Statement of Jenifer Zeigler Submitted to the Subcommittee on Human Resources of the House Committee on Ways and Means


My name is Jenifer Zeigler and I am a welfare policy analyst atthe Cato Institute. I want to thank the committee for allowing meto submit testimony on welfare reform reauthorization proposals. Inthis statement I will summarize my findings outlined in greaterdetail in Cato Policy Analysis no. 529, "Implementing WelfareReform: A State Report Card," (available at address current reauthorization proposals.

In summary, Congress should:

  • look to the states and evaluate how welfare reform has workedand how it can improve;
  • reauthorize the Personal Responsibility and Work OpportunityReconciliation Act;
  • strengthen welfare reform's work requirements;
  • avoid federal funding of private charities;
  • avoid federal marriage programs; and
  • ultimately, replace welfare with private charity.

In the early 1990s, welfare caseloads were at an historic highand out-of-wedlock births were skyrocketing. States decided to takeaction and applied for waivers from the federal welfare program,seeking flexibility to serve their neediest citizens in a differentway. Based on success at the state level, Congress recognized itwas time to overhaul welfare on the federal level. Looking to thestates for examples of successful reform, in 1996 the PersonalResponsibility and Work Opportunity Reconciliation Act (PRWORA) wassigned into law, and the nation waited to see if welfare reformwould truly "end welfare as we know it." Block grant funding andadministrative devolution gave the states a chance to move beyondpilot programs and prove that they could transition people offwelfare more efficiently and effectively than the federalgovernment. As a result, caseloads have dropped by more thanhalf.

Since 2002, Congress has been debating the reauthorization ofPRWORA, and there are a variety of perspectives on the directionwelfare reform should now take. Once again, the federal governmentneeds to look to the states to see what has worked, and what hasnot. "Implementing Welfare Reform: A State Report Card" emphasizesthe positive policy choices made by states regarding welfare reformimplementation-choices that encourage personal responsibility andself-sufficiency.

Strong structural reforms in a state's welfare system-includingtime limits, sanctions, and narrow definitions of work activity-laythe foundation for successful reorganization. Pilot programs,waivers, and the flexible guidelines of the block grant systemallow states to experiment with programs and make policy decisionsthat best serve their citizens. It is important for Congress toreview and compare the structural reforms that states haveimplemented and the quantitative results those programs haveproduced.

Looking at Where We Have Been

Welfare reform has allowed states the flexibility to spend moneyand implement programs that will help recipients escape welfare's"cycle of dependence." The idea behind welfare reform was toprovide recipients with job experience for a better transition intothe job market, rather than to give them cash handouts for doingnothing. With job skills and an incentive to hurry off the rolls(time limits), families have been leaving welfare in recordnumbers.

The report card grades each state on program and performancemeasures. It is just as important to evaluate the programs a statehas instituted (structural reforms) as it is the results of thosereforms (quantitative results). It is necessary that states reducecaseloads and poverty rates, but if they are not establishing soundwelfare policies that will sustain self-sufficiency, manyrecipients will never completely escape the system.

The states with the highest grade ranked in the top third of thestates in both structural reforms and quantitative measures. Thosestates recognized that it is important to reduce rolls and rates inthe short term (high quantitative results score) as well as preparefor the long term by implementing strong work policies, timelimits, sanctions, and family caps (high structural reformsscore).

It is not surprising to see Wisconsin receive an "A" (along withIdaho, Ohio, and Wyoming), since much of PRWORA was modeled on theWisconsin Works (W2) system, one of the first innovations in statewelfare reform in the 1990s. Seven states earn a "B," there are 20"C" states, and 11 "D" grades. Nine states receive failing gradesfor their implementation of welfare reform. The jurisdictionsreceiving "F"s are the District of Columbia, Maine, Missouri,Nebraska, New Hampshire, North Dakota, Rhode Island, Utah, andVermont (which received the lowest of the failing grades, includingthe lowest grade on implementation of structural reforms requiredfor a successful state welfare program).

Structural Reforms

Family Caps
PRWORA authorized states to impose a family cap, which would denyincreased TANF benefits to women on welfare who have additionalchildren. Twenty-three states have established such caps.1 Family caps showrecipients that welfare is a temporary safety net, not a subsidyfor a life of dependency. If a family is not making it on its own,creating another mouth to feed is not the path toself-sufficiency.

Because a family cap is an elective policy, states can decidewhether or how best to implement it. Family cap policies vary fromstates that do not give any cash increase for an additional child,to states that do not halt incremental cash adjustments, but reducethe level, to states that technically have a family cap policy, butrather than reduce the incremented benefit, issue payment in theform of a voucher or to a third party payee.

