a-6d display the changes in work and welfare status for individuals between 1999-00 and 2001. Overall, the most prevalent transitions are to the “work-only” group in 2001. Most in the “work-only” group in 1999-00 (72%) maintained their “work-only” status in 2001, and almost half of those who were combining work and TANF in 1999-00 (44%) made this transition to “work-only” by 2001. The majority of those in the “TANF-only” group in 1999-00 were still not employed in 2001. Over one-third (38%) remained in the “TANF only” group in 2001, while 30% had neither work nor TANF to depend upon in 2001. Similarly, nearly half of the “no-work/no-TANF” group in 1999-00 (48%) were still without work or TANF in 2001.


Figure 6a: Transition from work-only status in 1999-00 (2001 status among those in the work-only group in 1999-00; this group made up 30% of respondents in 1999-00)

72%

16%

3%

8%

Work Only TANF Only No work/ No TANF Work + TANF

2001 Status

Source: IFS survey data Figure 6b: Transition from combined work and TANF status in 1999-00 (2001 status among those in the work-and-TANF group in 1999-00; this group made up 21% of respondents in 1999-00)

44%

14%

18%

24%

Work Only TANF Only No Work/ No TANF Work + TANF

2001 Status

Source: IFS survey data


Figure 6c: Transition from TANF-only status in 1999-00 (2001 status among those in the TANF-only group in 1999-00; this group made up 32% of respondents in 1999-00)

10%

30%

21%

38%


2001 Status

Source: IFS survey data Figure 6d: Transition from neither work nor TANF in 1999-00 (2001 status among those in the no-work/no-TANF group in 1999-00; this group made up 17% of respondents in 1999-00)

29%

15%

48%

7%


2001 Status



Family formation Between the 1999-00 and 2001 interviews, there was no change in the marriage rate. Overall, only 10% of respondents were married in 2001. There was no change for the overall sample and for Cook County in the marriage rate from 1999-00 to 2001, although it declined slightly (-2 percentage points) Downstate (see Table 10). There was a significant increase in non-marital cohabitation (+4 percentage points) during this time period, although only 8% of respondents reported this type of living arrangement. Table 10: Marriage and cohabitation

2001


All

(n=1,182)

Cook County (n=1070)

Downstate

(n=112)

All Cook County

Downstate

Currently married 10% 10% 15% 0 0 -2% Current non-marital cohabitation

8%

8%

10%

+4%***

+5%**

+3%

***p<.001, **p<.01 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data There was a decline in childbearing among the IFS sample from calendar year 1999 to 2000. Considering all of the children born to or fathered by respondents as of the 2001 interview, 9% of respondents said they had (or fathered) a child in 2000, down from 12% in 1999 (see Table 11).25 There was little difference observed in this trend when “out-of-wedlock” births were analyzed. Pregnancy rates (i.e., the percentage of respondents who reported being pregnant at each survey wave) declined slightly between survey waves. Table 11: Births

2000

Change from 1999

All

(n=1,161)

Cook County (n=1,050)

Downstate

(n=111)

All

Cook County

Downstate

Births to biological parents

9% 9% 9% -3%* -3%* -2%

Births to unmarried biological parents (“out-of-wedlock” births)

8% 9% 7% -3%* -2%* -2%

Pregnancies 5% 5% 5% -1% -1% -1% *p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data 25 Children’s birth dates were used to assess changes in birth rates from 1999 to 2000, as opposed to relying on survey self-reports of whether a child was born in the year prior to the 1999-00 interview, and in between the 1999-00 and 2001 interviews. We believe the former strategy yields the most accurate results.


Changes in births and pregnancies over this two-year period do not afford a true assessment of whether welfare policies have affected the trends depicted above, since we cannot determine whether the declines observed reflect natural “life course” patterns in fertility for this age range, or whether respondents based decisions about fertility on welfare policy incentives and disincentives. These findings should therefore be interpreted with caution. Given that the original IFS sample was a cross-section of the welfare population (and therefore over-represented “long-term” recipients of welfare, who were more likely to be older and to already have multiple children as of the first survey), the observed decline in births are somewhat expected. In order to assess the effectiveness of policies such as the “family cap,” more detailed analyses of the relationship between births, welfare status, and work status were performed. If it is assumed that those on welfare are more aware of TANF policies and therefore more likely to consider the “family cap” disincentive, we would expect welfare receipt to reduce the likelihood of future births. However, it may also be the case that those who are unable to leave welfare are also more disadvantaged, and for various reasons, more likely to have additional children irrespective of policy influences. In a separate analysis (see Table 12), we assessed whether births in 2000 were associated with several 1999 characteristics, including whether the respondent gave birth to a child in 1999, marital status, cohabitation, age, family size, and work and welfare status. In these bivariate analyses, cohabitation, being less than 25 years of age, and receiving welfare were all associated with a greater likelihood of a birth in 2000; being 35 or older and working were associated with a decrease in the likelihood of having a subsequent birth. Giving birth in 1999, being married, being 25-34 years of age were not associated with subsequent births.


Table 12: Births, by 1999 characteristics

1999 Characteristics

Gave Birth in 2000 (n=1,161)

Gave birth Yes No

13% 9%

Married Yes No

7% 10%

Cohabiting (unmarried) Yes No

17%* 9%

Under 25 years old Yes No

18%** 7%

25-34 years old Yes No

10% 9%

35 or older Yes No

2%** 13%

More than 2 children Yes No

9% 10%

Working Yes No

5%** 14%

Receiving TANF Yes No

11%* 7%

**p<.01, *p<.05 (statistical significance for differences between groups; Independent samples t-test) Source: IFS survey data Interestingly, when a multivariate analysis (not shown) was conducted predicting births in 2000 using the above 1999 characteristics, the only statistically significant findings were for family size (more children reduced the likelihood of having a birth in 2000), and giving birth in 1999 (which increased the likelihood of having a birth in 2000). In other words, when age and marital, cohabitation, work, and welfare status are held constant, smaller family sizes and recent births emerge as significant predictors of childbearing. This suggests that it may be a relatively small group of respondents continuing to have additional children under welfare reform, and that family planning decisions are hinged, at least in part, on the current family size. These findings suggest that trends in childbearing in the IFS are likely to reflect “life course” patterns, as opposed to childbearing disincentives embedded in policy, although more in-depth analyses are needed to further explore this question. Summary The purpose of this section was to answer the question, Have the main goals of PRWORA been realized in Illinois? The evidence presented here shows that decreases in welfare dependence were achieved, and to a slightly lesser extent, increases in work were also achieved. The family formation goals do not appear to have been met in Illinois. There has been no increase in marriage between surveys, and a significant increase in non-marital cohabitation. Although evidence shows that births have declined in the IFS


sample, this trend seems to be more heavily influenced by recent childbearing and family size, rather than welfare and work status.


Life after welfare reform: How are families faring? This section includes a wide range of indicators of child and adult well-being. We begin with income, poverty, and material hardship, and then go on to examine various dimensions of work, health, and child care that affect family well-being. We conclude this section by assessing satisfaction with welfare reform. Income and poverty Although poverty reduction was not an explicit goal of PRWORA, it is important to examine poverty and income for this population of families with children. The IFS survey asks respondents to report income received from employment and other benefits (including TANF) in the previous tax year, while family size is measured at the time of the interview. For this reason, the income ratio and the proportion of families living below the Federal Poverty Line (FPL) could only be computed from the 2001 survey data, which reports income from 1999 and 2000. Table 13 shows the mean and median annual income for respondents in 1999-00. It also shows the increases in the mean and median from 199826. Mean income increased significantly in both regions27. Overall, however, incomes were quite low. Three-quarters of all respondents reported income levels less than $17,500 in the previous tax year (not shown). For this sample, the mean income in 1999-00 was $11,812, while the median was only $8,750. Downstate respondents reported higher incomes than Cook County residents. The median annual income for Downstate respondents in 1999-00 was $12,250, compared to only $8,750 for Cook County respondents. To place these incomes in context, the mean and median annual incomes for the state of Illinois in 2000 were $70,398 and $55,19828, respectively. The Self-Sufficiency Standard for Illinois provides another point of comparison. This standard measures how much income is needed for a full-time working family of a given composition (number of adults and number and ages of children) in a given place to adequately meet its basic needs without public or private assistance. The standard takes into account housing, child care, food, transportation, heath care, taxes, and miscellaneous costs, as well as several tax credits. According to the Self-Sufficiency Standard, the adequate annual income for an adult with one infant and one preschoo-age child living in the Chicago Metropolitan Statistical Area (MSA)29 in 2001 was $40,884. For a comparable family in the Peoria-Pekin MSA it was $33,90330.

26 These numbers are not adjusted for inflation. The general rate of inflation in the Midwest from 1998-2000 was 5.6%. (Consumer Price Index, Bureau of Labor Statistics, 2002) 27 The measure of income in the IFS survey was improved for the 2001 interview. This change in the income measure could have artificially inflated the increase in income from 1998 to 1999-00. 28 Census 2000 Supplementary Survey (www.census.gov) 29 Excluding downtown and selected north side areas (high cost areas). 30 Pearce, 2001.


Table 13: Changes in mean and median annual income

1999-00 (tax year preceding

2001 interview)

Change from 1998 (tax year preceding 1999-00 interview)

Respondent earnings and other benefits combined

All (n=1,068)

Cook County (n=977)

Downstate (n=91)

All Cook County

Downstate

Mean annual income $11,812 $11,747 $12,510 +$4,337*** +$4,320*** +$4,526*** Median annual income $8,750 $8,750 $12,250 +$2,500 +$2,500 +$6,000 ***= p<.001 (Statistical significance for differences in mean between time periods; Chi-square test. Note: Income does not include EITC tax refunds. Numbers are not adjusted for inflation. Source: IFS survey data The data shown in Table 14 reflect the median income category reported by respondents for the previous calendar year31. This value was then divided by the number of children under the age of 18 that respondents were caring for in 1999-00, plus the respondent (e.g., if a respondent was caring for 3 children, the respondent’s reported income level was divided by 4). The annual income-to-family-member ratio was $2,917 overall. Almost three-quarters of the sample (76%) were living below the Federal Poverty Line in 1999-00. Reflecting the increased overall income for Downstate respondents (shown in Table 13), Downstate respondents also enjoyed a slightly higher income ratio ($3,125) compared to their Cook County counterparts ($2,917). Table 14: Income ratio and poverty

1999-00 annual income

All (n=1,151)


Downstate (n=108)

Income ratio (respondent’s annual income divided by the number of minor children in household, plus respondent)

$2,917

$2,917

$3,125 Percent of families below the Federal Poverty Line

76%

76%

73%

Note: The 2000 Federal Poverty Line was used because most respondents reported income from 2000. Source: IFS survey data

31 Incomes were recorded in ranges. In order to calculate means, medians, and ratios, the midpoint value of the range was selected.


Material hardship Between the 1999-00 and 2001 interviews there was an overall decrease in the number of families who reported experiencing any of the material hardships listed in Table 15, although the majority still reported some hardship. Fifty-six percent of respondents in 2001 reported experiencing any of these problems, compared to 64% in 1999-00, a statistically significant decrease of eight percentage points. Although there was no change for the overall sample in the number of families experiencing severe hardship (three or more of these problems), there was an increase for downstate families, from 19% in 1999-00 to 27% in 2001. Overall, utility shutoffs saw the biggest declines. The number of respondents reporting that their phone, gas, or electricity service had been shut off, for example, decreased significantly, although phone shutoffs remained a major problem, affecting 40% of respondents in 2001. Most housing-affordability problems increased. The proportion of families who were not able to pay their full rent or mortgage in one or more months increased by 13 percentage points, going from 24% in 1999-00 to 37% in 2001. Evictions also increased by 13 percentage points, from 4% in 1999-00 to 17% in 2001. There were no significant changes in unmet medical or dental needs.


Table 15: Material hardships experienced since last interview/ in past year

2001


All (n=1,182-

1,183)

Cook County

(n=1,070)


All Cook County

Downstate

Phone service turned off/ were without phone service for any reason

40% 40% 38% -8%*** -8%*** -4%

Could not pay full rent/ mortgage

37% 37% 37% +13%*** +12%*** +21%***

Evicted from home 17% 17% 18% +13%*** +13%*** +15%*** Respondent or child needed to see a dentist but couldn’t afford to go

15% 15% 18% -2% -1% -2%

Respondent or child needed to see a doctor but couldn’t afford to go

12% 12% 12% -2% -1% -2%

Respondent or child needed to fill a prescription but couldn’t afford it

12% 12% 15% -1% -1% +1%

Moved in with others to reduce expenses

6% 5% 13% -3%* -4%** +7%

Respondent had a family member/ friend move in to reduce expenses

6% 6% 5% +2%** +2% +1%

Gas/ electricity turned off

2% 2% 3% -12%*** -13%*** -7%

One or more of the above hardships

56% 56% 55% -8%*** -8%*** -3%

Three or more of the above hardships

23% 22% 27% 0 -1% +8%

***= p<.001; **= p<.01; *= p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data There were modest but significant declines in food insecurity and homelessness between the 1999-00 and 2001 interviews. Nine percent of respondents in 2001 reported that they “sometimes” or “often” did not have enough of the kinds of food they wanted to eat, down from 10% in 1999-00 (see Table 16). Homelessness declined from 7% in 1999-00 to 5% in 2001 (see Table 17). This includes families who stayed in a shelter, car,


abandoned building, “on the streets,” or temporarily (less than two weeks) doubled up with a friend or relative. The proportion of respondents reporting problems with housing conditions also declined significantly, from 36% in 1999-00 to 30% in 2001. There was no change in the overall proportion of families living in crowded housing conditions (more than two persons per bedroom). Seventeen percent of respondents reported crowded conditions in 2001, with Cook County families (18%) being much more likely than Downstate families (9%) to experience this problem. Table 16: Food insecurity since last interview/ in past year

2001


Family “sometimes” or “often”….

All (n=1,178-1,182)

Cook County (n=1,066-

1,070)

Downstate (n=112)

All Cook County

Downstate

Did not have enough of the kinds of food they wanted to eat.

9% 9% 7% -1%** -1%** -2%

Worried that children were not eating enough because there was not enough money for food.

10% 10% 8% -3%** -3%** -4%

***= p<.001; **= p<.01(statistical significance for differences between time periods; Chi-square test) Source: IFS survey data


Table 17: Homelessness and problems with housing conditions since last interview/ in past year

2001


All (n=1178-1182)

Cook County

(n=1066-1070)


All Cook County

Downstate

Homeless 5% 5% 5% -2%* -2%* +1% Lived in two or more places since last interview

23% 23% 26% -1% -1% +1%

High “crowdedness ratio” (more than two persons per bedroom)

17% 18% 9% 0 +1 0

Any problems with housing conditions (insects, plumbing problems, no heat, etc.)

30% 30% 29% -6%** -6%** -2%

**= p<.01; *= p<.05(statistical significance for differences between time periods; Chi-square test) Source: IFS survey data There was a significant increase in the number of respondents who “somewhat” or “strongly” agreed that they could “generally afford to buy the things we need,” from 61% in 1999-00 to 67% in 2001 (see Table 18). Other measures of perceived financial situation saw no change. Half of respondents felt their financial situation was better than it had been “in a long time” and 83% said they worried “a lot” about having enough money in the future. There was some regional variation, with downstate respondents reporting somewhat more positive perceptions of their financial situation. Twenty-two percent of all respondents said that they were saving some of their money (not shown; question only asked in 2001).


