Economic Status of Welfare Recipients REPORT #02-05

OFFICE OF THE LEGISLATIVE AUDITOR STATE OF MINNESOTA

PROGRAM EVALUATION REPORT Economic Status of Welfare Recipients

JANUARY 2002

Photo Credits: The cover and summary photographs as well as those on pages 7, 16, and 30 were provided by Lifetrack Resources; BlueCross BlueShield of Minnesota furnished the photograph on page 14; and the picture on page 11 was taken by staff of the Office of the Legislative Auditor.

Centennial Building, 658 Cedar Street, St. Paul, Minnesota 55155 Tel: 651/296-4708 Fax: 651/296-4712 E-mail:legislative.auditor@state.mn.us TDD Relay: 651/297-5353 Website: www.auditor.leg.state.mn.us OFFICE OF THE LEGISLATIVE AUDITOR STATE OF MINNESOTA James Nobles, Legislative Auditor

January 28, 2002 Members Legislative Audit Commission
In the last decade, Minnesota has fundamentally changed its programs to assist low-income families. In May 2001, the Legislative Audit Commission directed OLA to study how well these programs support needy families, encourage work, and discourage dependency. We focused on the economic status of current and former recipients of the state's primary "welfare" program— the Minnesota Family Investment Program (MFIP). We found that Minnesota's new approach to welfare has achieved many positive results. To a significant degree, the system now rewards work instead of dependency. A majority of families that were on MFIP in 1998 had a family member working in 2000. Their earnings and the government assistance they received— in cash and non-cash resources— lifted most of these families above the federal poverty level. But we also found that some families on MFIP are not working or earning enough to leave the program.

This report was researched and written by John Patterson (project manager), Adrienne Howard, and Dan Jacobson. We received the full cooperation of the Department of Human Services, Department of Revenue, and other agencies from which we collect data and information.

Sincerely, /s/ James Nobles /s/ Roger Brooks James Nobles Roger Brooks Legislative Auditor Deputy Legislative Auditor

Table of Contents
SUMMARY
INTRODUCTION
GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES
Program Descriptions
Recent Changes in Government Assistance and Tax Credit Programs
2. THE ECONOMIC STATUS OF CURRENT AND FORMER MFIP RECIPIENTS
1998 MFIP Families' Economic Status in 2000
Incentives to Work
Dependency on Government Assistance
Comparing Minnesota With Other States
Other Considerations
3. VARIATION IN ECONOMIC STATUS BY DEMOGRAPHIC FACTORS 49
Variation in Economic Status
Family Structure
Number of Children
Education
Citizenship
Race
Region
Sanction Status
APPENDIX A: Methodology and Sample Design
APPENDIX B: Historical Program Data
APPENDIX C: Major Assumptions in the Incentives Model
FURTHER READING
AGENCY RESPONSE
RECENT PROGRAM EVALUATIONS
Back Cover
List of Tables and Figures
Tables Page 1.1 Caseloads, Expenditures, and Average Benefits for Government Assistance and Tax Credit Programs, 2000 5 1.2 MFIP Income Standards, October 1, 1999 to September 30, 2000 6 1.3 Program Expenditure Changes, 1989-99 21 2.1 Factors for Computing Net Income 25 2.2 U.S. Census Bureau's Poverty Thresholds, 2000 26 2.3 Percentage of Cases Under the Federal Poverty Threshold, For Various Definitions of Income, 2000 27 2.4 Average Incomes, 2000 28 2.5 Cash and Non-Cash Resources at Various Employment and Wage Levels, 2000 and 1988 35 2.6 Percentage of Cash Resources by Source, 2000 38 2.7 Percentage of Cases Receiving Various Cash Resources and Paying Taxes, 2000 39 2.8 Non-Cash Resources as a Percentage of Total Resources, 2000 40 2.9 Percentage of Cases Receiving Various Non-Cash Resources, 2000 41 2.10 Poverty Rates for Former Welfare Recipients From Selected State Studies 44 3.1 Average Resources by Change in Family Structure, 2000 51 3.2 Average Resources by Number of Children, 2000 52 3.3 Average Resources by Education of Parent, 2000 53 3.4 Average Resources by Citizenship Status, 2000 54 3.5 Average Resources by Race of Family Head, 2000 55 3.6 Average Resources by Region, 2000 56 3.7 Average Resources by Sanction Status, 2000 57 A. 1 Income Sources and Tax Liabilities Included and Not Included in Analysis 60 A. 2 Types of Families in Working Sample 62 A. 3 Comparison of Samples 66 B. 1 Monthly Average Cases/Recipients for Government Assistance and Tax Credit Programs in Minnesota, 1988-2000 (in Thousands) 68 B. 2 Monthly Average Cases/Recipients for Government Assistance and Tax Credit Programs in Minnesota, per Thousand Households, 1988-2000 69 B. 3 Annual Government Payments for Government Assistance and Tax Credit Programs in Minnesota, 1988-2000 (in Millions of Year 2000 Dollars) 70 B. 4 Annual Average Benefit Per Case/Recipients for Government Assistance and Tax Credit Programs in Minnesota, 1988-2000 (in Year 2000 Dollars) 71 C. 1 Incentives Model Assumptions
Figures Page 1.1 Total Expenditures for Government Assistance and Tax Credits in Minnesota, 1989-99 19 2.1 Annual Total Resources After Out-of-Pocket Child Care Expenses at Various Hourly Wages 32 2.2 Annual Cash Resources After Out-of-Pocket Child Care Expenses at Various Hourly Wages 33 2.3 Annual Non-Cash Resources After Out-of-Pocket Child Care Expenses at Various Hourly Wages 33 2.4 Annual Total Resources Excluding Medical Subsidies 36 2.5 Government Assistance and Privately Generated Income as a Percentage of Total Resources 37 2.6 Self-Reported Financial Situation of Former MFIP Recipients 46 2.7 Self-Reported Qualify of Life for Former MFIP Recipients 47 2.8 Minnesota Unemployment Rates and MFIP Caseloads 47 3.1 Income Distribution in 2000 of Families on MFIP in 1998 50

viii ECONOMIC STATUS OF WELFARE RECIPIENTS
Summary Major Findings
Under the Census Bureau's official measure of income, which only includes cash income, 45 percent of families that were on MFIP in 1998 had incomes above the federal poverty threshold in 2000. When both cash and non-cash resources were included, the rate increased to 80 percent (p. 27 of the full report). Families that left MFIP by 2000 had a higher economic status than the families that remained on the program (p. 30).
In general, current and former MFIP families that had one parent, three or more children, or a family head without a high school degree tended to have fewer economic resources as a percentage of the poverty level in 2000 (p. 49).
Families that moved off MFIP were not very dependent on government assistance. Families still on MFIP received 82 percent of their resources from the government, while former recipients only received 13 percent (p. 37).
Compared with a decade ago, Minnesota's package of government assistance is more focused on assisting low-income working families than providing financial assistance to those without jobs. Since 1988, the state and federal governments have increased the incentive for welfare recipients to work, imposed work requirements and time limits, and expanded programs that assist low-income working families (p. 20).
In 2000, the financial status of MFIP recipients increased significantly when they started working at minimum wage. After that, however, the financial incentive to work additional hours or find a better paying job diminished. As earnings increased, government assistance declined by a nearly equivalent amount (p. 32).

Minnesota's system of government assistance for low-income families provides substantial support but also encourages recipients to become more self-sufficient.
Report Summary Minnesota has fundamentally changed its welfare system in the last decade. Most importantly, the Legislature replaced the Aid to Families with Dependent Children (AFDC) program with the Minnesota Family Investment Program (MFIP) in 1997. Four goals of MFIP are to (1) encourage participants to work, (2) prevent long-term dependency, (3) help families increase their income, and (4) bring families out of poverty. 1

While MFIP expects families to become less dependent on welfare, some legislators question the ability of some families to meet their financial needs. This question will not be completely answered until families lose cash assistance under the program's five-year time limit and the state goes through a full recession. Nevertheless, so far, MFIP and other government assistance programs have brought many families above the federal poverty threshold and have given recipients an incentive to work and leave MFIP. Yet, some families remain in poverty and on MFIP.

We based this analysis on a sample of 1,159 families that were on MFIP in 1998 and their economic status in calendar year 2000. In 2000, 403 of these families were on MFIP the entire year, 384 were on MFIP part of the year, and 372 were off MFIP the entire year. To determine net income, we examined 28 income sources and 3 tax liabilities. (See Table 2.1 on page 25 of the full report for a complete listing.) However, we excluded medical and child care subsidies from some analyses for methodological reasons. In addition, we were unable to obtain data on financial support from family and friends and "cash" jobs from the underground economy.

MFIP and Other Government Assistance Programs Have Brought Many Families Above the Federal Poverty Threshold

The federal government determines poverty thresholds to measure the income needed to meet a family's basic needs. However, the current thresholds have been criticized for being too low. Despite these criticisms, the thresholds are still the most widely cited measure of poverty. Forty-five percent of 1998 MFIP families had incomes above the poverty threshold in 2000, using the federal government's official measure of income (which only includes cash income). 2 When non-cash supports are added to cash income, 80 percent of these families had incomes above the poverty threshold. Non-cash resources contributed significantly to the economic status of these families.

Minnesota's government assistance programs lift many, but not all, families above the federal poverty threshold. The families who fare more poorly tend to have one parent, three or more children, or a family head who lacks a high school degree. Interestingly, families with one parent in 1998 that became two-parent families by the end of 2000 fared the best. Families that remained one-parent families had similar economic situations with those that were two-parent families in 1998 and remained that way. These two-parent families had larger families, less education, and a larger percentage of disabled adults than other MFIP families, which probably offset the economic advantage of having a second potential wage earner.

While non-cash benefits are not part of the official poverty calculation, they contribute significantly to improving the economic status of MFIP recipients.

1 While state law does not provide explicit goals for the program, these goals are stated in the MFIP employment services manual or implied in the program's design. 2 The official measure of income includes cash resources before taxes but excludes capital gains.

MFIP Gives Families Incentives to Work and Leave the Program
For MFIP recipients to have an incentive to work and leave the program, they need to understand that they will have a better financial situation when they are off the program. MFIP recipients from 1998 who left the program by 2000 had greater average resources (cash and non-cash) than those recipients still on welfare in 2000 ($23,511 annually compared with $21,811). Former MFIP recipients also had fewer economic needs than current recipients because the former recipients had smaller families (3.0 people on average versus 3.9). Consequently, their average resources (cash and non-cash) were 166 percent of the poverty threshold, compared with 130 percent for current recipients. Nevertheless, former recipients were not conclusively better off than current recipients because we excluded two non-cash resources from our analysis— child care subsidies and private and public medical resources. Generally, families on MFIP receive greater support from these benefits than former recipients because MFIP recipients tend to have fewer cash resources.

In 2000, an MFIP family's financial incentive to work was initially strong but tapered off as earnings increased. If an MFIP family with an unemployed single parent and two children took full advantage of Minnesota's government assistance programs, it would have had $22,055 annually in cash and non-cash resources. 3 If this parent worked full-time at minimum wage, the family's resources would have increased to $29,783. However, at this point, the financial incentives to work more hours and find a better job became less clear. If the parent increased his or her hourly wage all the way up to $20 an hour, the family's resources would have hardly changed, remaining at about $30,000. As the hourly wage increased, the family's cash resources increased, but non-cash resources decreased by a nearly equivalent amount.

While the financial incentive to leave MFIP is not completely clear with respect to increasing a recipient's economic resources, MFIP's work requirements and time limits make the decision much clearer. MFIP requires most adult recipients to participate in at least 30 hours of work-related activities each week. In addition, families face a five-year lifetime limit on MFIP cash assistance. If the parents in these families are deemed employable and do not work at least 25 hours per week, their families will lose cash assistance once they reach the five-year limit.

In addition, as families work and leave MFIP, they become far less dependent on government assistance. In 2000, former recipients received only 13 percent of their total cash and non-cash resources through government assistance, compared with 82 percent for full-year recipients.

SUMMARY xi Families that left MFIP received 13 percent of their resources from government assistance in 2000, while families still on MFIP received 82 percent.

3 This portion of the analysis includes medical subsidies. It also incorporates child care by subtracting the family's out-of-pocket expenses for child care (total cost less the government subsidy) from total resources.

Welfare Recipients Have a Greater Incentive to Work Today Than in 1988
In 1988, a family that had a single parent and two children and took full advantage of government assistance programs would have had essentially the same economic resources if the parent was unemployed or working full-time at $16 per hour. The family's financial situation would not have consistently improved until the parent earned at least $18 per hour. (The 1988 wage figures are adjusted for inflation and expressed in year 2000 dollars.)

Since 1988, the state and federal governments have made several policy changes that give welfare recipients more incentive to increase their earnings. First, the state disregards a higher percentage of a family's earnings when calculating its MFIP grant than under AFDC. Second, MFIP has work requirements and a time limit on cash assistance that did not exist under AFDC. Third, the state and federal governments created and expanded programs that support low-income working families rather than the unemployed. MinnesotaCare, the federal Earned Income Tax Credit, the Minnesota Working Family Tax Credit, and child care assistance grew from only 7 percent of government assistance expenditures in 1989 to 27 percent in 1999. 4

In the last decade, Minnesota has placed a greater emphasis on assisting low-income working families than on helping non-working families.

4 These percentages exclude housing programs because we were unable to get historical numbers for Minnesota from the U.S. Department of Housing and Urban Development.
Introduction
Over the last decade, Minnesota's system of government assistance programs has fundamentally changed. Most significantly, the Legislature replaced the Aid to Families with Dependent Children (AFDC) program with the Minnesota Family Investment Program (MFIP) in 1997. Under AFDC, recipients were not required to work and could remain in the program indefinitely. Under MFIP, the state now requires most adult recipients to participate in work-related activities and limits many families to five years of cash assistance. Unless adult recipients are ill, incapacitated, or hard to employ or work at least 25 hours a week, their families will lose MFIP cash assistance after five years.

In addition to requiring families to work and move off MFIP, the state has tried to lift welfare recipients above the federal poverty level. Families are eligible to receive MFIP until their earnings reach about 120 percent of the poverty level. In addition, the state has created or expanded programs for the working poor— including MinnesotaCare (a health care program for low-income Minnesotans), the Working Family Tax Credit (a refundable income tax credit for low-income Minnesotans with earnings) and child care assistance.

The full ramifications of these changes will not be known until families start losing their MFIP benefits under the five-year limit and the economy goes through a full recession. Nevertheless, the Legislative Audit Commission asked our office in May of 2001 to provide a snapshot of the economic status of current and former MFIP recipients and an analysis of the state's system of government assistance programs. In our study, we posed the following questions:

What government assistance programs are available to lower-income Minnesotans?
How has Minnesota's package of government assistance programs changed in the last decade?
How well are current and former MFIP recipients doing financially?
How does the financial standing of former recipients in Minnesota compare with that of former recipients in other states?
To what extent does the state's package of government assistance benefits encourage MFIP recipients to work more and increase their earnings?
How dependent are current and former MFIP families on government assistance?
Over the last decade, Minnesota has fundamentally changed its assistance programs for low-income families.
To what extent is there variation in economic status among individual MFIP families?
What demographic characteristics are associated with MFIP families' economic status?

To address these questions, we examined income received in the year 2000 by a sample of 1,159 families that were on MFIP in 1998. In 2000, 403 of these families were on MFIP the entire year, 384 were on MFIP part of the year, and 372 had left MFIP. To determine net income, we collected administrative data concerning 28 sources of income and 3 tax liabilities. (See Appendix A for more details about our sample design and methodology.) In addition, we (1) reviewed state and federal statutes, rules, and program documents, (2) created a model of how these income sources interact with each other as earnings increase, and (3) reviewed similar analyses done in other states. While we analyzed how these programs operated and how 1998 MFIP recipients were doing financially in 2000, we also gathered information about welfare recipients in 1988 for comparison. We chose 1988 as the comparison year because we already had information about the use of public assistance benefits by AFDC recipients in 1988 from an earlier report by our office. 1

While other studies have assessed the economic status of former welfare recipients, this report is unique in its scope. During our national literature review, we found that most other studies did not collect administrative data for as many types of cash and non-cash resources as we did. In addition, the other studies typically used only one source for earnings data— either wage records from the unemployment insurance program or income tax records. Both of these sources are incomplete. For example, wage records for the unemployment insurance program exclude self-employment income, while tax records exclude low-income individuals that do not file a return. In contrast, our study used three sources for earnings data: (1) tax records, (2) unemployment insurance records, and (3) MFIP records. Within our time and resource constraints, we felt it was important to get as broad and complete a picture as possible. As this report will show, excluding non-cash resources from the analysis (as some of the other studies did) provides a very different picture of the economic status of current and former MFIP recipients than when they are included.

Chapter 1 describes each of the government assistance programs and how they have changed since 1988. Chapter 2 examines the economic status of current and former MFIP recipients and their sources of income. We also compare the economic status of current MFIP recipients with former recipients and assess the financial incentives that families have to increase their earnings and leave MFIP. Chapter 3 describes the family characteristics associated with the economic status of current and former recipients. Finally, the appendices provide more detail about our methodology and assumptions.

Unlike most other studies that assess the economic status of welfare recipients, we included non-cash resources in our analysis.

1 Office of the Legislative Auditor, Use of Public Assistance Programs by AFDC Recipients (St. Paul, February 1989).
Government Assistance Programs for Low-Income Families
SUMMARY
Minnesota provides a diversity of economic assistance to lower-income residents, including current and former MFIP recipients. The package of government assistance benefits is composed of cash and non-cash assistance and tax credits. Over the past decade, the state has placed a greater emphasis on benefits for working families rather than non-working families. MFIP has stronger work incentives than AFDC did, including work requirements, time limits, and more generous financial incentives. In addition, between 1989 and 1999, government expenditures for government assistance programs increased 45 percent after adjusting for inflation. Increases in expenditures in four programs supporting working families (child care assistance, MinnesotaCare, the federal Earned Income Tax Credit, and the Minnesota Working Family Credit) accounted for much of this increase.

Low-income Minnesotans are financially supported by a variety of goverment assistance and tax credit programs. Because these benefits are available to, and in many cases targeted towards, current and former recipients of the Minnesota Family Investment Program (MFIP), they are an important part of understanding the economic status of these recipients. In this chapter, we address the following questions:
What government assistance programs are available to lower-income Minnesotans?
How has Minnesota's package of government assistance programs changed in the last decade?
We answered these questions by examining state and federal statutes and rules, program documents, and funding and caseload data for 1988 through 2000.

PROGRAM DESCRIPTIONS
In this section, we discuss the significant features of the state's economic assistance programs for low-income families, including eligibility requirements, benefit determination, and program size. Since our analysis of the economic status of MFIP recipients relies on data from the year 2000, we focus on the programs as they existed in that year. However, we also note major program changes instituted by the Legislature in 2000 and 2001. To examine program changes over the last decade, we also discuss significant changes in the programs since 1988. In reviewing the programs, we found that:

To assess the economic status of current and former MFIP recipients, more than MFIP must be considered. MFIP is only one part of a much larger system of government assistance benefits and tax credits in Minnesota.

Minnesota offers three types of economic assistance to low-income families: (1) cash assistance, (2) non-cash assistance, and (3) tax refund programs. Cash assistance programs, such as MFIP and Supplemental Security Income, provide a cash grant that recipients can spend any way they wish, while non-cash assistance programs, such as the Food Stamp program and Medical Assistance, provide a subsidy that recipients can use only for a specific good or service. With "refundable" tax credits, such as the federal Earned Income Tax Credit, low-income recipients receive an income tax "refund" even if they pay no income taxes (including withholdings). Table 1.1 shows caseloads, expenditures, and average benefits in 2000 for these programs in Minnesota. Based on total annual payments, MFIP and SSI are the largest cash assistance programs, and Medical Assistance and housing assistance (Section 8 and public housing) are the largest non-cash assistance programs. Of the refundable tax programs, the federal Earned Income Tax Credit and the state Property Tax Refund are the largest. Below we discuss program requirements for each of the assistance and tax refund programs in more detail.

Cash Assistance Programs Under the various cash assistance programs, recipients receive a cash grant and are generally free to spend the money as they wish. We include MFIP food assistance, which is not a cash grant, in this section because it is part of the MFIP program along with MFIP cash assistance.

Minnesota Family Investment Program (MFIP) In 1996, Congress replaced the Aid to Families with Dependent Children (AFDC) entitlement program with the Temporary Assistance for Needy Families (TANF) block grant. 1 In response, the Minnesota Legislature enacted legislation in April of 1997 to implement MFIP on a statewide basis as Minnesota's TANF program. 2 Unlike the AFDC program that provided cash assistance only, MFIP combines cash and Food Stamps into a single program.

MFIP assists low-income families with children. A family is eligible if its monthly net income (income less deductions) is less than the "standard of need" for its family size. The standards of need for 2000 are presented in Table 1.2. The standard of need is divided into cash and food portions, and each year the state has increased the food portion of the grant but has not changed the cash portion. For example, for 2002, the food portion for a family of three is now $299, but the cash portion is still $532. In fact, the $532 cash benefit standard has not changed since 1986. 3

Minnesota provides a wide variety of assistance to low-income families.

1 Public Law 104-193, The Personal Responsibility and Work Opportunity Reconciliation Act of 1996. 2 Laws of Minnesota (1997), ch. 85, art. 1. This law is codified in Minn. Stat. §256J.
GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES 5
Table 1.1: Caseloads, Expenditures, and Average Benefits for Government Assistance and Tax Credit Programs, 2000
Monthly Monthly Total Annual Annual Average Average Average Payments Benefit per Year Cases Recipients (in Millions) Case/Recipient Cash Assistance Minnesota Family Investment Program a SFY 38,833 N/A $193 $4,981 General Assistance b SFY 9,117 N/A 51 5,604 Emergency and Diversionary Assistance SFY 1,433 N/A 14 c Supplemental Security Income d CY N/A 64,084 272 4,200 Minnesota Supplemental Aid b SFY N/A 35,851 75 2,102 Non-Cash Assistance Food Stamps e SFY 94,231 N/A $171 $1,810

Child Care Assistance SFY 26,257 N/A 159 6,056 Medical Assistance (MFIP and other families only) f SFY N/A 244,111 656 2,689 MinnesotaCare SFY N/A 109,031 158 1,447 General Assistance Medical Care SFY N/A 23,347 128 5,471 National School Lunch (free lunch) g SFY N/A 156,407 52 h 404 i National School Lunch (reduced-price lunch) g SFY N/A 63,633 j 326 i Women, Infants, and Children k FFY N/A 90,093 33 363 Public Housing and Section 8 (1999 data) FFY 78,647 N/A 283 3,603 Energy Assistance FFY 84,115 N/A 64 414 Refundable Tax Credits (1999 data) Earned Income Tax Credit CY N/A 210,724 $305 $1,450

State Working Family Tax Credit CY N/A 203,032 89 437 State Dependent Care Tax Credit CY N/A 35,893 12 327 State K-12 Education Credit CY N/A 57,962 21 369 Property Tax Refund (homeowners) CY N/A 197,500 64 324 Property Tax Refund (renters) CY N/A 268,000 112 417

NOTE: Data are for state fiscal year, federal fiscal year, or calendar year depending on the program. SFY = State Fiscal Year. FFY = Federal Fiscal Year. CY = Calendar Year. N/A = Not Applicable. Cases or recipients are shown depending on whether the respective program provided benefits to a household or an individual. For programs that provide a monthly benefit, the average annual benefit is the total annual payments divided by the monthly average cases or recipients. Cases or recipients for the National School Lunch, Energy Assistance, and tax credit programs are annual caseloads rather than monthly average cases. a MFIP includes federal and state cash assistance. b General Assistance data includes the GA segment of Group Residential Housing, and Minnesota Supplemental Aid data includes the MSA segment of Group Residential Housing. c The monthly average benefit for Emergency and Diversionary Assistance was $831 per case in 2000. An annual average benefit is not presented because Emergency and Diversionary Assistance may not be received for longer than a few months.

d SSI data includes all recipients, not just those in families with children. e Food Stamps includes regular Food Stamp receipt, MFIP food assistance, and Minnesota Food Assistance. f Data for Medical Assistance is only for families with children and excludes data for the aged, blind, and disabled. g Data are for public schools only. h Payments include the aggregate payments allocated by the federal government for both free and reduced-price lunches. State funding for all lunches and additional federal funding for paid lunches are excluded. i The average benefit includes federal and state funds and is based on the assumption that students received the meal each school day for a nine-month school year.

j Expenditures are included with the free lunches listed above. k WIC data shows net costs after the infant-formula rebate.