Teens at Home
PRWORA requires unmarried mothers under the age of 18 to remain inschool and live with an adult. That was a priority in welfarereform since, by the early 1990s, half of unwed teen mothers wouldgo on welfare within one year of the birth of their first child andan additional 25 percent are on welfare within five years.2 Nearly 55 percent ofwelfare expenditures are attributable to families that begin with ateen birth.3

High school dropouts are roughly three times more likely to endup in poverty than are those who obtain at least a high schooleducation.4 Ifdropouts do find jobs, their wages are likely to be low. Wages forhigh school dropouts have declined (in inflation-adjusted terms) by23 percent during the past 30 years.5 And the economic impact is intergenerational.Children whose parents have not completed high school are far morelikely to live in poverty than children whose parents are moreeducated. Simply put, more education equals less poverty.6

TANF allows high school attendance to fulfill the workrequirement for minor teen mothers, who are supposed to remain in aparent's home while finishing school. All states are required toimplement this policy, but the specific guidelines are at thediscretion of the each state. Unfortunately, many states havecreated broad definitions and extensive exceptions that make thefederal law ineffective. Examples include 17 states that exempt ateen who has lived away from her family for a year or is"successfully living on her own."7 Just how "successful" is a teenager living onher own if she has an out-of-wedlock pregnancy and needs welfareassistance?

Work Policy
Getting a job as a solution to poverty may seem like common sense.Granted, not every job pays a wage that will catapult a family intothe middle class. However, every job provides job experience, andthat leads to a better job. Maybe today's minimum-wage, serviceindustry employee is not on a track for management. But he isshowing that he is a reliable worker who can learn and performduties, something a future employer will value.

PRWORA's addition of work requirements to TANF benefits was oneof the most substantial changes to the welfare system. Workexperience is the most effective way to move recipients off ofwelfare and into the job market, and at a lower cost than educationor job-training programs. 8

By 2002, half of each state's eligible caseload had to beengaged in "work-related" activities at least 30 hours per week.The Department of Health and Human Services (HHS) divides jobs thatqualify for work participation credit into 14 categories forreporting purposes. Of those allowable work activity categoriesunder TANF, only half are activities in which the recipient isactually working: subsidized and unsubsidized employment (publicand private), community service, on-the-job training, and workexperience. Unfortunately, states permit too much participationunder the remaining activities: job search, job skills training,adult basic education/English as a Second Language (ESL) classes,education directly related to employment, and vocational training.These should not be considered actual work activities because theyare educational and do not provide actual work experience.

Additionally, caseload-reduction credits essentially releasedstates from their participation rate obligations. Without credits,only three states would have met their single-parent participationrequirements. Through credits, 19 states were able to reduce theirwork requirement to zero. Absent waivers, exemptions, and credits,the national participation rate for recipients in actual workactivities is less than 30 percent.9

States have made it very hard on themselves by not striving tomeet the work requirement guidelines, regardless of credits. Withweakened economies and tighter budgets, states must scramble tofigure out how to create jobs for welfare recipients to meet workrequirements, and how to fund the administrative oversight suchregulations require.

Since PRWORA eliminated the welfare entitlement, states have beenfree to put conditions on the receipt of benefits. Thirty-fourstates and the District of Columbia have used this authority toestablish diversion programs that prevent potential welfarerecipients, particularly those considered able to work or who haveanother potential source of income, from ever entering thesystem.10

Generally, diversion programs fall into one of three categories.Most common are diversion programs that provide "lump sum payments"in lieu of welfare benefits.11 Those programs assist families facing animmediate financial crisis or short-term need. The family is givena single cash payment in the hope that the immediate problem can betaken care of without the need to go on welfare. In fact, a familyis usually precluded from going on welfare for a period of time,after accepting a diversion payment.

Most states do not restrict how lump sum payments may be used;they have been used to pay off back debts, as well as forchildcare, car repairs, medical bills, rent, clothing, and utilitybills. Recipients may also use lump sum payments towardwork-related expenses, such as purchasing tools, uniforms, andbusiness licenses. A few states restrict the use of lump sumpayments to job-related needs, although that definition can beinterpreted broadly. For example, even moving expenses for a newjob may qualify. 12

Another common diversion approach is a "mandatory applicant jobsearch," used by 27 states. Under this approach, welfare applicantsare required to seek employment before they become eligible forbenefits. In most cases, the state will assist with the job searchby providing job contacts and leads, access to a "resource room"where applicants can prepare résumés and conduct jobsearches, or classes in job search skills. The state may alsoprovide childcare and transportation assistance.

