Preparing for Welfare Block Grants: Issues Facing California

James N. Dertouzos and Robert F. Schoeni

Although reform legislation has not been finalized, it is clear that the 104th Congress intended to consolidate separate federal programs within a variety of broad domestic policy areas, including welfare, Medicaid, transportation, and job training. The elimination of categorical programs and increasing reliance on block grants pose significant challenges to state and local governments, which will need to develop new capabilities to establish and administer programs, allocate funds, evaluate policy options, and demonstrate program effectiveness. For example, in some versions of welfare reform, administrative burdens and associated costs might be shifted to states and communities at the same time that overall funding levels for programs are reduced. Longer-term budgetary pressures would increase further if the revision of "entitlements" provisions means that federal funding levels become less responsive to increased needs of the local populations. For example, legal noncitizen populations in the United States less than five years would not be eligible for certain welfare payments, or overall funding levels may not adjust on the basis of changes in the numbers of eligible populations. As a result, states will wish to adopt programs that do not induce relocation of individuals seeking higher benefit levels. Finally, if state budgets are fixed in the short run and the demand for assistance increases, states will need to ration limited funding by lowering benefits, limiting the duration of program eligibility for individual families, or denying benefits to some population segments.

Despite these challenges, the devolution of government and the emphasis on local control over the design and execution of social programs create a unique opportunity to reinvent the manner in which such services are provided. In particular, states and communities have an opportunity to restructure and coordinate the current array of aid programs to make them more consistent and effective in serving the needs of families. If they become less encumbered by federal restrictions, states will be able to tailor social service efforts, make policy tradeoffs that are based on cost-benefit criteria, and approach regional and local problems in a strategic rather than piecemeal fashion.

For this to happen, state and community leaders will have to think about social services in a whole new way. They will have to abandon "stove-pipe" approaches to policymaking, restructure social service agencies, and collaborate across traditional policy domains in order to make sound decisions. In addition, they will have to develop institutional mechanisms for creating and sharing information, establishing oversight, and evaluating alternative policy options. The greatest challenge, however, may come from the need to define overall strategic goals, without which an integrated strategy cannot be designed and implemented effectively.

In this paper, we first provide a brief description of the existing welfare program as well as other social services that have administrative links (for example, via eligibility requirements) or otherwise serve the same or overlapping populations. We then review the most recent versions of legislation and describe the key programmatic attributes that are likely to emerge. Next, we identify some of the most important issues that states and communities will have to address in response. We conclude with a discussion of some of the more interesting options already adopted by some states and review the available evidence concerning the likely efficacy of alternative approaches to welfare reform.

Current System of Assistance

Established in 1935, Aid to Families with Dependent Children (AFDC) provides cash grants to needy children under the age of 18 whose families cannot meet their basic needs. When AFDC was established, assistance was meant primarily for single mothers who had been widowed or whose husbands had abandoned them. At that time, it was expected that a mother's primary responsibility was to work in the home, including raising her children, and the AFDC cash grant allowed a single mother to do just that. Today, the norms and expectations are quite different. While only 25 percent of working-age women were employed outside the home in 1930, current levels of labor force participation are about 70 percent (Bureau of Labor Statistics, 1996). Moreover, today over 50 percent of mothers participate in the formal labor market within six months of childbirth (Klerman and Leibowitz, 1993). Perhaps more important, the prevailing wisdom is that rising caseloads and long average duration of welfare dependence are directly the result of incorrect personal incentives that stem from the design of social programs. Thus, it is widely believed that single mothers (or both parents in two-parent households) should not be perennially insulated from the need to support their family via cash assistance, but instead be encouraged to work outside the household and take responsibility for their own well-being.

Families qualify for AFDC under one of two categories: AFDC-FG (i.e., Family Group), which consists of single-parent families, and AFDC-U (i.e., Unemployed Parent), which consists of two-parent households in which at least one parent is unemployed. Currently, in California, only about 15 percent of the AFDC caseload qualifies for assistance under the AFDC-U category. However, cash grants are typically higher to those in the AFDC-U category because these families are larger and therefore have greater need. As a result, assistance to AFDC-U households accounts for about 25 percent of all cash grants made through the AFDC program. In total, approximately 907,000 households, or 2,660,000 individuals, received AFDC assistance in California in August 1995, which is 8.3 percent of the state's population (California Department of Social Services, 1995a).

Eligibility and the amount of assistance received under AFDC is determined by a number of factors that are set by each state. In California, the amount of assistance increases by roughly $100 per month for each additional child in the family, and the benefits decrease (on a dollar-for-dollar basis) with the amount of labor market income earned by the mother. Applicants are not eligible if they hold assets of more than $1000 (excluding their home and the value of their car up to $1500), and many children are not eligible if both the mother and father live with the child.

In California, the maximum benefit for a mother with two children is $607 per month in cash assistance plus an additional grant of about $214 in food stamps, totaling approximately $9852 annually. Although these combined benefits still leave the family's income about 15 percent below the poverty line, they are higher than the benefits in all but three states. The median maximum combined AFDC and food stamp payments across all states is $7932 annually. However, California is a state with a relatively high cost of living. Compared with New York, a state that probably has a similar cost of living, California's benefits are about the same.[1] It is worth noting, however, that while in the average state the maximum benefit has decreased by 47 percent (in real terms) over the past 25 years, California's benefits have eroded by only 13 percent, which is the lowest decline registered among all states. However, between 1989-1990 and 1995-1996, the maximum benefit fell by 26 percent, so that changes in benefits in the last five years have been distinct from the changes in the earlier period (U.S. House of Representatives, 1995).