SOURCE: Office of the Legislative Auditor analysis of data from (1) Department of Human Services, (2) Department of Children, Families, and Learning, (3) Department of Economic Security, (4) Department of Health, (5) Department of Revenue, (6) U.S. Social Security Administration, and (7) U.S. Internal Revenue Service.

MFIP encourages work by allowing families to keep a portion of their earnings without losing their grant. For example, MFIP benefits are calculated by disregarding 38 percent of a family's earnings and then subtracting the remaining income from the "family wage level" listed in Table 1.2. (The "family wage level" is defined as 110 percent of the standard of need.) 4 The difference is the grant. 5

MFIP's expansion of work requirements and creation of sanctions for noncompliance also reflect welfare reform's new emphasis on encouraging work and discouraging dependency. MFIP requires most adult recipients to participate in employment, job search, or job training activities. Single parents are required to participate in these activities for at least 30 hours per week, and parents from two-parent families are required to participate for at least 55 hours in combination. 6 Certain groups of recipients are exempt from participating in employment and training activities. 7 Unless exempted, recipients who fail to comply with work activities or other program requirements are subject to sanctions. In 2000, for the first occurrence of noncompliance, a family's grant was reduced by 10 percent of the standard of need, while for the second or subsequent occurrences, the grant was reduced by 30 percent. 8 Starting in March 2002, counties will have the option to disqualify a family from receiving the MFIP grant after the sixth or subsequent occurrences of noncompliance. 9

Table 1.2: MFIP Income Standards, October 1, 1999 to September 30, 2000
Cash Food Standard Family Family Size Portion Portion of Need Wage Level

1 $250 $108 $358 $394 2 437 192 629 692

3 532 257 789 868 4 621 313 934 1,027 5 697 364 1,061 1,167 6 773 434 1,207 1,328 7 850 469 1,319 1,451 8 916 538 1,454 1,599 9 980 608 1,588 1,747 10 1,035 681 1,716 1,888 (Each additional member) 53 74 127 140

SOURCE: Minnesota Department of Human Services.

MFIP provides both cash and food assistance.

3 Minnesota Department of Human Services, AFDC Manual (St. Paul, December 1988), VII-D( ii); and Minnesota House of Representatives, Research Department, Minnesota Welfare: A Guide for Legislators (St. Paul, January 1985), 4.

4 For determining eligibility for new applicants, the earned income disregard is 18 percent, and income after the disregard is compared with the standard of need, rather than the "family wage level." 5 Minn. Stat. (1998 and Supplement 1999), §256J. 21; and Minn. Stat. (Supplement 1999), §256J. 35.

6 Minnesota Department of Human Services, Minnesota Family Investment Program Employment Services Manual (St. Paul, 2000), 3.1.30.

7 Minn. Stat. (Supplement 1999), §256J. 56.

MFIP further emphasizes work and reduces dependency by establishing a lifetime limit of 60 months for cash assistance. In 2000, most caregivers in Minnesota were subject to a lifetime limit of 60 months of cash assistance. The time limit clock began running in Minnesota on July 1, 1997; however, MFIP recipients in certain circumstances, such as caregivers 60 years or older, could have their clock stopped. 10 Also, in 2001, the Legislature made certain groups eligible for cash benefits beyond the 60-month limit. These include recipients deemed to be ill, incapacitated, or hard to employ. In addition, single parents who participate in work-related activities for at least 30 hours per week and are employed 25 hours per week will also receive an extension. The extension applies to two-parent families if the parents participate in work-related activities a total of 55 hours per week and are employed 45 hours per week. 11

The structure and eligibility requirements of the MFIP program are a significant departure from the AFDC program in 1988. AFDC placed less emphasis on incentives to work by not having a time limit on benefits or as stringent work requirements. For example, AFDC only required recipients to register with the statewide job service in most counties without any obligation to participate. In the few counties operating the Work Incentive Program (WIN), AFDC recipients were required to actually participate in employment and training services. Nevertheless, the exemptions for these programs were very broad, including parents providing full-time care for a child under the age of six. A few other voluntary employment programs were available in some counties. As Chapter 2 will show, financial incentives for recipients to work were also smaller under the AFDC program. AFDC permitted a disregard of one third of earned income for the first four months and $30 per month for the first year, compared with the indefinite deduction of 38 percent allowed under MFIP. 12

GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES 7
MFIP requires most adult recipients to participate in work-related activities.

MFIP limits many families to 60 months of cash assistance.

8 Minn. Stat. (1998 and Supplement 1999), §256J. 46. 9 Laws of Minnesota (1Sp2001), ch. 9, art. 10, sec. 33. Counties will have the option to eliminate both the cash and food portions of the grant or to pay the recipient's rent directly to the landlord up to the amount of the cash portion and then eliminate the remaining cash and food portion of the grant. 10 Minn. Stat. (1998 and Supplement 1999), §256J. 42. 11 Laws of Minnesota (1Sp2001), ch. 9, art. 10, sec. 28. Employed participants must be in compliance with program requirements (not in sanction status) during at least 10 of the 12 months immediately preceding the 61 st month on assistance, including the 60th month.

As shown in Table 1.1, the average monthly number of families receiving MFIP cash assistance in 2000 was 38,833. The average annual cash portion of the grant was $4,981. Total payments for cash assistance were $193 million statewide. Federal funds accounted for 55 percent of the total and state funds for 45 percent.

Supplemental Security Income and Minnesota Supplemental Aid Supplemental Security Income (SSI) is a federally funded and administered program that aids low-income aged (65 years of age or older), blind, or disabled adults and blind or disabled children. 13 A recipient's income after allowable deductions must be less than the maximum monthly SSI benefit, which was $512 for an individual living alone and $769 for a couple living alone in 2000. The SSI program requires states to provide supplemental assistance to certain groups of SSI recipients and permits them to provide the same assistance to others. The Minnesota Supplemental Aid (MSA) program provides assistance to aged, blind, and disabled adults and to blind children. MSA recipients must have monthly gross income (income before deductions) below 300 percent of the SSI benefit and monthly income after deductions below the MSA benefit standards. In 2000, the MSA monthly benefit standards were $573 for individuals living alone and $860 for married couples living alone. 14 Both the SSI and MSA grants are calculated by subtracting the recipient's net income (income after deductions) from the respective benefit standards. 15

General Assistance General Assistance (GA) is a state-funded program providing cash assistance to low-income individuals who are not eligible for MFIP and not receiving SSI or MSA. Single adults, childless couples, and adult children living with their parents are eligible for GA. While families with children were eligible for family general assistance in 1988, these families are currently covered under the MFIP program. GA recipients must meet at least one of a list of categorical requirements indicating that they are unable to work. The income standards for GA have not changed between 1988 and 2002. The monthly income standard is $203 for a single adult and $260 for a married couple with no children. The GA grant is accessed November calculated by subtracting monthly income after deductions from the GA income standards. 16

Supplemental Security Income assists low-income disabled and elderly individuals.

12 Minnesota Department of Human Services, AFDC Manual (St. Paul, 1988), IV-G, VIII-D-I; and Minnesota House of Representatives, Research Department, Minnesota's Public Assistance Programs: A Guide for Legislators (St. Paul, December 1988), 64-67.

13 Individuals receiving SSI or MSA are not eligible to receive benefits from the MFIP program, but other family members may still be eligible.

14 The $860 standard applies to married couples who became eligible for MSA after January 1, 1994. Married couples living alone who became eligible prior to 1994 have a standard of $875. MSA standards differ for individuals and married couples who live with others. 15 Committee on Ways and Means, U.S. House of Representatives, 106 th Congress, 2d Session, 2000 Green Book: Background Material and Data on Programs within the Jurisdiction of the Committee on Ways and Means (Washington, DC: United States Government Printing Office, 2000), 228-229, 232-234, http://www.access.gpo.gov/congress/wm001.html,

29, 2001; Minnesota Department of Human Services, Combined Manual (St. Paul, 2000), 0013.09, http://www.dhs.state.mn.us/ecs/reguproc/cm/master01.htm, accessed November 28, 2001; Minn. Stat. (1998 and Supplement 1999), §256D. 435; Minn. Stat. (1998), §256D. 45; and Minnesota Department of Human Services, unpublished document, COLA Increase (2000).

Emergency Assistance The state's Emergency Assistance program offers aid to meet the emergency needs of pregnant women and families with children under the age of 21. In 1988, Emergency Assistance covered a broader list of expenses than the current program. 17 The state now limits Emergency Assistance to rental payments to avoid eviction, payments of arrearages on mortgages, damage or utility deposits, moving expenses, home repairs, and utility costs. Emergency assistance is typically not available for more than one 30-day period in 12 consecutive months. Generally, Emergency Assistance is the amount needed to cover the emergency less the income and assets available to the family to cover the emergency need. In addition to the Emergency Assistance program, other government assistance programs (such as General Assistance, SSI, and Minnesota Supplemental Aid) offer emergency benefits above the general monthly grant in specific circumstances. 18

Diversionary Assistance Families that face a temporary loss of income or are unable to obtain or retain employment and are at risk of going on MFIP may receive state Diversionary Assistance. The assistance must resolve the emergency and divert the family from MFIP. The grant can be up to four times the monthly MFIP standard of need, and families are then ineligible for MFIP for a corresponding period of time. 19

Non-Cash Assistance Programs Non-cash programs provide specific benefits to the recipients, such as food, housing, or medical benefits. Unlike cash assistance, recipients cannot spend these benefits as they wish.

Food Stamps The federal Food Stamp program assists low-income families in purchasing food. Most eligible households must have net income (income after deductions) below 100 percent of the federal poverty level and gross income (income prior to

GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES
The Emergency and Diversionary Assistance programs provide short-term grants for financial emergencies.

16 Minn. Stat. (1986 and Supplement 1987), §256D. 01, §256D. 05; Minn. Stat. (1998 and Supplement 1999), §256D. 05; Minnesota Department of Human Services, General Assistance Manual (St. Paul, 1988), 5.4.2; and Minnesota Department of Human Services, Combined Manual (2000), 0014.06, 0020.18.

17 In addition to the expenses currently covered, Emergency Assistance in 1988 covered storage costs, specific furnishings and household appliances, clothing, food, drinking water, necessary medical and dependent care, transportation necessary for employment or for medical care, and other items. Minn. Rules (1987), §9500.2820.

18 Minn. Stat. (1998 and Supplement 1999), §256J. 48; and Minnesota Department of Human Services, Combined Manual (2000), 0004.01, 0004.18.

19 Minn. Stat. (1998 and Supplement 1999), §256J. 47. 21deductions) below 130 percent of the federal poverty level. 20 In 2000, a family of three must have had monthly net income below $1,157 and monthly gross income below $1,504. Persons receiving General Assistance or Supplemental Security Income automatically meet the income test for Food Stamps. MFIP recipients also automatically meet this test but, as mentioned above, receive their food assistance through the MFIP program. 21 Unless exempt, all adult recipients must register for work, accept suitable employment, and participate in the Food Stamp employment and training program. In 1988, the employment and training program was offered in only selected project areas. 22

The United States Department of Agriculture (USDA) bases the Food Stamp benefit on its food stamp allotment— the amount needed to purchase an adequate amount of food for a family. The USDA adjusts the allotment each year for inflation. In 2000, the maximum monthly food allotment for a family of three was $335. The program assumes that a family should contribute 30 percent of its monthly income to the purchase of food. Consequently, a family's benefit level is determined by subtracting 30 percent of net monthly income from the monthly food allotment. The difference is the benefit. As shown in Table 1.1, the average annual benefit in 2000 was $1,810, and the average monthly caseload (including MFIP recipients receiving food assistance) was 94,231. The federal government funds the Food Stamp program, and total expenditures in Minnesota were $168 million, with an additional $3 million from the state for the non-citizens who are ineligible for federal food assistance. 23

National School Lunch Program The National School Lunch program is a federal program that subsidizes the cost of school lunches. The program provides a small subsidy for all student lunches, and provides lunches to low-income students on a free or reduced-price basis. Students with household incomes less than 130 percent of the federal poverty level qualify for free meals, and students with incomes between 130 and 185 percent of the federal poverty level qualify for reduced-price meals. All public school districts in Minnesota and over one-third of non-public schools participate in this program. The state requires public schools that participate to provide subsidized breakfasts if at least 33 percent of their lunches are served at a free or reduced price. Students eligible for free or reduced-price school lunch are also eligible for free or reduced-price breakfast if their school participates in the program. 24

As shown in Table 1.1, approximately 156,000 students in public schools were approved for free meals, and 64,000 were approved for reduced-price meals during the 1999-2000 school year. Twenty-four percent of all lunches served were free lunches, and nine percent were at a reduced-price. Federal funds allocated specifically to free and reduced-price lunches in Minnesota amounted to $52 million in school year 1999-2000. 25

The Food Stamp program helps low-income families purchase food.

20 Households with disabled or elderly members must meet only the net income test. 21 Families ineligible for both the cash and food portion of MFIP on the basis of income may still be eligible for food benefits under the Food Stamp program.

22 Committee on Ways and Means, 2000 Green Book, 871-873; Minnesota Department of Human Services, Combined Manual (2000), 0028.06.12; and Minnesota Department of Human Services, Food Stamp Program Manual (St. Paul, 1988), IV-C-1.

23 Committee on Ways and Means, 2000 Green Book, 876-878; and Minnesota Department of Human Services, Combined Manual (2000), 0022.12.01.

24 Committee on Ways and Means, 2000 Green Book, 957; and Minnesota Department of Children, Families, and Learning, Meals for Learning: School Nutrition Programs Fiscal 1998 Annual Report (St. Paul, 1998), 3-4.

Women, Infants, and Children The Special Supplemental Food Program for Women, Infants, and Children (WIC) is intended to supplement the diet of certain groups of low-income individuals who are nutritionally at risk. The program is funded primarily at the federal level, and is administered by the state Department of Health through local WIC agencies. 26 Infants under age one, children under age five, and pregnant and postpartum women are eligible if they meet income requirements and are determined to be nutritionally at risk. Recipients must have family income at or below 185 percent of the federal poverty level. In 2000, the monthly income limit for a family of three was $2,140. 27 Recipients receive vouchers for specific types of food to meet their nutritional needs, in addition to nutrition education and screening. 28

Section 8 and Public Housing Both Section 8 and public housing are federally funded programs that provide rental assistance to low-income families. The United States Department of Housing and Urban Development (HUD) establishes regulations and distributes funding for the programs, and local authorities administer them. Section 8 provides rental subsidies to families living in privately-owned housing, while the public housing program provides publicly-owned housing. "Project-based" Section 8 links a family's assistance to a particular housing unit, while "tenant-based" Section 8 allows a family to choose any available privately-owned housing as long as the housing meets program requirements.

GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES

Section 8 is a federally-funded housing program.
The National School Lunch and WIC programs also provide nutritional assistance.

25 The state also contributes funds for all school lunches and breakfasts in Minnesota. Total payments in school year 2000 were $8.2 million. Minnesota Department of Children, Families, and Learning, unpublished documents, "FNS Meal Counts by Category by Year, Public School Information Only" and "Meal/Milk Reimbursements Report, 7/1999 through 6/2000," received August, 2000. 26 The local agencies are typically county public health agencies. Minnesota Department of Health, Local WIC Project Listing; http://www.health.state.mn.us/divs/fh/wic/local.html; accessed November 12, 2001.

27 Indicates the guidelines in effect from April 1, 1999 through March 31, 2000. Minnesota Department of Health, Minnesota WIC Program Operations Manual (St. Paul, 2000), Exhibit 5-A, "Minnesota WIC Program Income Guidelines."

28 Minnesota Department of Health, Minnesota W. I. C. Program Operations Manual (2000), Sections 5.2.2, 5.2.4, 5.3.

Public housing and project-based Section 8 require that a family's income be below 80 percent of an area's median family income at the time of application. 29 Tenant-based Section 8 requires family income to be below 50 percent of the median, with a few exceptions in which the family income must be below 80 percent of the median. However, the incomes of families participating in the programs are generally much lower due to several additional limitations. For example, in 2000, HUD required local authorities to reserve at least 40 percent of their public housing and project-based Section 8 units and 75 percent of their tenant-based assistance for families with incomes below 30 percent of the median. 30

Under these programs, families generally pay up to 30 percent of their income for rent, with the federal government paying for the rest. However, for Section 8 overall rents are generally capped at a "fair market rent" established by HUD. As shown in Table 1.1, the state had about 79,000 Section 8 and public housing units and received about $283 million in subsidies in 1999. 31

Energy Assistance The Low-Income Home Energy Assistance Program (LIHEAP) is a federally funded block grant program that assists low-income families in paying their energy bills. Recipients must be financially vulnerable to increases in home heating costs and have incomes less than 50 percent of the state's median income. (The income limit was 135 percent of the federal poverty level in 1988.) 32 Depending on family size, income, and type of fuel used, the benefit ranged between 18 percent and 98 percent of a household's heating costs in 2000. The minimum annual payment per household was $100 while the maximum annual payment was $950. 33

Medical Assistance Medical Assistance (MA), Minnesota's Medicaid program, is a federal-and state-funded program that provides health care services for low-income families with children and the aged, blind, and disabled. The federal government establishes broad guidelines regarding eligibility and health care services provided through the program, but permits states some flexibility. 34

In 2000, low-income families could have received between $100 and $950 annually in home heating assistance.

29 HUD determines median incomes for various areas in each state. 30 42 U.S. C. S., §§ 1437a, 1437f, 1437n (2000); 24 C. F. R. §5.607 (2000); 24 C. F. R. §5.653 (2000); 24 C. F. R. §982.201 (2000); and Committee on Ways and Means, 2000 Green Book, 942-945.

31 United States Department of Housing and Urban Development. 32 If applied in 2000, the 135 percent of poverty standard would be a more restrictive standard than 50 percent of the state's median income.

33 Minnesota Department of Jobs and Training, Minnesota Energy Assistance Plan for Fiscal Year 1988 (St. Paul, August 1987), 15-16; Minnesota Department of Economic Security, Energy Programs Manual (St. Paul, 2000), 3-4; and Minnesota Department of Economic Security, unpublished document, "EAP 2000 Income Eligibility Guideline Table, 90 Day Table."

34 Committee on Ways and Means, 2000 Green Book, 889-892.

The income limits for MA vary by category of eligibility. In 2000, parents in households with income below 133 1/3 percent of the last AFDC income standards (a $709 monthly limit for a family of three) were eligible. In contrast, children under age two in households with incomes below 280 percent of the federal poverty level (a $3,239 monthly limit for a family of three) were eligible. 35 Individuals who do not initially meet the MA income limits may become eligible by "spending down," or incurring medical expenses that would reduce their remaining income to a level equal or less than the income limit. In both 1988 and 2000, families receiving AFDC/MFIP were automatically eligible for MA; however, beginning on July 1, 2002, the state will base MA eligibility for MFIP recipients on their income. 36 Between 1988 and 2000, eligibility categories and the types of services provided were expanded and the income limits were raised. For example, in 1988, the household income limit for all children, including children under two, was 133 1/3 percent of the AFDC standard, which was a $709 limit for a family of three. When adjusting for inflation, this income limit is 213 percent less than the $3,239 limit for children under two in 2000. 37

As Table 1.1 shows, in 2000, approximately 244,000 individuals from families with children were eligible for MA. Total MA payments for these people were about $656 million (52 percent from the federal government and 48 percent from the state). 38 For families with children, the annual average benefit per person was $2,689.

General Assistance Medical Care General Assistance Medical Care (GAMC) is a state-funded program that provides access to health care for certain groups of low-income individuals. Individuals must be ineligible for MA and either receive General Assistance or meet the income requirements for GAMC. The income standard for GAMC is the same standard used for parents of children under the MA program (133 and 1/3 percent of the AFDC standard). From April 1999 to March 2000, the monthly limit was $467 for a single adult. As with MA, individuals may "spend down" to meet the eligibility requirements for GAMC. Most, but not all, of the services provided under MA are available to GAMC recipients. 39

GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES
Medical Assistance is jointly funded by the state and federal government.

35 Standards given here are those that were in effect between April 1, 1999 and March 31, 2000. 36 Laws of Minnesota (1Sp2001), ch. 9, art. 2, sec. 15. 37 Minnesota Department of Human Services, Health Care Programs Manual (St. Paul, 2000), 0913.03, http://www.dhs.state.mn.us/hlthcare/ReportsManuals/ManualCounty/master01.htm, accessed November 13, 2001; Minnesota Department of Human Services, unpublished document,

1999 MinnesotaCare, Minnesota MA, and GAMC Income and Asset Guidelines; and Minnesota House of Representatives, Research Department, Minnesota's Public Assistance Programs: A Guide for Legislators (St. Paul, December 1988), 23.

38 Minnesota Department of Human Services, Reports and Forecasts Division, February 2001 Forecast (St. Paul, 2001), 19.

39 Minn. Stat. (1998 and Supplement 1999), §256D. 03; Minnesota Department of Human Services, Health Care Programs Manual (2000), 0907.25, 0907.25.03, 0907.25.05; and Minnesota Department of Human Services, unpublished document, 1999 MinnesotaCare, Minnesota MA, and GAMC Income and Asset Guidelines.