Finally, eight states have programs designed to encouragewelfare applicants to use "alternative resources" before receivingTANF benefits. Those programs generally do not have specificguidelines but amount to caseworkers encouraging would-beapplicants to seek help from family, private charity, or othergovernment programs.13 Even in states with alternative resourcereferral programs, this approach is the least used, possiblybecause it is poorly understood by potential recipients andrequires extensive caseworker involvement.

In Utah and Virginia, the states that have the most extensivediversion-tracking information, between 81 and 85 percent of thoseinitially diverted do not subsequently reapply for TANF.14 HHS should consider amethod of awarding states credit for participating in diversionprograms. If states are being rewarded for moving recipients offthe roles, then they should similarly be encouraged to keep peoplefrom ever entering the system.

Time Limits
Before welfare reform, pride and self-determination were the mainforces driving recipients off welfare. Unfortunately, many werecomfortable with the lifestyle that welfare benefits provided andsaw no need to work their way out of the system. They had been toldwelfare benefits were an entitlement, and with no end in sight,some dependents made welfare a way of life.

In an effort to deter such "career recipients," PRWORA setlimits to how long someone can receive welfare. The federal TANFprogram imposes a lifetime limit of 60 months (5 years). States canreduce that period or continue to support recipients after thattime with their Maintenance of Effort (MOE) money or other statefunds. Because caseloads include on-again-off-again recipients,many are just now reaching the overall five-year moratorium on aid.As recipients begin to hit the federal time limit, states arestruggling with the decision to kick families off the rolls orcontinue benefits out of scarce state funds. Eighteen states havebeen spared the dilemma as they were granted waivers before PRWORAthat allow for the exclusion of all or part of their caseloads fromtime limits. Many states have implemented categorical exemptionsfor various recipients, choosing to continue funding with their ownmoney.15

Obviously, it is not enough for states to just promulgate newwelfare policies-those policies must be enforced. If welfarerecipients fail to meet work requirements or violate other areas ofa state's welfare policy, penalties must be imposed. Modestsanctions tend to deduct only the adult portion of the TANFbenefit, sparing any children in the household and thereby onlyminimally reducing the benefit. States with the most stringentsanctions withhold the entire TANF benefit upon the firstviolation. Then there are sanction policies that fall along thespectrum, allowing multiple violations as benefits are graduallyreduced or withheld.16

Michael New, postdoctoral fellow at the Harvard-MIT data center,evaluated the effectiveness of sanctions in a Cato Institute PolicyAnalysis entitled "Welfare Reform That Works." New found that astate's sanction policy could affect caseload decline by as much as20 percent, through both the indirect effect of encouragingrecipients off the rolls and the direct effect of ending theireligibility.17Not only is there a relationship between state sanction policy andcaseload decline, New found, but that relationship is constant overseveral years.18

Sanctions are not successful because they throw recipients offwelfare; rather they serve as a threat of actual consequences forfailing to meet requirements or reaching time limits. Only aboutsix percent of those leaving welfare have done so due to sanctionenforcement.19However, there is a wide variation among states as to thepercentage of their caseloads affected by sanctions. For example,in an average month in 1998, almost 30 percent of case closures inNorth Carolina were due to sanctions, while less than 1 percent ofclosures in California, Oklahoma, and Nebraska were related tosanctions. 20

Quantitative Results

Caseload Reductions
The greatest decline in welfare rolls occurred in the first twoyears following the enactment of welfare reform. Caseloads began tolevel out in most states by 1998, and some states that experiencedthe most significant initial declines began to see caseloads inchback up. New Mexico, for example, reduced its rolls by almost halfin the first two years following reform, and then had a nearly 25percent increase in 1999. Delaware, Tennessee, and Wisconsin alsosaw their caseloads increase after initial declines.21 As the economy began toslow in 2001 and 2002, the era of declining caseloads came to aclose. In 2002, 26 states experienced higher caseloads than theyear before, although all state caseloads remained significantlybelow prereform levels. 22

Poverty Rate and Child Poverty Rate
Poverty rates mirrored the success of caseload reductions asnational poverty rates declined every year after reform until 2001.Even though 2002's slow economy caused a minor uptick in povertyrates, they continue to remain well below prereform rates.23 Most significant,poverty rates declined for women, children, and minorities, groupsthat were thought to be most at risk. Many critics of welfarereform issued dire predictions, forecasting at the time PRWORA waspassed that more than a million children would be thrown intopoverty.24Instead, child poverty rates declined from 20.5 percent in 1996 to16.2 percent in 2000, the lowest level in more than 20 years.25