Table 18: Perceived financial situation

2001


Respondent “somewhat” or “strongly” agrees that….

All (n=1,182)

Cook County

(n=1,070)

Downstate (n=112)

All Cook County

Downstate

My financial situation is better than it’s been in a long time.

50% 49% 59% 0 0 0

I can generally afford to buy the things we need.

67% 66% 74% +6%*** +5%*** +5%

I worry a lot about having enough money in the future.

83% 83% 81% 0 0 +5%

***= p<.001(statistical significance for differences between time periods; Chi-square test) Source: IFS survey data There was a striking 20 percentage point increase in the number of respondents who said they went to a church or charity for clothes or financial help at some point in the past year or since their last interview (see Table 19). Twelve percent reported this kind of help in 1999-00, compared to 32% in 2001. There were very slight increases in food pantry, soup kitchen, and government crisis assistance use. One-third of respondents said they had to borrow money from friends or family to pay bills. Table 19: Use of charities and crisis assistance since last interview/ in past year

2001 Change from 1999-00 to 2001 All

(n=1,182-1,183)

Cook County

(n=1,070)


All Cook County

Downstate

Used a food pantry or soup kitchen

17% 17% 17% +1% +1% +2%

Went to a church or charity for clothes or financial help

32% 32% 30% +20%*** +20%*** +15%**

Received help from a government crisis assistance program

9% 8% 11% +2% +1% +1%

Had to borrow money from friends/family to pay bills

33% 33% 33% -2% -3% +2%

***= p<.001; **= p<.01(statistical significance for differences between time periods; Chi-square test) Source: IFS survey data


Job characteristics As a group, employed respondents made significant gains in hourly wages, although the overall median wage for 2001 was $8.00 per hour (see Table 20). The average wage increased by $1.12 between 1999-00 and 200132. Although wages increased in both areas of the state, regional differences persisted. Cook County respondents reported an hourly median wage of $8.08 in 2001, compared to $7.00 for downstate respondents. Table 20: Median and mean hourly wage at current job

2001 Change from 1999-00 to 2001 All

(n=589) Cook

County (n=519)

Downstate (n=70)

All Cook County

Downstate

Median $8.00 $8.08 $7.00 +$1.00 +$0.83 +$0.50 Mean $8.54 $8.67 $7.54 +$1.12 *** +$1.17 *** +$0.72 * Note: Statistical significance was only assessed for the mean hourly wage, not the median. ***= p<.001; *= p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Table 21 shows the change in hourly wages for individuals who were working in 1999-00 and 2001. Most saw some increase in wages (73%), although 17% reported a decrease in hourly wage. Table 21: Changes in hourly wage for individuals between 1999-00 and 2001 (among those employed in 1999-00 and 2001) Hourly Wage All

(n=399) Cook County

(n=346) Downstate

(n=54) Increased 73% 74% 72% Decreased 17% 16% 21% Stayed the same 10% 10% 7% Source: IFS survey data Paralleling wage increases, more working respondents were receiving employer benefits in 2001 than in 1999-00, although the majority still did not receive the benefits listed in Table 22. The largest gain was in employer-sponsored health insurance, which increased 10 percentage points, from 13% of workers in 1999-00 to 23% in 2001. Only 18% received a retirement program through their job in 2001 and 29% had paid sick days. 32 These numbers are not adjusted for inflation.


Table 22: Employer benefits received at current job (among workers)

2001


All (n=612-

635)

Cook County (n=547-

552)

Downstate (n=73-74)

All Cook County

Downstate

Health insurance for self

23% 24% 20% +10%*** +11%*** +9%*

Dental insurance for self

26% 26% 19% +5%** +4%* +4%

Retirement program 18% 18% 18% +5%*** +5%** +7% Paid sick days 29% 30% 28% +3% +3% +6% Unpaid sick days or personal leave days

29% 29% 28% +7%*** +7%*** +8%

***= p<.001; **= p<.01 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Among workers, there was a significant increase in the number of respondents working in the service sector, with an equally significant drop in manufacturing jobs (see Table 23). The proportion of respondents working in service jobs—which include health services, social services, and hotels/lodging—increased from 57% in 1999-00 to 64% in 2001. Meanwhile, manufacturing jobs dropped from 6% to 3%. Overall, services (64%) and retail trade (19%) were the most commonly reported industries in 2001. Table 23: Industry (of current job)

2001


All (n=623)

Cook County (n=549)

Downstate (n=74)

All Cook County

Downstate

Services 64% 64% 66% +7%*** +7%*** +11% Retail trade 19% 18% 21% -2% -2% -6% Transportation, communications, and other public utilities

9% 10% 3% +1% +1% -1%

Finance, Insurance, and Real Estate

5% 5% 4% 0 0 -1%

Manufacturing 3% 3% 4% -3%*** -3%*** -1% Other 1% 0% 2% -2%*** -2%*** -2% ***= p<.001 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data In the two-year period under study, there was a slight but significant trend towards more part-time work and less full-time work, although the majority of employed respondents


(59%) reported that they were working full-time in 2001 (see Table 24). The mean number of hours worked per week declined from 33 in 1999-00 to 32 in 2001, and the proportion working 30 or more hours per week declined from 75% to 71%. Downstate respondents tended to work more hours (mean=34 per week) than Cook County respondents (mean=32 per week). About half of all working respondents (51%) said they would like to work more hours, representing a 6 percentage point decline from 1999-00. Table 24: Full and part-time work

2001


All (n=591-

612)

Cook County (n=548-

552)

Downstate (n=74)

All Cook County

Downstate

Part-time (currently working 1-29 hours/week)

29% 30% 26% +4%* +5%* +4%

Full-time (currently working 30-40 hours/week)

59% 59% 59% -4%* -4% -4%

Currently working 41+ hours/week

12% 11% 15% 0 -1% 0

Mean number of hours worked per week

32 32 34 -1 hour -1 hour 0

Would like to work more hours

51% 50% 52% -6%** -7%** -2%

**= p<.01; *= p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Table 25 shows the changes in hours worked for individuals who were employed in 1999-00 and 2001. The majority (62%) continued as full-time workers. Slightly more individual workers decreased their hours from full-time to part-time (15%), compared to those who moved from part-time to full-time status (12%). Part of the decrease in hours is due to fewer hours worked among new workers who were working at the time of the 2001 interview, but not at the time of the 1999-00 interview. Sixty-four percent of these new workers were working full time in 2001, compared to 76% of those who were already working in 1999-00 (not shown).


Table 25: Changes in work status for individuals between 1999-00 and 2001 (among those employed in 1999-00 and 2001) All

(n=393) Cook County

(n=341) Downstate

(n=52) Went from part-time to full-time

12% 12% 12%

Went from full-time to part-time

15% 15% 10%

Continued as part-time

11% 11% 10%

Continued as full-time

62% 62% 68%

Source: IFS survey data There were no significant changes in the length of time it took respondents to get to work, including the time it took them to drop off their children at child care (when applicable). The mean number of minutes for the overall sample in 2001 was 43, while the median was 30 (see Table 26). It took Cook County residents much longer to get to work (mean=45 minutes) than Downstate residents (mean=25 minutes). Eighteen percent reported a one-way commute that was more than 60 minutes long. Table 26: Length of time it takes to get to work, including dropping off children for child care

2001


All (n=599)

Cook County (n=529)

Downstate (n=70)

All Cook County

Downstate

Median number of minutes

30 40 20 0 +3 0

Mean number of minutes

43 45 25 0 0 -3

More than 60 minutes 18% 20% 3% +1% +1% -2% Source: IFS survey data There were overall declines in job satisfaction between 1999-00 and 2001, although 73% of respondents in 2001 still reported that they were satisfied with their job (“very satisfied” or “somewhat satisfied”; see Table 27). The proportion of respondents saying they were “very satisfied” with their job dropped 10 percentage points, from 37% in 1999-00 to 27% in 2001. Downstate respondents reported greater satisfaction than their Cook County counterparts; 40% of Downstate workers were “very satisfied,” compared to 25% of Cook County workers in 2001.


Table 27: “How satisfied are you with your current main job?”

2001


All (n=620)

Cook County (n=546)

Downstate (n=74)

All Cook County

Downstate

Very satisfied 27% 25% 40% -10%*** -11%*** 0 Somewhat satisfied 46% 47% 44% +3% +4% +3% Somewhat dissatisfied 13% 13% 11% +4%** +4%** +1% Very dissatisfied 14% 15% 5% +3% +3% - 4% ***= p<.001; **= p<.01 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data In 2001, 41% of respondents said they received a promotion or pay raise from their current employer in the past year. This represents a 9 percentage point increase from 1999-00 when 32% of respondents reported a promotion or raise. Table 28: Received a promotion or pay raise from current employer in past year

2001


All (n=586)

Cook County (n=516)

Downstate (n=70)

All Cook County

Downstate

41% 40% 46% +9%*** +9%*** +3% ***= p<.001 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data In 2001, respondents had been at their current or most recent jobs for an average of 18 months, and they reported holding an average of 1.45 jobs since their last interview (see Table 29). Job retention increased slightly from 1999-00 to 2001. By 2001, respondents had held their jobs for an average of 3 more months, compared to 1999-00, and they held slightly fewer jobs.


Table 29: Job retention

2001


All (n=760-

771)

Cook County (n=677-

694)

Downstate (n=77-83)

All Cook County

Downstate

Average number of months at current or most recent job

18 months

18 months

19 months

+3 months

+3 months

+3 months

Average number of jobs held in past year/ since last interview (currently and recently employed)

1.45 jobs

1.43 jobs

1.66 jobs

-.09 jobs

-.06 jobs

-.18 jobs

Source: IFS survey data Physical and mental health There was no change in self-reported health status for respondents and a slight improvement in health status for their children. Overall in 2001, 22% of respondents described their health as “fair” or “poor,” compared to only 5% of children (see Table 30). Nineteen percent of respondents reported experiencing symptoms of depression within the past week, down 4 percentage points from 23% in 1999-00 (see Table 31). To measure depressive symptoms, respondents were asked questions related to mood and functioning in the past week33. Table 30: Overall health status of respondents and their children

2001 Change from 1999-00 to 2001 All Cook

County Downstate All Cook

County Downstate

Respondent: self-rated health status (n=1,182)

Good, very good, or excellent

78% 78% 81% 0 0 +4%

Fair or poor 22% 22% 19% 0 0 -4% Child: parent-rated health status (n=2,965)

Good, very good, or excellent

95%

95%

95%

+3%***

+3%***

+2%

Fair or poor 5% 5% 5% -3% -3% - 2% ***p<.001 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data

33 Depressive symptoms were assessed using a 12-item version of the Center for Epidemiological Studies Depression Scale (CES-D) developed by Ross, Mirowsky, and Huber (1983).


Table 31: Depression

2001


All (n=1,159)

Cook County

(n=1,047)

Downstate (n=111)

All Cook County

Downstate

Any depression 19% 19% 21% -4%** -4%* -6% Mild depression 4% 4% 5% 0 0 0 Moderate depression 7% 7% 8% -3%* -3%* -2% Severe depression 8% 8% 8% -1% 0 -4% Note: Respondents were asked to identify symptoms of depression experienced over the past week only. **= p<.01; *= p<.05 (statistical significance for differences between time periods; Paired t-test) Source: IFS survey data Health insurance Significant increases in employer-sponsored health insurance were not enough to offset a large decline in Medicaid receipt, resulting in an increase in the number of uninsured adults. Although Medicaid remained the most common form of coverage, Medicaid rates fell 11 percentage points, from 73% of adults in 1999-00 to only 62% in 2001 (see Table 32). Meanwhile, the proportion of respondents with health insurance through their employer rose 5 percentage points, from 8% in 1999-00 to 13% in 2001. The proportion of adults with no health coverage rose 6 percentage points, resulting in one-quarter of the sample being uninsured at the time of the 2001 interview. Only a fraction of respondents reported being covered through their spouse or partner or through some other source. Table 32: Respondents’ current health insurance coverage

2001


All (n=1,182)

Cook County

(n=1,070)

Downstate (n=113)

All Cook County

Downstate

Medicaid 62% 62% 56% -11%*** -11%*** -16%* No coverage 25% 24% 27% +6%*** +5%*** +9% Employer-sponsored 13% 13% 14% +5%*** +5%*** +6%* Spouse or partner’s plan

1% 1% 4% 0 0 +1%

Other 1% 1% 3% -1% -1% +2% ***= p<.001; *= p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Changes in health insurance status for children were more positive and less dramatic than for adults. Increases in KidCare and employer-sponsored coverage helped to offset declines in Medicaid, resulting in a decrease in the number of uninsured children (see Table 33). Medicaid was also the most common form of coverage for children, although it declined from 80% in 1999-00 to 78% in 2001. More children were covered by


KidCare (4%) or through an employer-based plan (7%) in 2001 than in 1999-00 (2% and 5%, respectively). Overall, 9% of children were uninsured in 2001, compared to 13% in 1999-00. Few children were covered by their parent’s spouse or partner, their father, or some other source of coverage. Table 33: Children’s current health insurance coverage

2001


All (n=1,170)

Cook County

(n=1,058)

Downstate (n=111)

All Cook County

Downstate

Medicaid 78% 78% 77% - 2%*** -1%***

-4%

KidCare 4% 3% 6% +2% +1% +3% No coverage 9% 9% 10% - 4%** - 4%** 0% Employer-sponsored 7% 7% 7% +2%* +2% +3% Spouse or partner’s plan

2% 2% 5% - 1%

0%

- 1%

Child(ren)’s father’s plan

2% 2% 3% N/A N/A N/A

Other 2% 2% 2% 0% 0% -1% Note: Respondents were asked if any of their children were covered by these health insurance options. Respondents were not asked about their child(ren)’s father’s plan in 1999-00. ***= p<.001; **= p<.01; *= p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Parents were asked if they or any of their children experienced a gap in health insurance coverage during the past year. Reflecting the trends in health insurance coverage in general, more parents reported gaps in 2001 than in 1999-00, while fewer children experienced gaps (see Table 34). Overall, 30% of parents and 19% of children experienced a gap in coverage in 2001. Parents had a significant increase in “long gaps” (four or more months without coverage), from 14% in 1999-00 to 21% in 2001.


Table 34: Gaps in health coverage for respondents and their children, during past year

2001


All

Cook County

Downstate

All Cook County

Downstate

Respondents (adults) (n=1,182)

Any gap in coverage 30% 30% 34% +2% +2% +7% Long gap (4+ months) 21% 21% 22% +7%*** +7% *** +7% Children (n=1,172) Any gap in coverage 19% 19% 20% - 4%** - 5%** +1% Long gap (4+ months) 11% 11% 12% +1% +1% +3% Note: Respondents were asked if any of their children experienced a gap in health insurance. **= p<.01; *= p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Child care Respondents were asked “What was the main type of child care arrangement you used for this child last week?” for each of their children under age 12. There was very little change in the main type of child care arrangements across surveys. Care from an extended family member (e.g., grandparent, aunt, etc.) or sibling (“informal care”) was by far the most common child care arrangement, representing the arrangements of 41% of all children in 2001 (see Table 35). The next most common form of care was with the respondent her/himself, which increased slightly to 38% in 2001. Eight percent of children overall were in center-based care in 2001 (daycare center, preschool, Head Start, or after-school program), with Downstate children (11%) being much more likely than Cook County children (7%) to receive this type of care. There was a slight increase in the stability of child care arrangements over the two-year period. Overall, 90% of children in 2001 had one or fewer child care arrangements during the previous 12 months (see Table 36).