MinnesotaCare Established in 1992, MinnesotaCare is a state-and federal-funded health insurance program for low-and moderate-income individuals and families in Minnesota. The program provides coverage to individuals who do not have access to employer-provided health insurance and who typically have income in excess of the limits for Medical Assistance and General Assistance Medical Care. The MinnesotaCare program replaced the Children's Health Plan, a state-funded program begun in July 1988 that provided primarily preventive health care services to lower-income children. With a few exceptions, MinnesotaCare provides access to the services covered under the MA program. 40

As with the MA program, income limits for MinnesotaCare vary by the category of applicant. For example, families with children and pregnant women must have monthly income at or below 275 percent of the federal poverty level (a $3,181 monthly limit for a family of three in 2000). As with private health insurance, MinnesotaCare recipients must pay premiums. Children with household incomes below 150 percent of the federal poverty level have a monthly premium of only $4. Other participants pay a premium based on a sliding fee that varies with family size and income. In 2000, the monthly premium for a family of three with gross annual income of $18,000 was $46. 41

As shown in Table 1.1, the monthly average number of individuals receiving MinnesotaCare was 109,031 in state fiscal year 2000. Net annual payments from the state and federal governments were $158 million. Average annual government payments per enrollee were $1,447.

Child Care Assistance There are three primary child care subsidy programs in Minnesota: (1) the MFIP Child Care Program, (2) the Transition Year Child Care Program, and (3) the Basic Sliding Fee Program. These programs are funded through state funds and the federal government's Child Care Development Fund and TANF block grant. The MFIP Child Care Program provides assistance to MFIP recipients for the purpose of employment, job search, and training. 42 The Transition Year Child Care Program provides assistance to families exiting MFIP for one year as long as their income remains below 75 percent of the state's median income. However, assistance is available only for the purposes of job search and employment. 43 The Basic Sliding Fee Program provides assistance to Minnesotans with incomes at or below 75 percent of state median income ($38,168 annually for a family of three in state fiscal year 2000) for the purposes of employment, job search, or training. Because the Basic Sliding Fee Program is not an entitlement program, counties have waiting lists for their programs. The monthly average number of families on waiting lists statewide has decreased in recent years (6,906 in 1999 compared with 2,764 in 2000).

MinnesotaCare provides health insurance coverage to many low-income families.

Minnesota has three primary child care subsidy programs.

40 Minn. Stat. (1998 and Supplement 1999), §256L; Minn. Stat. (Supplement 1987), §256.936; and Minnesota House of Representatives, Research Department, Minnesota Welfare: A Guide to Public Assistance Programs in Minnesota (St. Paul, January 1991), 40.

41 Standards listed were in effect between April 1, 1999 and March 30, 2000. Minnesota Department of Human Services, unpublished document, 1999 MinnesotaCare, Minnesota MA, and GAMC Income and Asset Guidelines; and Minnesota Department of Human Services, unpublished document, MinnesotaCare Premiums for April 1999 through March 2000.

The benefit received by a family under all three programs depends upon the rate charged by the child care provider and the family's co-payment amount. All the programs require families to make co-payments if their income exceeds 75 percent of the federal poverty level. Co-payments vary based on family size and income (a family of three with gross annual income of $18,000 paid $48 per month in state fiscal year 2000). The maximum provider rate that a county may subsidize is the 75 th percentile rate for similar child care arrangements in that county. Families may choose providers with higher rates but must pay the difference in addition to the co-payments. 44

The structure of the child care assistance programs has changed significantly since 1988. In 1988, the program for AFDC recipients and the Basic Sliding Fee Program were consolidated within one program, and no Transition Year program was in place. At that time, the program placed special emphasis on certain groups of AFDC caregivers and post-secondary education students by granting counties set-aside funds for these groups. In addition, AFDC recipients were able to deduct a portion of their child care expenses from their income when determining AFDC benefits. 45 Finally, as we will describe later in this chapter, child care assistance has grown dramatically over the last decade.

Refundable Tax Credit Programs Beside cash and non-cash assistance, the federal and state governments subsidize low-income families with "refundable" tax credits. These credits allow qualifying filers to receive a payment from the federal or state government based on their earnings or types of expenses incurred, such as child care expenses. Filers can receive "refundable" credits even when they pay no income taxes (including withholdings). In contrast, the federal and state governments offer several "non-refundable" credits that filers can use only to offset their tax liability. In this chapter, we do not discuss the "non-refundable" credits.

GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES
Minnesota and the federal government also subsidize families through the tax system.

42 The 2000 Legislature made MFIP families participating in social services activities eligible for MFIP child care assistance. Minn. Stat. (2000), §119B. 05.

43 In 2000, assistance under this program was available only to families exiting MFIP due to increased income from employment or child or spousal support. The 2000 Legislature removed this limitation. 44 Minn. Stat. (1998 and Supplement 1999), §119B. 011 -. 16; Minnesota Department of Children, Families, and Learning, CFL Issues State Fiscal Year 2000 Co-Payment Schedules and Announces Child Care Allocations, Bulletin #99-003 (St. Paul, 1999); and Minnesota Department of Children, Families, and Learning, unpublished document, Waiting List History, received August 27, 2001.

45 Minn. Stat. (Supplement 1987), §268.91.

Federal Earned Income Tax Credit The Earned Income Tax Credit (EITC) is a refundable federal income tax credit intended to increase the income of low-income working families. The maximum credit and the program income limit depend on whether filers have zero, one, or two or more children. In 1988, only filers with children were eligible for the credit, and the maximum credit and income limit did not depend on the number of children. 46

At first, the credit increases as earnings increase, up to a maximum credit. In tax year 2000, the maximum credit was $353 for filers with no children, $2,353 for filers with one child, and $3,888 for filers with two or more children. The credit then stays at the maximum as earnings increase until a phase-out income threshold is reached. For example, the phase-out began at $12,700 in 2000 for filers with two or more children. At this point, the credit declines as earnings increase until the filer is no longer eligible. The maximum income level at which filers were eligible for the credit in 2000 was $10,380 for filers with no children, $27,413 for filers with one child, and $31,152 for filers with two or more children. 47

As shown in Table 1.1, approximately 210,000 filers in Minnesota claimed the EITC in tax year 1999. Total payments to Minnesota filers were $305 million, and the average annual credit received was $1,450. Based on an analysis by the Department of Revenue of the Minnesota Tax Incidence Database, 71 percent of income tax filers who met the income eligibility requirements for the EITC in Minnesota in 1998 claimed the credit. 48

The federal Earned Income Tax Credit subsidizes low-income working families.

The maximum federal Earned Income Tax Credit was $3,888 in 2000.

46 U.S. Department of the Treasury, Internal Revenue Service, Package X, Volume 1, Informational Copies of Federal Tax Forms, Instructions to Federal Form 1040 (1988), 39-40, 72-73; and 26 U.S. Code §32 (1988).

47 U.S. Department of the Treasury, Internal Revenue Service, 2000 1040 Instructions (2000), 40-49; and 26 U.S. C. S. §32 (2000).

48 U.S. Department of Treasury, Internal Revenue Service, "Selected Historical and Other Data Tables," Statistics of Income (SOI) Bulletin, 20, no. 4 (Spring 2001): 274; and Analysis by Minnesota Department of Revenue, Tax Research Department. Number of dependents was used as a proxy for number of qualifying children in this analysis.

Federal Additional Child Tax Credit Certain tax filers with three or more children may qualify for the additional child tax credit from the federal government. The credit is based on a rather arcane formula that includes the filer's (1) tax savings from the federal $500 per child tax credit (a non-refundable tax credit), (2) Social Security and Medicare payroll tax, and (3) Earned Income Tax Credit. As Chapter 2 will show, very few welfare families receive this credit.

Minnesota Working Family Tax Credit First implemented in 1991, the Minnesota Working Family Tax Credit (WFTC) is the state's version of the federal EITC. As of January 2000, only ten other states offered a state earned income tax credit. Minnesota's credit is refundable like the federal EITC and requires filers to be eligible for the federal EITC; thus, all federal eligibility requirements and income limits apply to the state credit. When first established, the credit was a fixed percentage of the federal credit (10 percent in 1991 and 1992, and 15 percent between 1993 and 1997). In 1998, the credit was restructured as a percentage of earned income. 49 In 2000, the credit was calculated in much the same way as the federal EITC. The income limits were essentially identical, but the maximum credits were lower. The maximum credit was $88 for filers with no children, $706 for filers with one child, and $1,360 for filers with two or more children. 50

Table 1.1 shows that approximately 203,000 individuals claimed the Working Family Credit in tax year 1999. Payments for the credit totaled $89 million, and the average annual credit received was $437. Based on an analysis by the Department of Revenue of the Minnesota Tax Incidence Database, 68 percent of 1998 Minnesota tax filers who were eligible for the credit based on their income actually claimed the credit. 51

Minnesota Dependent Care Credit Minnesota taxpayers may receive a refundable state income tax credit for child or dependent care expenses if incurred for the purpose of employment or obtaining employment. The credit can equal up to 30 percent of expenses, and claimed expenses could not exceed $2,400 for one dependent and $4,800 for two or more dependents in 2000. Thus, the maximum credit was $720 for one dependent and $1,440 for two or more dependents. Because the state decreases the maximum expenses that filers can claim as their income increases, families with incomes over $31,690 were ineligible for the credit in 2000. The maximum income for the state credit was $24,000 in 1988. 52

GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES
The state implemented its own earned income tax credit in 1991.

49 Minnesota House of Representatives, Research Department, The Federal Earned Income Tax Credit and The Minnesota Working Family Credit (St. Paul, January 2000), 7.

50 Minnesota Department of Revenue, Schedule M-1WFC (St. Paul, 2000); Minnesota Department of Revenue, 2000 Minnesota Individual Income Tax Forms and Instructions (St. Paul, 2000), 16-17; and Minn. Stat. (2000), §290.0671.

51 Analysis by Minnesota Department of Revenue, Tax Research Department. Number of dependents was used as a proxy for number of qualifying children in this analysis.

52 Minnesota Department of Revenue, Schedule M-1CD (St. Paul, 2000); Minn. Stat. (1988), §290.067; and Minn. Stat. (2000), §290.067.

Minnesota Education Tax Credit Minnesota income tax filers can also claim a refundable credit for education expenses for children in kindergarten through twelfth grade. Qualifying expenses include computer equipment (up to $200 per family), tutoring, books, and other materials, but tuition for non-public schools may not be claimed. In 2000, households with incomes of $33,500 or less could have claimed expenses up to a maximum of $1,000 per child and $2,000 per family. The maximum credit began to phase-out at $33,500, and families were no longer eligible at an income above $37,500. In tax year 2002, the credit may be taken against only 75 percent, rather than 100 percent, of expenses. 53

Minnesota Property Tax Refund Since 1975, the Minnesota Property Tax Refund has provided tax relief for low-and moderate-income households who pay property taxes that are high relative to their income. In tax year 2000, homeowners with household income below $71,700 and renters with household income below $41,820 were eligible for the credit. Filers are eligible for a refund if their property tax exceeds a certain percentage of their income. To determine the property taxes paid by renters, the state assumes that property taxes are a fixed percentage of rent (19 percent in 2000). 54 If it is determined that a household paid "excessive" property taxes relative to its income, it will receive a refund for a portion of the excess. In addition to the general property tax refund, homeowners whose property taxes increase by a certain percentage over the previous year (12 percent in 2000) and by a certain magnitude ($100 or more in 2000) are eligible for the "targeted refund," regardless of their income level. 55

In 2000, the maximum refund amount was $510 for homeowners and $1,190 for renters. The 2001 Legislature changed the structure of the property tax refund for homeowners by increasing the income limit from $71,700 to $80,180 and increasing the maximum refund from $510 to $1,500. In 1988, for both homeowners and renters, the maximum income level to qualify for the credit was $35,000, and the maximum refund allowed was $1,100. 56

As shown in Table 1.1, 197,500 homeowners and 268,000 renters filed for the refund for tax year 1999. 57 Payments for the property tax refund totaled $176 million. The average annual refund for homeowners was $324, and the average annual refund for renters was $417. Based on an analysis of the Minnesota Tax Incidence Database by the Department of Revenue, 25 percent of all households (not just those who were eligible) in Minnesota with income below $70,000 filed for the property tax refund in 1998. 58

Minnesotans that pay high property taxes relative to their income can receive a refund.

53 Minnesota Department of Revenue, Schedule M-1ED (St. Paul, 2000); Minnesota Department of Revenue, 2000 Minnesota Individual Income Tax Forms and Instructions (St. Paul, 2000), 13; Minn. Stat. (2000), §290.0674; and Laws of Minnesota (1Sp2001), ch. 5, art. 9, sec. 11.

54 In 1988, the amount of property taxes deemed to have been paid by renters was calculated as the actual amount of property taxes attributable to their rental unit. The amount could not exceed 50 percent of the rent paid by the filer during the year. Minn. Stat. (1988), §290A. 03, subd. 11. 55 Minn. Stat. (2000), §290A. 01-. 27; and Minnesota Department of Revenue, 2000 Minnesota Property Tax Refund Forms and Instructions (St. Paul, 2000), 2 and 7-10.

56 Minnesota Department of Revenue, 2000 Minnesota Property Tax Refund Forms and Instructions (St. Paul, 2000), 12-22; Laws of Minnesota (1Sp2001), ch. 5, art. 4, sec. 2; and Minnesota Department of Revenue, 1988 Minnesota Form M-1PR (St. Paul, 1988).

57 Data for tax year 1999 is based on income for 1999.

RECENT CHANGES IN GOVERNMENT ASSISTANCE AND TAX CREDIT PROGRAMS

To analyze changes in the size of Minnesota's government assistance and tax credit programs in the last decade, we examined aggregate expenditures (direct program payments) between 1989 and 1999, as shown in Figure 1.1. 59 (Because of data availability issues, Figure 1.1 excludes expenditures for housing programs.) We found that:

Between 1989 and 1999, government expenditures on economic assistance benefits in Minnesota increased 45 percent after adjusting for inflation.

GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES 19

$0.0 $0.5 $1.0 $1.5 $2.0 $2.5 $3.0

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Federal Government County Government State Government

Cumulative Total Figure 1. 1: Total Expenditures for Government Assistance and Tax Credits in Minnesota, 1989-99 Billions in 2000 dollars

State Fiscal Year Source: Office of the Legislative Auditor analysis of data from (1) Department of Human Services, (2) Department of Children, Families, and Learning, (3) Department of Economic Security, (4) Department of Health, (5) Department of Revenue, (6) U.S. Department of Agriculture, (7) U.S. Social Security Administration, and (8) U.S. Internal Revenue Service.

The state has taken over most county funding of government assistance for low-income families.

58 Analysis by Minnesota Department of Revenue, Tax Research Department. The definition of household used in the database differs from the U.S. Census definition. For example, the 1996 Tax Incidence Study Database included 2,193,971 households while the U.S. Census reported 1,763,000 households. Minnesota Department of Revenue, 1999 Minnesota Tax Incidence Study (St. Paul, March 1999), 18; http://www.taxes.state.mn.us/reports/fiscal/incid99.html; accessed November 15,

2001. 59 Expenditure figures for some federal programs (Energy Assistance and WIC) are recorded for federal fiscal year, and figures for tax credits and Supplemental Security Income (SSI) are recorded for calendar year. Expenditures for each state fiscal year were estimated for these programs.

The allocation of government assistance and tax credit expenditures among the federal, state, and county governments has also changed between 1989 and 1999. As shown in Figure 1.1, state spending increased by 60 percent and federal spending increased by 48 percent. In contrast, spending by counties decreased by 95 percent. The large decline in county spending reflects the state takeover of most county economic assistance funding during this period. 60

We also analyzed the extent to which each program contributed to the 45 percent increase in total spending between 1989 and 1999. As shown in Table 1.3, six programs accounted for the majority of the increase. These programs were (1) Medical Assistance, (2) the federal Earned Income Tax Credit, (3) Supplemental Security Income, (4) MinnesotaCare, (5) child care assistance, and (6) Minnesota's Working Family Tax Credit. In contrast, AFDC/MFIP had a 47 percent decline in spending for cash assistance between 1989 and 1999. (For Figure 1.1, we combined MFIP food assistance with Food Stamps.) Expenditures for other programs remained relatively constant. (Appendix B provides detailed data on cases, expenditures, and average benefit for each program for the years 1988 through 2000.)

When examining program changes more closely, we found that:
Compared with 1988, Minnesota's package of government assistance benefits in 2000 is more focused on assisting low-income working families than providing financial resources to families without jobs.
Several policy changes in Minnesota's basic cash assistance programs reflect this shift in emphasis between 1988 and 2000. With respect to families with no income, the cash grant has not changed since 1986. The cash grant for a family of three was $532 in both 1988 and 2000. Adjusting for inflation, this is a 32 percent decline. With respect to families with earnings, MFIP has a more generous earned income disregard than did AFDC. In determining the cash grant, AFDC disregarded $30 per month and then 33 percent of the remaining earned income. The $30 disregard was available for only the first year a recipient received earnings, and the 33 percent disregard was available for only the first four months of earnings. In contrast, MFIP disregards 38 percent of earned income indefinitely. In addition, as discussed in the MFIP section of this chapter, MFIP introduced the "family wage level," which is a 10 percent increase in the program's income standard for families with earnings. 61

In their funding decisions, the federal and state governments have also placed a greater emphasis on programs for low-income working families. In the analysis of total government assistance expenditures discussed above, four of the six programs that account for the majority of the overall increase (child care assistance, MinnesotaCare, the federal Earned Income Tax Credit, and the state Working Family Tax Credit) are designed to provide assistance to low-income working families. In state fiscal year 1989, these four programs accounted for only 7 percent of government assistance expenditures. (The Working Family Credit had not yet been established, and the Children's Health Plan, now MinnesotaCare, had just begun operation in 1989.) By 1999, the four programs accounted for 27 percent of total expenditures.

The state has not adjusted the basic AFDC/MFIP cash grant since 1986.

60 Minn. Stat. (2000), §256.025. 61 Minnesota Department of Human Services, AFDC Manual (1988), VIII-D; Minnesota Department of Human Services, Combined Manual (2000), 0018.18, 0020.09.

In the next chapter, we analyze the use of these programs by a sample of current and former MFIP recipients. This analysis further illustrates the policy shift toward assisting low-income working families.

GOVERNMENT ASSISTANCE PROGRAMS FOR LOW-INCOME FAMILIES
Table 1.3: Program Expenditure Changes, 1989-99
Expenditure Percentage Percentage of Growth Change Change in Total Expenditures

Medical Assistance (MFIP and other families only) a $242,434,690 65% 30% Earned Income Tax Credit 226,882,281 259 28 Supplemental Security Income 142,675,809 111 18 MinnesotaCare/Children's Health Plan 140,307,620 10,295 17 Child Care Assistance b 122,708,593 355 15 Working Family Tax Credit 86,539,033 N/A 11 Minnesota Supplemental Aid c 25,573,439 54 3 State K-12 Education Tax Credit 18,375,565 N/A 2 National School Lunch (free and reduced federal funds) d 14,074,918 39 2 Property Tax Refund 10,366,576 6 1 General Assistance Medical Care 9,646,902 8 1 Women, Infants, and Children e 5,430,450 19 1 Food Stamps f 3,245,199 2 0 Emergency and Diversionary Assistance -798,291 -6 0 State Dependent Care Tax Credit -2,352,443 -16 0 General Assistance f -12,059,357 -20 -1 Energy Assistance -33,602,267 -42 -4 AFDC/MFIP (cash assistance) f -192,735,042 -47 -24

All Programs $806,713,674 45% 100%

NOTE: Expenditures were adjusted for inflation to year 2000 dollars using the All-Urban Consumer Price Index. All figures are based on state fiscal years. N/A= Not available. a Medical Assistance excludes expenditures for the aged, blind, and disabled. b Costs include state-funded administrative expenses at the county level and part of county-funded administrative expenses in addition to direct program costs. c Minnesota Supplemtnal Aid includes the MSA segment of Group Residential Housing.

d Payments only include the aggregate payments allocated by the federal government for both free and reduced-price lunches. State funding for all lunches and additional federal funding for paid lunches are excluded. Data are for public schools only.

e WIC expenditures are net costs after the infant-formula rebate received by states. f Data for MFIP includes expenditures for MFIP cash assistance and Family General Assistance, and excludes MFIP food assistance. Food Stamps includes expenditures for regular Food Stamps, MFIP food assistance, and Minnesota Food Assistance. General Assistance data excludes Family General

Assistance and includes the GA segment of Group Residential Housing. SOURCE: Office of the Legislative Auditor analysis of data from (1) Department of Human Services, (2) Department of Children, Families, and Learning, (3) Department of Economic Security, (4) Department of Health, (5) Department of Revenue, (6) U.S. Department of Agriculture, (7) U.S. Social Security Administration, and (8) U.S. Internal Revenue Service.

The Economic Status of Current and Former MFIP Recipients

SUMMARY
Minnesota's system of government assistance appears to be meeting several public policy goals. First, earnings and the state's package of cash and non-cash benefits lifted most 1998 MFIP families above the Census Bureau's poverty threshold in 2000. Over three-quarters of these families were above the poverty threshold, and the average income for all these families was about $22,000. This included families that left MFIP and those still on it. Second, to some extent, the system of benefits rewards work instead of dependency. On average, former MFIP families had significantly more cash income than current recipients ($22,153 compared with $13,674), but this difference diminished significantly when non-cash resources were included in the comparison ($23,511 compared with $21,811). Third, most families (88 percent) had at least one member working at some point during 2000. Fourth, some families moved off MFIP and became less dependent on welfare. About one-third of the families were off MFIP for all of 2000, and this group was far less dependent on government assistance than families that remained on MFIP for all of 2000. The average former MFIP family received only 13 percent of its cash and non-cash income through government assistance, while the average current recipient received 82 percent. While many families have had success under MFIP, some families are not working or earning enough to leave the program.

AS described in Chapter 1, government assistance programs in Minnesota underwent a fundamental change in the last decade. Under the new system, the state expects and encourages families to work and move off the Minnesota Family Investment Program (MFIP). According to documents from the Minnesota Department of Human Services (DHS), MFIP's goals are to:

1. Encourage and enable all families to find employment, 2. Help families increase their income, and 3. Prevent long-term dependence on welfare as a primary source of income. 1

1 Department of Human Services, Minnesota Family Investment Program Employment Services Manual (St. Paul, 2001), 1.1.
In addition, while the goal of lifting families above the federal poverty level is not explicitly stated in state law or DHS documents, it is implicit because families are eligible to continue receiving MFIP until their earned income equals about 120 percent of the poverty level. The state is trying to accomplish these goals by (1) imposing work requirements and time limits for MFIP participants, (2) allowing families to keep a greater share of their earned income without decreasing their MFIP grant, and (3) providing resources for families moving off MFIP, such as MinnesotaCare, child care assistance, and earned income tax credits.