Teen Birth Rate
For many women, having a child out of wedlock leads to a lifetimeof poverty. Once on welfare, single mothers often find it verydifficult to escape. Although the average recipient remains onwelfare for less than two years,26 by the early 1990s almost 40 percent of allnever-married mothers on welfare remained on the rolls for 10 yearsor longer.27

Teen mothers now account for roughly 29 percent of allout-of-wedlock births. That figure, however, may understate theseverity of the problem. Women who give birth out of wedlock asteens frequently go on to have additional children out of wedlock.More than a third of all out-of-wedlock births to mothers agedtwenty and over are to women who had their first child as unwedteenager.28

Teenage birth rates peaked nationally at 61.8 in 1991 and havefallen by 27 percent in the past decade.29 It is essential that states continueto reduce teenage pregnancy if there is to be any hope of endingwelfare dependence. If states can dissuade young women from givingbirth out of wedlock in their teenage years, more women willcomplete school and have a better chance for a self-sufficientfuture. Reduction in births to teenagers is an important measurebecause it shows whether states are laying the groundwork to breakthe cycle of welfare dependence.

Looking at Where We Are Going

The greatest result welfare reform could produce would be theelimination of the welfare system. Colonial America had only amodest government safety net. Churches, charities, and thecommunity-known as "civil society"-took the lead in providingassistance to those in need. These entities had the freedom todistinguish between the "deserving" and "undeserving" poor. Thedeserving poor included those who, although normallyself-sufficient, had experienced temporary setbacks due tosickness, accident, or loss of employment during a recession. Thedeserving poor also included those incapable of self-sufficiency,such as the elderly and orphans. The undeserving poor were thosewho could be self-sufficient but elected not to work, or who madepoor choices that were an obstacle to employment.30

Early U.S. welfare law was modeled after English Poor Law. Thatlaw established four basic principles for government charity: (1)care for the poor was a public responsibility; (2) care for thepoor was a local matter; (3) public relief was denied toindividuals who could be cared for by their families; and (4)children of the poor could be apprenticed to farmers and artisanswho would care for them in exchange for work.31 As with civil society'sassistance, the themes were personal responsibility andself-sufficiency. If you were able-bodied, you should be working.If you could not work, then assistance was best delivered on thelocal level to ensure effectiveness and accountability.

Unfortunately, the United States did not maintain its modestsafety net. Politicians learned that the promise of social programswins elections, and the economic repercussions of such programs arefor the next president to worry about. As each presidentattempted to shower more "compassion" on those in need, the numberof needy continued to rise. For many, the satisfaction of earning asalary was vanquished by the temptation to draw a check for doingnothing.

Welfare reform is a step in the right direction, attempting toreverse the growth of a federal welfare state that had beenexpanding for decades. PRWORA removed the entitlement to cashassistance and now sends the message that welfare is meant to betemporary, not a way of life. As welfare administration continuesto devolve from the federal government to the states, andeventually to more local levels, communities will effectivelyassume responsibility for the welfare system. Those localities,held accountable by local residents and voters, will begin to findinnovative ways to meet the needs of the poor, using charitableorganizations and encouraging civil-society solutions rather thanrelying on government.

Corrupting Charity

Just because something is a good idea does not mean it should bea government program. In the case of faith-based organizations,government involvement can easily kill the very entity it is tryingto nurture. During the past decade, the federal government hasrecognized the successful results that come from social servicesdelivered by civil society, including religious organizations. Itis the charity's autonomy and flexibility that allows for itssuccess, yet these characteristics are threatened by the red tapeand liability that come with government funding.

Many faith-based organizations lack the manpower, financialresources, and technical knowledge to deal with mountains ofpaperwork, much less sorting out all of the new rules andregulations.32Religious entities succeed because of their focus on theindividuals they are serving; their strength lies in their care forothers, not their careful reading of the FederalRegister.

Faith-based initiative money is certainly a temptation for thoseserving the needy. If they are serving many now, how many morecould they serve with more funding? Unfortunately, federal fundingis not reliable, and faith-based organizations are susceptible tomission creep-following the subsidies and rewriting their missionto fit whatever grant is popular that year.33 Essentially, through funding, thegovernment can kill a successful charity, forcing it to change fromwhatever service it was successfully offering or to shut down dueto lack of funding. Faith-based organizations are crucial membersof civil society that need to replace the federal welfaresystem, not be dependent on it.