Table 35: Main type of child care arrangement used last week

2001


All (n=2,117)

Cook County

(n=1,920)

Downstate (n=197)

All Cook County

Downstate

Daycare center, preschool, Head Start, or after-school program

8%

7%

11%

+1%

+1%

-2%

Extended family member or sibling (in child’s home or relative’s home)

41%

40%

43%

-1%

-1%

+1%

Nonrelative (in child’s home or other home)

7%

7%

6%

-2%*

-1%

-5%

Child stayed home alone

<1%

<1%

<1%

0

0

0

Child stayed with respondent’s partner/spouse

6%

6%

6%

+1%

+2%*

-2%

Child is always with respondent

38%

39%

31%

+1%

0

+7%

Other <1% <1% 2% 0 0 +1% ***= p<.001; **= p<.01; *= p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Table 36: “How many different child care arrangements did child have during past 12 months?”

2001


All (n=2,174)

Cook County

(n=1,977)

Downstate (n=197)

All Cook County

Downstate

0-1 90% 91% 85% +2% *** +2%*** +9%*** 2-3 8% 7% 10% - 3% *** - 3%*** -12%*** 4 or more 2% 2% 5% 0 0 +3% ***= p<.001 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Parents were asked if they had experienced any of a series of specific child care concerns during the past year or since their last interview. Results for parents of children under age 12 are displayed in Table 37. Overall, 23% of parents reported that they had at least one concern about child care. Quality (15%) and cost (11%) were the most commonly reported problems. There were significant decreases in the proportion of parents reporting these child care concerns, indicating that parents were more satisfied with their


child care options in 2001 than they were in 1999-00. This trend was driven by decreased concerns among Cook County parents, however, while the Downstate sample experienced no significant changes. Table 37: Child care problems experienced since last interview/ in past year (among parents of at least one child under age 12)

2001


All (n= 999)

Cook County (n= 906)

Downstate (n= 93)

All Cook County

Downstate

Concerned about quality

15% 16% 12% -4 -4 0

Provider too far away 7% 8% 5% -7** -6** 0 Provider not dependable

8% 7% 9% -7*** -9*** +1

Cost too much 11% 11% 13% -9*** -10*** +2 No relatives or friends who could help with child care

8% 8% 8% -8*** -8*** -3

Afraid caretaker might harm child or arrangement unsafe

7% 7% 4% -7* -8* -2

Couldn’t find child care during work hours

9% 9% 10% -8*** -9*** -2

Couldn’t arrange regular child care when needed

8% 8% 8% -8*** -9*** -3

Couldn’t arrange emergency child care when needed

9% 9% 7% -7*** -8*** -4

Other child care concerns

1% 1% 1% N/A N/A N/A

1 or more of the above problems

23% 23% 24% -12*** -13*** +1

***= p<.001; **= p<.01 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Information on use of child care subsidies is provided on pages 54-55. Satisfaction with welfare The majority of respondents who were still receiving TANF reported positive feelings about their welfare worker. Overall, 78% “somewhat” or “strongly” agreed that their welfare worker treated them with dignity and respect, up seven percentage points from


72% in 1999-00 (see Table 38). More respondents also reported that their worker took time to explain program rules. This increased 11 percentage points, from 68% in 1999-00 to 79% in 2001. Table 38: Satisfaction with current welfare worker: percent who “somewhat” or “strongly agree”

2001


All (n=529)

Cook County (n=496)

Downstate (n=33)

All Cook County

Downstate

Treated me with dignity and respect

78%

78%

81%

+7%

+6%

+9%

Takes the time to explain program rules

79%

80%

75%

+11%*

+10%

+12%

Note: Only those still on TANF were asked these questions in 2001; all 1,363 respondents were asked this question in 1999-00. ***= p<.001; **= p<.01; *= p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Support for welfare reform policies, already quite high in 1999-00, increased in 2001. The majority of respondents “somewhat” or “strongly” agreed that time limits (74%) and work requirements (92%) were a good idea (see Table 39). Overall, support for time limits increased 13 percentage points, while support for work requirements increased four percentage points. Respondents who were still receiving TANF did not differ significantly from those who were already off TANF when responding to these questions (not shown). Seventy-five percent of TANF leavers “somewhat” or “strongly” agreed that time limits are a good idea, compared to 71% of those still receiving TANF. Similarly, 92% of TANF leavers “somewhat” or “strongly” agreed that work requirements are a good idea, compared to 93% of those still receiving TANF.


Table 39: Attitudes about welfare reform: percent who “somewhat” or “strongly agree”

2001


It’s a good idea to…. All (n=1,180)

Cook County

(n=1,068)

Downstate (n=112)

All Cook County

Downstate

Limit the amount of time people can stay on [cash welfare or TANF].

74%

73%

77% +13%***

+13%***

+14%

Require people on [cash welfare or TANF] to find a job and work.

92%

91%

95% +4%***

+3%***

+6%

***= p<.001; **= p<.01; *= p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Respondents were also asked, “If the money and medical coverage were the same, do you prefer working or getting welfare?” Almost all respondents (96%) said they preferred working. Responses to this question did not vary significantly by region or between time periods.


Making work pay in Illinois: Use of work supports and other services Earnings disregards and stopping the clock The “stopped clock” is a unique provision of Illinois TANF that allows parents who work at least 30 hours per week and still qualify for TANF to stop their 60-month time-limit “clock” while they continue receiving TANF benefits. (A description of the “stopped clock” policy and its applications for employment, education, and care-giving responsibilities is provided on page 3.) Combined with earnings disregards, this policy helps to support parents who are working but earning low wages, and prevents or delays “hitting the time limit.” It also allows the recipient to “bank” months of TANF eligibility to insure against future risk. Table 40 shows that the majority of the IFS respondents (84%) had their clocks stopped by this policy at some point between July 1997 and February 2001. This means that these respondents had their clocks stopped while they still continued to receive TANF. The vast majority of the respondents had their clocks stopped because they were working, although some may have been attending postsecondary education programs. Other “stopped clock” provisions (see page 3) had not yet been implemented in the study period. The data presented in this section, therefore represent use of the “stopped clock” due to employment34. Among those who had their clocks stopped for these reasons, 10 months was the average amount of time the clock was stopped, and about one-quarter (26%) used the option for more than half of the months they were receiving TANF. Downstate respondents appear to be making better use of this option. More Downstate families (93%) than Cook County families (83%) had their clocks stopped, and among those who did use did use the option, Downstate parents used it longer. Almost half (45%) of Downstate “stopped clock” users received this benefit for more than half of their time on TANF, compared to 24% of Cook County “stopped clock” users.

34 It was not possible to identify the proportion of respondents that had their clocks stopped due to participation in postsecondary education. Given that only 3% of the total IFS sample, regardless of TANF status, was in postsecondary education, we can assume that over 97% of these “stopped clock” cases were for employment, not education.


Table 40: Respondents who had their clocks stopped (while receiving TANF and working)

2001

Of 44 possible months counted toward 60-month lifetime limit:

All

Cook County

Downstate

Percent who had their clock stopped at any point (n=1,123)*

84%

83%

93%

Of those using stopped clock option, average number of months “saved” (n=946)

10 months

10 months

13 months

Of those using stopped clock option, percent who have used it more than 50% of time on TANF (n=946)**

26%

24%

45%

*p<.05; **p<.001(statistical significance for differences between regions, Chi-square test and Paired t-test) Source: Illinois Department of Human Services (administrative records) The analyses reported above were restricted to those who consented to administrative data access (n=1,123). However, the same analysis was conducted using data on the original sample (n=1,899) (without linking to the survey data). There were no significant differences in the observed means or medians, suggesting that the group that consented to having their administrative records reviewed did not differ significantly from the overall sample. Table 41 identifies some characteristics that differentiate “stopped clock” users from those who did not have their clocks stopped. “Stopped clock” users were significantly more likely to have participated in a job training program (18%, compared to 12% among non-users). “Stopped clock” users had an average of 2.7 children, compared to 2.4 children among other respondents. Although this difference may seem small, it was highly statistically significant (see Table 41).


Table 41: Characteristics of respondents who had their clocks stopped (while receiving TANF and working)35

2001 Selected characteristics:

Never used stopped clock option

(n=178)

Used stopped clock option

(n=945) Rent is government subsidized 26% 23% Currently working 46% 54% Participated in job search program* 12% 18% Average cumulative number of months on TANF prior to 1999-00 survey***

71 months

80 months

Married 11% 10% Average age of youngest child 5.4 years 5.4 years Any children with health conditions limiting their activities

14% 17%

Number of children as of 2001 survey (under age 18)***

2.4 children 2.7 children

Have chronic health problems 11% 13% Have high school degree 69% 71% Unaware of earnings disregard policy

37% 32%

***p<.001; *p<.05 (Statistical significance for differences between “never used” and “used stopped clock”; Chi-square test and Paired t-test) Sources: Stopped clock use and average cumulative number of months on TANF were provided by Illinois Department of Human Services (administrative records). All other data displayed in this table is from the IFS survey. Because it is important to discern whether the “stopped clock” policy has a positive effect on those who use it, we examined the relationship between “stopped clock” status and material hardship. Table 42 displays the percent of respondents who have experienced a list of hardships by three groups—those who had never used the “stopped clock” option, those who used it for less than half the time they were on TANF, and those who used it more extensively. It is difficult to discern a pattern from these findings. Tests of statistical significance indicate that the group that used the “stopped clock” for 1-50% of the time they were on TANF experienced more material hardship than either those who never used the “stopped clock” and those who used it for more than half of their months on TANF. For example, 59% of those who had a “stopped clock” for less than half of their time on TANF experienced one or more material hardship between the 1999-00 and 2001 interviews, compared to 57% of those who never used it and 47% who used it for more than half of their time on TANF. There was a similar pattern for having phone service turned off, not being able to pay full rent or mortgage, and having a family member or friend move in to reduce expenses. The decreased material hardship among

35 Tables 40-42 include all respondents for whom data is available, not just those who were still receiving TANF in 2001.


longer term “stopped clock” users (more than 50% of their time on TANF) may indicate that the policy is associated with positive outcomes if used for an extensive period of time. Table 42: “Stopped clock” use and material hardship 2001 Selected Material hardships

Never used “stopped

clock” (n=178)

Used stopped clock 1-50%

of months on TANF

(n=695)

Used stopped clock > 50%

of months on TANF

(n=250) Phone service turned off** 36% 43% 32% Could not pay full rent/ mortgage* 33% 41% 32% Evicted from home 18% 17% 14% Respondent or child needed to see a dentist but couldn’t afford to go

18% 15% 14%


14% 12% 11%


14% 11% 13%


3% 6% 6%

Respondent had a family member/ friend move in to reduce expenses*

1% 8% 3%

Gas/ electricity turned off 1% 2% 1% One or more of the above hardships**

57% 59% 47%


22% 24% 19%

Can generally afford to buy the things we need (somewhat/strongly agrees)*

64% 65% 75%

Went to church or charity for clothes/financial help

29% 34% 27%

Worried that children were not eating enough because not enough money for food (sometimes/often)

10%

10%

9%

Homeless 4% 6% 4% Lived in two or more places since last interview

18% 26% 20%

***p<.001; **p<.01; *p<.05 (Statistical significance for differences between three “stopped clock” status groups; ANOVA test) Sources: Stopped clock use and average cumulative number of months on TANF were provided by Illinois Department of Human Services (administrative records). All other data displayed in this table is from the IFS survey.


Food stamps and Medicaid Trends in food stamps and Medicaid use are displayed on pages 21-22. Use of these benefits among the IFS sample dropped between 1998 and early 1999, and then leveled off, declining gradually between 1999 and January 2001. According to administrative records, as of January 2001, 59% of the original IFS sample was receiving food stamps and 61% was receiving Medicaid (see Figure 5, page 21). Between September 1999 and February 2001, there was an upward trend in the proportion of sample members who were receiving food stamps and Medicaid without TANF. In September 1999, 14% of the sample was receiving this benefit combination, while in February 2001, 27% was receiving it (see Table 7, page 22). Other work supports Table 43 depicts an increase in child care subsidy receipt between 1999 and 2000. These analyses were restricted to respondents with any reported earnings in the 3rd quarter of each year (i.e., July, August, or September), and with at least one child under the age of 12 as of the first survey. No additional criteria were available to determine “need” for child care at each of these time points. The proportion of each group receiving a child care subsidy in September of the same year, according to IDHS administrative records, is presented below. It appears that, at least among workers, the proportion with child care support is increasing over time. Table 43: Child Care Subsidy Receipt

2000

Change from 1999

All

(n=510)

Cook County (n=457)

Downstate

(n=53)

All

(n=526)

Cook County (n=445)

Downstate

(n=81) Receiving child care subsidy in September

55% 57% 42% +10% +11% +2%

No tests for statistically significant differences were conducted since the subgroups of workers differed in each calendar year. Source: Illinois Department of Human Services (administrative records) Non-workers may also be in need of child care. For example, those who are participating in training programs, educational programs, or simply taking time to look for work may also require child care support. Using self-reported data from the IFS surveys, an analysis was conducted to determine whether respondents were receiving support for child care expenses for all, some, or none of their children who were in need of child care. Using these data, a different trend emerges that is perhaps more accurately based on real child care needs. There has been a decline in the percentage of respondents receiving subsidies “from the welfare office” for all of their children who need care, and a corresponding increase in the percentage who say they are not receiving subsidies for any of their children who need care.


Table 44: Self-reported use of and need for the child care subsidy

2001


“Is the welfare office currently helping you to pay for childcare expenses on…”

All (n=1,182)

Cook County

(n=1,070)

Downstate (n=113)

All Cook County

Downstate

All of my children who need it

17% 17% 13% - 3%* -3% - 8%

Some of my children who need it

6% 6% 8% +1%* +1% +4%

None of my children who need it

30% 29% 40% +3% +4%* - 4%

Not applicable (None of my children need child care right now)

47% 48% 39% - 1% - 2% +8%

Source: IFS survey data *p<.05 (statistical significance for differences between time periods; Chi-square test) One caveat related to Table 44 is that some respondents may not be aware that their children’s care arrangements are subsidized, such as when the subsidy check is sent from the welfare office directly to the child care provider. Survey data were used to measure receipt of housing assistance. Nearly one-third of respondents (30%) reported receiving some kind of housing subsidy in 2001. Fourteen percent said they were living in a public housing development and 16% had a rent voucher (such as Section 8). There were no significant changes from 1999-00 to 2001. Downstate residents were more likely than Cook County residents to be receiving housing assistance, including public housing and vouchers. Table 45: Current receipt of housing assistance

2001


All (n=1,183)

Cook County

(n=1,070)

Downstate (n=113)

All Cook County

Downstate

Any housing subsidy 30% 29% 36% -2% -2% -5% Currently live in a public housing development

14%

13%

20%

0

0

-3%

Currently receive a rent voucher, such as Section 8

16%

15%

19%

+1%

+1%

-2%



Most IFS respondents reported that they did not receive formal child support. A little over one-third (35%) of the respondents in the sample had one or more children with papers filed to establish paternity or a child support order in 2001, a decrease of 12 percentage points from 1999-00. Altogether, only 16% of respondents actually received formal child support payments on behalf of one or more of their children (see Table 46). In contrast, more than half (56%) of the respondents received informal support for one or more children in 2001. This includes any money or supplies (e.g., clothes, diapers, food, toys) that the child’s other parent gives to the respondent to help cover expenses. Between 1999-00 and 2001, there was an overall increase in the number of children receiving informal support (7%). The 12 percentage point decrease in children covered by a child support order or with paternity papers filed may warrant further research. One possible explanation for this finding has to do with the sharp decreases in TANF use. Cooperation with child support enforcement is required for TANF eligibility; families must comply in order to receive benefits. Since fewer families are now receiving TANF, it is now possible that as new children are born, fewer of them are the subjects of formal child support orders or paternity agreements. Table 46: Current receipt of child support

2001


All (n=1,150)

Cook County

(n=1,042)

Downstate (n=108)

All Cook County

Downstate

Child is covered by a child support order, or papers have been filed to establish paternity

35%

34%

51% -12%***

-12%***

-4% ***

Other parent actually pays child support once or twice a year or more (of total sample)

16% 15% 24% -1% -1% +3%***

Respondent receives informal child support from the child’s father

56%

56%

58% +7%***

+4%***

+7% ***

***p<.001 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Seventy-one percent of those who worked for pay at some point in the previous tax year said they received the Earned Income Tax Credit (EITC) in 1999-00. This is an 11 percentage point increase from 1998, indicating increasing awareness and/or use of the EITC. Downstate respondents (78%) were more likely to say they received the credit than were Cook County respondents (71%).