While MFIP expects families to become less dependent on welfare, some legislators question the ability of certain families to meet their financial needs. Two years ago, our office attempted to address this issue but limited the analysis to three sources of income (private income, MFIP, and the earned income tax credits). 2 Yet, as Chapter 1 of this report showed, Minnesota has a wide array of programs from which poor families can receive support. Consequently, the Legislature asked us to evaluate the economic status of current and former MFIP families and their incentives to increase their earnings, taking into account the wide array of government assistance that is available. Specifically, this chapter addresses the following questions:

How well are current and former MFIP recipients doing financially?
To what extent does the state's package of government assistance benefits encourage MFIP recipients to work more and increase their earnings?
How dependent are current and former MFIP families on government assistance?
How does the financial standing of former recipients in Minnesota compare with that of former recipients in other states?
To answer these questions, we developed a random sample of 1,159 families that received MFIP sometime between May and October 1998. In 2000, 403 of these families were on MFIP the entire year, 384 were on MFIP part of the year, and 372 were off MFIP the entire year. For each of these families, we collected data about the amount of income and benefits received and taxes paid in 2000. Table 2.1 lists the types of income and taxes included in our analysis. 3 We combined the data to determine average net incomes and poverty rates for the three types of MFIP families— full-year, part-year, and former participants. Appendix A provides more details about our sample and methodology.

We examined a sample of 1,159 current and former MFIP families.

2 Joel Alter and Dan Jacobson, Office of the Legislative Auditor, memorandum to Representative Lee Greenfield, Poverty Status of MFIP Recipients, March 22, 2000.

3 Unfortunately, we could not get administrative data for Section 8 housing, public housing, National School Lunch, and the Women, Infants, and Children program. As a proxy, we used information that a portion of the families in our sample reported to the Department of Human Services (DHS) as part of a longitudinal study that the department is carrying out. We chose to base our sample largely on DHS' longitudinal sample because DHS has been collecting a wealth of information about these people. Minnesota Department of Human Services, Minnesota Family Investment Program Longitudinal Study: Baseline Report (St. Paul, August 1999); and DHS, Minnesota Family Investment Program Longitudinal Study: One Year After Baseline (St. Paul, December 2000).

THE ECONOMIC STATUS OF CURRENT AND FORMER MFIP RECIPIENTS 25 Table 2.1: Factors for Computing Net Income CASH RESOURCES

Privately Generated Income
Earnings
Child Support
Workers' Compensation
Unemployment Insurance
Other (e. g. interest, dividends, capital gains, veterans benefits, pensions, and Social Security)
Cash Assistance
MFIP Cash
Supplemental Security Income
Minnesota Supplemental Aid
General Assistance
Diversionary Assistance
Emergency Assistance
Refundable Tax Credits and Other Refunds
Federal Earned Income Tax Credit
Federal Additional Child Tax Credit
Minnesota Working Families Tax Credit
Minnesota Dependent Care Tax Credit
Minnesota Education Tax Credit
Minnesota Property Tax Refund

NON-CASH RESOURCES
Nutrition
MFIP Food
Food Stamps and State Food Program
National School Lunch
Women, Infants, and Children
Housing
Section 8 Housing
Public Housing
Energy Assistance Medical
Medical Assistance
MinnesotaCare
General Assistance Medical Care Other
Child Care

TAX LIABILITIES
Federal Income Taxes
State Income Taxes
Social Security and Medicare Payroll Taxes

SOURCE: Office of the Legislative Auditor.

We collected data on 28 sources of income and 3 tax liabilities. Income from sources not reported in Table 2.1 were not included in our analysis. For example, we did not collect information about financial support from relatives and friends or income from "cash" jobs in the underground economy.

1998 MFIP FAMILIES' ECONOMIC STATUS IN 2000

In this section, we compare a family's income to the poverty threshold established by the United States Bureau of the Census. The threshold is supposed to indicate whether a family has the financial resources to meet its basic needs. It was developed in the early 1960s by determining the cost of a minimum diet and then multiplying that cost by three to allow for expenditures on other goods and services. The Census Bureau adjusts the threshold each year for inflation. Over the years, the threshold has been criticized for being too low. 4 For example, the Jobs Now Coalition estimates that a Minnesota family with one parent and two children needed an annual income before taxes of $34,032 in 2000 to meet its basic needs. 5 In contrast, the Census Bureau's official poverty threshold for a family of three was $13,738. Other researchers have suggested thresholds between those two. An evaluation of the various measures of poverty is beyond the scope of this study; we used the Census Bureau's official threshold because it is the most widely referenced benchmark for poverty. (Table 2.2 shows the poverty thresholds for 2000.) Nevertheless, when we report that a family has an income above the Census Bureau's poverty threshold, the family may still have serious difficulty meeting its basic needs. We cannot determine if a family is actually out of poverty because an appropriate measure of poverty is a matter of considerable debate.

Table 2.2: U.S. Census Bureau's Poverty Thresholds, 2000 Family Size Poverty Threshold 1 $8,794 2 11,239

3 13,738 4 17,603 5 20,819 6 23,528 7 26,754 8 29,701 9 and more 35,060

SOURCE: United States Bureau of the Census, Poverty in the United States: 2000( Current Population Reports, Consumer Income, P60-214) (Washington DC: U.S. Government Printing Office, September 2001), 5.

An appropriate measure of poverty is a matter of considerable debate.

4 A report by the Census Bureau outlines the weakness of the current poverty measure and evaluates some alternatives— United States Bureau of the Census, Experimental Poverty Measures: 1990 to 1997 (Current Population Reports, Consumer Income, P60-205) (Washington, DC: United States Government Printing Office, 1999), 2.

5 Herbert Cederberg, Kevin Ristau, and Bruce Steuernagel, The Cost of Living in Minnesota (St. Paul: Jobs Now Coalition, May 2001), 3.
As discussed early in this chapter, an implicit goal of MFIP is to bring families above the federal poverty level, which would theoretically allow them to meet their basic needs. We found that:

About half of 1998 MFIP families had cash incomes below the official U.S. poverty threshold in 2000.
Far fewer of these families had total incomes (cash and non-cash) below the poverty threshold in 2000.
Tables 2.3 and 2.4 show poverty rates and average incomes in 2000 for 1998 MFIP families under four alternative definitions of income. In its official poverty measure, the Census Bureau defines income as all cash income (excluding capital gains) before taxes. The Census Bureau also uses alternative measures of income in some analyses, including (1) all cash income after taxes, (2) all cash and non-cash income after taxes, and (3) all cash and non-cash income after taxes but excluding medical subsidies. 6 Clearly, the definition of income has significant ramifications for assessing a family's economic status. For example, the poverty rate for all types of MFIP cases declines from 55 percent under the official definition of income to 7 percent when all cash and non-cash resources are included, and average income increases from $16,905 to $27,554.

THE ECONOMIC STATUS OF CURRENT AND FORMER MFIP RECIPIENTS 27

Table 2.3: Percentage of Cases Under the Federal Poverty Threshold, For Various Definitions of Income, 2000 Alternative Definitions of Income All Cash All Cash and All Cash and Resources (Except All Cash Non-Cash Non-Cash Status of 1998 Capital Gains) Resources Resources, Excluding Resources, Including MFIP Cases in 2000 Before Taxes a After Taxes b Medical Subsidies c Medical Subsidies d All Types of MFIP Cases 55% 48% 20% 7% Full-Year MFIP Cases 83% 77% 25% 1% Part-Year MFIP Cases 56 41 18 9

Former MFIP Cases 26 23 16 13

NOTE: The two definitions of income that only include cash resources are based on our full sample of current and former MFIP recipients. N equals 403 for full-year recipients, 384 for part-year recipients, and 372 for former recipients. The two definitions of income that include non-cash resources are based on a partial sample. The families in this partial sample completed DHS' longitudinal survey for 2000. N equals 199 for full-year recipients, 213 for part-year recipients, and 184 for former recipients.

a This is the measure of income used in the federal government's official poverty measure. b The same as the official poverty measure but includes capital gains, tax liabilities, and tax credits. c The same as the total cash after taxes measure but includes non-cash resources, excluding medical subsidies. d The same as the total cash after taxes measure but includes all non-cash resources, including medical subsidies.

SOURCE: Office of the Legislative Auditor.

The Census Bureau's official measure of "income" only includes cash resources.

6 United States Bureau of the Census, Income, Poverty, and Valuation of Noncash Benefits: 1994 (Current Population Reports, Consumer Income, P60-189) (Washington, DC: United States Government Printing Office, 1996), xviii – xxiii.

How best to define income is a subject of considerable debate in the economics and public finance literature. While non-cash resources such as medical and housing subsidies provide an economic benefit to recipients, it is unclear how much. Valuing non-cash assistance at the cost of providing it is probably an overstatement. Under prevailing economic theory, cash resources are more valuable to recipients than non-cash resources because the recipients can spend the cash as they wish. 7 For example, if the state offered a recipient of Medical Assistance a choice between participating in the program or receiving $165 in cash per month (the average cost to subsidize an MFIP recipient in 2000), most poor people would probably choose the cash. With the cash, they can spend the subsidy on things other than medical care, and if they need medical care, they could probably still receive free indigent care through hospital emergency rooms. In fact, the Census Bureau estimates that Medical Assistance has no "cash" value for people in poverty. 8 Other non-cash benefits have a "near-cash" value, such as food stamps. 9 Because people would buy about the same amount of food if they were on or off the Food Stamp program, they see food stamps as being very similar to cash. Consequently, it is more appropriate to value these "near-cash" benefits at their cost.

While it is beyond the scope of this report to determine the most appropriate definition of income, we chose to focus on cash and non-cash resources excluding medical subsidies for two reasons. First, placing an appropriate "cash" value on medical benefits is problematic. Second, we did not have access to data on medical benefits provided by employers. This can be a sizable benefit for some people. For example, according to a DHS survey, 47 percent of former MFIP families received employer-based insurance for at least one of its members, while only 2 percent of full-year MFIP families received this benefit. 10 By excluding all medical benefits in our primary measure of income, we treat all families consistently. In most cases, our tables present data based on all four definitions of income.

Table 2.4: Average Incomes, 2000 Alternative Definitions of Income All Cash All Cash and Non-All Cash and Non-Resources

(Except All Cash Cash Resources, Cash Resources, Average Status of 1998 Capital Gains) Resources Excluding Medical Including Medical Household MFIP Cases in 2000 N Before Taxes a After Taxes b Subsidies c Subsidies d Size All Types of MFIP Cases 1,159 $16,905 $17,639 $22,364 $27,554 3.5 Full-Year MFIP Cases 403 $12,732 $13,674 $21,811 $29,316 3.9 Part-Year MFIP Cases 384 15,962 17,428 21,946 27,697 3.5

Former MFIP Cases 372 22,399 22,153 23,511 25,615 3.0

a This is the measure of income used in the federal government's official poverty measure. b The same as the official poverty measure but includes capital gains, tax liabilities, and tax credits. c The same as the total cash after taxes measure but includes non-cash resources, excluding medical subsidies. d The same as the total cash after taxes measure but includes all non-cash resources, including medical subsidies.

SOURCE: Office of the Legislative Auditor.

It is difficult to place a "cash" value on some non-cash benefits.

7 Edgar K. Browning and Jacquelene M. Browning, Public Finance and the Price System (third edition) (New York: MacMillan Publishing, 1987), 278.

8 United States Bureau of the Census, Income, Poverty, and Valuation of Noncash Benefits: 1994, footnote 15 on page xxi.

9 United States Bureau of the Census, Experimental Poverty Measures, 7.

We made one modification to our sample when determining the poverty rates shown in Table 2.3 under the measures of income that include non-cash resources. We were unable to get administrative data for the (1) Section 8 housing, (2) public housing, (3) National School Lunch, and (4) Women, Infants, and Children (WIC) programs, but we received participation data for these programs for just over half of our sample from a survey that DHS carried out in 2000. The poverty rates that include non-cash resources only apply to this limited sample. Nevertheless, with some caution, we used the limited sample as a proxy for the entire sample. The limited DHS sample is a random sample of one-parent MFIP families from 1998 that the department selected for a five-year longitudinal study. In contrast, our sample is representative of all MFIP families from 1998. In creating our sample, we started with DHS' sample but added MFIP families not represented, such as two-parent families. On average, families in DHS' limited sample were smaller and had lower incomes than our complete sample, but the average incomes as a percentage of the poverty threshold for the two samples were statistically the same. Finally, we were able to estimate the cash and non-cash incomes shown in Table 2.4 for the complete sample by using information about housing, National School Lunch, and WIC subsidies from the limited sample as a proxy for the complete sample. Appendix A provides more details on our sample and methodology.

While 80 percent of the limited sample had cash and non-cash resources (excluding medical) above the federal poverty threshold, critics argue that the poverty threshold understates the resources required to meet a family's basic needs. As a higher measure of need, we also examined the percentage of families with incomes above 200 percent of the official poverty threshold. Only 17 percent of the limited sample had resources above the 200 percent level. In Chapter 3, we discuss the variation in economic resources for our sample in more detail.

INCENTIVES TO WORK As discussed at the beginning of the chapter, MFIP has the goal of encouraging people to work and increase their earnings. If working more and moving off MFIP will increase people's economic status, they will have an incentive to do so. In this section, we examine these incentives with two types of analyses. First, we compare the financial status in 2000 of current MFIP recipients as a group with the status of former recipients as a group. Second, we examine the incremental changes in cash and non-cash resources that an individual faces when deciding to start working and later increase his or her hourly wage. We discuss each type of analysis in turn.

THE ECONOMIC STATUS OF CURRENT AND FORMER MFIP RECIPIENTS
Seventeen percent of our sample had cash and non-cash resources above 200 percent of the poverty threshold.

10 Office of the Legislative Auditor analysis of DHS' "MFIP Longitudinal Study: Two Year Follow-Up Survey," questions I3 and L3. N = 184 for former recipients, and N = 199 for current recipients.
Comparing the Economic Status of Current and Former MFIP Recipients in 2000

To have an incentive to work and get off MFIP, recipients need to understand that they will have a higher economic status when they leave the program. We found that:

In 2000, families that were on MFIP in 1998 but left the program by 2000 had cash incomes (after taxes) substantially higher than families still on MFIP.
However, the income advantage of former MFIP families over current recipients narrowed considerably when non-cash resources were also counted as income.

As Tables 2.3 and 2.4 from the previous section illustrated, when examining cash resources after taxes, full-year MFIP recipients had an average income of $13,674 and a poverty rate of 77 percent. In contrast, former recipients had an average income of $22,153 and a poverty rate of only 23 percent. However, full-year recipients received more non-cash resources than former recipients did. Consequently, by including non-cash resources without medical subsidies in the definition of income, the financial situation for the two groups became more similar. The poverty rate for full-year recipients dropped to 25 percent, while the rate for former recipients dropped to 16 percent. This is only a 9 percentage-point difference. 11 Similarly, the average cash and non-cash incomes increased to $21,811 for full-year MFIP recipients and $23,511 for former recipients. Nevertheless, former recipients still appear to be better off financially. On average, their total resources excluding medical subsidies were 166 percent of the poverty level, compared with 130 percent for full-year recipients. 12

In this report, we try to give a complete and consistent picture of each family's financial status, but there are some notable gaps. As discussed earlier, we did not include medical benefits because we lacked data on employer-provided insurance and the "cash" value that recipients placed on medical benefits. Also, we did not include child care subsidies because of consistency issues. People who are not working, looking for work, or attending training or school do not need child care. Once people start one of these activities and need child care, they may receive a child care subsidy but will also have a new expense. Consequently, it would be misleading to include child care subsidies without including the new child care expenses. However, we lacked adequate child care expense data. 13 If we had been able to include data on medical and child care subsidies, the financial situation for all types of MFIP cases would have improved, but current MFIP families would have likely seen the biggest improvement. Generally, lower-income families receive higher medical and child care subsidies.

MFIP recipients have a financial incentive to work and leave the program.

The economic status of former MFIP recipients appears to have been significantly higher than the status of current recipients.

11 The poverty rates are statistically different at a 0.05 significance level. 12 The resources as a percentage of the poverty level are statistically different at a 0.05 significance level.

Child care costs and subsidies can have a sizable impact on a family's financial situation. Based on child care data that we collected from seven large counties, 31 percent of 1998 MFIP families in these counties received government child care subsidies at some point during 2000. 14 On average, the families received the subsidy for eight months of the year. Many other families chose to use alternative child care arrangements, such as family and friends. While MFIP families are entitled to child care subsidies, only 30 percent of the full-year MFIP families received a subsidy. 15

Based on DHS' longitudinal survey, the average family that received a government subsidy had annual out-of-pocket expenses of $509 for child care and received an annual subsidy of $7,983. 16 As described in Chapter 1, child care subsidies decline as family income rises. Consequently, former MFIP families, with their higher than average cash incomes, would receive smaller than average subsidies, while full-year MFIP families, with their lower cash incomes, would have larger than average subsidies. In addition, some families receive free child care services from family and friends. According to the DHS survey data, average annual out-of-pocket expenses were $1,266 annually for former MFIP families and $262 for full-year MFIP families, regardless of whether they received a subsidy or not. 17 We did not include these expenses in our aggregate analysis because DHS' survey primarily applied to one-parent families, and we suspected that two-parent families had significantly different child care needs and expenses.

Considering the data gaps, we cannot definitely say that the financial situation of former MFIP families in 2000 was better than full-year recipients, but the former recipients clearly had more cash income. Consequently, the financial incentive to work more and move off MFIP depended on how the families valued non-cash resources. The greater the value current recipients placed on non-cash resources the more similar their financial situation was to former recipients, and the less incentive current recipients had to work more.

Child care expenses can have a significant impact on the economic status of low-income families.

13 While we also considered excluding the Minnesota Dependent Care Tax Credit because it is a type of child care subsidy, we decided to include it because it is also an income tax benefit provided to low-income families. Nevertheless, the inclusion or exclusion would have absolutely no effect on our findings.

14 These counties are Anoka, Dakota, Hennepin, Olmsted, Ramsey, St. Louis, and Washington and accounted for 68 percent of our overall sample. 15 Not all families need child care, such as those with older children or parents not participating in work-related activities. 16 Minnesota Department of Human Services, data collected by the department for its five-year longitudinal study (N = 147). DHS collected data about a single month, and we multiplied them by 12 for an annual figure. 17 Office of the Legislative Auditor analysis of DHS' "MFIP Longitudinal Study: Two Year Follow-Up Survey," question K20. N = 128 for former recipients, and N = 103 for current recipients. Again, we multiplied DHS' monthly data by 12 to get an annual figure.

Incentives Faced by Individual Families The aggregate analysis from above examined how well a group of individuals at a certain stage of moving off MFIP were doing financially compared with another group at a different stage. We also examined the incentives an individual faces when he or she is making the decision to work or increase his or her earnings. Figures 2.1 through 2.3 illustrate this analysis. We included information for 2000 and 1988 to show how economic incentives have changed over the last decade. We discuss each year in turn.

We found that:
Non-working MFIP recipients have strong economic incentives to work. For MFIP recipients working full-time, however, there is little economic gain when their hourly wage increases beyond minimum wage.

As Figure 2.1 shows, a single parent who had two children and started working full-time at minimum wage in 2000 would have increased his or her family's total economic resources from $22,055 to $29,783. But, after that, the family's income would have remained largely unchanged (at about $30,000) as the parent's wage level increased up to $20 per hour. As the family's cash resources increased (Figure 2.2), its non-cash resources decreased by a nearly equivalent amount (Figure 2.3).

0 5,000 10,000 15,000 20,000 25,000 30,000 35,000

No Work 5.15 Half-Time 5.15 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00

Hourly Wage in 2000 Year Dollars

2000 1988

NOTE: The cash and non-cash resources include: (1) after tax earnings, (2) MFIP cash and food assistance, (3) federal Earned Income Tax Credit, (4) Minnesota Working Family Tax Credit, (5) Minnesota Dependent Care Tax Credit, (6) Minnesota Property Tax Refund, (7) Food Stamps, (8) Medical Assistance, (9) MinnesotaCare, (10) Section 8, (11) Energy Assistance Program, (12) Women, Infants, and Children program, and (13) National School Lunch. In addition, 1988 and 2000 wages are equivalent after adjusting for inflation. In the chart, all dollar figures are expressed in 2000 dollars. Unless stated otherwise, the person worked full-time.

Figure 2. 1: Annual Total Resources After Out-of-Pocket Child Care Expenses at Various Hourly Wages Annual Cash and Non-Cash Resources in Year 2000

SOURCE: Office of the Legislative Auditor. After MFIP recipients worked full-time at minimum wage, their resources remained largely unchanged as their hourly wage increased in 2000.
THE ECONOMIC STATUS OF CURRENT AND FORMER MFIP RECIPIENTS
33 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000

No Work 5.15 Half-Time 5.15 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00

Hourly Wage in Year 2000 Dollars

2000 1988 Figure 2. 2: Annual Cash Resources After Out-of-Pocket Child Care Expenses at Various Hourly Wages Annual Cash Resources in Year 2000 Dollars

NOTE: The cash resources include: (1) after tax earnings, (2) MFIP cash assistance, (3) federal Earned Income Tax Credit, (4) Minnesota Working Family Tax Credit, (5) Minnesota Dependent Care Tax Credit, and (6) Minnesota Property Tax Refund. In addition, 1988 and 2000 wages are equivalent after adjusting for inflation. In the chart, all dollar figures are expressed in 2000 dollars. Unless stated otherwise, the person worked full-time.

SOURCE: Office of the Legislative Auditor.

0 5,000 10,000 15,000 20,000 25,000 30,000 35,000

No Work 5.15 Half-Time 5.15 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00

Hourly Wage in Year 2000 Dollars

2000 1988

NOTE: The non-cash resources include: (1) MFIP food assistance, (2) Food Stamps, (3) Medical Assistance, (4) MinnesotaCare, (5) Section 8, (6) Energy Assistance Program, (7) Women, Infants, and Children program, and (8) National School Lunch. In addition, 1988 and 2000 wages are equivalent after adjusting for inflation. In the chart, all dollar figures are expressed in 2000 dollars. Unless stated otherwise, the person worked full-time.

Figure 2. 3: Annual Non-Cash Resources After Out-of- Pocket Child Care Expenses at Various Hourly Wages Annual Non-Cash Resources in Year 2000

SOURCE: Office of the Legislative Auditor.

As the hourly wage of MFIP recipients increased, their cash resources increased, but their non-cash resources decreased.

Similar to our aggregate analysis of the MFIP sample from the previous section, this analysis of an individual's incentives indicates that families that earned enough to get off MFIP had greater cash resources than those still on it, but the difference in resources diminished significantly when non-cash resources are also included. The economic incentive of people on MFIP (aside from work requirements and time limits) to increase their earnings and move off MFIP depends on how they value the non-cash resources that they receive. People who place a similar value on cash and non-cash resources have less incentive to increase their earnings than people who strongly prefer cash.

We created Figures 2.1 through 2.3 by assuming that a single-parent and two children (one in preschool and one in elementary school) participated in all the government assistance programs listed in Table 2.5 for which they were eligible. 18 We also subtracted out-of-pocket expenses for child care from the cash resources. The family only had to pay $5 per month for child care when the parent worked full-time at minimum wage; however, at $22 per hour, the family lost its child care subsidy and paid $587 per month. Appendix C provides details about our assumptions.