Federal Marriage Programs

Another area where Congress should resist the urge to "do good"is the marriage initiative. We all agree that marriage is a goodidea. Social science shows that marriage is good for society. Butas I previously mentioned, not every good idea should be federallyfunded. Often what is good for society needs to be promotedprivately, not forced onto society by the government.

Additionally, promoting marriage as a solution to poverty is aninsult to those who are struggling to escape poverty. Who, exactly,are these women supposed to marry? In areas of high poverty (andaccompanying crime and unemployment), there are relatively fewmarriageable men.34 Studies show that the fathers of childrenborn out of wedlock are not men who will lift single mothers out ofpoverty-more than a third lacked a high school diploma, 28 percentwere unemployed, and another 20 percent had incomes of less than$6,000 per year. In addition, roughly 38 percent had criminalrecords.35

If Congress wants to encourage marriage, it should start byremoving the disincentives to marriage. The current welfare system,as well as our tax code, erect barriers to marriage by reducingbenefits and/or increasing tax liability if a couple weds. Beforethe government starts spending new money on incentives, itshould fix current programs to reflect its pro-marriage agenda.Additionally, research shows that financial difficulty is one ofthe leading causes of divorce. Congress should focus its resourceson encouraging a dynamic economy, through lower taxes and lessregulation of business. Job security, higher wages, and a lightertax burden would go a long way toward securing maritalstability.


Congress needs once again to look to the states and evaluatewhat has worked under welfare reform. We need to keep moving in thedirection of devolution and innovation, placing more control in thehands of local government and encouraging civil society to play abigger role in helping the neediest members of the community.Congress can help the states with their own dependency problem byweaning states off federal funding. Without the strings that comewith federal dollars, states would have even greater flexibility tobe innovative and efficient. Partnering with local nonprofits andcommunity organizations, states could encourage a shift in thesafety net back to civil society, where it belongs.


1. U.S. Department of Health andHuman Services, TANF Fifth Annual Report to Congress, February2003, sec. II, "Trends in Caseloads and Expenditures," p. II-5.

2. G. Adams and R. C. Williams,Sources of Support for Adolescent Mothers (Washington:Congressional Budget Office, 1990), pp. 49-51.

3. Richard Wertheimer and KristinMoore, "Childbearing by Teens: Links to Welfare Reform," UrbanInstitute, New Federalism Issues and Options for the States, SeriesA, no. A-24, August 1998.

4. U.S. Census Bureau, unpublishedtabulations from the Survey of Income and Program Participation,2000,

5. Lawrence Mishel, Jared Bernstein,and John Schmitt, The State of Working America, 2000-2001 (Ithaca,NY: Cornell University Press, 2001), p. 153.

6. Uri Bronfenbrenner et al., TheState of Americans (New York: Free Press, 1996), pp. 176-77.

7. Sources: State PolicyDocumentation Project, "Minor Living Arrangement: Eligibility andExemptions," February 1999,; and 07Alaska Administrative Code 45.227 "Assistance to a Minor Parent,"2003.

8. U.S. Department of Health andHuman Services, "HHS Releases Evaluation of Welfare-to-WorkStrategies," November 7, 2001. The press release summarizes 26separate studies. A complete list of those studies can be found

9. U.S. Department of Heath and HumanServices, "Temporary Assistance for Needy Families ProgramInformation Memorandum," TANF-ACF-IM-2004-03, December 27,2004.

10. Kathleen A. Maloy et al., "ADescription and Assessment of State Approaches to DiversionPrograms and Activities under Welfare Reform," George WashingtonUniversity Center for Health Policy Research, August 1998, TableI-1.

11. Although more states haveauthorized lump sum payments than any other type of diversionprogram, the U.S. Department of Health and Human Services reportsthat those programs are rarely used in practice. Kathleen Maloy etal., "Diversion as a Work-Oriented Welfare Reform Strategy and ItsEffect on Access to Medicaid: An Examination of the Experience ofFive Local Communities," U.S. Department of Health and HumanServices, Office of the Assistant Secretary for Planning andEvaluation, March 1999, pp. 8-9. Utah, Virginia, and Montana appearto have the most extensive experience with the concept.