Table 47: Earned Income Tax Credit (EITC) receipt

1999-00 (tax year preceding 2001

interview)

Change from 1998 (tax year preceding 1999-00

interview)

All

(n=1182)

Cook County (n=1070)

Downstate (n=113)

All Cook County

Downstate

Worked for pay (at some point in previous tax year)

71%

69%

83%

+9%***

+9%***

+1%

Received tax refund (among those who worked for pay)

79%

78%

89%

+8%***

+9%***

+4%

Received Earned Income Tax Credit (among those who worked for pay)

71%

71%

78%

+11%***

+13%***

+4%

***= p<.001; **= p<.01(statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Participation in job readiness, job training, work experience, basic education, and secondary education programs declined between 1999-00 and 2001, while participation in postsecondary education saw no change (see Table 48). The biggest decline came in job readiness and job training programs. Participation in these programs dropped 14 percentage points, from 36% in 1999-00 to only 22% overall in 2001. Although Cook County saw a larger decline, Cook County respondents were still more likely than Downstate respondents to participate in these programs in 2001. Only a fraction of respondents took part in a work experience program (4%), adult basic education or secondary education program (4%), or postsecondary education program (3%) in 2001.


Table 48: Participation in job training, work experience, and education programs since last interview/ in past year 2001


All (n=1,182)

Cook County

(n=1,070)

Downstate (n=113)

All Cook County

Downstate

Any job readiness or job training program

22% 23% 14% -14%*** -14%*** -9%

Any work experience placement program

4% 4% 5% -3%** -3%** -1%

Any adult basic education or secondary education program (ABE, ESL, GED, or high school diploma)

4% 4% 5% -2%** -2%* -2%

Postsecondary education (for associate’s, bachelor’s, or graduate degree)

3% 3% 5% 0%

0% 0%

***p<.001, **p<.01, *p<.05 (statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Despite the low participation in education programs, 12% of respondents reported that they had completed a high school degree or GED at some time between their 1999-00 and 2001 interviews36 (12% of Cook County respondents and 10% of Downstate respondents) (not shown). Knowledge of work supports and other policies Table 49 displays the percentages of IFS respondents who were aware of several welfare reform policies at the point of their 2001 surveys, and changes in policy knowledge from 1999-00 to 2001. Respondents were asked during each interview whether “the months you work at least 30 hours per week and also receive welfare count toward the time limit” (“stopped clock”), whether “you can keep receiving any of your cash welfare benefits if you are working” (Work Pays), whether “you get to keep receiving Medicaid if you get a job,” whether “you get to keep receiving food stamps if you get a job,” and whether “a person’s cash welfare/TANF grant increases when a new baby is born into the family” (family cap). Respondents who answered “no” or “it depends” to the “stopped clock” and family cap questions were considered to have answered correctly. The answers “yes” and “it depends” were counted as accurate responses to the other questions. 36 The discrepancy between the number of respondents reporting that they had participated in a GED or high school diploma program compared to those who said they had completed their GED or high school diploma since their last interview could be due to respondents reporting receipt of one of these degrees prior to the time period about which they were asked, or due to students who completed their GED through home study, and not through a formal course.


There was an overall increase in policy knowledge across survey waves. This is particularly true in Cook County, in terms of the “stopped clock.” Downstate, there was a notable increase in knowledge about the earnings disregard policy. Some of this increased knowledge may stem from experience—as more respondents work, they learn that it is possible to keep at least a portion of their benefits, and that the months they spend working a 30-35 hours results in fewer months counted toward the lifetime limit on assistance. Overall, knowledge of these policies was fairly high. Most respondents in 2001 were aware of the Work Pays program, transitional Medicaid, and continuing food stamps in 2001. Far fewer respondents were aware of the family cap policy. Just 32% of the full sample were aware that one’s TANF grant may not increase with the birth of a new child. This differed significantly by region: 31% of the respondents in Cook County were aware of this policy, and 41% of Downstate respondents were aware of this policy. (This question was not asked in the 1999-00 survey.) Table 49: Knowledge of welfare policies 2001


All (n=1,182)

Cook County

(n=1,070)

Downstate (n=113)

All Cook County

Downstate

Knowledge of “stopped clock” policy

44%

45%

31%

+22%***

+22%***

+6%

Knowledge of earnings disregards (Work Pays)

60%

60%

67%

+13%***

+13%***

+24%***

Knowledge of transitional Medicaid (after leaving TANF)

88%

88%

92%

+11%***

+12%***

+8%

Knowledge of continuing food stamps (after leaving TANF)

79%

78%

89%

+7%***

+7%***

+8%

Knowledge of the family cap

32% 31% 41% NA NA NA

***p<.001, **p<.01, *p<.05 (statistical significance for differences between time periods) Source: IFS survey data The Family Cap question was not asked in the 1999-00 survey. Most valued benefit Respondents were asked, “If you could pick one thing, what would you say is most important to your family’s well-being?” Over half the sample (57%) chose Medicaid or free health coverage in 2001, up nine percentage points from 1999-00. Housing


assistance (19%) and food stamps (14%) were also chosen by many respondents. Cash assistance, child care assistance, and transportation assistance were most valued by far fewer families. Downstate respondents (71%) were more likely than Cook County respondents (55%) to select Medicaid, while Cook County respondents (20%) saw a greater need for housing assistance compared to their Downstate counterparts (8%). Table 50: “If you could pick one thing, what would you say is most important to your family’s well-being?”

2001


All

(n=1,180)

Cook County

(n=1,068)

Downstate (n=112)

All Cook County

Downstate

Medicaid or free health care coverage

57%

55%

71%

+9%

+7%

+17%

Housing assistance 19% 20% 8% - +1% -7% Food stamps 14% 14% 11% -2% -2% -2% Cash assistance 7% 7% 4% +5% +5% +5% Child care assistance 3% 3% 3% -1% -1% -2% Transportation assistance

1%

1%

2%

-

-

-1%

***= p<.001; **= p<.01(statistical significance for differences between time periods; Chi-square test) Source: IFS survey data Does it pay to move from welfare to work? The architects of welfare reform intended that the kinds of work supports discussed in this section, along with the opportunity to earn more income through work than is provided by monthly TANF benefits, were intended to “make work pay.” An important research question is therefore, “Does it pay to move from welfare to work in Illinois?” Because this study gathers income information retrospectively (for the previous calendar year), it is not yet possible for us to determine how work and welfare use in 1999-00 affected total income in 2001. We can, however, assess the relationship between work and welfare status in 1999-00 on material hardship reported in 2001. Table 51 reports the results of this analysis. A clear pattern emerges from these results, showing that work is protective against material hardship. Those who were relying on TANF only, or had neither work nor TANF in 1999-00 consistently reported more material hardships in 2001 than did those who were working. For example, 47% of those in the TANF-only group in 1999-00 and 40% of those in the no-work/no-TANF group said they “could not pay their full rent or mortgage” in 2001. This compares to 27% of the work-only group and 34% of the work TANF group. Similarly, homelessness was much more common among the no-work/ no-TANF group (10%) than among the work-only (3%) or work-and-TANF group (4%) in 2001. Again exemplifying the overall pattern, experiences of “any material hardship” (one or more of the problems listed in the first portion of Table 51), was highest among


the no-work/no-TANF group (67%) and the TANF-only group (63%), and lowest for the work-only (46%) and work-and-TANF (52%) groups. These findings indicate that work does indeed “pay” and that staying on welfare alone does not seem to protect families from material hardship. Families who have neither welfare nor work as a source of support are particularly vulnerable. It is also important to note that although work seems to prevent some material hardship, levels of hardship were still high even among those who were working and off TANF; 46% of the work-only group experienced some hardship in 2001.


Table 51: Material hardship by work and TANF status

Characteristics in 1999-00

Material hardship outcomes in 2001

No-work/ No-TANF (n=204)

TANF only

(n=375)

Work and TANF

(n=243)

Work only

(n=355) Phone service turned off*** 43% 48% 36% 31% Could not pay full rent/ mortgage***

40%

47%

34%

27%

Evicted from home 21% 15% 17% 16% Respondent or child needed to see a dentist but couldn’t afford to go**

23%

12%

14%

15%


17%

10%

11%

12%

Respondent or child needed to fill a prescription but couldn’t afford it**

18%

10%

13%

9%

Moved in with others to reduce expenses**

3%

9%

7%

4%

Respondent had a family member/ friend move in to reduce expenses

8%

7%

5%

4%

Gas/ electricity turned off 2% 2% 2% 2% One or more of the above hardships***

67%

63%

52%

46%


29%

23%

22%

19%

Can generally afford to buy the things we need (somewhat/strongly agrees)

67%

65%

63%

72%

Went to church or charity for clothes/financial help

36%

34%

31%

27%

Worried that children were not eating enough because not enough money for food (sometimes/often)*

8%

14%

8%

8%

Homeless** 10% 6% 4% 3% Lived in two or more places since last interview

2%

4%

2%

1%

*p<.05; **p<.01; ***p<.001 (statistical significance for differences between work/welfare groups; ANOVA test) Source: IFS survey data.


Time limits and grant closures (“sanctions”) Who is most likely to reach the lifetime limit? The Illinois 60-month time limit began in July 1997. A person who was receiving TANF continuously starting July 1997, without using the “stopped clock” option,37 will reach their time limit on cash assistance in July 2002. We were able to examine 44 of the 60 possible months counted towards the time clock (e.g., use of TANF months from July 1997 to February 2001). Of the 44 months examined, IFS respondents used an average of 20 months, and nearly one-quarter (23%) used up more than half of their possible time on TANF (see Table 52). There was significant regional variation for these findings. Although no large difference emerged between regions in the average number of months of TANF use, Downstate respondents had only used an average of 14 TANF months toward their time limit, compared to an average of 21 months for Cook County respondents. One-quarter of Cook County residents had “used up” more than half of their months, compared to only 7% of their Downstate counterparts. This indicates that a larger proportion of Cook County respondents than Downstate respondents could reach the time limit in 2002 and 2003. Table 52: Moving towards the 60-month time limit, 1997-2001 As of February 2001

Of 44 possible months counted toward 60-month lifetime limit:

All (n=1,123)


Downstate (n=109)

Average number of months on TANF from July 1997 to February 2001*

29

29

27

Average number of months counted toward lifetime limit**

20

21

14

Percent of respondents who have used up more than half of their time limit (e.g., 30 months or more)**

23%

25%

7%

**p<.01; *p<.05 (statistical significance for differences between regions; Paired t-test) Source: Illinois Department of Human Services (administrative records) Note: Use of the “stopped clock” option accounts for the differences between the first and second rows. There were many significant differences between those who have used less than half of their TANF months and those who have already used half or more of their TANF months toward the lifetime limit (see Table 53). Those closer to reaching the time limit were more likely to have chronic health problems, more children, and a child with a limiting health condition in 1999-00. They also were more likely to have received TANF for a longer period of time, receive government housing subsidies, and have participated in a job training program. Those who had used up less than half of their TANF months were 37 Descriptions of the “stopped clock” policy are provided on page 3.


more likely to be working and to have a high school degree. It seems that those who “used up” more time had problems that limited labor force participation. Table 53: Risk characteristics of respondents for reaching the time limit

2001

Selected characteristics (from 2001 interview):

Used less than half of months toward limit

(n=860)

Used half or more of months toward limit

(n=264)

Rent is government subsidized*** 21% 31% Currently working*** 60% 27% Participated in job search program***

12% 32%

Average cumulative number of months on TANF between February 1989 and the 1999-00 survey***

75 months

90 months

Currently married 11% 7% Average age of youngest child 6 years 5 years Any children with health conditions limiting their activities***

15% 23%

Number of children as of 2001 survey (under 18)***

2.6 children 2.9 children

Have chronic health problems*** 11% 18% Have high school degree*** 73% 59% Unaware of earnings disregard policy

33% 32%

***p<.001 (statistical significance for differences between time limit status groups; Chi-square test and Paired t-test) Sources: Number of months counted towards the lifetime limit was provided by the Illinois Department of Human Services (administrative records). All other data displayed in this table are from the IFS survey. How are families affected by losing TANF? Although the Illinois Department of Human Services recently incorporated an indicator of sanction status into its administrative data system, the reliability and validity of this indicator has not yet been examined by the IFS research team. Consequently, administrative data on TANF case closures were used to approximate full-grant "sanctions" for the present analyses. Although this measure is not synonymous with full-grant sanctions, the information on case closures did allow us to identify closures associated with client "non-cooperation." Since sanctions are typically instituted for reasons of client “non-cooperation,” it is possible that many of the same clients are affected by these two situations (sanctions and full-grant closures for non-cooperation). Furthermore, we argue that the “non-cooperation” case closures, also referred to here as full-grant losses, are linked to grant losses that may operate in the same way as a full-grant sanction regardless of whether a formal sanction was actually applied, at least in terms of how a family experiences the grant loss.


To better characterize the nature of the case closures, an assessment was conducted of the original IFS sample (n=1,899) in terms of the case closures they experienced in 1999. Of 505 case closures for "non-cooperation" in 1999, 70% (n=354) were generated because a client failed to keep an appointment, and an additional 13% (n=68) were generated because a client failed to provide paperwork (e.g., a redetermination or earnings report, verification of employment, etc.). In 1999 and 2000, one third of the IFS sample experienced a full-grant loss for non-cooperation (see Table 54). These closures were significantly more common for Cook County families (34%) than for Downstate families (23%). Given the small difference in the number of months on TANF between Cook and Downstate respondents (see Table 54), this does not seem to be an artifact of the difference of length in welfare use. Table 54: Percent of respondents who experienced a full-grant loss due to “non-cooperation”

1999/2000

All (n=1,123)


Downstate (n=109)

Full-grant loss for non-cooperation *

33% 34% 23%

*p<.05 (statistical significance for differences between regions; Chi-square test) Source: Illinois Department of Human Services (administrative records) Table 55 displays the characteristics of those who experienced a full-grant loss compared to those who did not. Nonclosed respondents were more likely to be working and receiving a rent subsidy. They had received TANF for more months as of the 1999-00 survey; nonclosed respondents received TANF for an average of 82 months, compared to 72 months among those who received a full-grant loss. On average, respondents who had experienced a full-grant closure had younger children than nonclosed respondents, although nonclosed respondents had slightly more children.