As we discussed earlier in this chapter, there is considerable debate within the economics and public finance literature about how to value non-cash resources, especially medical benefits. In Figure 2.4, we present the same information as in Figure 2.1 but without medical benefits. The overall picture is about the same, but the family would have less income for all wage levels below $20 per hour. Under this scenario, the family's financial situation would improve as earnings increase at the lower wages, then level off, and improve again at the higher levels. Nevertheless, the information in Figure 2.4 is consistent with the overall finding that the incentive to work is initially strong and then tapers off.

MFIP families have incentives to work besides financial incentives. The state requires most adult MFIP participants to participate in 30 hours of work-related activities per week (including employment, job search, training, and/or education). In 2000, those that did not comply faced up to a 30 percent sanction against their MFIP grant. (Starting in March of 2002, counties will have the option of disqualifying from MFIP assistance a family that does not comply with its work requirements.)

There is also a five-year lifetime limit on MFIP cash benefits. In 2000, the limit applied to most families. However, in 2001, the Legislature decided to allow extensions of MFIP cash benefits for certain families that reach the time limit, including families with various hardships or in which the parent works at least 25 hours per week.

MFIP recipients have work requirements and a lifetime limit on cash assistance.

18 For each of the last couple of years, DHS has prepared a similar analysis. In our analysis, we assumed that the family did not receive workers' compensation, unemployment insurance, child support, Supplemental Security Income, Minnesota Supplemental Aid, Emergency Assistance, Diversionary Assistance, General Assistance, the Education Tax Credit, public housing, or General Assistance Medical Care.
THE ECONOMIC STATUS OF CURRENT AND FORMER MFIP RECIPIENTS 35
Table 2.5: Cash and Non-Cash Resources at Various Employment and Wage Levels, 2000 and 1988
Hourly Wage $5.15 Not Half-Year Working Time $5.15 $6.00 $8.00 $10.00 $12.00 $14.00 $16.00 $18.00 $20.00 $22.00
Cash Resources Earnings 2000 $0 $5,356 $10,712 $12,480 $16,640 $20,800 $24,960 $29,120 $33,280 $37,440 $41,600 $45,760 1988 0 5,356 10,712 12,480 16,640 20,800 24,960 29,120 33,280 37,440 41,600 45,760 Federal and State Taxes 2000 0 0 0 0 -94 -318 -811 -1,490 -2,031 -2,648 -3,574 -4,493 1988 0 0 0 0 -319 -1,148 -1,695 -1,401 -2,344 -3,276 -4,353 -5,300 Payroll Taxes 2000 0 -410 -819 -955 -1,273 -1,591 -1,909 -2,228 -2,546 -2,864 -3,182 -3,501 1988 0 -402 -804 -937 -1,250 -1,562 -1,874 -2,187 -2,499 -2,812 -3,124 -3,437 Earned Income Tax Credit 2000 0 2,150 3,888 3,888 3,059 2,175 1,301 427 0 0 0 0 1988 0 751 1,272 1,272 1,039 624 210 0 0 0 0 0 Working Family Tax Credit 2000 0 535 972 972 1,330 1,055 633 200 0 0 0 0 1988 0 0 0 0 0 0 0 0 0 0 0 0 Dependent Care Credit 2000 0 0 17 17 103 138 251 288 0 0 0 0 1988 0 0 16 21 41 85 404 699 210 0 0 0 Property Tax Refund 2000 291 446 491 543 634 705 763 644 451 269 131 22 1988 e 383 556 491 563 635 684 739 569 556 410 213 64 Child Care Expenses 2000 0 0 -60 -60 -396 -576 -1,140 -2,064 -3,600 -5,328 -6,672 -7,042 1988 0 -3,025 a -52 -70 -140 -314 -1,555 -6,600 -6,600 -6,600 -6,600 -6,600 MFIP/AFDC Cash Grant 2000 6,384 4,010 689 0 0 0 0 0 0 0 0 0 1988 c 9,293 7,790 0 b 0 b 0 0 0 0 0 0 0 0
Total Cash Resources 2000 6,675 12,087 15,890 16,885 20,003 22,388 24,047 24,897 25,554 26,869 28,303 30,746 1988 9,676 11,026 11,634 13,329 16,646 19,169 21,189 20,199 22,602 25,163 27,736 30,487
Non-Cash Resources MFIP Food Assistance 2000 $3,084 $3,084 $3,084 $2,677 $98 $0 $0 $0 $0 $0 $0 $0 1988 0 0 0 0 0 0 0 0 0 0 0 0 Food Stamps 2000 0 0 0 0 0 0 0 0 0 0 0 0 1988 d 1,729 1,968 2,031 1,560 565 0 0 0 0 0 0 0 Women, Infants, and Children 2000 363 363 363 363 363 363 363 0 0 0 0 0 1988 545 545 545 545 545 545 545 0 0 0 0 0 National School Lunch 2000 404 404 404 404 404 326 326 0 0 0 0 0 1988 422 422 422 422 422 308 308 0 0 0 0 0 Medical Assistance 2000 5,952 5,952 5,952 5,952 5,952 0 0 0 0 0 0 0 1988 4,273 4,273 4,273 4,273 0 0 0 0 0 0 0 0 MinnesotaCare 2000 0 0 0 0 0 4,716 3,876 3,360 2,616 1,764 0 0 1988 0 0 0 0 0 0 0 0 0 0 0 0 Section 8 Housing 2000 5,477 4,582 3,989 3,666 2,519 1,325 246 0 0 0 0 0 1988 e 5,588 4,207 4,797 4,272 3,180 2,231 1,355 1,620 372 0 0 0 Energy Assistance 2000 100 100 100 100 100 100 100 0 0 0 0 0 1988 f 0 0 0 0 0 0 0 0 0 0 0 0
Total Non-Cash Resources 2000 15,380 14,486 13,893 13,163 9,436 6,830 4,911 3,360 2,616 1,764 0 0 1988 12,558 11,415 12,068 11,072 4,713 3,084 2,208 1,620 372 0 0 0
Total Resources 2000 $22,055 $26,573 $29,783 $30,048 $29,439 $29,218 $28,958 $28,257 $28,170 $28,633 $28,303 $30,746 1988 22,234 22,441 23,701 24,401 21,359 22,253 23,397 21,820 22,975 25,163 27,736 30,487

NOTE: 1988 and 2000 wages are equivalent after adjusting for inflation. All figures are expressed in year 2000 dollars. a Although the family had out-of-pocket child care expenses of $3,025, its AFDC grant was also $3,025 higher than if the family did not have child care expenses. b The family was still eligible for AFDC at these wages levels, but the family's financial situation improved if it dropped AFDC in order to receive a higher child care subsidy. The resulting increase in child care subsidies was greater than the loss of AFDC assistance.

c The AFDC grant calucations ignores the $30 and 1/3 earned income disregard because the $30 disregard only applied to the first year of earnings and the 1/3 disregard only applied to the first four months of earnings. In this analysis, we tried to model long-term incentives. d Under our assumptions, food stamps initially increase as income increases due to interactions with other programs. e The Section 8 subsidy in 1988 did not phase out gradually on a consistent basis because of interactions with other programs. As the housing subsidy fluctuated as earnings increased, the family's rent payments also fluctuated, which affected its property tax refund.

f In 1988, families that received Section 8 and had their heating costs included in their rent were not eligible for energy assistance, while in 2000, they were eligible for a $100 subsidy for the year.

SOURCE: Office of the Legislatiave Auditor.
Currently, the state uses both carrot and stick approaches to encourage families to work. Families that work receive the carrot of improving their economic status, while families that do not participate in work-related activities face the stick of a sanction against their MFIP grant. In addition, families that do not work will eventually lose MFIP cash assistance altogether.

For comparison, we also examined the economic incentive to work in 1988. We found that:

The economic incentive to work in 1988 was lower than in 2000.
Figure 2.1 showed that a 1988 family's total cash and non-cash resources remained largely unchanged as the parent went from not working to working full-time at $16 per hour. Figure 2.2 showed that a family's cash income increased as its wage level increased. Thus, there may have been some incentive to work, but less than in 2000. For example, as illustrated in Figures 2.1 and 2.2, a family with no earnings would have had about the same amount of total resources (but more cash resources) in 1988 than in 2000. In contrast, a working family would have had more resources in 2000 than in 1988 at any wage level.

In addition, as discussed in Chapter 1, most adult AFDC recipients were not required to participate in work-related activities, such as employment, job search, or training/education. Furthermore, the state and federal governments did not have a time limit on AFDC assistance. Families could receive welfare benefits for an unlimited period without working.

0 5,000 10,000 15,000 20,000 25,000 30,000 35,000

No Work 5.15 Half-Time 5.15 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00

Hourly Wages in Year 2000 Dollars

2000 1988 Figure 2. 4: Annual Total Resources Excluding Medical Subsidies Annual Cash and Non-Cash Resources in Year 2000 Dollars

NOTE: The cash and non-cash resources include: (1) after tax earnings, (2) MFIP cash and food assistance, (3) federal Earned Income Tax Credit, (4) Minnesota Working Family Tax Credit, (5) Minnesota Dependent Care Tax Credit, (6) Minnesota Property Tax Refund, (7) Food Stamps, (8) Section 8, (9) Energy Assistance Program, (10) Women, Infants, and Children program, and (11) National School Lunch. In addition, 1988 and 2000 wages are equivalent after adjusting for inflation. In the chart, all dollar figures are expressed in 2000 dollars. Unless stated otherwise, the person worked full-time.

SOURCE: Office of the Legislative Auditor.

AFDC did not have a lifetime limit on benefits. There is one incentive to work that we did not include in our analysis for 1988. For the first year that AFDC recipients had earnings, $30 were disregarded when computing the AFDC grant. In addition, for the first four months of earnings, 33 percent of the earnings (after the $30 disregard) were also disregarded. We did not include these work incentives in our analysis because they were short term, especially the more substantial 33 percent disregard. In contrast, the 38 percent earnings disregard for 2000 does not have a time limit. In the analysis, we were trying to illustrate the long-term incentives that these families faced.

DEPENDENCY ON GOVERNMENT ASSISTANCE

As mentioned earlier in this chapter, a third goal of MFIP is to prevent long-term dependency on welfare. We found that:

About one-third of 1998 MFIP recipients were off MFIP for all of 2000, and this group was not very dependent on any government assistance.
As discussed earlier, 372 out of 1,159 families in our complete sample were off MFIP for all of 2000, while another 403 remained on MFIP the entire year. Figure 2.5 presents the breakdown of government assistance and privately generated income for these two groups. For former MFIP families, government assistance accounted for only 13 percent of their total resources (including medical), compared with 82 percent for full-year MFIP families. While we earlier focused on total resources excluding medical subsidies, Figure 2.5 includes medical benefits to show the resources that government devoted to the families, not the value the families placed on the benefits.

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

Former MFIP Families Full-Year MFIP Families Privately Generated Income Government Assistance

Figure 2. 5: Government Assistance and Privately Generated Income as a Percentage of Total Resources

NOTE: The figure applies to cash and non-cash resources (including medical subsidies) and the Department of Human Services' sample for its MFIP longitudinal study. N = 184 for former recipients and N = 199 for full-year recipients.

SOURCE: Office of the Legislative Auditor.

Percentage of Total Resources Only 13 percent of former MFIP families' resources came from government assistance in 2000.

The three primary sources of cash income after taxes for the entire sample were earnings (65 percent of the total), MFIP cash assistance (15 percent), and federal and state earned income tax credits (9 percent). Table 2.6 shows the complete breakdown. Similarly, these were the three sources of income with the highest percentage of recipients as shown in Table 2.7. However, the percentage of cash income provided by each source varied significantly by the type of MFIP case. For example, while most families from each type of MFIP case had earnings, earnings accounted for 87 percent of former MFIP families' cash income but only 33 percent for full-year MFIP families. While we did not analyze hourly wages or hours of work in this study, hours of work probably accounted for part of the difference. In an earlier welfare reform study, we found that during May 1999 only 28 percent of MFIP cases with an eligible adult had at least 20 hours of work per week. 19

Table 2.6: Percentage of Cash Resources by Source, 2000 Status of 1998 MFIP Cases in 2000 -Complete Sample a Former Partial-Year Full-Year All Cases MFIP Cases MFIP Cases MFIP Cases

Privately Generated Income
Earnings 64.9% 86.9% 64.2% 32.7%
Child Support 4.1 7.0 3.1 0.9
Workers' Compensation 0.2 0.2 0.1 0.2
Unemployment Insurance 0.8 1.3 0.6 0.3
Other 3. 7 2. 9 5. 3 3. 1
Subtotal 73.7% 98.4% 73.2% 37.2%
Cash Assistance
MFIP Cash Assistance 14.5% 0.0% 12.6% 38.6%
Supplemental Security Income 6.7 2.5 4.6 15.5
Minnesota Supplemental Aid 0.0 0.0 0.0 0.0
General Assistance 0.0 0.0 0.1 0.0
Diversionary Assistance 0.0 0.0 0.1 0.0
Emergency Assistance 0. 8 0. 2 1. 0 1. 7
Subtotal 22.2% 2.8% 18.4% 55.9%

Refundable Tax Credits
Federal Earned Income Tax Credit 7.0% 5.5% 9.8% 5.9%
Federal Additional Child Tax Credit 0.1 0.1 0.1 0.0
Minnesota Working Family Tax Credit 2.1 1.9 2.9 1.6
Minnesota Dependent Care Credit 0.2 0.3 0.2 0.1
Minnesota Educational Tax Credit 0.4 0.3 0.4 0.6
Minnesota Property Tax Refund 1. 1 0. 9 1. 1 1. 4
Subtotal 11.0% 9.0% 14.4% 9.6%

Tax Payments Before Refundable Credits
Federal Income Tax -1.0% -2.1% -0.5% -0.1%
State Income Tax -0.7 -1.3 -0.4 0.0
Social Security and Medicare Tax -5. 1 -6. 7 -5. 1 -2. 6
Subtotal -6.8% -10.2% -6.0% -2.7%
Average Cash Resources $17,639 $22,153 $17,429 $13,674
Number of Cases 1,159 372 384 403

a The "complete sample" refers to all 1,159 families in our sample.
SOURCE: Office of the Legislative Auditor.

For families in our sample, earnings, MFIP cash assistance, and earned income tax credits were the primary sources of cash income.

Non-cash resources accounted for 35 percent of the average family's total resources (including medical subsidies) as shown in Table 2.8. However, full-year MFIP recipients were far more dependent on non-cash income than former recipients. While full-year recipients received 54 percent of their total resources from non-cash sources, former families only received 13 percent from these sources. For the overall limited sample, Medical Assistance, Section 8, and MFIP food assistance were the primary sources of non-cash income. Table 2.9 shows the percentage of families receiving each type of non-cash income.

Table 2.7: Percentage of Cases Receiving Various Cash Resources and Paying Taxes, 2000
Status of 1998 MFIP Cases in 2000 -Complete Sample a Former Partial-Year Full-Year All Cases MFIP Cases MFIP Cases MFIP Cases

Privately Generated Income
Earnings 88% 94% 94% 76%
Child Support 33 50 36 13
Workers' Compensation 2 3 2 1
Unemployment Insurance 7 11 6 3
Other 31 39 34 22

Cash Assistance
MFIP Cash Assistance 66% 0% 94% b 100%
Supplemental Security Income 20 10 16 34
Minnesota Supplemental Aid 1 1 1 0
General Assistance 1 1 3 0
Diversionary Assistance 0 0 1 0
Emergency Assistance 15 5 20 21

Refundable Tax Credits
Federal Earned Income Tax Credit 59% 62% 70% 45%
Federal Additional Child Tax Credit 2 5 3 0
Minnesota Working Family Tax Credit 58 62 69 45
Minnesota Depedent Care Credit 11 20 10 4
Minnesota Educational Tax Credit 16 19 14 15
Minnesota Property Tax Refund 38 37 36 39

Tax Payments Before Refundable Credits
Federal Income Tax 15% 35% 10% 3%

State Income Tax 29 58 26 4 Social Security and Medicare Tax 88 94 94 76

Number of Cases 1,159 372 384 403 a The "complete sample" refers to all 1,159 families in our sample. b Some families on MFIP only receive food assistance.

SOURCE: Office of the Legislative Auditor.

Current MFIP families depend significantly on non-cash assistance.

19 Office of the Legislative Auditor, Welfare Reform (St. Paul, January 2000), 55.
When examining program participation rates, we also found that:
Some families did not take advantage of government assistance benefits for which they appeared eligible.

As described earlier in this chapter, an implicit goal of MFIP is to lift families above the federal poverty level. Families generally remain eligible for MFIP until their earned income reaches about 120 percent of the poverty level. 20 Consequently, if families took full advantage of MFIP and only left once they were above the poverty level, one would not expect former MFIP families to be in poverty under the Census Bureau's official definition. Yet, 26 percent of former families were below the federal poverty threshold in 2000. There are several possible explanations. These families may have:

chosen to leave MFIP because they received financial support from sources outside the scope of this study, including relatives, friends, and "cash" jobs in the underground economy;

left Minnesota and the MFIP program and received financial support and earnings in another state, which is not captured in our data, or left MFIP and lived in poverty. In addition, families did not take full advantage of the earned income tax credits. As Table 2.7 had shown earlier, 88 percent of our overall sample received earnings, but only 59 percent received the earned income tax credit. Some of the highest income families in our sample may have been ineligible for the credit, but nearly all of the full-year MFIP families with earnings should have been eligible. While 76 percent of these families had earnings in 2000, only 45 percent received the earned income tax credit. It is unclear why more families did not participate. Many of these families were not required to file income tax returns and may have been unaware that they could receive the tax credit even if they did not owe taxes. It is also possible that some families did not believe the hassle of filling out the tax forms was worth the benefit of the credit, especially if it was small.

Table 2.8: Non-Cash Resources as a Percentage of Total Resources, 2000
Status of 1998 MFIP Cases in 2000 -Limited Sample a Former Partial-Year Full-Year All Cases MFIP Cases MFIP Cases MFIP Cases

Nutrition
MFIP Food Assistance 6.1% 0.0% 5.9% 12.0%
Food Stamps 0.6 0.5 0.7 0.6
National School Lunch 1.7 1.1 1.9 2.0
Women, Infants, and Children 0. 5 0. 1 0. 5 0. 9
Subtotal 8.9% 1.7% 9.0% 15.4%

Housing
Section 8 6.3% 2.9% 5.4% 10.4%
Public Housing 1.2 0.2 1.2 2.1
Energy Assistance 0. 5 0. 4 0. 5 0. 7
Subtotal 8.0% 3.4% 7.1% 13.2%

Medical
Medical Assistance 17.7% 6.8% 20.1% 25.2%
MinnesotaCare 0.3 0.9 0.2 0.0 General Assistance Medical Care 0. 1 0. 1 0. 1 0. 0 Subtotal 18.1% 7.7% 20.4% 25.3%

Total Non-Cash Resources 35.0% 12.8% 36.5% 53.9%

Total Cash Resources 65.0% 87.2% 63.5% 46.1%

Average Cash and Non-Cash Resources $26,493 $26,155 $27,181 $26,058

Number of Cases 596 184 213 199 a The "limited sample" refers to the 596 families that responded to the Department of Human Services' longitudinal survey for the year 2000.

SOURCE: Office of the Legislative Auditor.

Not all eligible families participated in MFIP in 2000.

20 This assumes that the family has no unearned income, such as dividends and interest.

As discussed in Chapter 1, legislators are interested in how government assistance programs in 2000 compare with the past, including participation. In 1988, our office analyzed the use of public assistance benefits by recipients of Aid to Families with Dependent Children (AFDC). 21 Because the methodology for the

THE ECONOMIC STATUS OF CURRENT AND FORMER MFIP RECIPIENTS 41
Table 2.9: Percentage of Cases Receiving Various Non-Cash Resources, 2000
Status of 1998 MFIP Cases in 2000 -Limited Sample a Former Partial-Year Full-Year All Cases MFIP Cases MFIP Cases MFIP Cases

Nutrition
MFIP Food Assistance 69% 0% 99% b 100%
Food Stamps 23 15 32 21
National School Lunch 57 42 66 62
Women, Infants, and Children 22 8 22 35

Housing
Section 8 34% 23% 30% 49%
Public Housing 6 1 7 10
Energy Assistance 31 19 31 42

Medical
Medical Assistance 86% 55% 100% 100%
MinnesotaCare 6 15 4 1
General Assistance Medical Care 2 1 4 0

Number of Cases
596 184 213 199
a The "limited sample" refers to the 596 families that responded to the Department of Human Services' longitudinal survey for the year 2000. b Technically, this should be 100 percent, but the Department of Humans Services' data system records a few cases in which the family received cash assistance but not food assistance.

SOURCE: Office of the Legislative Auditor.

In addition, not all eligible families filed for earned income tax credits.

21 Office of the Legislative Auditor, Use of Public Assistance Programs by AFDC Recipients (St. Paul, February 1989).
1988 study was different from this current study, comparisons should be made with caution. In the previous study, we examined the use of public assistance programs in 1988 by a random sample of AFDC recipients from 1988. In contrast, in our current study, we examined the use of public assistance programs in 2000 by a random sample of MFIP recipients from 1998. Because of this two-year lag, the sample of families that were still on MFIP for all of 2000 represented longer-term recipients and excluded families that moved off MFIP quickly (the former and part-year recipients). In addition, for 1988, we looked at the use of programs in one month; for 2000, we examined participation at anytime during the year. Finally, for 1988, we examined recipients of AFDC cash assistance, while for 2000, we analyzed recipients of MFIP cash and food assistance.

Nevertheless, when comparing families that remained on MFIP for all of 2000 from our current study with the 1988 AFDC sample, we found that:

Even with differences in study methodology, participation rates in public assistance programs appear to be similar for welfare recipients in 2000 and 1988.
The 2000 and 1988 participation rates were respectively, 100 versus 100 percent for Medical Assistance, 100 versus 81 percent for food assistance, 62 versus 57 percent for National School Lunch, 35 versus 34 percent for the Women, Infants, and Children program, 42 versus 45 percent for the Energy Assistance Program, 59 versus 33 percent for Section 8 and public housing, and 30 versus 5 percent for child care. 22 While participation rates for most programs were similar, food, housing, and child care assistance programs appear to have substantial differences between 2000 and 1988. The difference in participation rates for food assistance may be explained by MFIP's merger of AFDC and Food Stamps into a single program. Consequently, MFIP recipients automatically get food assistance. In contrast, in 1988, cash and food assistance were two separate programs, each with its own application. It is possible that not everyone applied for both programs.

Differences in the 2000 and 1988 samples may explain some of the difference in housing participation rates. Our 1988 study found that longer-term AFDC families were more likely to receive a housing subsidy than new AFDC families because of long waiting lists for the subsidy. With long waiting lists, families recently on public assistance are likely to be still waiting for the subsidy. 23 As discussed above, our sample of MFIP recipients had been on MFIP for at least two years, while the 1988 sample included families that had been on AFDC for a short time. Thus, it is not surprising to find a higher participation rate in housing programs for the MFIP sample. Program and funding changes in housing programs between 1988 and 2000 could also explain some of the difference, but we did not have the data to assess this possibility.