12. U.S. Department of Health andHuman Services, Office of the Assistant Secretary for Planning andEvaluation, "A Description and Assessment of State Approaches toDiversion Programs and Activities," August 1998, chap. 2,

13. Ibid., chap. 3,

14. U.S. Department of Health andHuman Services, Office of the Assistant Secretary for Planning andEvaluation, "A Description and Assessment of State Approaches toDiversion Programs and Activities," August 1998, chap. 2,

15. Forty-six states have put inplace exemptions for parents or caretakers of children withdisabilities and others caring for a disabled family member.Forty-two states exempt women in cases of domestic abuse, and 26states exempt elderly recipients. Other states grant exemptions forindividuals making a "good-faith" effort to find work (23 states),parents with young children (22 states), recipients engaged in"work activities" (22 states), recipients enrolled in educationalor training programs (21 states), and families in areas of highunemployment (19 states). General Accounting Office, "WelfareReform: With TANF Flexibility States Vary in How They ImplementWork Requirements and Time Limits," pp. 16-18.

16. General Accounting Office,"Welfare Reform: State Sanction Policies and Number of FamiliesAffected," March 2000, pp. 44-47.

17. Michael New, "Welfare ReformThat Works: Explaining the Welfare Caseload Decline, 1996-2000,"Cato Institute Policy Analysis no. 435, May 7, 2002, p. 8.

18. Ibid., p. 6.

19. An additional 16 percent haveleft as a result of "state policies," which could include timelimits or other administrative regulations. U.S. Department ofHealth and Human Services, "Characteristics and FinancialCircumstances of TANF Recipients, FY 1998,"

20. General Accounting Office,"Welfare Reform: State Sanction Policies and Number of FamiliesAffected," pp. 52-53.

21. U.S. Department of Health andHuman Services, TANF Fifth Annual Report to Congress, sec. II,"Trends in Caseloads and Expenditures," Table 2:2:c, p. II-34.

22. U.S. Department of Health andHuman Services, "Change in TANF Caseloads"; U.S. Department ofHealth and Human Services, "TANF: Average Monthly Number ofRecipients-Fiscal Year 2001," February 2002,; and U.S.Department of Health and Human Services, "TANF Total Number ofFamilies and Recipients January-March 2002," November 2002,

23. U.S. Census Bureau, "Poverty inthe United States: 2002," September 2003, p. 1-3,

24. See, for example, SheilaZedlewski, "Potential Effects of Congressional Welfare ReformLegislation on Family Incomes," Urban Institute, 1996,

25. U.S. Census Bureau, "PovertyStatus of People by Age, Race, and Hispanic Origin: 1959-2000,"

26. U.S. Department of Health andHuman Services, Indicators of Welfare Dependence: Annual Report toCongress 2004, June 2004. p. II-31.

27. Barbara DaFoe Whitehead, "DanQuayle Was Right," Atlantic Monthly, April 1993, pp. 47-84.

28. Elizabeth Terry-Humen et al.,"Births Outside of Marriage: Perceptions vs. Reality," Child TrendsResearch Brief, Washington, April 2001, p. 2.

29. "Revised Birth and FertilityRates for the 1990s and New Rates for Hispanic Populations, 2000and 2001: United States," National Vital Statistics Reports 51, no.12 (August 4, 2003): 4.

30. Marvin Olasky, The Tragedy ofAmerican Compassion (Washington: Regnery, 1992), pp. 6-24.

31. Michael B. Katz, In the Shadowof the Poor House: A Social History of Welfare in America (NewYork: Basic Books, 1986), pp. 13-14.

32. The average church in theUnited States has a congregation of only 75 members. Less than 1percent of churches have congregations of more than 900, and lessthan 10 percent have congregations exceeding 250 people. Theaverage annual church budget is only $55,000. Mark Chaves,"Religious Congregations and Welfare Reform," Social Science andModern Society 38 (January-February 2001): 26.

33. Stanley Carlson-Thies,"Faith-Based Institutions Cooperating with Public Welfare: ThePromise of the Charitable Choice Provision," in Welfare Reform andFaith-Based Organizations, ed. D. Davis and B Hankins (Houston, TX:Baylor University, 1999), p. 38.

34. Kathryn Edin, "Few Good Men:Why Poor Mothers Don't Marry or Remarry," American Prospect, June2, 2000.

35. Sara McLanahan et al., "TheFragile Families and Child Well-Being National Baseline Report,"Princeton University, 2001; and Irwin Garfinkle et al., Fathersunder Fire: The Revolution in Child-Support Payments (New York:Russell Sage Foundation, 1998).

Jenifer Zeigler

Subcommittee on Human Resources
Committee on Ways and Means
The United States House of Representatives