Table 55: Characteristics of respondents who experienced a full-grant loss for “non-cooperation”

1999/2000

Selected characteristics (from 2001 IFS survey):

No full-grant loss in 1999/2000 (n=755)

Full-grant loss in 1999/2000

(n=368)

Rent is government subsidized* 25% 20% Currently working*** 57% 44% Participated in job search program 17% 18% Average cumulative number of months on TANF prior to 1999-00 survey (months)***

82 72

Currently married 9% 12% Average age of youngest child (years)*

6 5

Any children with health conditions limiting their activities

17% 15%

Number of children as of 2001 survey (less than 18) (number)*

2.7 2.5

Have chronic health problems 14% 10% Have high school degree 71% 68% Unaware of earnings disregard policy

31% 36%

***p<.001; *p<.05 (statistical significance for differences between sanction status groups; Chi-square test) Sources: Data on full-grant losses and average cumulative number of months on TANF were provided by the Illinois Department of Human Services (administrative records). All other data displayed in this table is from the IFS survey. Respondents who experienced a full-grant loss experienced more material hardship than those who did not. Closed families were significantly more likely than other families to experience the following problems: could not pay full mortgage or rent, evicted from home, had a family member or friend move in to reduce expenses, had gas or electricity turned off, or went to a church or charity for help (see Table 56). Furthermore, closed families were significantly more likely to experience any or three or more of the hardships listed in Table 56. Paradoxically, these closed families were more likely to “somewhat” or “strongly agree” that they could afford to buy the things they needed. Overall, however, closed families were at a distinct disadvantage.


Table 56: Full-grant losses and material hardship

2001

Selected material hardships (from 2001 IFS survey)

No full-grant loss in 1999/2000 (n=755)

Full-grant loss in 1999/2000

(n=368) Phone service turned off** 37% 45% Could not pay full rent/ mortgage***

34% 44%

Evicted from home* 15% 20% Respondent or child needed to see a dentist but couldn’t afford to go

14%

16%


11%

13%


12%

12%


6%

4%

Respondent had a family member/ friend move in to reduce expenses***

4%

9%

Gas/ electricity turned off*** 1% 4% One or more of the above hardships***

52% 63%

Three or more of the above hardships*

21%

26%

Can generally afford to buy the things we need (somewhat/strongly agrees)*

66% 68%

Went to church or charity for clothes/financial help***

29%

37%

Worried that children were not eating enough because not enough money for food (sometimes/often)

10%

10%

Homelessness 3% 6% Lived in 2 or more places since last interview

23%

27%

*p<.05; ***p<.001 (statistical significance for differences between sanction status groups; Chi-square test) Sources: Data on full-grant losses were provided by the Illinois Department of Human Services (administrative records). All other data displayed in this table is from the IFS survey.


Putting it all together: What helps families attain self-sufficiency? It is important to understand the relative impact of policies, work supports, and individual characteristics on the lives of current and former welfare recipients. Of particular interest is how these factors influence the intended objectives of the 1996 welfare reform legislation. Are more people working and off welfare because of these policies and work supports? Are people keeping jobs, earning more income, and experiencing less hardship? Are more people marrying as a result of these policies and work supports? What helps families attain self-sufficiency and avoid material hardship? Looking at data from two points in time (1999-00 and 2001) and using multivariate analyses helps to “tease out” factors that uniquely contribute to these outcomes. Table 57 presents the results of multivariate analyses (logistic and linear regressions) predicting six key outcomes, as measured in the 2001 IFS survey: work, hourly wages38 (among those employed in 2001), marriage, welfare use, job loss (among those working in 1999-00), and material hardship. This analysis allows us to look at the impact of selected policies (such as full-grant cancellations and the “stopped clock”), work supports (such as child care subsidies and employer-sponsored health insurance), and personal characteristics (such as education level and marital status) on these six outcomes. The results displayed in Table 57 also control for the effects of several other demographic, personal, and program activity characteristics measured prior to the 2001 interview. These include: race/ethnicity (African American vs. all other racial and ethnic groups), respondent’s age in years, marital status (currently married vs. not married) as of the first survey, number of children, age of the youngest child in the home, whether any children have a health condition that limits activity, whether the respondent has a health condition that limits activity, region (Cook County vs. downstate), number of months respondents have received cash welfare benefits since July 1997 (the month TANF was implemented in Illinois), participation in a job search or job training program in the year prior to the 1999-00 survey, and receipt of a rent subsidy or formal child support. These analyses also control for work in 1999-00. Some people have personal characteristics that make them more likely to work. Controlling for work in 1999-00 reduces this potential bias and allows us to examine the effects of work supports and other factors, independent of work status at the baseline time period. A similar strategy is used for welfare status. Number of months on TANF prior to February 2001 is also one of the control variables, allowing us to examine the impact of various factors independent of welfare history. In summary, the results displayed in Table 57 show that getting people into jobs with employer-sponsored health insurance reduces welfare use, keeps them working, and helps them to earn more money. People who take advantage of the state’s “stopped clock” option and child care subsidies are also more likely to keep working. People who work,

38 Hourly wages were examined because annual income was not available for 2001.


in turn, are less likely to experience material hardships. It is important to keep in mind that those who work, attain jobs with health insurance benefits, and use the “stopped clock” option and available child care subsidies may be more advantaged to begin with. Although the analyses controlled for certain characteristics that are presumed to be associated with greater human capital in general, there may be important unmeasured characteristics that partially or even fully explain our findings related to the stopped clock policy and other work supports. Even if we assume that this limitation exists, however, we can discern that for some individuals, these policies and supports are highly associated with self-sufficiency. Respondents who experience full grant losses for “non-cooperation” have, on average, lower earnings and more material hardship. This may, in part, reflect an underlying instability among the group of recipients that loses cash assistance for reasons of “non-cooperation,” a finding that deserves closer attention. Having a high school degree or GED helped respondents to stay employed and earn higher wages. Respondents with poor health were less likely to be working and more likely to experience material hardship. These findings indicate that access to human capital and quality health care promotes self-sufficiency and well-being.


Table 57: What helps families to meet the goals of welfare reform and attain self-sufficiency?

Outcomes in 2001 (n=1,123)

A Working

B Hourly Wages

(among those working in

2001, n=590)

C Married

D Receiving

TANF

E Job Loss

(among those working in 1999-00, n=561)

F Any

Material Hardship

Characteristics prior to 2001 interview

Odds Ratio Beta Odds Ratio Odds Ratio Odds Ratio Odds Ratio

“Stopped Clock” (number of months used “stopped clock,” July 1997 – February 2001; continuous)

1.07*** .01 1.01 1.00 .94*** 1.00

Full-grant loss for “non-cooperation” (during 1999 and 2000)

.83 -.64* 1.61 1.16 1.17 1.47**

Employer-sponsored health insurance

3.99** 1.37* 1.55 .20* .24** .83

Child care subsidies in 1999 or 2000

1.59** -.05 .50 .81 .71 .85

Has high school degree or GED

1.68** .95** .78 .91 .45** .82

Working 2.31*** 1.15 .85 .70* Married .73 -.52 .76 1.82 .90 Respondent has health problem

.42*** -.54 .96 1.17 1.60 2.22***

Constant .91 6.34*** .13** .03*** .67 1.46 *p<.05; **p<.01; ***p<.001. Shaded cells indicate variables omitted from model. Note: Analysis controls for the following characteristics in 1999-00: race/ethnicity, age, number of children, age of youngest child, any children with health problems, region, number of months receiving TANF (July 1997-February 2001), participation in job search/ job training program, receive government-subsidized rent, and receive formal child support for one or more children. Work and Wages (Columns A and B) Even after controlling for work status in 1999-00, respondents with the following characteristics were significantly more likely to be working in 2001: Had a job with employer-sponsored health insurance Received a child care subsidy Used the “stopped clock” (even after controlling for number of months on welfare) Had a high school degree or GED

Respondents who reported a health problem were less likely to be working in 2001.


Respondents with the following characteristics were significantly more likely to be earning higher hourly wages, relative to others who were working in 2001: Had a job with employer-sponsored health insurance Had a high school degree or GED

Respondents who experienced a full-grant loss for “non-cooperation” earned lower hourly wages. On average, those who lost their welfare grant for reasons of non-cooperation during 1999-00 were earning 64 cents less per hour in 2001 than those whose grants were not closed for non-cooperation. Marriage (Column C) There were no significant findings for marriage associated with the policies, work supports, and individual characteristics listed in Table 1. These factors do not appear to have a strong impact on whether or not respondents become or stay married. Receiving TANF (Column D) Respondents who were working in 1999-00 in a job with employer-sponsored health insurance were much less likely to be on welfare in 2001. Other factors listed in Table 1, however, did not appear to have a significant impact on whether or not a respondent was still receiving TANF in 2001. Job Loss (Column E) Respondents with the following characteristics were significantly less likely to lose their jobs in 2001: Had a job with employer-sponsored health insurance Had a high school degree or GED Used the “stopped clock” (This finding was robust even after controlling for the

number of hours worked per week.) Material Hardship (Column F) Respondents with the following characteristics were significantly more likely to experience material hardship in 2001: Health problem Experienced a full-grant loss for “non-cooperation”

Respondents who were working in 1999-00 were less likely to experience material hardship. Additional description of methods In the analyses, the coefficient for employer-sponsored health insurance should be interpreted as the effect of working in a job with employer-sponsored health insurance, and the coefficient for “work” should be interpreted as the effect of working without employer-sponsored health insurance. A similar interpretation can be given to the coefficient on child care subsidies (e.g., controlling for subsidies is the same as controlling for work activities plus child care subsidies, while the effects of work and job


training can be interpreted as participation in these activities without use of child care subsidies). Models A, C, D, E, and F are logistic regression analyses. The odds ratio is interpreted as the increase (greater than one) or decrease (less than one) in the likelihood of experiencing the relevant 2001 outcome for a person with a specific characteristic or work support/sanction, controlling for all other factors in the model. For dichotomous predictor variables, the reference group is comprised of those individuals who lack the specific characteristic or work support/sanction. For example, in Model A, respondents who were working in 1999-00 are 2.31 times more likely than non-workers in 1999-00 to be working in 2001, controlling for other factors. Those who were married in 1999-00 are .73 times as likely as unmarried respondents to be working in 2001, controlling for other factors. For continuous predictor variables, the odds ratios reflect the increased or decreased likelihood of the outcome for each additional unit of the predictor variable. Model B is a linear regression with 2001 hourly wages as the outcome. The “beta” coefficient is provided, which can be interpreted in the units of the hourly wage variable (e.g., dollars and cents). For continuous variables, the coefficient reflects the increase or decrease in hourly wage for each additional unit of the predictor variable. Of particular interest in these models are the effects of various work and income supports, and “sanctions” on the outcomes presented. We control for select demographic characteristics only to minimize “competing explanations” for the work support and sanction effects. Still, it is important to understand that the effects that emerge in relation to work supports and sanctions may be partially explained by other omitted characteristics of the sample members.


Discussion We began this report with five key questions regarding the success of welfare reform in Illinois (see page one). In this discussion we will draw upon the findings presented in this report to answer these questions. Have the federal goals of welfare reform been achieved in Illinois? As evidenced by the sharp declines in TANF caseloads between 1997 and 2001, the goal to decrease welfare dependence was successfully met in Illinois. Efforts to increase work were also largely successful, although nearly one-half of the IFS sample was still not working in 2001. Family formation goals have clearly been less successful, as marriage rates remained unchanged, although childbearing decreased slightly. How are families doing? There was little evidence of major ongoing threats to the health and well-being of most of the people interviewed for this study. Although many families experienced some instability or hardship, many of the most severe hardships, such as homelessness and food insecurity, decreased slightly over the two-year study period. Although the average income rose significantly between 1998 and 1999-00, incomes remained very low for these families; 76% were living in poverty in 1999-00, despite the fact that about half of the respondents were working. Between 1999-00 and 2001, respondents made significant gains in their hourly wages and in employer-sponsored benefits, although wages still remained fairly low and most working respondents did not receive benefits from their employer. Are families better off when they work and/or leave welfare? Work appears to be protective against material hardship; families who leave welfare for work seem to be doing better than families who stay on welfare or have neither work nor welfare to depend upon. However, it is important to note that while work seems to prevent some material hardship, almost half of the families who were working and no longer receiving TANF still experienced some hardship in 2001. Are families able to meet their children’s needs as they move from welfare to work? Looking across a broad range of indicators of well-being, families appeared to be doing slightly better in 2001 than in 1999-00. Measures of child well-being—such as having access to a doctor or dentist, having enough food to eat, overall child health status, health insurance coverage, and child care stability—all improved slightly during the two-year study period. Health insurance for adults and housing costs, however, stand out as increased trouble-spots in 2001. Although health insurance status improved for children, more parents


became uninsured and/or experienced gaps in coverage. One-quarter of respondents were uninsured in 2001 and 30% had experienced a gap in coverage during the past year. Evictions and inability to pay full rent or mortgage also increased significantly during the two-year study period, although homelessness did not increase. How do current and former welfare recipients feel about the welfare system and policies? Most respondents reported positive feelings about their welfare worker and approved many of the changes brought about by welfare reform. Already quite high in 1999-00, support for these policies increased in 2001, with 92% of respondents agreeing that it is a good idea to require welfare recipients to work. Overall, respondents expressed a strong preference for work over welfare receipt, which may explain their strong support for the new system. How are the more vulnerable families faring? Between 1999-00 and 2001, there was a troubling increase in the proportion of families who were relying upon neither work nor TANF. More than one-quarter of the sample was in this category in 2001. Although some in this group may be relying on family support or informal or unreported work, they were still much more likely to experience material hardship, including homelessness, than all other families. These families appear to be shut out of the welfare system, or are simply unwilling to comply with work requirements. As the unemployment rate climbs and time limits begin to hit, this group of families will probably continue to grow and requires greater attention from policymakers and social service providers. Who is most likely to hit the time limit? Almost all of the families in this sample have “saved” some of their TANF months by leaving welfare temporarily or permanently, and/or by using the “stopped clock” option. For this reason it is unlikely that there will be a large number of families losing benefits in July 2002. It is possible, however, to identify groups of families who are most at risk of hitting the time limit at some point within the next few years. Those with chronic health problems, more children, and a child with a limiting health condition appear to be most at risk of hitting the time limit. The unemployed and those without a high school degree also appear to be more vulnerable.

What is the impact of sanctions? We were not able to measure the impact of sanctions, per se. Instead, looking at a similar group of families—those who lost their full TANF grant due to “non-cooperation”—we find that families who lose their benefits involuntarily are more likely to experience material hardship and to earn lower hourly wages compared to other families. It is not possible to determine whether the grant losses caused the material hardship, or if families already prone to hardship are more vulnerable to grant closures due to “non-cooperation.” It does not appear, however, that grant closures prompt families to find work.