Participation rates for child care subsidy programs have increased substantially since 1988.

22 The participation rates for (1) National School Lunch, (2) Women, Infants, and Children program, and (3) housing subsidies actually apply to one month in 2000 rather than the entire year. See Appendix A for more details about our methodology.

Evidence suggests that programmatic and funding changes explain a lot of the difference in child care participation rates between 2000 and 1988. Between 1988 and 1999, the per capita caseload for child care subsidies increased by 188 percent, 24 and funding increased by 603 percent after adjusting for inflation. One of the major policy shifts in the last decade was the expansion of child care subsidies.

COMPARING MINNESOTAWITH OTHER STATES

Following welfare reform, policymakers and researchers from across the country have had a great deal of interest in the economic status of welfare recipients after they exit the programs. Consequently, many states have tracked and monitored the economic status of former welfare recipients. We reviewed studies from a selected group of states to provide a point of comparison for Minnesota. We also compared our findings with a national survey of low-income families conducted by the Urban Institute. We found that:

Former MFIP recipients appear to have fared better economically than former welfare recipients in other states. In part, this likely reflected Minnesota's strong economy and MFIP policies regarding eligibility and benefit levels.

To provide a more accurate comparison with other studies, we used a different sample than we used in the rest of this report. In this comparison sample, we selected families that (1) received MFIP cash assistance (thus, excluding families that received MFIP food assistance only) at some point between May and October of 1998, (2) left MFIP cash assistance before the end of 1998, and (3) remained off cash assistance for a period of at least two months. 25 We measured the economic status of these individuals for calendar year 2000. Thus, the economic status of these recipients was examined 12 to 18 months after their exit from the program. Some of these families had returned to MFIP cash assistance by 2000, and we kept them in our sample.

We compared the economic status of these "former" MFIP recipients with former recipients of cash assistance in a selected group of states, including Colorado, Iowa, Missouri, New Jersey, Washington, and Wisconsin. 26 It appears that former MFIP recipients are less likely to have an income below the poverty threshold than former recipients in these other states. Examining the pre-tax measure of household cash income used in the official poverty measure, 38 percent of former MFIP families had an income below the poverty threshold in 2000. As shown in Table 2.10, the percentage of former recipients below the poverty level in the comparison states was higher, ranging between 46 and 66 percent. 27

THE ECONOMIC STATUS OF CURRENT AND FORMER MFIP RECIPIENTS Researchers from across the country have assessed the economic status of former welfare recipients.

23 Office of the Legislative Auditor, Use of Public Assistance Programs by AFDC Recipients, 21. 24 The per capita refers to Minnesota households rather than individuals. 25 There were 263 families in this sample.

The 1999 National Survey of America's Families conducted by the Urban Institute also provides a useful standard of comparison. This survey is a nationally representative survey of the economic status of low-income families, including current and former welfare recipients. Former recipients were defined as individuals receiving welfare benefits at some point between 1997 and 1999 who were no longer receiving benefits at the time of the survey in 1999. Examining cash income before taxes, 52 percent of these former recipients had incomes below the poverty level, compared with the 38 percent in Minnesota. 28

Table 2.10: Poverty Rates for Former Welfare Recipients From Selected State Studies Period Between Leaving Welfare Data Source for and Assessing Period of Percent below Study Household Income Sample Economic Status Analysis Poverty Level

Minnesota 2000 tax, Recipients between May 12-18 months Year 38% unemployment, and and October of 1998 who MFIP records left prior to 1999 New Jersey a survey Clients from July 1997 26-33 months Month 46 to December 1998 from entry

Iowa b survey Recipients leaving in 8-12 months Month 47 spring of 1999 Washington b survey Recipients who left in 8-12 months Month 58 October of 1998 Missouri survey Recipients leaving during 24-36 months Month 58 fourth quarter of 1996 Colorado c 1999 tax records Recipients between July 1-18 months Year 65 1997 and December 1998 who left prior to 1999 Wisconsin c 1999 tax records Recipients leaving in first 9-12 months Year 66 3 months of 1998

a Measure of income includes two sources of non-cash income: Food Stamps and child care subsidies. b Analysis of single-parent cases only. c Measure of income excluded TANF benefits.

SOURCE: Berkeley Policy Associates, Evaluation of the Colorado Works Program: Second Annual Report (Denver, CO: Office of the Colorado State Auditor, November 2000), 83-89; Mathematica Policy Research, Inc., Iowa Families That Left TANF: Why Did They Leave and How Are They Faring? (Des Moines, IA: Iowa Department of Human Services, February 2001), 36; Midwest Research Institute, Chapter 2 -Household Income and Poverty (Jefferson City, MO: Missouri Department of Social Services, April 2000), 17-18; Mathematica Policy Research, Inc., Current and Former WFNJ Clients: How Are They Faring 30 Months Later (Trenton, NJ: New Jersey Department of Human Services, November 2000), 44; Wisconsin Legislative Audit Bureau, Wisconsin Works (W-2) Program (Madison, WI: Wisconsin Legislative Audit Bureau, April 2001), 43-45; Jean Du, Debra Fogarty, Devin Hopps, and James Hu, A Study of Washington State TANF Leavers and TANF Recipients: Findings from the April-June 1999 Telephone Survey (Washington Department of Social and Health Services, March 2000), 23.

26 While we examined additional state studies, we discuss here only those that presented findings in terms of the percentage of former recipients with incomes above or below the federal poverty level. 27 Because Colorado and Wisconsin excluded government cash assistance from their definitions of income, we also developed a narrow definition of pre-tax cash income that excluded government cash assistance, child support, workers' compensation, and unemployment insurance. Under this narrow definition, 46 percent of "former" MFIP recipients were in poverty in 2000, well below the 65 and 66 percent found in those two states.

In comparing these studies, one needs to consider several caveats. The studies used different (1) research designs, (2) time periods, (3) data sources for income (administrative data or surveys), (4) definitions of income, and (5) samples of former recipients. For these reasons, comparisons should be viewed with caution. In addition to the general limitations on cross-state comparisons, Minnesota's welfare policies may explain why former recipients of MFIP cash assistance appear to fare better than former recipients in other states. Because the income level at which MFIP recipients leave cash assistance (about 85 percent of the poverty level) is higher than most other states, it is not surprising that former MFIP recipients are doing better financially. 29 Finally, in 2000, Minnesota had a very strong economy relative to the rest of the country. Minnesota's unemployment rate was 3.3 percent, compared with 4.0 percent for the entire country, and the state's median household income was $50,865, compared with $42,148. 30

OTHER CONSIDERATIONS This report focuses on the economic status of current and former MFIP families in 2000. There is much more to a family's quality of life than just financial status. Although a single parent may be working, off MFIP, and above the federal poverty threshold, his or her family may not necessarily be better off. The parent may have to work long hours, have long commutes on a bus, and spend almost the entire day away from his or her children. Many of these considerations are qualitative issues and outside the scope of the data we collected. Nevertheless, in its 2000 survey of its longitudinal sample, the Department of Human Services asked former recipients to assess both their economic situation and quality of life.

According to DHS' survey, most former MFIP recipients said that their economic status and quality of life improved after leaving MFIP.

As Figure 2.6 shows, 59 percent of former recipients reported that they had more money left over after paying their bills when they were off MFIP than on, while only 22 percent reported the opposite. This economic improvement appears to translate into an improved quality of life. Figure 2.7 shows that 66 percent of former MFIP recipients said their life off MFIP was better than when on it. MFIP and other government assistance appear to (1) encourage families to work their way off welfare, (2) bring most families above the poverty threshold and, (3) provide a higher quality of life for former MFIP recipients. This assessment applies to current and former MFIP families as a group and does not reflect the conditions of all families.

Welfare recipients in Minnesota become ineligible for cash assistance at a higher income than in many other states.

28 Pamela Loprest, How Are Families That Left Welfare Doing? A Comparison of Early and Recent Welfare Leavers from the series "New Federalism: National Survey of America's Families" No. B-36 (Washington, DC: Urban Institute, 2001); http://newfederalism.urban.org/html/series_ b/b36/b36.html; accessed November 26, 2001. 29 Office of the Legislative Auditor, Welfare Reform, 11. The exit point for MFIP food assistance is about 120 percent of the poverty level.

30 Minnesota Department of Economic Security, unpublished table titled "Minnesota Unemployment Statistics, Seasonally Adjusted," http://www.mnworkforcecenter.org/lmi/laus/mn_ s_ adj. htm, accessed November 13, 2001; and United States Bureau of the Census, unpublished tabled titled "Historical Income Tables – Households," http://www.census.gov/hhes/income/histinc/h08.html, accessed December 10, 2001. 57

Our assessment of the economic status of current and former MFIP recipients applied to calendar year 2000. The economy has slowed down since then, but so far, it has not had a large impact on the MFIP caseload. Figure 2.8 shows that the unemployment rate in Minnesota increased from 3.3 percent to 3.6 percent between July 2000 and August 2001, while the MFIP caseload increased from about 42,000 families to 43,000. 31 If the unemployment rate continues to increase, low-income families will probably become more dependent on MFIP. While these families could face a substantial reduction in cash resources, the reduction in total resources should not be as great. This assessment, however, assumes that the families will continue to receive MFIP. After June 2002, the first Minnesota families will face the five-year time limit on MFIP cash assistance. If a parent is on MFIP for five years, is deemed employable, and works less than 25 hours per week, the family will lose its MFIP cash assistance. 32 Working 25 hours per week could be a substantial challenge for these parents during a recession.

0 10 20 30 40 50 60 70

More Money Off MFIP About the Same Less Money Off MFIP

Figure 2. 6: Self-Reported Financial Situation of Former MFIP Recipients

59

20 22

Percentage of Former MFIP Recipients

NOTE: The Minnesota Department of Human Services (DHS) asked former MFIP recipients if they had (1) more money, (2) about the same amount of money, or (3) less money left over after they paid their bills than when they were on MFIP. N = 272.

SOURCE: Office of the Legislative Auditor analysis of DHS' "MFIP Longitudinal Study: Two Year Follow-Up Survey," question D12.

If the current recession continues, the economic status of current and former MFIP families will likely decline.

31 The MFIP caseloads listed here include child-only cases, which are no longer considered MFIP cases. 32 Parents from two-parent families would need to work 45 hours in combination.

THE ECONOMIC STATUS OF CURRENT AND FORMER MFIP RECIPIENTS
47 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Jul-98 Jan-99 Jul-99 Jan-00 Jul-00 Jan-01 Jul-01 37,000
38,000 39,000 40,000 41,000 42,000 43,000 44,000 45,000 46,000 47,000

SOURCE: Minnesota Department of Economic Security, "Minnesota Unemployment Statistics: Seasonally Adjusted," (2001); http://www.mnworkforcecenter.org/lmi/laus/mn_ s_ adj.htm; accessed November 13, 2001; and Minnesota Department of Human Services.

Figure 2. 8: Minnesota Unemployment Rates and MFIP Caseloads Percentage Unemployed

Unemployment Rate

Month

MFIP Caseload MFIP Caseload NOTE: MFIP caseload numbers include child-only cases, which are no longer considered MFIP cases.

0 10 20 30 40 50 60 70

Life is Better Off MFIP Life is About the Same Life is Worse Off MFIP Back on MFIP

Figure 2. 7: Self-Reported Quality of Life for Former MFIP Recipients

66 59

19 13

Percentage of Former MFIP Recipients

NOTE: The Minnesota Department of Human Services (DHS) asked former MFIP recipients if their life is (1) better, (2) about the same, or (3) worse now that they are off MFIP. N = 273.

SOURCE: Office of the Legislative Auditor analysis of DHS' "MFIP Longitudinal Study: Two Year Follow-Up Survey," question D13.

Most former MFIP recipients reported that they were better off after leaving the program.
Variation in Economic Status by Demographic Factors

SUMMARY
Although MFIP appears to be achieving some public policy goals, some families continued to have incomes below the federal poverty threshold in 2000. In general, families had less economic resources as a percentage of the poverty threshold if they had one-parent, three or more children, or a family head who was an American Indian or had not graduated from high school. Interestingly, one-parent families from 1998 that became two-parent families by 2000 had the best financial situation, even better than families that had two parents for the entire period. For our sample of 1998 MFIP recipients, U.S. citizenship and region were not associated with economic status in 2000.

In Chapter 2, we reported that most 1998 MFIP families had cash and non-cash incomes above the federal poverty threshold in 2000, but a sizeable minority had incomes below the threshold. Thus, it is important to understand what type of families are not doing well economically, and how they differ from families achieving greater economic success. In this chapter, we focus on how various demographic groups fared economically in 2000. In particular, we address the following questions:

To what extent is there variation in economic status among individual MFIP families?

What demographic characteristics are associated with MFIP families' economic status?
To answer these questions, we analyzed the cash and non-cash resources in 2000 for our sample of 1998 MFIP recipients. We included families who remained on MFIP as well as those that left MFIP. To make comparisons among different demographic groups, we relied primarily on two measures of income: (1) cash resources after taxes as a percentage of the federal poverty threshold, and (2) total resources excluding medical subsidies as a percentage of the federal poverty threshold. It is important to recognize that the results presented in this chapter reflect the experience in 2000 of families that received MFIP assistance in 1998. These results may differ significantly from a demographic analysis of the entire state population.

VARIATION IN ECONOMIC STATUS Figure 3.1 illustrates how total resources, measured as a percentage of the poverty threshold, varied among families that were on MFIP in 1998. 1 Throughout this chapter, we excluded medical subsidies from our measure of total resources. As we explained in Chapter 2, some families may have received support from sources outside the scope of our report, particularly families that left MFIP and had extremely low incomes by our measure. We found that:

Total resources as a percentage of the federal poverty threshold varied considerably in 2000 among families that were on MFIP in 1998.

Total resources of current and former MFIP recipients averaged about 147 percent of the poverty threshold, but 10 percent had resources less than 75 percent of the poverty threshold and 10 percent had more than 216 percent. The highest percentage was 470 percent. In the remainder of this chapter, we examine which types of families did better or worse economically in 2000.

0 2 4 6 8 10 12 14 16 18

Figure 3. 1: Income Distribution in 2000 of Families on MFIP in 1998 Percentage of Families

SOURCE: Office of the Legislative Auditor.

0 40 80 120 160 200 240 280 320 360 400 440 480

Total Resources Excluding Medical as a Percentage of the Poverty Threshold

Eighty percent of families on MFIP in 1998 had cash and non-cash incomes above the poverty threshold in 2000.

1 As we explained in Chapter 2, we were able to obtain data on subsidies for housing, National School Lunch, and Women, Infants, and Children programs only for our limited sample of 596 families that responded to the Department of Human Service's longitudinal survey. As a result, we restricted our analysis of the income distribution to this limited sample. When we compared the distribution of the limited sample with the complete sample based on sources for which we had complete data, we found that they were similar. This suggests that the results shown in this section would not change very much if we had complete information for the entire sample. In the remainder of this chapter, we use the full sample to estimate the economic status of different demographic groups. Appendix A explains how we estimated the subsidies under programs for which we lacked complete data.

FAMILY STRUCTURE To analyze how economic status varies by family structure, we grouped families by the number of parents living with the family in 1998 and in 2000. Table 3.1 compares the economic resources of one-and two-parent families. We found:

One-parent families in 1998 who became two-parent families by 2000 were financially better off than families who remained one-parent families. They were also better off than families that had two parents throughout the period.

Eleven percent of families with one parent in 1998 became two-parent families by the end of 2000. On average, the overall resources for these families were 169 percent of the poverty threshold, compared with 144 percent for those who remained as one-parent families. 2 Families who became two-parent families had substantially more cash resources because they earned about 2.5 times as much as did one-parent families ($23,800 compared with $9,400).

Families that recently became two-parent families were also financially better off than families that had two parents throughout this period. 3 In fact, the economic standing of families that remained as two-parent families appears to be similar to that of families that remained as one-parent families. These two-parent families had a larger percentage of disabled adults, larger families, and less education than other MFIP families. 4 Apparently, these financial disadvantages offset the advantage of having two potential wage earners.

VARIATION IN ECONOMIC STATUS BY DEMOGRAPHIC FACTORS 51 Table 3.1: Average Resources by Change in Family Structure, 2000 Resources as a Percentage of the Poverty Threshold Total Resources Type of Case N Cash Income Except Medical One-parent family in 1998 Remained one-parent family in 2000 846 113 144 Became two-parent family in 2000 114 152 169 Two-parent family in 1998 Remained two-parent family in 2000 148 115 146 Became one-parent family in 2000 51 77 111 TOTAL 1,159 116 145

SOURCE: Office of the Legislative Auditor.

One-parent families on MFIP that became two-parent families significantly improved their economic situation because of higher earnings.

2 This difference was statistically significant at the 0.05 level. 3 This difference was statistically significant at the 0.05 level.

NUMBER OF CHILDREN As Table 3.2 shows, families' economic status varied with the number of children living with the family in 2000. We found that:

Families on MFIP in 1998 with one or two children tended to be financially better off in 2000 than families with three or more children.

The average cash income received by families with one or two children was 127 percent of the poverty threshold in 2000, compared with 97 percent for families with three or more children. When non-cash resources were included, families with one or two children still had the economic advantage. On average, the total resources of families with one or two children were 155 percent of the poverty threshold, whereas families with at least three children had resources averaging 131 percent of their poverty threshold. 5 Also, small families were less dependent on MFIP than large families. The percentage of 1998 MFIP families that stayed off MFIP for all of 2000 ranged from 40 percent for one-child families to 10 percent for families with five or more children.

Families with one or two children were financially better off because their poverty threshold was lower, not because they had higher earnings. It was simply much easier for small families to earn enough money to get well above the poverty threshold. For example, a single parent with one child needed to earn $22,478 to be at 200 percent of the poverty threshold, while a single parent with five children needed to earn $47,056. 6

Table 3.2: Average Resources by Number of Children, 2000
Resources as a Percentage of the Poverty Threshold Total Resources Number of Children N Cash Income Except Medical

None 55 96 114 1 -2 719 127 155 3 or more 385 97 131 TOTAL 1,159 116 145

SOURCE: Office of the Legislative Auditor.

Families with one or two children were economically better off than larger families.

4 In 2000, 39 percent of families that had two parents in both 1998 and 2000 had an adult member on SSI (a federal disability program), compared with 12 percent for families who had one parent and 7 percent for families that changed to two parents. Families with two parents in both 1998 and 2000 had an average family size of 5.4, compared with 3.0 for one-parent families and 4.5 for families that changed to a two-parent family. Finally, 43 percent of families with two parents in 1998 and 2000 were headed by a high school graduate, compared with 66 percent for families that had one parent in 1998. 5 These differences were statistically significant at the 0.05 level.

These differences in economic status were not caused by differences in government assistance benefits. As discussed in Chapter 1, most government assistance programs provide greater assistance to larger families. In our sample, families with five or more children received an average of $32,300 in government assistance benefits, about 3.6 times as much as one-child families ($9,000).

EDUCATION When we compared the economic status of families by the education level of the family head, we found that:

Families on MFIP in 1998 that were headed by a high school graduate tended to be financially better off in 2000 than families headed by non-graduates.
Table 3.3 shows that families headed by a parent with some college education had total resources that averaged 169 percent of the poverty threshold, compared with 149 percent for high school graduates and 133 percent for families headed by non-graduates. 7 These differences were primarily due to differences in average earned income, which ranged from $14,900 for families with some college education to $9,000 for non-high school graduates. 8 In addition, on average, families headed by high school graduates had smaller families. 9

VARIATION IN ECONOMIC STATUS BY DEMOGRAPHIC FACTORS

Table 3.3: Average Resources by Education of Parent, 2000 Resources as a Percentage of the Poverty Threshold Total Resources Education of Family Head N Cash Income Except Medical

Non-high school graduate 436 100 133 High school graduate 593 122 149 Some college 130 142 169 TOTAL 1,159 116 145

NOTE: The family head is the parent who applied for MFIP. SOURCE: Office of the Legislative Auditor.

Families headed by high school graduates were better off because of higher earnings.

6 As a result, although the average earned income for small families was slightly less than that of large families, it was a much higher percentage of poverty than it was for large families. 7 The family head is the parent who applied for MFIP. The relationship between education and total resources was statistically significant at the 0.05 level. 8 Average earnings of families headed by a high school graduate ($12,400 in 2000) were in between the averages of non-graduates and parents with some college. 9 Average family sizes were 3.8 for families headed by non-graduates, 3.3 for families headed by high school graduates, and 3.2 for families headed by parents with some college education.

CITIZENSHIP About 12 percent of MFIP cases were headed by a non-U.S. citizen in 1998. These families differed from families headed by a U.S. citizen on several of the factors we found to be associated with economic status. On the one hand, non-U.S. citizens had more children and much less formal education than their U.S. citizen counterparts, two economic disadvantages. On the other hand, families headed by a non-citizen were more likely to be two-parent families, an economic advantage. Table 3.4 shows how economic status varied by citizenship status. We found that:

When resources were measured as a percentage of the poverty threshold, families headed by a non-U.S. citizen had financial situations similar to families headed by a U.S. citizen.

Although families headed by a non-U.S. citizen had about 30 percent more resources than U.S. citizens had, the average poverty threshold for non-citizen families was also about 30 percent higher because of their larger families. As a result, when resources were measured as a percent of poverty, the economic conditions of both groups were very similar in terms of cash and overall resources. 10

RACE Table 3.5 summarizes the economic status of 1998 MFIP families by the race of the parents. We found that:

Based on cash resources, 1998 MFIP families headed by a white parent were better off financially in 2000 than non-white families, but the gap narrowed when non-cash resources were included.

Table 3.4: Average Resources by Citizenship Status, 2000
Resources as a Percentage of the Poverty Threshold Citzenship Status Total Resources of Family Head N Cash Income Except Medical

Non-U.S. Citizen 143 112 146 U.S. Citizen 1,016 116 145 TOTAL 1,159 116 145

NOTE: The family head is the parent who applied for MFIP. SOURCE: Office of the Legislative Auditor.

Non-citizens on MFIP in 1998 were as well off as citizens in 2000, even though they had less education and more children.

10 The differences were statistically insignificant at the 0.05 level.
When measured as a percent of the poverty threshold, white families had higher average cash incomes than black/African-American, Asian, and American-Indian families. 11 Families headed by a white parent had higher cash resources primarily because they had higher earned income ($13,000 compared with $9,800 for non-white families as a whole). White families also tended to be smaller than non-white families. 12

When non-cash resources were also included, the differences among white, black/African-American, and Asian families became small. 13 Black/African-American and Asian families received more government assistance benefits than white families because they had lower earned income and had larger families.