What kinds of obstacles to self-sufficiency do families face? Overall, poor health and low levels of education appear to be major obstacles to leaving welfare and getting and keeping a job. After controlling for a variety of personal characteristics, health problems, lack of access to employer-sponsored health insurance, and having less than a high school degree or GED, were associated with not working, earning lower wages, and still receiving TANF. Full-grant closures for “non-cooperation” were also associated with negative outcomes. Increases in the proportion of adults who have no health insurance or have experienced gaps in coverage, and low levels of participation in education programs, do not seem to bode well for these more vulnerable families who struggle with health problems and low levels of education. Improved access to health care and education may help these families to attain self-sufficiency. Other problems that affect large numbers of families include difficulty paying for housing, poor housing conditions, low wages, poor access to employer-sponsored benefits, depression, and concerns about child care quality and cost. These problems may deter some adults from successfully moving from welfare to work. Affordable housing, living wage jobs with benefits, mental health services, and affordable quality child care may help these families to attain self-sufficiency. Do welfare reform policies and work supports make work pay in Illinois? Looking at the relationship between work and welfare status in 1999-00 and material hardship about one year later, we find that work does indeed “pay.” That is, those who were working in 1999-00 were less likely to experience hardship in 2001 than those who were not working in 1999-00. (We are not able to determine at this time, however, to what extent work helps to lift families from poverty.) Illinois provides a wide array of work supports, including earnings disregards, the “stopped clock” option, and child care subsidies. Despite fairly high rates of awareness of these policies, many respondents were not receiving these resources. It is not clear how many respondents simply do not want or need these supports, and how many cannot access them, despite unmet need. With the exception of the child care subsidy, use of these work supports and other services appeared to be decreasing over the two-year study period. Even after controlling for several personal characteristics including work in 1999-00, receipt of a child care subsidy and use of the “stopped clock” option appear to promote and support employment. Even after controlling for work status in 1999-00, those who received a child care subsidy or took advantage of the “stopped clock” in 1999-00 were more likely to be working in 2001. In addition, longer term use of the “stopped clock” option was associated with decreased material hardship. Widely accessible child care subsidies and the “stopped clock” option (with its accompanying income disregards) emerge as critical elements of the Illinois TANF program that appear to truly help families move toward self-sufficiency.


What is the overall impact of the moderate Illinois approach to welfare reform? A strong economy, coupled with the welfare reforms implemented in 1997, has produced a significant decrease in welfare receipt without an accompanying rise in material hardship in Illinois. In this report, it is not possible to disentangle the effects of welfare reform policies from the effects of the strong economy and low unemployment rates in Illinois during the study period. Recent declines in the economy will allow us to examine the impact of the moderate Illinois approach to welfare reform in more difficult conditions. The next wave of the Illinois Families Study survey, which is being conducted from February to September 2002, will help to elucidate these effects. Overall, the combination of economic and reform factors appears to have promoted slightly improved well-being for the families in this study. Increased wages and earnings, high levels of job satisfaction, approval of many reform policies, and a strong preference for work over welfare receipt indicate that many families have enthusiastically taken advantage of new opportunities and are improving their own lives through work. The mismatch between the sharp decline in TANF use and the moderate increase in work, however, indicates that more work still needs to be done to ensure that all families in Illinois have the means to attain self-sufficiency.


Conclusions and policy implications Work supports are key to the success of welfare reform The welfare reforms of 1996 ushered in a new “social contract” that promised to reward those who left welfare and got jobs. For the most part, Illinois welfare recipients have kept their end of the bargain. As of 2001, about three-quarters of the IFS sample were no longer receiving TANF, slightly more than half were working, and 80% had worked at some point since 1998. Welfare reform emphasized the recipient’s responsibilities if she was to receive benefits. There has been less attention to what the government owes the recipient who is acting responsibly. Given a labor market with low wages and few employer benefits (health insurance, child care, paid family leave, etc.), what is the obligation of the government to those who act responsibly? If many of the jobs available pay a meager wage, what should the government do to reward independence and help maintain it? Reduction of the welfare rolls in Illinois shifts the discussion from responsible recipient behavior to responsible government action. Thanks in part to several strong work supports available in Illinois, we found that work does indeed “pay” for many families. Material hardship and poverty, however, continue to be prevalent in this population and those leaving welfare often enter jobs with low wages and no benefits. Policymakers therefore have an obligation to follow through on their end of the bargain by insuring that working and leaving welfare are sustainable and translate into concrete gains for parents and their children. A comprehensive network of work supports for low-income families is the key to meeting this state obligation. In Illinois, three work supports stand out as being particularly critical for promoting work and self-sufficiency: 1) the “stopped clock,” 2) child care subsidies, 3) Medicaid and KidCare. These supports appear to be very helpful to families, although it is important to note that not all families who need Medicaid and child care subsidies are receiving them. The “stopped clock” Illinois’ unique “stopped clock” option for employed TANF recipients helped adults

to get and keep jobs and longer term use of the policy appeared to protect families from material hardship.

Those most vulnerable to reaching the 5-year TANF time limit include parents caring for a child with a health problem. The recent implementation of a “stopped clock” option for Illinois TANF recipients who provide constant in-home care for a medically dependent child or spouse should help to protect these vulnerable families from being cut off welfare.

Child care subsidies The Illinois child care subsidy appears to promote work and job retention. Illinois has invested a great deal in creating a child care subsidy that ties eligibility to

income, rather than TANF status. This allows for a “seamless” provision of child


care assistance for working poor families who may move on and off welfare. The program also has no waiting lists and is not time-limited. This approach to providing child care assistance appears to be helpful for those Illinois families who can access the program. Many parents who said they needed the subsidy, however, were not receiving it, indicating possible problems with accessibility.

The people interviewed for this study were largely satisfied with their child care arrangements, and most who use child care rely on care from relatives. We were not able to assess the quality of child care arrangements or their impact on children. High rates of parent satisfaction and stability, however, indicate that recent improvements to the child care system in Illinois appear to be a good start to providing needed support to working families and their children.

Medicaid and KidCare (S-CHIP) Health problems appear to be a major barrier to employment and self-sufficiency.

For this reason, it is extremely important that current and former TANF recipients have continued access to health insurance through public programs such as Medicaid.

Those who did have access to employer-sponsored health insurance were more likely to keep their jobs. This indicates that health insurance supports work, possibly by keeping workers and their children healthy, or by helping to “make work pay.”

Medicaid or free health coverage was identified by more than half of all respondents (57%) as the benefit most important to their family’s well-being. Despite the great need for Medicaid, use of this benefit among the IFS sample dropped considerably from 1998 to 2001. For adults, Medicaid losses appear to be responsible for an increase in the number of parents who have no insurance coverage or unstable coverage. Although coverage by employer-sponsored insurance is increasing, most do not have access to health insurance through their job. Public programs are therefore the most viable source of coverage for these low-income parents.

Thanks to more generous income eligibility guidelines for children (compared to adults), declines in Medicaid have been less severe for children. Increased KidCare take-up also appears to be buffering many children from the risks of not having medical coverage.

Other work supports Other work supports, such as food stamps, housing assistance, transportation assistance, and child support are also critical. Given the low incomes and prevalence of material hardship in this population, these supports help working poor families to “stretch” their paychecks to meet their children’s basic needs, and help to make work a viable alternative to staying on welfare. In the face of apparent need, the low take-up of these programs indicates that burdensome eligibility requirements, administrative difficulties, and lack of awareness may be deterring some families from receiving them. The Work Pays program (TANF earnings disregards) and the state and federal Earned Income Tax Credit (EITC) are also critical for promoting and supporting work and self-sufficiency.


Family well-being looking better, but still needs improvement There was little evidence of major ongoing threats to the health and well-being of most of the people interviewed for this study. Looking across a broad range of indicators of well-being, families appeared to be doing slightly better in 2001 than in 1999-00. Although many families experienced some instability or hardship, many of the most severe hardships, such as homelessness and food insecurity, decreased slightly over the two-year study period. Three problems, however, stand out as obstacles to work, self-sufficiency, and well-being: 1) poor health and lack of health insurance, 2) low levels of education, and 3) difficulty paying for housing. Poor health and lack of health insurance In 2001, almost one-quarter of the adult sample reported fair or poor health, and

almost one-fifth reported symptoms of depression. We found that poor health was an obstacle to work and was associated with material

hardship. Even after controlling for a variety of other personal characteristics, health problems were associated with not working and experiencing material hardship.

Having a child with a health problem that limits their activities was associated with being at higher risk of reaching the 5-year time limit.

One-quarter of adults in the sample had no health insurance in 2001 and 30% had experienced a gap in coverage within the past year. Both of these problems increased from 1999-00 to 2001, indicating a troubling trend in health insurance coverage for low-income parents.

Even after controlling for work and personal characteristics, lack of access to employer-sponsored health insurance was associated with job loss and staying on TANF.

Low levels of education About 40% of the sample did not have a high school diploma or GED in 1999-00. Even after controlling for other characteristics, having less than a high school

diploma or GED was associated with lower wages and job loss. Despite the Illinois TANF program’s generous allowances for pursuing secondary

and postsecondary education, only a fraction of IFS respondents were participating in these programs, and there was a significant decrease in participation in basic and secondary education from 1999-00 to 2001.

Difficulty paying for housing Between 1999-00 and 2001, there was an alarming increase in the percent of

respondents who said they could not pay their full rent or mortgage or were evicted at some point since their last interview.

Although there was no commensurate increase in homelessness or crowding, the above finding indicates that housing costs may pose a major obstacle to self-sufficiency. This is not surprising given the high cost of housing relative to the low incomes earned by most in this sample.


Family formation Welfare reform does not appear to have had a significant impact on family formation in Illinois. Marriage rates did not increase in the IFS sample and births declined modestly. Clearly, there are factors beyond welfare policy that influence these trends.


Recommendations Given that the welfare reforms implemented in Illinois in 1997, coupled with a strong economy, have produced a significant decrease in welfare receipt without an accompanying rise in material hardship, we recommend that states look carefully at the Illinois example. Illinois has developed a strategy that “makes work pay” and reduces welfare dependency. We recommend several policy improvements at the state and federal levels to achieve the Illinois result. Furthermore, we do not recommend that TANF be dramatically altered at this time. Instead, we recommend specific improvements and alterations to existing policies. Recommendations to federal policymakers for TANF reauthorization Time limits The “stopped clock” policy for employed TANF recipients has been successfully

implemented in Illinois and should be adopted nationally. The “stopped clock” policy for TANF recipients who are participating in post-

secondary education, providing in-home care for a medically dependent child, or caring for a sick relative should also be encouraged by federal legislation.

Medicaid and S-CHIP Public programs are the most viable source of health insurance coverage for most

current and former welfare recipients. The federal government should provide incentives to states to make Medicaid more

accessible to parents by: Raising income eligibility cutoffs Expanding Transitional Medicaid Assistance (TMA) beyond 6-12 months after

TANF exit The federal government should facilitate and encourage S-CHIP waivers that allow

states to expand their S-CHIP benefits to parents. The federal government should reward states for high enrollment in Medicaid and S-

CHIP. Outreach and marketing efforts should be encouraged to increase awareness of these programs.

Child care Child care assistance should receive the highest possible levels of financial support

from the federal government. States should be rewarded for high take-up rates for their child care subsidies. The Illinois approach to providing child care subsidies with eligibility tied to income

(instead of TANF status) should be encouraged in other states that have more fragmented child care programs.


Education and training The federal government should require states to: Include secondary and postsecondary education as a legitimate work activity Allow TANF recipients to combine work and school activities to meet the weekly

number of hours specified by the work requirement Make sure recipients are aware of available educational opportunities

Work requirements Given the demands of single parenting and recent declines in the economy, requiring

a 40-hour work week is not recommended. The current requirement of 30-35 hours per week appears to be working well and should be maintained. Requiring a full 40-hour week may discourage some recipients and could be counterproductive.

Recommendations to Illinois policymakers Restore child care funding in fiscal year 2003. Maintain child care subsidies as a high

spending priority. Increase take-up of the child care subsidy. Expand income eligibility cutoffs for Medicaid coverage for adults and extend the

provision of Transitional Medicaid Assistance to at least 12 months, if not longer, for TANF leavers. (Eliminate the 6-month re-certification.)

Extend KidCare to parents through the FamilyCare program. Raise the TANF monthly cash grant. This may help to decrease the high levels of

material hardship among TANF recipients. (Concurrent maintenance or expansion of existing work supports will ensure that work still “pays more” than TANF receipt.)

Ensure that TANF and non-TANF working poor families have access to emergency assistance funds. This will help to alleviate material hardship and may help parents to stay in the work force (e.g., through emergency transportation assistance).

Enact policies that promote affordable housing. Increase the earnings eligibility cutoff for exiting TANF. Ensure that families who

leave TANF due to earnings have an income above the Federal Poverty Line.


References The Annie E. Casey Foundation. 2000 Kids Count Data Book: State Profiles of Child Well-Being. 2000. Center on Budget and Policy Priorities. A Hand Up: How State Earned Income Tax Credits Help Working Families Escape Poverty In 2001. 2001. Center on Hunger and Poverty, Tufts University. Are States Improving the Lives of Poor Families?: A Scale Measure of State Welfare Policies. 1998. Lee, B., Goerge, R., and Dilts, J. Outcomes for the Income Maintenance Caseload After Receipt: Caseload Dynamics, Employment and Earnings in Illinois, 1995-1999. Chapin Hall Center for Children discussion paper. 2001. Lee, B., Goerge, R., and Dilts, J. Outcomes for the Income Maintenance Caseload During Receipt: Caseload Dynamics, Employment and Earnings in Illinois, 1991-1999. Chapin Hall Center for Children discussion paper. 2000. Lewis, D., George, C., and Puntenney, D. Welfare Reform in Illinois: Recent Efforts in the Context of the National Debate. Institute for Policy Research working paper, Northwestern University. 1995. National Center for Children in Poverty, Mailman School of Public Health, Columbia University. Child Poverty in Illinois. Fact sheet. 2000. Pearce, D. The Self-Sufficiency Standard for Illinois. Wider Opportunities for Women and Women Employed. 2001. Voices for Illinois Children. Illinois Kids Count 1999-2000. 1999. WELPAN. TANF Spending Patterns in the Midwest. The New Face of Welfare. 2002.


Appendices


Appendix A: The Illinois TANF Program Table 58: Characteristics of the Illinois TANF program Eligibility (at intake) A family must include:

a child under age 18; a child age 18 who is a full-time student in secondary school; or a woman with no other children who documents that she is pregnant regardless of

month of pregnancy. Amount of monthly gross earnings above which an applicant cannot qualify for cash assistance: Family of 1*: $302 Family of 2: $368 Family of 3: $467 Family of 4: $504

$90 of applicant’s earnings disregarded in determining eligibility. Asset Limits: Family of 1*: $2000 Family of 2: $3000 $50 addition for each additional family member

All two-parent families are eligible using the same criteria as single parent families.

Income disregards Family exits program when gross monthly earnings exceed: Family of 1*: $636 Family of 2: $834 Family of 3: $1131 Family of 4: $1242

Maximum Benefit Levels39

All Families (Excluding Child-Only Assistance Units): Family of 1*: $212 Family of 2: $278 Family of 3: $377 Family of 4: $414

Child-Only Assistance Units: 1 Recipient: $102 2 Recipients: $201 3 Recipients: $249 4 Recipients: $319

Benefit Calculation

A recipient’s benefit is the difference between his/her countable income and the maximum benefit. The countable income is his/her gross monthly income less 67% of earnings.

* Family of One = Pregnant woman 39 Maximum benefit levels for the counties with the highest cash benefits and the most recipients in Illinois (Boone, Champaign, Cook, DeKalb, DuPage, Kane, Kankakee, Kendall, Lake, McHenry, Ogle, Whiteside, Winnebago, Woodford).


Family Cap Assistance payments to families will not increase when a child is born unless: The baby is born within ten months after application is filed. If conception occurs during a month when family received TANF and family did not

receive TANF for a period of nine consecutive months any time following conception The baby is the first child of a child in the assistance unit Conception was result of documented incest or rape

Work Requirements/ Work Activities

The recipient must find paid work or participate in one of the following activities: Job search Work experience Self-employment Work First (state employment and training program) Community service Basic education Vocational training Education beyond high school Foster parenting Other self-sufficiency activities such as:

- alcohol or substance abuse treatment - mental health treatment - domestic violence counseling

Time limits Lifetime time limit of 60 months: No extensions (unless working or attending school; see exemptions below) Does not apply to teen parents under 18 who have their own grant (months start

counting toward limit when parent turns 18) Exemptions Work Exemptions:

Recipients whose youngest child is less than one year old Recipients age 60 or older An adult in a family with a child-only assistance unit

Time Limit Exemptions: Adult is employed at least 30 hours per week (one-parent family) or 35 hours a week

(two-parent family) Adult from one-parent family is attending a postsecondary degree program full time and

maintains at least a 2.5 GPA. Adult provides constant in-home care for a medically dependent child Adult provides care for a disabled child or spouse

Sanctions

The sanctions for failure to participate in work or child support enforcement activities, or failure to meet school attendance requirements, are as follows: 1st Instance: 50% reduction of cash assistance until family complies with requirements

or three months have passed; if still no cooperation after three months, all cash assistance is terminated.

2nd Instance: 50% reduction of cash assistance for a minimum of three months; if still no cooperation after three months, all cash assistance is terminated.

3rd Instance: Cash assistance is discontinued for a minimum of three months; cash assistance will not be reinstated until family has complied with requirements

Recipients who are not engaged in work activities within 24 months will have their cash assistance terminated unless agreed upon support services were not provided.


Participation of eligible legal residents who are not citizens of the U.S.

Same as citizens

Child Care Benefits

All families with income below 50% of 1997 state median income ($1818 per month for family of three), or 160% of federal poverty level, are eligible regardless of prior TANF status. A two-parent family may receive assistance only if the second parent is unavailable or unable to care for children. The following families may receive subsidies: If adult from eligible family (see above) is working If TANF recipient is participating in any work-related activity lf adult from eligible family attends a two- or four-year college degree program; S/he

must either work 10 hours per week or be involved in a combination of work and educationally required, work-like activities (i.e., student teaching) for a total of 20 hours per week.

If adult from eligible family attends adult basic education, English as a Second Language, GED (general education development) classes, or a non-degree vocational training program - No work requirement - Two-year time limit on eligibility

Medical Benefits (Medicaid)

All TANF recipients are entitled to coverage. Families that are no longer eligible for TANF due to earnings: Retain coverage for six months Can retain coverage for an additional six months if family income is below 185% of

federal poverty level (FPL)


Appendix B: Welfare policies and caseload trends in Illinois and other Midwestern states

Table 59: Comparison of six Midwestern state TANF programs

Illinois Indiana Michigan Minnesota Ohio Wisconsin Welfare program name1

TANF TANF/IMPACT Family Independence Program (FIP)

MN Family Investment Program (MFIP)

Ohio Works First (OWF)

Wisconsin Works (W-2)

Time Limit1 60 months 24 months None 60 months 36 months 60 months

“Stopped Clock” option2

Yes No No No No No

Family Cap1 Yes Yes No No No Yes Sanctions1

1st: 50% cash reduction until cooperation; 2nd: 50% cash reduction for 3 months; 3rd: cash cut-off >3months

1st: Cut 2 months of adult cash grant; 3rd: Cut cash grant at least 3 years.

Terminated from FIP & food stamps if not complying within 60 days. On FIP, 25% reduction non-comply

1st: 10% grant cut; 2nd: Vendor pays shelter cost with 30% reduction of grant

1st:: Lose 1 month 2nd: Lose 3 months 3rd: Lose cash & ineligible for Medicaid

Grant reduced $5.15 for each hour not worked plus additional sanctions

Asset Limit1 <$3,050 for Family of 3 <$1,500; car <$5,000

<$3,000; no car limit

<$2,000; car and asset <$7,500

No asset limit <$2,500; car <$10,000

Cash Diversion Program1

No No No Yes County discretion No

Work Requirements1 Paid work or work activities such as job search, education, or self-sufficiency

Job-ready or no-Job-ready tracks; 20-25 hours of TANF work

Work upon state determination; No community service option

Employment & range of work activities; plus education option. TANF recipients who have received assistance for 2 months and are not working are required to participate in community service employment.

Employment & range of work activities; plus education option

4-tiered work activity: unsubsidized work, trial jobs, community jobs, W-2 transition

Domestic Violence Exemption1

No* No Yes Yes No Yes

* Illinois plans to implement an exemption rule in 2002.

Second annual report from the Illinois Families Study

89

Illinois Indiana Michigan Minnesota Ohio Wisconsin Transitional Medicaid Assistance (TMA) Eligibility3

Available for up to 12 months (6 months after leaving TANF, plus additional 6 months if family income is below 185% FPL).

Available for up to 12 months.





Medicaid eligibility for children (as % of FPL)4

Infants born to women eligible at time of birth:

200% Age 0-1: 185% Age 1-18: 133%

150% Age 0-1: 185% Age 1-18: 150%

Age 0-2: 280% Age 1-18: 275%

200% 185%

State Children’s Health Insurance Program (S-CHIP) eligibility (as % of FPL)5

185% 200% 200% Not Available (No separate S-CHIP program; Minnesota

Care)

Not Available (No separate S-CHIP

program; Healthy Start Healthy Families)

Not Available (No separate S-CHIP

program; BadgerCare)

Child Care Subsidy Eligibility (as % of FPL for family of three)6

160% (50% of 1997 state median income) Subsidy is available regardless of TANF status.

143% TANF recipients are eligible by virtue of their TANF status. Other applicants must be at or below 143% FPL. 12-month limit after leaving TANF. Once enrolled, families remain eligible until their income exceeds 181% FPL (at county discretion).

188% Subsidy available regardless of TANF status.

275% (Transitional Child Care program) MFIP recipients in approved activities are eligible for MFIP child care subsidy. Transitional Child Care available for up to 12 months after leaving MFIP. Families may be eligible for the non-guaranteed Basic Sliding Fee program after leaving Transitional Child Care.

182% (Transitional Child Care) OWF participants with children under 13 years old are eligible for OWF child care subsidy. Transitional Child Care is available for up to 12 months after leaving OWF. Families may be eligible for non-guaranteed child care subsidy after leaving Transitional Child Care.

185% Subsidy available regardless of TANF status. Once enrolled, families remain eligible until their income exceeds 200% FPL.

90 W

elfare Reform in Illinois: Is the m

oderate approach working?

Illinois Indiana Michigan Minnesota Ohio Wisconsin Postsecondary education counted as a work requirement6

Yes (2-year and 4-year degree programs) 36-month limit (as stand-alone activity; varies by case; for approved programs only) No other work activity required; must maintain 2.5 GPA

Yes (2-year degree program, when accompanied by part-time work activity) No (4-year degree program) 12-month limit

Yes (2-year and 4-year degree programs, when accompanied by part-time work activity) 12-month limit

Yes (2-year degree program) No (4-year degree program) 24-month limit

County discretion (may be allowed if combined with work activity) 24-month limit (maximum; county discretion)

No

State EITC7 Yes (Non-refundable tax

credit)

No* No Yes (Refundable tax credit)

No Yes (Refundable tax credit)

* Indiana offers an alternative “earned income” tax credit.

1. Source: Compiled by Midwest Partners 2. Source: Personal communication with state officials 3. Source: Welfare Information Network (www.welfareinfo.org) 4. Source: American Academy of Pediatrics (www.aap.org) 5. Source: Kaiser Family Foundation (www.statehealthfacts.org) 6. Source: Center for Law and Social Policy (www.spdp.org) 7. Source: Center on Budget and Policy Priorities (www.cbpp.org)

Second annual report from the Illinois Families Study

91


Table 60: Maximum monthly benefits and earnings levels in six Midwestern states (for a family of three; as of October 2000) Illinois Indiana Michigan Minnesota Ohio Wisconsin Maximum monthly welfare benefit

$377 $288

$459 $801 $373 $673

Food stamps for family getting maximum welfare

$341 $341 $333 N/A $399 $250

Total income (cash + food stamps)

$718 $629 $792 $801 $712 $923

Income as percent of poverty

59% 52% 65% 66% 58% 76%

Earnings level at which eligibility for welfare ends

$1,131 $1,179 $774 $1,400 $996 $673

Earnings at TANF exit as a percent of poverty

93% 97% 63% 115% 82% 55%

Source: U.S. Department of Health and Human Services; compiled by Midwest Partners Table 61: Standard of need and AFDC benefits for six Midwestern states, January 1994 AFDC benefit Standard of need Benefit/Standard of

need Illinois 367 890 .41 Indiana 288 320 .90 Michigan 459 551 .83 Minnesota 532 532 1.00 Ohio 341 879 .39 Wisconsin 517 647 .80 Source: Committee on Ways and Means, US House of Representatives, 1994 Green Book Table 62: AFDC/TANF caseload declines in six Midwestern states, 1994-2000 Source: US Department of Health and Human Services; compiled by Midwest Partners Illinois Indiana Michigan Minnesota Ohio Wisconsin AFDC/TANF caseload change, FY 1994- September 2000

-69% -49% -69% -38% -63% -78%

September 2000 TANF caseload

75,147 37,788 69,919 39,005 91,654 17,185

Change in food stamp usage, FY 1994- FY1999

-27% -38% -36% -45% -49% -47%


Appendix C: IFS Study Regions Rationale for studying these counties TANF families who live in the following counties have been selected for this study: Cook, St. Clair, Peoria, Fulton, Knox, Marshall, Woodford, Tazewell, and Stark. These nine counties represented 75% of the Illinois TANF caseload when the sample was drawn in 1998. Cook and St. Clair counties were selected because they are the counties with the largest shares of TANF recipients in Illinois. Cook County alone represented approximately 67% of the 1998 TANF caseload, and St. Clair County made up 4% of the TANF caseload. Peoria County was selected for several reasons: 1) it represents the third largest share of TANF recipients in the state (2%); 2) it includes a mid-size city rather than a larger urban area (as is true of both Cook and St. Clair counties); and 3) researchers felt it was important to represent central Illinois in the study sample. Peoria is not intended to be a “representative” mid-size county, as the wide variation in demographic and TANF caseload characteristics among other mid-size Illinois counties means there is no “typical” mid-size county. Two important characteristics should be noted, however, that distinguish Peoria County from other similar-sized counties. It has a slightly higher poverty rate and a greater proportion of African-American residents than other mid-size counties. The great majority of Illinois counties (approximately 85%) have populations less than 100,000. Given this fact, researchers felt it was important to include a number of smaller counties in the sample. Although a random sample of smaller counties would have been ideal, budget constraints did not allow such a design. Instead, we included the ring of six counties that surrounds Peoria County: Fulton (1998 population 38,788), Knox (56,070), Marshall (12,791), Woodford (34,576), Tazewell (127,602), and Stark (6,396). When these rural counties are compared (as a group) to all other counties with populations less than 100,000 in the state, there are minimal differences across a number of welfare caseload characteristics. The only exception is that the rural counties are less likely to include African-American residents than other small Illinois counties (7% versus 16%, respectively). Given the larger representation of African-American TANF recipients in Peoria, along with the efficiency gained by selecting counties in close proximity to one of our other selected sites, this difference was not determined to be sufficient cause for revising the sampling plan. Demographic characteristics Child poverty and TANF caseloads The child poverty rate varies widely by study county. Seventeen percent of Illinois children were living in poverty in 1997, a decrease of 2 percentage points from 1995. Four of the study counties (Cook, St. Clair, Peoria, and Fulton) had child poverty rates that exceeded the overall state rate, while the remaining five study counties—all of them


rural—had rates at or below the state level. Dropping 2 and 3 percentage points since 1995, St. Clair (25%) and Cook (23%) counties continued to have particularly large proportions of children living in poverty (see Figure 3). Figure 7. Percent of children in poverty by study county, 1995 and 1997

9%

13%

14%

11%

19%

21%

22%

27%

26%

19%

Woodford

Tazewell

Stark

Marshall

Knox

Fulton

Peoria

St. Clair

Cook

State of Illinois

Stud

y C

ount

y

Percent of Children

17%

23%

22%

25%

20%

19%

13%

14%

13%

8%

Source: Illinois Kids Count 1999-2001, Voices for Illinois Children, 2001. Variations in TANF caseloads mirror differences in the child poverty rate and overall population size. Cook County, for example, had a monthly average of 41,282 TANF grantees in 2000, compared to only two grantees in Stark County (see Table 62). The state of Illinois and all study counties experienced steep declines in TANF caseloads between 1996 and 1999 (see Figure 8). The most urban county (Cook) experienced caseload declines from 1996 to 1999 that were slower than that of the state overall (-52%). The second-most urban county (St. Clair) also experienced a caseload decline from 1996 to 1999 that was slower than that of the state overall. The rural counties saw their caseloads diminish at an astonishing rate, including a 93% decline in Stark County and a 84% decline in Fulton County (see Figure 8).

1995 1997


Table 63: Number of TANF grantees (monthly average) by study county, 1996, 1999, and 2000 1996 1999 2000 State of Illinois 169,869 82,057 51,480 Cook 110,856 61,149 41,282 St. Clair 6,619 3,594 1,971 Peoria 3,259 1,559 855 Fulton 457 74 27 Knox 650 176 109 Marshall 98 26 26 Stark 42 3 2 Tazewell 1,003 177 65 Woodford 157 40 9 Source: Illinois Kids Count 1999-2001, Voices for Illinois Children, 2001. Figure 8. Percent change in TANF grantees by study county, 1996-1999

-75%

-82%

-93%

-74%

-73%

-84%

-52%

-46%

-45%

-52%

-100% -90% -80% -70% -60% -50% -40% -30% -20% -10% 0%

Woodford

Tazewell

Stark

Marshall

Knox

Fulton

Peoria

St. Clair

Cook

State of Illinois

Stud

y C

ount

y

Percent Change

Source: Illinois Kids Count 1999-2001, Voices for Illinois Children, 2001.


Race and ethnicity As mentioned above, the rural counties are predominantly white, while the more urban counties (Cook, St. Clair, and Peoria) have significant African-American populations (see Table 64). Cook County had the largest proportion of persons of Hispanic ethnic origin, at 19.9% in 2000 (see Table 65). Table 64: Race by study county, 2000* Black White Other race State of Illinois (n=12,419,293)

15.1% 73.5% 9.4%

Cook (n=5,376,741) 26.1% 56.3% 15.0% St. Clair (n=256,082) 28.8% 67.9% 3.3% Peoria (n=183,433) 16.1% 79.4% 4.5% Fulton (n=38,250) 3.6% 95.1% 1.3% Knox (n=55,836) 6.3% 89.9% 6.7% Marshall (n=13,180) 0.3% 98.2% 1.5% Stark (n=6,332) 0.1% 98.6% 1.3% Tazewell (n=128,485) 0.9% 97.4% 1.7% Woodford (n=35,469) 0.3% 98.5% 1.2% *Percents do not add up to 100% due to rounding. Source: U.S. Census Bureau, 2000. Table 65: Hispanic origin by study county, 2000* Hispanic Origin State of Illinois (n=12,419,293)

12.3%

Cook (n=5,376,741) 19.9% St. Clair (n=256,082) 2.2% Peoria (n=183,433) 2.1% Fulton (n=38,250) 1.2% Knox (n=55,836) 2.5% Marshall (n=13,180) 1.0% Stark (n=6,332) 0.9% Tazewell (n=128,485) 1.0% Woodford (n=35,469) 0.7% *Percents do not add up to 100% due to rounding. Source: U.S. Census Bureau, 2000.