By all of the measures of income we used, families headed by an American-Indian parent were financially worse off than families headed by a white, black/African-American, or Asian parent. Our results need to be interpreted cautiously because our data did not include federal or tribal programs that provide benefits only to American Indians. However, based on discussions with staff of the Department of Human Services, we suspect that these programs are relatively small. On average, cash resources of American-Indian families were 88 percent of the poverty threshold, 17 percentage points below the next lowest average (for Asian families). But when all resources were included, families headed by an American-Indian parent were 25 percentage points below the next lowest average. 14 Although American-Indian families had less cash resources than black/African-American and Asian families, they received less non-cash benefits. While our data are not conclusive, they suggest that American-Indian families were less likely to obtain the benefits for which they were eligible.

VARIATION IN ECONOMIC STATUS BY DEMOGRAPHIC FACTORS
Table 3.5: Average Resources by Race of Family Head, 2000
Resources as a Percentage of the Poverty Threshold Total Resources Race of Family Head N Cash Income Except Medical

American Indian 104 88 117 Asian 122 105 142 Black/African American 300 109 146 White 583 126 151

TOTAL a 1,159 116 145 NOTE: The family head is the parent who applied for MFIP. This table excludes estimates for families whose head was classified as Hispanic because our small sample size made these estimates unstable. a Total figures include families headed by a Hispanic parent.

SOURCE: Office of the Legislative Auditor.

American Indians appear to be worse off than other racial groups.

11 Families headed by a white parent had more cash resources in 2000 (as a percent of poverty threshold) than did families headed by a Hispanic parent, but the difference was not statistically significant. Our sample only had 50 families headed by a Hispanic parent. 12 In 2000, white families that were on MFIP in 1998 had an average family size of about 3.1, compared with 3.8 for non-white families.

13 In fact, the differences were not statistically significant.

REGION We compared the economic status of current and former MFIP families who lived in the Twin Cities metropolitan area with families who lived in outstate Minnesota. We found:

Families on MFIP in 1998 who lived in the Twin Cities metropolitan area had more resources in 2000 than families living in outstate Minnesota, but it is not clear that they were financially better off because of regional cost-of-living differences.

As Table 3.6 shows, the average total resources of families in the Twin Cities area were 152 percent of the poverty threshold, about 12 percent higher than families in outstate Minnesota. This difference reflects the fact that residents of the Twin Cities area received more benefits from cash assistance and housing programs. The regional difference in earned income was small.

The 12 percent difference in resources does not demonstrate that current and former MFIP recipients who lived in the Twin Cities area were financially better off because the Twin Cities area appears to have a higher cost of living. A 1989 report by our office estimated that the cost of living in the Twin Cities area was about 12 percent higher than in outstate Minnesota. 15 Also, a recent report by the Jobs Now Coalition estimated that the 1999-2000 cost of living in the Twin Cities metropolitan area was about 20 percent higher. 16

Table 3.6: Average Resources by Region, 2000 Resources as a Percentage of the Poverty Threshold

Total Resources Region N Cash Income Except Medical

Twin Cities Metropolitan Area (7-county) 703 118 152 Outstate Minnesota 456 112 136 TOTAL 1,159 116 145

SOURCE: Office of the Legislative Auditor.

Families in the Twin Cities had higher incomes, but faced a higher cost of living.

14 The differences between American Indians and other races were statistically significant at the 0.05 level. 15 Office of the Legislative Auditor, Statewide Cost of Living Differences (St. Paul, 1989), 11. 16 Herbert Cederberg, Kevin Ristau, and Bruce Steuernagel, The Cost of Living in Minnesota, 1999-2000 (St. Paul, Jobs Now Coalition, 2001), 33.

SANCTION STATUS Under MFIP, counties must impose sanctions if recipients do not comply with the program's work requirements. In 2000, Minnesota law allowed monthly sanctions of 10 percent for initial noncompliance and 30 percent for subsequent noncompliance. During an average month in 2000, about 10 percent of MFIP families were sanctioned.

Critics of MFIP sanctions contend that some people are inappropriately sanctioned and that sanctions place MFIP families in real financial hardship. In our previous study on welfare reform, however, many employment services staff told us that clients who remain in sanction for long periods likely have other sources of income— allowing them to more easily absorb the financial impact of a sanction. 17 These other sources may include "cash" jobs in the underground economy or support from family or friends, sources we were not able to measure. Although our study cannot resolve this controversy, we present what we do know about the financial status of families who received four or more sanctions during 2000. When making a comparison with other MFIP families, we restricted our analysis to families who were on MFIP for at least six months during 2000.

As shown in Table 3.7, the total financial resources of these sanctioned MFIP families averaged 104 percent of the poverty threshold, compared with 138 percent for other MFIP families. 18 Sanctioned families received smaller MFIP grants because of their sanctions, but the main reason for the lower income was smaller average earnings ($4,100 compared with $6,900). It is not surprising that sanctioned families reported smaller earnings because most families go into sanction for not complying with work requirements.

Table 3.7: Average Resources by Sanction Status, 2000 Resources as a Percentage of the Poverty Threshold Total Resources Sanction Status N Cash Income Except Medical

4 or more sanctions in 2000 79 65 104 Less than 4 sanctions in 2000 567 95 138 TOTAL 646 91 134

NOTE: This table only includes families that were on MFIP for at least six months during 2000. SOURCE: Office of the Legislative Auditor.

Sanctioned families had lower incomes primarily because they had less earnings.

17 Office of the Legislative Auditor, Welfare Reform (St. Paul, 2000), 112. 18 This difference was statistically significant at the 0.05 level.

Methodology and Sample Design APPENDIX A

In our study, we analyzed the economic status of 1,159 current and former MFIP families in calendar year 2000. All of these families received MFIP at some point between May and October 1998. In 2000, 403 of these families were on MFIP the entire year, 384 were on MFIP part of the year, and 372 were off MFIP the entire year. We used MFIP's definition of a family, which requires a child to be in it. 1 Thus, the family is (1) a minor child, (2) his or her minor siblings (including adopted or step), (3) the parents of these children (including adoptive, step, or unmarried), and (4) other children (such as nieces, nephews, and grandchildren of the parents) who were included in the MFIP assistance unit. In addition, all family members must live in the same household. However, we made two adjustments to MFIP's definition. First, we included immediate family members who were ineligible for MFIP, such as undocumented non-citizens and those receiving Supplemental Security Income (SSI). Second, while all the families had at least one child in 1998, not all of them had a child in 2000. In some cases, all the children became adults or moved out of the household by 2000. In total, these 1,159 families had 4,024 members in 2000.

In determining economic status, we tried to assess each family's ability to meet its daily living expenses— food, shelter, clothing, health care, and child care. Table A. 1 lists the income sources and tax liabilities for which we did and did not collect data. For the 28 income sources and 3 tax liabilities that we used to compute after tax income, we collected and merged data from 12 different data systems, including the Minnesota Department of Humans Services' MAXIS system and state income tax records.

We did not examine economic resources that:

Were designed to help pay for extra services (beyond daily living expenses) required by individuals with special needs— such as the Family Support Program,
Tried to increase the earnings potential of recipients— such as MFIP employment services— rather than subsidize daily living expenses,
Had eligibility requirements that current and former MFIP recipients could not meet— such as Refugee Cash Assistance,
Had very small average benefits per recipient— such as the Telephone Assistance program, or
Had data that were not readily available or could not be easily estimated.

1 Pregnant women are also eligible, even if they do not have any other children.

SAMPLE DESIGN As stated above, our analysis was based on a random sample of families who were on MFIP between May and October 1998. 2 We chose this time period to build our sample off of a sample that the Minnesota Department of Human Services (DHS) is using for a longitudinal study concerning current and former MFIP recipients. DHS is collecting a rich database of self-reported information that we used to supplement the administrative data that we collected.

Table A. 1: Income Sources and Tax Liabilities Included and Not Included in Analysis Income Sources Included Income Sources Not Included Earnings Child Support

Workers' Compensation Unemployment Insurance Other (e. g. interest and dividends) MFIP Cash Assistance Supplemental Security Income Minnesota Supplemental Aid General Assistance Diversionary Assistance Emergency Assistance MFIP Food Assistance Food Stamps and State Food Program National School Lunch Women, Infants, and Children Section 8 Housing Public Housing Energy Assistance Medical Assistance MinnesotaCare General Assistance Medical Care Child Care Assistance Federal Earned Income Tax Credit Federal Additional Child Tax Credit Minnesota Working Families Tax Credit Minnesota Dependent Care Tax Credit Minnesota Education Tax Credit Minnesota Property Tax Refund

Family Support Program Consumer Support Program Relative Custody Assistance Program Adoption Assistance Program Foster Care Refugee Resettlement Program Refugee Cash Assistance Refugee Medical Assistance United States Repatriate Program Telephone Assistance Plan Social Services Mothers and Children Program Child Care Resource and Referral MFIP Employment Services Self-Employment Investment Development Education Grants (Pell and State) Rental Assistance for Family Stabilization Group Residential Housing Family Homeless Prevention

Tax Liabilities Included Federal Income Taxes State Income Taxes

Social Security and Medicare Payroll Taxes

SOURCE: Office of the Legislative Auditor.

2 We developed a proportional, stratified, random sample.

DHS' Recipient and Applicant Samples The Department of Human Services designed its study to track two groups of MFIP families between 1998 and 2003. The new-applicant group was a random sample of 985 one-parent families who first applied between May and October 1998. The recipient group was a random sample of 843 one-parent MFIP families who were not new applicants. 3 Since DHS sampled a higher proportion of new applicants than other recipients, our sample included all of the families in the recipient sample and a subset of the new-applicant sample. We selected this subset of 33 new applicants so that the proportion of new applicants in our sample would be the same as it was in the MFIP population.

Additions to the DHS Sample To obtain a representative sample of MFIP families from 1998, we supplemented the DHS sample by adding groups they excluded, including (1) two-parent families, (2) families that participated in MFIP field trials, (3) one-parent families in which the parent was ineligible for MFIP, and (4) child-only cases with a relative caregiver. For each of these groups we selected a random sample in the same proportion used in the original DHS recipient sample. Overall, we added 489 families, including 227 two-parent families, 97 one-parent families who participated in the MFIP field trials, 116 one-parent families in which the parent was ineligible for MFIP, and 49 families with a relative caregiver.

Cases Excluded from the Analysis We dropped 206 cases from our sample because of a lack of income data, including 69 families that lived in another state for at least part of the year and 17 families for which we lacked a social security number for a parent. We excluded all 62 cases that did not have a parent in the household because we lacked good income information about these cases. Finally, we dropped 58 families that met all of the following four criteria:

1. Did not file a 2000 tax return in Minnesota, 2. Had no wage record for 2000 in Minnesota's unemployment system, 3. Had less than six months of information during 2000 from any combination of MFIP, General Assistance, Supplemental Security Income, Minnesota Supplemental Aid, workers' compensation, or unemployment insurance payments, and

4. Had less than $1,000 in reported child support benefits for 2000. We suspect that any family meeting all of these criteria left Minnesota or had financial support from sources for which we lack data. As Table A. 2 shows, our final sample size was 1,159 families, or 85 percent of our original sample.

APPENDIX A 61

3 For both the new-applicant sample and the recipient sample, DHS selected six separate random samples, one for each month during this six-month period. We followed the same procedure when we extended the sample to groups excluded by DHS.

Tracking Changes in Family Composition To accurately measure each family's economic status and to understand changes in income, we tracked changes in family composition. Our analysis was based on family members living in the household as of December 2000. 4 If a family split into different households, we used the family living with the MFIP applicant, typically the mother. If we did not have any information as of December 2000, we used the last known family composition prior to December 2000.

We used a variety of sources to track the composition of families, particularly families that left MFIP. If a family was on MFIP in December 2000, we used MFIP records to determine its composition. For families that left MFIP, we first looked at records from Medical Assistance and MinnesotaCare to identify who was living with the family in December 2000. Because of data limitations, we only used MinnesotaCare records to identify new family members, not to identify people who no longer lived with the family.

We also examined income tax returns and property tax refund returns to identify new spouses and children. However, tax records do not indicate whether the spouse is actually living with the family. If MFIP or Medical Assistance (MA) records indicated that the spouse was not living with the family, we assumed that this information was correct. If we did not have current information from MFIP or MA, we included the applicant's spouse if they filed a joint tax return. Finally, if there was conflicting information or unusual circumstances, we used the DHS longitudinal survey to help resolve questions about family composition.

DATACOLLECTION For most sources, we received administrative data from the relevant state agency about the income and benefits received by all Minnesota recipients in 2000. This electronic data also included each recipient's social security number (or another unique identifier). In the case of programs that provided assistance to a family or household (such as energy assistance) rather than an individual, we received the social security number of the family member applying for the assistance. We then identified the income and benefits for the people in our sample by matching their social security numbers with these databases. Because our unit of analysis was a family, we then aggregated for each family the income received by its members.

Table A. 2: Types of Families in Working Sample Number of Cases Percentage

One-parent families in DHS longitudinal sample 774 67% Families not in DHS longitudinal sample One-parent families in MFIP field trials 87 8% One-parent families without the parent eligible for MFIP 99 9 Two-parent families 199 17

TOTAL 1,159 100%

SOURCE: Office of the Legislative Auditor.

4 We excluded children who became 19 by the end of 2000.

We obtained data on privately-generated income from a variety of sources. Specifically, earnings data came from three sources: (1) income data reported to the Department of Human Services for MFIP or MA eligibility, (2) federal and state income tax returns, and (3) wage data reported by employers to the state's unemployment insurance system. Each of these sources has incomplete income information. For example, human service records have little income data after a family leaves MFIP. Income tax data are incomplete because many low-income people do not file tax returns. In addition, people who left Minnesota may file in another state but not in Minnesota. Wage data from the unemployment system are incomplete because certain types of employees are not included in the data system, including the self-employed, workers classified as independent contractors, and federal workers (such as post office employees). Finally, income reported under each of these sources may be under-reported. Because of these limitations, we used whichever source had the highest income. DHS and tax records also had data on unearned income, including interest and dividends, capital gains, and social security benefits. 5 For individuals for whom we had no income tax data, no unemployment data, and DHS data for only part of the year 2000, we projected their annual income based on their average monthly income during the months that they were a member of an active MFIP case.

We had to collect some of the data from federal tax returns by hand. Unlike state tax data, our office does not have the authority to access federal tax records collected by the Internal Revenue Service. However, the federal tax forms that Minnesotans submit with their state tax forms are considered state data, not federal. For the people who submitted their 2000 tax returns electronically with the state, we received federal income tax data electronically. However, for people who filed paper tax returns, we pulled their federal returns and entered the relevant data by hand. We received all the data from the state tax forms electronically, whether the filer submitted an electronic or paper return.

In our analysis, we also subtracted the Social Security and Medicare payroll taxes that each person paid in 2000. Because we did not have access to the actual amount of taxes paid by individuals, we estimated the amount based on each person's wages/salaries and self-employment income. In 2000, the tax was a straight percentage of wages/salaries (7.65 percent) and self-employment income (15.3 percent).

APPENDIX A 63

5 The unemployment system only reports wage income, while tax returns and DHS have broad measures of income, including self-employment income and unearned income. As a result, we used the highest income from the following three sources: (1) DHS, (2) tax returns, and (3) wage income from unemployment system plus non-wage income from DHS or tax returns, whichever is higher. While DHS data and tax returns also had unemployment insurance payments, workers' compensation, and child support payments, we obtained data on these income sources directly from the state agency involved and disregarded any of these payments included on tax returns or reported to DHS.

To estimate medical subsidies, we first determined how many months during 2000 that each person in our sample was enrolled in MA, MinnesotaCare, or General Assistance Medical Care (GAMC). For MA and GAMC, we estimated the annual subsidy by multiplying the number of enrolled months by the average monthly public subsidy for all MFIP recipients covered by Medical Assistance. 6 For MinnesotaCare, we estimated the annual subsidy by subtracting the premium paid by each family member from the actual payment made by the state to the health insurer on behalf of the family member.

In the case of four non-cash programs (Section 8, public housing, National School Lunch, and the Women, Infants, and Children program), we were unable to obtain administrative data on the amount of assistance received. Fortunately, for 596 of our sample families, DHS collected self-reported participation data for these four programs through its longitudinal survey for the year 2000. In the case of housing, DHS also collected information about each family's rent and utility payments. From this information, we estimated the amount of assistance received in 2000.

For each family participating in Section 8, we assumed that its unsubsidized gross rent was the area's "fair market rent." For each county or metropolitan area of the state, the U.S. Department of Housing and Urban Development (HUD) establishes "fair market rents," which are set at the 40 th percentile of rents charged in the area. Typically, these are the maximum rents that HUD will subsidize. We made this assumption for two reasons. First, because Section 8 families only pay 30 percent of their income for rent, they face in most cases no additional cost by renting a more expensive apartment as long as it is within the fair market rent. Second, according to DHS' longitudinal survey for 2000, average gross rents paid by families not in subsidized housing were 92 percent of the area's fair market rent. We suspect that families receiving Section 8 live in housing that costs at least as much as the unsubsidized housing rented by similar families. We then calculated the Section 8 subsidy by subtracting each family's actual rent payment from the area's fair market rent. Because DHS' longitudinal survey only applies to one month of 2000, we multiplied the monthly subsidy by 12.

We used the same process to estimate the public housing subsidy, but rather than using fair market rents, we used "flat rents" to approximate the unsubsidized gross rent. Each Public Housing Authority determines a "flat rent" for its buildings, which is supposed to reflect the units' market value. While we were unable to get flat rents for all areas of the state, we estimated that flat rents were about 90 percent of fair market rents in 2000 based on information from the Minneapolis and St. Paul public housing authorities. We then assumed each family's public housing subsidy was the difference between 90 percent of the area's fair market rent and the family's actual rent payment.

In the DHS longitudinal survey, families also indicated if their children participated in the National School Lunch program. For each school-aged child in these families, we estimated an annual subsidy by assuming a monthly subsidy of $44.93 for free lunches and $36.21 for reduced-price lunches and nine months of participation. 7 Based on each family's income, we determined if the children were eligible for a free lunch, rather than a reduced-price lunch.

6 We estimated the cost for adults and children separately, based on the average costs for MFIP adults and MFIP children. To estimate the average cost for GAMC enrollees, we applied the average MA cost of MFIP enrollees rather than an average GAMC cost because the GAMC population is much sicker than the MFIP population.

For the Women, Infants, and Children (WIC) program, we used self-reported participation data from DHS' survey and then estimated the portion of calendar year 2000 that each family member was eligible based on the children's ages. For example, if a child was born on July 1, 2000, the mother was eligible for the entire year (six months as a pregnant woman and six months as a new mother) and the infant was eligible for six months. For all eligible individuals, we assumed a monthly benefit of $30.28, the statewide average benefit for 2000.

ANALYSIS We then aggregated the income data for each family to determine its income compared with the U.S. Bureau of the Census' official poverty threshold. As discussed in Chapter 2, we compared four measures of income with the poverty threshold. The measures were (1) cash income before taxes (the Census Bureau's official measure), (2) cash income after taxes, (3) cash and non-cash resources excluding medical subsidies, and (4) all cash and non-cash resources. As described in Chapter 2, we settled on cash and non-cash resources without medical subsidies as an estimate that provided a consistent measure for all families in our sample. Using this data, we assessed the economic status of various groups from our sample, based on MFIP participation, family type, race, region, and other demographic factors.

Because we only had data for Section 8, public housing, National School Lunch, and WIC for the families that responded to DHS' longitudinal survey for 2000 (which accounted for a little over half of our complete sample), we sometimes used the limited DHS sample as a proxy for the complete sample. To assess the validity of this assumption, we examined how well the limited DHS sample represented the complete sample. While families that responded to DHS' survey had smaller average incomes and household sizes than the rest of the sample, the two portions of the sample had statistically the same average income as a percentage of the federal poverty threshold, as shown in Table A. 3. 8

In our analyses, we estimated average total resources for the complete sample although we did not have complete data for all of it. As mentioned above, we lacked data on housing, National School Lunch, and WIC subsidies for families that did not take part in or respond to DHS' longitudinal survey for 2000. Because both portions of the sample had similar resources as a percentage of the federal poverty threshold, we assumed that the families that did not respond to DHS' survey had the same average housing, National School Lunch, and WIC subsidy as a percentage of the poverty threshold as the families that responded to the survey did (17 percent). We then applied this percentage to the average poverty threshold for families that did not respond to the survey ($16,439) to compute an average housing, National School Lunch, and WIC subsidy of $2,803.

APPENDIX A 65

7 We calculated the average monthly subsidy by multiplying the program's meal reimbursement rate by 21.79 school days per month. This estimate assumes that eligible children receive the free or reduced-price meal each school day, which is likely an overstatement. 8 The various measures of average income as a percentage of the federal poverty thresholds for the two portions of our sample were statistically the same at a 0.05 significance level.

When we compared the average incomes for the various groups in our sample, we assumed that within each group the families that did not respond to DHS' survey had the same average housing, National School Lunch, and WIC subsidy as a percentage of the federal poverty threshold as the families that did respond. For example, we calculated that the average subsidy as a percentage of the poverty threshold was 26 percent for families that responded to the survey and were on MFIP for all of 2000. We then assumed the same average percentage for the "non-survey" families that were also on MFIP for the entire year.

Table A. 3: Comparison of Samples Average Income as a Percentage of the Federal Poverty Threshold Food, Cash, Food, Energy, Energy, and Cash Income and Medical Medical Resources Sample N After Taxes Subsidies After Taxes

DHS sample families that completed the 2000 survey 596 118% 43% 161% "Non-survey" families 563 113 44 157

Complete Sample 1,159 116% 44% 159%

NOTE: We have complete data for all families in our sample for cash, food, energy, and medical assistance. SOURCE: Office of the Legislative Auditor.

Historical Program Data APPENDIX B:

In Chapter 1, we discussed a variety of program data for the government assistance and tax credit programs in Minnesota. Table 1.1 presented caseload, expenditure, and average benefit data for programs for the year 2000. Figure 1.1 showed the trend in program expenditures between 1989 and 1999. In this appendix, we provide detailed program data for the years 1988 through 2000. Table B. 1 shows the number of cases (in thousands), Table B. 2 shows the number of cases per thousand Minnesota households, Table B. 3 shows annual expenditures, and Table B. 4 presents the average annual benefit provided by each program.

Table B. 1: Monthly

Average Cases/Recipients for Government

Assistance and Tax Credit Programs in Minnesota, 1988-2000 (in Thousands)

Percentage Change Year 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 1988-1999

NOTE: Caseload and recipient data are for calendar year, state fiscal year, or federal fiscal year depending on the program.

SFY = State

Fiscal Year. FFY= Federal Fiscal Year.

CY = Calendar

Year. N/A = Not available.

a Data for MFIP excludes cases receiving food assistance only and includes Family General Assistance cases. Food Stamps includes MFIP food assistance cases, regular Food Stamp cases, and Minnesota Food Assistance cases. General Assistance data excludes Family General Assistance cases and includes the GA segment of Group Residential Housing.

b The number of cases decreases from 1,661 to 1,379, a decline of 17 percent.

c Minnesota Supplemental Aid includes the MSA segment of Group Residential Housing.

d Medical Assistance data excludes the aged, blind, and disabled.

e Shading indicates years during which a program was not yet in existence.

f Data are for public schools only.

g Caseload data are the number served on an annual basis.

h WIC caseload data excludes participants funded by the state.