Appendix D: IFS Research Team, Advisory Groups, and Collaborative Partners Principal Investigators Dan A. Lewis Professor, School of Education and Social Policy Faculty Fellow, Institute for Policy Research Northwestern University 2040 Sheridan Road Evanston, IL 60208-4100 847-491-3715 [email protected] James H. Lewis Director, Institute for Metropolitan Affairs Roosevelt University 430 South Michigan Avenue, Suite 846 Chicago, IL 60605-1394 312-341-3541 [email protected] Paul Kleppner Director, Office of Social Policy Research Professor, Political Science and History Northern Illinois University 138 North 3rd Street DeKalb, IL 60115 815-753-1309 [email protected] Stephanie Riger Professor, Psychology and Women’s Studies University of Illinois at Chicago 1152 Behavioral Sciences Building 1007 Harrison Street Chicago, IL 60607-7137 312-413-2300 [email protected] Bong Joo Lee Research Fellow Chapin Hall Center for Children, University of Chicago 1313 East 60th Street Chicago, IL 60637 773-256-5156 [email protected]


Robert Goerge Research Fellow Chapin Hall Center for Children, University of Chicago 1313 East 60th Street Chicago, IL 60637 773-256-5137 [email protected] Northwestern University Amy Bush Stevens, IFS Project Coordinator Institute for Policy Research 2040 Sheridan Road Evanston, IL 60208-4100 847-491-5889 [email protected] Lisa Altenbernd, Graduate Research Assistant Irene Carvalho, Graduate Research Assistant Morgan Ward Doran, Graduate Research Assistant Marla McDaniel, Graduate Research Assistant Amber Stitziel Pareja, Graduate Research Assistant Natalie Phillips-Hamblett, Program Assistant University of Wisconsin—Madison Kristen Shook Slack, Assistant Professor School of Social Work 1350 University Ave. Madison, WI 53706 608-263-3671 [email protected] Alan Puckett, Graduate Research Assistant Lynette Renner, Graduate Research Assistant Matt Smith, Graduate Research Assistant Marya Sosulski, Graduate Research Assistant Joan Yoo, Graduate Research Assistant Metro Chicago Information Center Woody Carter, Director of Research Danae Corado, Data Quality Supervisor Joyce Franklin, Data Entry Clerk Tim Frye, Research Analyst Patricia Gross, Deputy Director Carole Moody, Senior Field Supervisor


Thais Seldess, Director of Field Operations Olivia Shaw, Director of Staff & Support Services D. Garth Taylor, Executive Director Rong Zhang, Senior Programmer Interviewers Cook County Delores Crawford Almeda Moore Esther Davis Cordell Rainey Abby Hernandez Jack Sloan Teresa Johnson Pearl Washington Eugene Lloyd Lynda Wells Sam Lopez Peoria region Pam Christianson Toby Gudeman Shirley Daniels Dennis McQuellin Barbara Fletcher Laurie Shoemaker Gretchen Flynn Ann Throckmorton St. Clair County Delores Foster Jacqueline Rice Gerald Hayes Lois Shaw Rosalyn Johnson Anthony Taylor Karen Militello Delores Owens Telephone Interviewers Keith Bibbs Diashun Taylor Precious Collins Sharon Thomas Michael McCrae Gezelle Wright Clarence Reed Legislative Advisory Committee Representative Barbara Flynn Currie Senator Miguel del Valle Representative Rosemary Mulligan Senator Steven Rauschenberger Senator David Syverson IDHS Steering Committee Marilyn Okon, Bureau Chief, Program Design and Evaluation, IDHS David Gruenenfelder, Manager, Program Design and Evaluation, IDHS Alan Whitaker, Research Analyst, Program Design and Evaluation, IDHS


Appendix E: IFS supplemental studies Qualitative Component Northwestern University received funding from The Searle Fund to conduct a qualitative supplement to the Illinois Families Study. The purpose of the qualitative component was to gain a deeper understanding of the process families undergo as they move toward self-sufficiency. In-depth interviews and participant observation will help researchers gain insight into the daily achievements and struggles of the families affected by welfare reform. This research seeks to explore the “building blocks of independence” as experienced by parents and their children. Specific areas of inquiry include work, identity, relationships, education, and child well-being. Researchers interviewed 69 Chicago-area participants in the summer of 2000, and completed follow-up interviews with 58 of those participants during the summer of 2001. The findings are currently being analyzed and summarized, and three selected profiles are included in Appendix F. Child Well-Being Supplement Northwestern University received funding from the National Institutes of Health to conduct a supplemental study of child well-being. Jane Holl of the Institute for Health Services Research and Policy Studies at Northwestern University serves as the lead investigator on this study and works in collaboration with the other IFS researchers. This supplemental study focuses on the etiology of two specific forms of child neglect: physical neglect and supervisory neglect. Welfare reform requires parents to spend more time in formal employment, subjects them to time limits and sanctions on the receipt of income assistance, and may result in unstable health care coverage for their families. Assessing whether such changes result in a differential rate of child maltreatment in general, and different types of child neglect in particular, is the focus of this study. The specific aims of the project are to assess the relationships between child neglect and (1) employment, (2) income, and (3) health care coverage, with a particular focus on how changes in these predictors relate to the incidence or onset of neglect. The project will identify the factors that mediate or moderate such relationships and expand the knowledge about causal pathways leading to specific forms of child neglect. It will also include baseline assessments of child development so that future studies may assess the developmental impacts of these forms of child neglect. This study involves a subsample of children in the IFS (550 children younger than five years old at the second interview). Parents of these children participated in an additional in-person interview after the main IFS survey during Wave 2. These families will be interviewed annually for a total of five years. The interview focuses on child development and temperament, parenting beliefs, household accident risks, adequacy of


basic needs, and health care. Medical chart reviews will be conducted to assess adequacy of health care and to identify additional risks for neglect. Quarterly administrative data from the Illinois Department of Employment Security’s Unemployment Insurance database, the New Hire registry, and all social service agency registries (Food Stamps, Medicaid, Child Abuse) will be linked for each subject. Data from a continuously integrated database of children’s services in Illinois (maintained by Chapin Hall) that depicts a full network of relationships between individual children and public services (e.g., child protection, juvenile justice, Medicaid, special education, and mental health services) also will be linked.


Appendix F: Three families tell their stories Selected case studies from the IFS qualitative component These three case studies are based upon selected in-depth qualitative interviews conducted with 58 IFS respondents during the summers of 2000 and 2001. Renita’s Story: A Plan for the Future Strong social support from her church and family, a good educational background, and benefits from her job helped Renita to make a fairly smooth transition from welfare to work. Having successfully reentered the labor force and stabilized her life, this mother of two is now eager to meet new goals she has set for herself and her family: getting a driver’s license, a car, and a new apartment. After six years on welfare, Renita Walker* had several factors in her favor as she began the transition to work two years ago. Deeply religious, she had a strong social network in her Baptist church, two years of college under her belt, intermittent help from her daughter’s father, and a plan for the future. She also has a full-time job. Employed as a secretary for the past 20 months, Renita now earns $11.25 an hour with medical benefits. She is completely off welfare, independent, and happy about it. Renita left college and went on welfare when her second child was born. She felt it was important to stay home and care for her baby and young son, now ages 5 and 7. All of her cash benefits went to rent, which she supplemented with babysitting and “doing people’s hair.” Food stamps covered the groceries and her mother helped out with phone bills. One temporary job lasted nine months but caused her benefits to be cut. “In the end I would have been better off getting aid,” she recalls, claiming she had more difficulty making ends meet when she was working. Like others transitioning from welfare to work, Renita found her benefits erratic, but coped by learning to budget—saving in the better months to cover the lean months. And though she was “barely making it” a year ago she was still able to cover her church tithes, bus fare, clothes, lights, gas, and food. Renita’s formerly working-class neighborhood is now largely populated by welfare recipients, single parents, and drug abusers. Strangers freely roam in and out of her building and the sounds of gunshots is not uncommon. After two break-ins, she felt compelled to spend some of her hard-earned cash to install bars on her doors and rear windows. Thus, her first priority is to move to a better neighborhood. She is saving for a down payment on a townhouse that she hopes will allow her son and daughter to have separate bedrooms. Currently, they share the dining room. She is also taking driving lessons and saving for a car. Getting around drains her cash, she says. Bus fare alone costs a minimum of $20 a week.


Renita gets some help from her son’s father who visits, pays half the costs of school supplies, and sometimes takes both children on little excursions. Her daughter’s father, who gets paid “under the table,” is obligated to pay child support but doesn’t. When she was on public aid, he gave her a total of $85 over a six-month period. The church provides both religious and social sustenance for the family, which faithfully attends Bible classes, services, and outings. Renita also has three close friends who round out her support system. Down the road, she is considering marriage, but not before she can “grow emotionally and financially so that I have something to bring to the table.” Once she moves to a nicer neighborhood, she plans to get her degree so she can become a probation officer and boost her salary. *Not her real name. Some circumstances have been altered to protect the anonymity of the respondent. Margaret’s Story: A Home of Her Own When she turned her back on free but miserable living conditions a year ago, this mother of five began a downward spiral in which she lost her housing, her job, her savings, and her child care. Her story shows how fragile success can be for a struggling single parent, and just how critical affordable housing can be for families moving toward self-sufficiency. Just a year ago, Margaret Poole* had a full-time job at a hotel, free rent at her boyfriend’s parent’s house, and subsidized child care. She could count on food stamps and KidCare health insurance for her children, and up to $398 a month in cash benefits. Even so, she was barely making ends meet. Twelve months later, this single mother of five, ages 3 to 16, has no job, is just months from losing her unemployment insurance, and will have to reapply for public aid. Though the court has ordered child support, she received only $50 last year and has scant hope of more, as the father of her children is unemployed. Yet Margaret is surprisingly happier now, and her new-found independence has made her optimistic about the future. She is anxious to get a part-time job and resume her education. Her landlord wants to rehab his property so she may be forced to move again, but she looks forward to a nicer home in the suburbs. She does not view welfare as a long-term solution to her problems. Margaret turned her life around when she decided to leave the unhappy, strife-ridden home she and her children shared with their father and his parents in a rundown neighborhood. The decision to leave him and find a new home was costly. First, she gave up the child care the grandmother was providing: “If I hadn’t gotten that, child care [costs] alone would kill me,” she said. And the instability and homelessness that followed meant she had to quit a job that was affording her not only a salary, but stock options she was counting on for her children’s education. Her two-hour commute to her job had never been easy. But the distances from the shelters were even longer and Margaret was forced to give up her job to care for her


children. Whether working, looking for work, or staying home with the kids, she has little time for herself, for fun, or for sleep. Like many single mothers with big families, housing was a major stumbling block for Margaret. “When [landlords] hear five kids, they want to run,” she said. Until she found her new apartment, the family lived in three homeless shelters, sometimes dragging their meager possessions through the streets in plastic garbage bags. Even shelters were hard to find; several family shelters were unwilling to take in a 16-year-old boy. “They really need more shelters,” she said, “and they need more ways to help people find affordable housing.” Margaret found cash benefits undependable and confusing when she was working; some months she got them, some she didn’t. A “good week” at work when she got several shifts meant she didn’t qualify for benefits, while a “bad week” at work meant she would receive benefits a few months later. When she lost her Medicaid she had no medical coverage because she could not afford the $50 per week for employer-provided medical and dental insurance. Now that she is not working, she still can’t make sense of the regulations: “When I was working, public aid gave me more, and now that I’m unemployed, they cut my food stamps.” Caseworker assistance, too, was erratic. In one year Margaret had three different caseworkers, none of whom ever helped, she says. In one instance, a caseworker’s error caused her benefits to be cut off. The only agency that really came through was Employment Services, which linked her to the hotel job she eventually had to give up. Though her life seems chaotic and difficult, her five children are bright and healthy and do well in school, and that is all-important to her: “I tell the kids, thank God, that at least we’ve got a house and we’re not hungry.” *Not her real name. Some circumstances have been altered to protect the anonymity of the respondent. Nina’s Story: Just Getting By When it was time to find a job, Nina had to overcome a history of domestic abuse, eight years of welfare dependence, health problems, negligible skills, and few social supports. A year later, she is scared and struggling, but is persevering and hopeful about the future. While she is no longer receiving cash assistance or food stamps and her job provides no benefits, a Section 8 housing voucher, subsidized child care, and Medicaid coverage help her to get by. Long-term welfare recipient Nina Morales*, age 45, endured an abusive marriage before she struck out on her own with her 5-year old daughter and was compelled by TANF rules to take classes or find a job. Frightened at the prospect of entering the labor market after eight years, she has nonetheless met both assignments head-on—working full-time for a year, and closing in on a teaching certificate that may eventually secure her a better job.


Her pay right now is low, her cash benefits and food stamps are gone, and she is struggling to survive. Not surprisingly, rent, utility costs, groceries, transportation, and child care co-pays eat up virtually all of her earnings from her $7.25-an-hour daycare job. She gets no child support from her estranged husband and thinks a court fight would not be worth the effort. She also receives no benefits from her job. And she cannot afford a phone. Nina also battles health problems. She has missed many days of work and classes visiting doctors for chronic leg pain, and fears she may lose her job. Though she still has a medical card (Transitional Medicaid) she worries that too may be taken away. She has few social supports. Far from her family in Mexico and a brother in New Jersey, Nina feels very isolated, and depends on neighbors for some sense of security. Once she decided to end her relationship with her abusive husband, Nina sought refuge in a battered women’s shelter. A worker at the shelter helped Nina find her Section 8 apartment that costs her $238 a month. But the neighborhood feels unsafe; gangs and drug dealers are rampant; and arsonists and vandals keep her on edge: “I’m scared to go to sleep at night. Our building does not have good security. I literally sleep with my doors blocked. Sometimes I pace back and forth throughout the night because I’m so scared,” she says. Finding a job was also a “scary prospect,” for a homemaker who could not even type a resume. “I hadn’t worked in years and I had no experience,” she recalls. “I thought they would look at me and laugh.” However, the volunteering she’d done at her daughter’s daycare center “looked real good when I was applying for jobs,” she says. At the end of a workday that sometimes lasts until 6 PM, she goes off to classes three nights a week while her five-year-old tags along and stays in the school’s evening child care program. They don’t get home until 10:00 pm on these nights. Her three-plus years of diligence is about to pay off. With a semester to go, she will earn an associate’s degree in early childhood education and a teaching assistant certificate even sooner. Nina thinks welfare recipients lose benefits too abruptly when they get a job and they should get to keep them until they know their work life is stable: “They shouldn’t just throw you off and tell you ‘that’s it.’ Especially because I had not worked in so long. Then for them to cut me from food stamps was harsh without yet knowing if my job was even secure.” Though her life is hard right now, Nina sees many bright spots in her life: a healthy daughter, a job, good neighbors, and the ability to pay her bills. “I’m happy I’m making it, you know. Everything I’ve done, I did it on my own. It may not be that much yet, but I did it all by myself.”

*Not her real name. Some circumstances have been altered to protect the anonymity of the respondent.

welfare reform in illinois - northwestern university · secretary of the illinois department of...

Documents