SOURCE: Office of the Legislative Auditor analysis of data from (1) Department of Human Services, (2) Department of Children, Families, and Learning, (3) Department of Economic Security, (4) Department of Health, (5) Department of Revenue, (6) U.S. Department of Agriculture, (7) U.S. Social Security Administration, and (8) U.S. Internal Revenue Service.

HISTORICAL PROGRAM DATA 69 Table B. 2: Monthly

Average Cases/Recipients for Government

Assistance and Tax Credit Programs in Minnesota, per Thousand Households, 1988-2000

Percentage Change

Year 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

1988-1999

NOTE: State household data are for calendar years. Caseload and recipient data are for calendar year, state fiscal year, or federal fiscal year depending on the program.

SFY = State

Fiscal Year. FFY= Federal

Fiscal Year. CY = Calendar

Year. N/A = Not available.

a Data for MFIP excludes cases receiving food assistance only and includes Family General Assistance cases. Food Stamps includes MFIP food assistance cases, regular Food Stamps, and Minnesota Food Assistance cases. General Assistance data excludes Family General Assistance cases and includes the GA segment of Group Residential Housing.

b Cases per thousand households decrease from 1.03 to 0.74, a decline of 28 percent.

c Minnesota Supplemental Aid includes the MSA segment of Group Residential Housing.

d Medical Assistance data excludes the aged, blind, and disabled.

e Shading indicates years during which a program was not yet in existence.

f Data are for public schools only.

g Caseload data are the number served on an annual basis.

h WIC caseload data excludes participants funded by the state.

SOURCE: Office of the Legislative Auditor analysis of data from (1) Department of Human Services, (2) Department of Children, Families, and Learning, (3) Department of Economic Security, (4) Department of Health, (5) Department of Revenue, (6) U.S. Department of Agriculture, (7) U.S. Social Security Administration, and (8) U.S. Internal Revenue

Table B. 3: Annual

Government

Payments for Government Assistance and Tax Credit Programs in Minnesota, 1988-2000 (in Millions of Year 2000 Dollars)
Percentage Change

Year 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

1988-1999

MFIP/AFDC cash assistance

NOTE: Expenditures were adjusted for inflation to year 2000 dollars using the All-Urban Consumer Price Index. Data are for calendar year, state fiscal year, or federal fiscal year depending on the program.

SFY = State

Fiscal

Year. FFY= Federal

Fiscal Year. CY = Calendar

Year. N/A = Not available.

a Data for MFIP includes expenditures for MFIP cash assistance and Family General Assistance, and excludes MFIP food assistance. Food Stamps includes expenditures for regular Food Stamps, MFIP food assistance, and Minnesota Food Assistance. General Assistance data excludes Family General Assistance and includes expenditures for the GA segment of Group Residential Housing. b Minnesota Supplemental Aid includes expenditures for the MSA segment of Group Residential Housing.

c Medical Assistance excludes expenditures for the aged, blind, and disabled.

d Shading indicates years during which a program was not yet in existence.

e Costs include state-funded administrative expenses at the county level and part of county-funded administrative expenses in addition to direct program costs.

f Payments only include the aggregate payments allocated by the federal government for both free and reduced-price lunches. State funding for all lunches and additional federal funding for paid lunches are excluded. Data are for public schools only.

g WIC expenditures are net costs after the infant-formula rebate received by states. Additional funds provided by the state between 1990 and 1997 are excluded. Annual state expenditures between 1993 and 1997 averaged $1.4 million.

SOURCE: Office of the Legislative Auditor analysis of data from (1) Department of Human Services, (2) Department of Children, Families, and Learning, (3) Department of Economic Security, (4) Department of Health, (5) Department of Revenue, (6) U.S. Department of Agriculture, (7) U.S. Social Security Administration, and (8) U.S. Internal Revenue Service.

HISTORICAL PROGRAM DATA 71 Table B. 4: Annual

Average Benefit Per Case/Recipient for Government Assistance and Tax Credit Programs in Minnesota, 1988-2000 (in Year 2000 dollars)

Percentage Change

Year 1988

1989 1990 1991 1992 1993 1994 1995 1996

1997 1998 1999

2000 1988-1999

MFIP/AFDC cash assistance per case a

SFY $7,684

$7,311 $7,067 $6,702 $6,347 $6,045 $5,749 $5,587 $5,363 $5,058

N/A $5,221

$4,981

-32%

General Assistance per case a

SFY 3,580 3,479 3,386 3,326 3,211 3,591 3,913 3,795 3,916 4,124 $4,367

4,826 5,604

35

Emergency and Diversionary Assistance per case

SFY ************** SSI per recipient CY 3,635 3,743 3,903 4,199 4,560 4,369 4,437 4,395 4,415 4,176 4,242 4,223 4,200 16 Minnesota Supplemental Aid per recipient b SFY 3,139 3,730 3,804 3,491 3,214 2,550 2,336 2,154 2,091 2,158 2,077 2,108 2,102 -33 Food Stamps per household a SFY 1,831 1,849 1,998 2,075 2,213 2,122 2,051 2,005 1,947 1,898 N/A 1,857 1,810 1 Child Care Assistance per family SFY 2,709 2,878 3,379 3,830 4,334 4,331 4,346 4,428 4,611 4,640 4,967 5,758 6,056 113 Medical Assistance per recipient (MFIP and other families only) c SFY 1,655 1,744 1,790 1,710 1,850 1,845 1,860 2,143 2,094 2,145 2,242 2,454 2,689 48 MinnesotaCare/Children's Health Plan per enrollee SFY d 211 330 332 374 348 425 605 772 890 929 1,330 1,447 N/A General Assistance Medical Care per recipient SFY 3,155 3,305 3,349 3,204 3,573 3,543 3,429 3,340 3,844 4,019 4,595 5,080 5,471 61 National School Lunch free lunch per approved student e SFY 424 421 420 417 413 406 406 401 400 397 401 405 404 -4 National School Lunch reduced-price lunch per aproved student e SFY 310 312 316 318 316 313 315 313 314 313 319 324 326 5 National School Lunch free breakfast per approved student e SFY 219 217 225 223 223 220 230 226 225 223 224 226 224 3 National School Lunch reduced-price breakfast per approved student e SFY 133 135 146 148 151 150 161 160 161 161 163 165 165 25 WIC per recipient f FFY 546 410 421 391 402 393 384 375 382 373 383 377 363 -31 Energy Assistance per household FFY 527 458 484 493 452 428 543 474 353 494 332 295 414 -44 Earned Income Tax Credit per recipient CY 717 719 717 973 1,058 1,123 1,095 1,309 1,412 1,431 1,484 1,499 N/A 109 State Working Family Tax Credit per recipient CY d 99 105 168 183 201 217 219 411 451 N/A N/A State Dependent Care Tax Credit per recipient CY 392 371 360 352 354 352 366 361 358 350 339 338 N/A -14 State K-12 Education Credit per recipient CY d 391 381 N/A N/A Property Tax Refund per recipient (homeowners) CY 323 293 299 301 307 306 354 355 357 345 339 335 N/A 4 Property Tax Refund per recipient (renters) CY 475 451 403 417 416 403 400 398 389 407 432 431 N/A -9 NOTE: Expenditures were adjusted for inflation to year 2000 dollars using the All-Urban Consumer Price Index. Expenditure data are for calendar year, state fiscal year, or federal fiscal year depending on the program. SFY = State Fiscal Year. FFY= Federal Fiscal Year. CY = Calendar Year. N/A = Not available. For programs that provide a monthly benefit, the annual average benefit is the total annual payments divided by the monthly average cases or recipients. *An annual average benefit is not shown because Emergency and Diversionary Assistance may not be received for longer than a few months. a The average benefit for MFIP is the average for cash assistance only (including family General Assistance data). The average benefit for Food Stamps includes regular Food Stamps, MFIP food assistance, and Minnesota Food Assistance. The average benefit for General Assistance excludes Family General Assistance and includes the GA segment of Group Residential Housing. b Minnesota Supplemental Aid includes the MSA segment of Group Residential Housing. c The average benefit is for families with children only and does not include expenditures for the aged, blind, and disabled. d Shading indicates years during which a program was not yet in existence. e The average benefit includes federal and state funds and is based on the assumption that students received the meal each school day for a nine-month school year. Data is for public schools only. f The average benefit excludes state funding and is based on net costs after the infant-formula rebate received by states. SOURCE: Office of the Legislative Auditor analysis of data from (1) Department of Human Services, (2) Department of Children, Families, and Learning, (3) Department of Economic Security, (4) Department of Health, (5) Department of Revenue, (6) U.S. Department of Agriculture, (7) U.S. Social Security Administration, and (8) U.S. Internal Revenue Service.

Major Assumptions in the Incentives Model APPENDIX C

In Figures2.1 through 2.4andTable 2.5,we showed the resources available to a family at various wagelevels.  We created this model to illustrate the incentives that low-income parents face when making the decision to increase their earnings. TableC.1lists the major assumptions that we used in this model.
Table C. 1: Incentives Model Assumptions 2000 1988 Assumption ($ figures are in 2000 dollars) ($ figures are in 1988 dollars)

Parents One parent One parent Children One preschooler and one child in elementary school One preschooler and one child in elementary school
Annual full-time employment 2080 hours 2080 hours
Income Earnings and government assistance Earnings and government assistance
Monthly Medical Assistance costs Average for MFIP recipient— $165 Average for AFDC recipient— $82
Monthly MinnesotaCare costs (subsidy and premium)
Program average -$141 per recipient N/A

Monthly child care costs Statewide average for family day care—$587 for both children. Preschooler received full-time care. School-aged child received full-time care in the summer and 2.5 hours of care during the school year.

Statewide median for family day care -$378 for both children. Same care assumptions as 2000.

Monthly Rent Statewide weighted average "fair market rent" for a two bedroom apartment—$592 per month.

Statewide weighted average "fair market rent" for a two bedroom apartment—$456 per month

Energy Assistance Utilities were included in rent. Utilities were included in rent. Monthly National School Lunch benefit

Average monthly cost per child in Minnesota—$45 for free lunches and $36 for reduced-priced. School-aged child received the benefit for nine months of the year.

Average monthly cost per child in Minnesota—$32 for free lunches and $24 for reduced-priced. Same months of benefit as 2000.

Monthly WIC benefit Average monthly cost per recipient in Minnesota—$30. Preschooler received benefits entire year.

Average monthly cost per recipient in Minnesota—$31. Preschooler received benefits entire year.

Property taxes as a percentage of rent 19 percent— amount specified in the property tax refund instructions for 2000.

19 percent

SOURCE: Office of the Legislative Auditor.

Further Reading

Acs, Gregory, and Pamela Loprest. Initial Synthesis Report of the Findings from ASPE's "Leavers" Grants. Washington, DC: Urban Institute, January 2001. http://aspe hhs.gov/hsp/leavers99/synthesis01/index. htm.

Ahn, Jay, Shon Kraley, Debra Fogarty, Faith Lai, and Laurie Deppman. A Study of Washington State's TANF Leavers and TANF Recipients: Welfare Reform and Findings from Administrative Data. Washington Department of Social and Health Services, February 2000.

Berkeley Policy Associates. Evaluation of the Colorado Works Program: Second Annual Report. Denver, CO: Office of the Colorado State Auditor, November 2000.

Cederberg, Herbert, Kevin Ristau, and Bruce Steuernagel. The Cost of Living in Minnesota. St. Paul: Jobs Now Coalition, May 2001.

Committee on Ways and Means, U.S. House of Representatives. 106 th Congress, 2d Session. "Appendix L. Monitoring the Effects of Pre-and Post-TANF Welfare Reform Initiatives." 2000 Green Book: Background Material and Data on Programs within the Jurisdiction of the Committee on Ways and Means. Washington, DC: United States Government Printing Office, 2000.

Congress, 2d Session. Background Material and Data on

Programs within the Jurisdiction of the Committee on Ways and Means: 1988 Edition. Washington, DC: United States Government Printing Office, 1988.

Du, Jean, Debra Fogarty, Devin Hopps, and James Hu. A Study of Washington State TANF Leavers and TANF Recipients: Findings from the April-June 1999 Telephone Survey. Washington Department of Social and Health Services, March 2000.

Loprest, Pamela. How Are Families that Left Welfare Doing? A Comparison of Early and Recent Welfare Leavers. Series B, N. B-36. Washington, DC: Urban Institute, April 2001.

Mathematica Policy Research, Inc. Current and Former WFNJ Clients: How Are They Faring 30 Months Later? Trenton, NJ: New Jersey Department of Human Services, November 2000.

Iowa Families that Left TANF: Why Did They Leave and How Are They Faring? Des Moines, IA: Iowa Department of Human Services, February 2001. 87Midwest Research Institute. Chapter 2-Household Income and Poverty, Interim Report. Jefferson City, MO: Missouri Department of Social Services, April 2000.

Minnesota Department of Human Services. Combined Manual. St. Paul, November 2001. http://www.dhs.state.mn.us/ecs/reguproc/cm/master01.htm.

Minnesota Family Investment Program Longitudinal Study: Baseline Report. St. Paul, August 1999.

Minnesota Family Investment Program Longitudinal Study: Report on Recipient Sample Six Months After Baseline, Second Report in a Series. St. Paul, February 2000.

Minnesota Family Investment Program Longitudinal Study: Report on Applicant Sample Six Months After Baseline, Third Report in a Series. St. Paul, June 2000.
Minnesota Family Investment Program Longitudinal Study: One Year After Baseline, Fourth Report in a Series. St. Paul, December 2000.

Minnesota Family Self-Sufficiency and Medical Programs: November 2001 Forecasts. St. Paul, 2001.

Minnesota House of Representatives, Research Department. Basic Needs Budgets for Custodial and Noncustodial Parents: A Research Report. St. Paul, October 1999.

Minnesota Family Assistance: A Guide to Public Programs Providing Assistance to Minnesota Families. St. Paul, January 2000.

Minnesota's Public Assistance Programs: A Guide for Legislators. St. Paul, December 1988.

Office of the Legislative Auditor. Use of Public Assistance Programs by AFDC Recipients. St. Paul, February 1989.

Welfare Reform. St. Paul, January 2000.

Panel on Poverty and Family Assistance. Measuring Poverty: A New Approach. Washington, DC: National Academy of Science, 1995.

Tout, Kathryn, Karin Martinson, Robin Koralek, and Jennifer Ehrle. Recent Changes in Minnesota Welfare and Work, Child Care, and Child Welfare Systems. Washington, DC: Urban Institute, July 2001.

Tweedie, Jack, Courtney Jarchow, and Andrea Wilkins. Tracking Recipients After They Leave Welfare. Denver, CO: National Conference of State Legislatures, April 2001. http://www.ncsl.org/Home/tabid/118/Default.aspx?statefed/welfare/leavers01.htm.
United States Bureau of the Census. Experimental Poverty Measures: 1990 to 1997. Current Population Reports, Consumer Income, P60-205. Washington, DC: United States Government Printing Office, June 1999.

Income, Poverty, and Valuation of Noncash Benefits: 1994. Current Population Reports, Consumer Income, P60-189. Washington, DC: United States Government Printing Office, April 1996.

Poverty in the United States: 2000. Current Population Reports, Consumer Income, P60-214. Washington, DC: United States Government Printing Office, September 2001.

United States General Accounting Office. Welfare Reform: Information on Former Recipients' Status. Washington, DC: United States General Accounting Office, April 1999.

Welfare Information Network. http://www.welfareinfo.org. Wisconsin Legislative Audit Bureau. Wisconsin Works (W-2) Program. Madison, WI: Wisconsin Legislative Audit Bureau, April 2001.

444 Lafayette Road North ! Saint Paul, Minnesota ! 55155 ! An Equal Opportunity Employer Minnesota Department of Human Services January 15, 2002 James R. Nobles, Legislative Auditor Office of the Legislative Auditor Centennial Office Building 658 Cedar Street St. Paul, MN 55155

Dear Mr. Nobles:
Thank you for the opportunity to review and comment on your report, "Economic Status of Welfare Recipients." We concur with all the report's major findings; they are consistent with conclusions we have drawn from our own studies and analyses. The report confirms that Minnesota has made great strides in transforming its welfare system into a system that encourages families to become self-sufficient. More importantly, it confirms that Minnesota has fulfilled its promise that families who transition from welfare to work will not live in poverty. We are also pleased to see that your analysis finds former welfare recipients in Minnesota are better off than former welfare recipients in comparison states. Specifically, we concur with the following findings from the report: · The vast majority of families that have left MFIP are no longer in poverty. This finding is consistent with the findings of the DHS Longitudinal Study. We also agree that when non-cash benefits are considered, most families on MFIP also have total resources above the federal poverty line.

· Families who have left MFIP support themselves and rely very little on government assistance. This is a truly striking finding that is confirmed by the DHS Longitudinal Study. Only 13% of MFIP leavers income was from government sources. Families who leave MFIP support themselves.

· Over the last decade Minnesota has dramatically shifted its spending for low-income families to investments aimed at helping families become self-sufficient. In fact, your analysis found spending on the state's two main cash assistance programs – AFDC/MFIP and General Assistance – dropped $205 million annually in real dollar terms between 1988 and 2000. At the same time, the state and federal governments have dramatically increased investments in programs like tax credits and child care assistance which provide supports to working families.
· The financial status of families on MFIP improves dramatically when they go to work.
Following from the point above, Minnesota now has a system that provides significant financial incentive for individuals to leave welfare and go to work. However, similar to your analysis, we have also found that there is less financial incentive for a family to move from a relatively low-paying job to a higher-paying job. We do have two concerns related to how Health Care benefits and costs are presented in the report. First, while we agree that it is difficult to quantify the value of health care benefits to the individual, we think you could have done more to demonstrate how health care benefits impact the well-being of families. For example, in our Longitudinal Study, we simply identify whether families that have left welfare have or do not have health care coverage. By not including some similar measure more systematically throughout your report, we believe you have missed an opportunity to more fully describe the economic circumstances of families who have left welfare.

Second, including increases in medical program expenditures in your analysis of changes in aggregate spending levels may overstate the level of change in spending on low-income families over the last decade. Most of the programs included in your analysis were clearly expanded to support low-income working families. While this was certainly an objective in most expansions of health care coverage, such expansions were also made to help further important public health goals. The inclusion of health care costs in your analysis is also problematic because you use a general inflation adjustment rather than the health care inflation adjustor, which is much higher. Given that increases in health care expenditures make up almost half of the increased costs you identify over the last decade, we think some more sensitivity to how these costs are represented is warranted.

Thank you again for the opportunity to review the report. We think on balance it makes a strong contribution to the public understanding of how Minnesota is doing in supporting low-income families as they transition off of welfare. Sincerely, /s/ Michael O'Keffe Michael O'Keefe Commissioner

Recent Program Evaluations
Game and Fish Fund Special Stamps and Surcharges, Update, January 1994 94-01
Performance Budgeting, February 1994 94-02 Psychopathic Personality Commitment Law, February 1994 94-03
Higher Education Tuition and State Grants, February 1994 94-04
Motor Vehicle Deputy Registrars, March 1994 94-05
Minnesota Supercomputer Center, June 1994 94-06
Sex Offender Treatment Programs, July 1994 94-07
Residential Facilities for Juvenile Offenders, February 1995 95-01
Health Care Administrative Costs, February 1995 95-02
Guardians Ad Litem, February 1995 95-03
Early Retirement Incentives, March 1995 95-04
State Employee Training: A Best Practices Review, April 1995 95-05
Snow and Ice Control: A Best Practices Review, May 1995 95-06
Pollution Control Agency's Use of Administrative Penalty Orders, Update July 1995 95-07
Development and Use of the 1994 Agency Performance Reports, July 1995 PR95-22
State Agency Use of Customer Satisfaction Surveys, October 1995 PR95-23
Funding for Probation Services, January 1996 96-01 Department of Human Rights, January 1996 96-02
Trends in State and Local Government Spending, February 1996 96-03
State Grant and Loan Programs for Businesses February 1996 96-04
Post-Secondary Enrollment Options Program, March 1996 96-05
Tax Increment Financing, March 1996 96-06
Property Assessments: Structure and Appeals, A Best Practices Review, May 1996 96-07
Recidivism of Adult Felons, January 1997 97-01
Nursing Home Rates in the Upper Midwest, January 1997 97-02
Special Education, January 1997 97-03
Ethanol Programs, February 1997 97-04
Statewide Systems Project, February 1997 97-05
Highway Spending, March 1997 97-06
Non-Felony Prosecution, A Best Practices Review, April 1997 97-07
Social Service Mandates Reform, July 1997 97-08
Child Protective Services, January 1998 98-01
Remedial Education, January 1998 98-02
Transit Services, February 1998 98-03
State Building Maintenance, February 1998 98-04
School Trust Land, March 1998 98-05 9-1-1
Dispatching: A Best Practices Review, March 1998 98-06
Minnesota State High School League, June 1998 98-07
State Building Code, January 1999 99-01
Juvenile Out-of-Home Placement, January 1999 99-02
Metropolitan Mosquito Control District, January 1999 99-03
Animal Feedlot Regulation, January 1999 99-04
Occupational Regulation, February 1999 99-05
Directory of Regulated Occupations in Minnesota, February 1999 99-05b
Counties' Use of Administrative Penalties for Violations of Solid and Hazardous Waste Ordinances, February 1999 99-06
Fire Services: A Best Practices Review, April 1999 99-07 State Mandates on Local Governments, January 2000 00-01
State Park Management, January 2000 00-02
Welfare Reform, January 2000 00-03
School District Finances, February 2000 00-04
State Employee Compensation, February 2000 00-05
Preventive Maintenance for Local Government Buildings: A Best Practices Review, April 2000 00-06
The MnSCU Merger, August 2000 00-07
Early Childhood Education Programs, January 2001 01-01
District Courts, January 2001 01-02
Affordable Housing, January 2001 01-03
Insurance for Behavioral Health Care, February 2001 01-04
Chronic Offenders, February 2001 01-05
State Archaeologist, April 2001 01-06
Recycling and Waste Reduction, January 2002 02-01
Minnesota Pollution Control Agency Funding, January 2002 02-02
Water Quality: Permitting and Compliance Monitoring, January 2002 02-03
Financing Unemployment Insurance, January 2002 02-04
Economic Status of Welfare Recipients, January 2002 02-05
State Employee Health Insurance, February 2002 02-06
Teacher Recruitment and Retention, Research Summary, March 2002 02-07
Local E-Government: A Best Practices Review, April 2002 02-08
Managing Local Government Computer Systems: A Best Practices Review, April 2002 02-09

Evaluation reports can be obtained free of charge from the Legislative Auditor's Office, Program Evaluation Division, Room 140, 658 Cedar Street, Saint Paul, Minnesota 55155, 651/296-4708. Full text versions of recent reports are also available at the OLA web site: https://www.auditor.leg.state.mn.us