In California, the federal, state, and local governments all play a role in the AFDC program. The federal government prescribes overall rules regarding eligibility, benefit standards, and administrative requirements for the program. The state provides additional administrative guidelines, sets benefit levels, and allocates state and federal money to counties. California is one of a minority of states in which local governments also play a significant role in welfare delivery, with counties being responsible for implementing and delivering the services to applicants.

The cost of the program in California is shared by all three levels of government. In California, the federal government pays 50 percent of the costs of the benefit payments and administration of the program.[2] California's 50 percent contribution is the highest in the nation, with other states averaging 46 percent. States with the lowest per-capita income receive as much as 80 percent of the benefit costs from the federal government. Out of expenditures not reimbursed by the federal government, counties contribute 5 percent of benefit payments (primarily for special assistance) and 30 percent of local administrative costs. The state pays for 100 percent of the administrative expenses incurred at the state level that are not reimbursed by the federal government. In total, government expenditures on AFDC in California were about $6.7 billion in 1995, with about 50 percent, 46 percent, and four percent derived from federal, state, and local funds, respectively (U.S. House of Representatives, 1995).

The AFDC program serves a wide array of people in different predicaments. About 75 percent of all new episodes of welfare dependence were precipitated by a personal "event" such as the birth of a child (30 percent) or the loss of a spouse (45 percent were the result of a divorce or separation). The average age of single mothers on AFDC in California is 31, and only 3.3 percent of the mothers are currently teenagers. However, national estimates suggest that almost one-half of all AFDC recipients had a teen birth (Maynard, 1994). Forty-one percent of current participants have just one child, and only 11.2 percent have more than three children, which is counter to the myth that AFDC recipients have unusually large numbers of children (California Department of Social Services, 1995b). About one-half of AFDC recipients nationally have less than a high school degree, and 40 percent did not work at all in the previous two years (Ellwood, 1986).

One-quarter of California's population is foreign-born and about 17 percent are not U.S. citizens, which has unique ramifications for the state's AFDC program. About 10 percent of persons in the AFDC-FG assistance units are not U.S. citizens, while among AFDC-U assistance units the share is 31.2 percent. In addition, approximately 30 percent of recipients do not speak English as their primary language. Thirty percent of recipients are (non-Hispanic) white, while 37 percent are Hispanic, 18 percent are black, and the remainder come from a variety of other ethnic backgrounds (California Department of Social Services, 1995b).

AFDC is the primary source of cash income for most participants, with only 17 percent of California recipients reporting cash income from other sources (California Department of Social Services, 1995b). Some mothers physically cannot work--an estimated 19 percent report that a disability limits their ability to work (Ellwood, 1986). This is an important factor when considering that the proposed federal legislation requires that at least 50 percent of AFDC-FG recipients work, and that many of these families may not be eligible for aid after receiving assistance for five years in total over their lifetimes.[3] At the same time, about one-half of AFDC recipients use the program for more than five years during their lifetimes, and 30 percent are on for eight or more. Just 30 percent of all recipients are enrolled for less than two years over their lifetimes (Ellwood, 1986). In addition, counter to popular belief, although caseloads have increased nationally, the length of time that participants remain enrolled in AFDC has, if anything, decreased between the early 1970s and late 1980s (which is the latest available data), and there is no evidence that the duration of welfare use among long-term users has increased (Hoynes and MaCurdy, 1993).

Most AFDC participants are also affected by other government programs. All states are required to maintain a job opportunities and basic skills program, whose purpose is to facilitate transition off welfare and into the workforce. In California, this program is called Greater Avenues for Independence, and the monthly GAIN caseload in 1995 was estimated at 78,000 (California Department of Social Services, 1996). Families receiving AFDC are automatically eligible for Medi-Cal, which is California's version of Medicaid. Although the bulk (about 75 percent) of the $89 billion cost of Medicaid goes to paying for the long-term care of disabled and elderly populations, this program also provides basic health care for 26 million poor individuals nationally, including all AFDC recipients. Some 85 to 90 percent of California AFDC recipients participate in the federal food stamp program, increasing the basic package of assistance (i.e., cash assistance plus food stamps) by about 30 percent (California Department of Social Services, 1995b). Child care assistance is also available, with the government covering up to 75 percent of the regional market rate for child care (California Department of Social Services, 1996).

Although the federal government sets basic AFDC guidelines, since 1962 the law (Section 115 of the Social Security Act) has given states the option to apply for waivers that have allowed them to experiment with alternative AFDC structures. As a result, welfare reform has been implemented, albeit at a limited level, in a large number of states. California, in particular, has been granted a number of waivers. For example, one waiver increases the asset limit from $1000 to $2000, permits families to own an automobile valued at $4500 instead of just $1500, and allows families to retain up to $5000 in a restricted savings account. Money from the savings account can be withdrawn only for the purchase of a home, education of children in the assistance unit, or to start a new business. Another waiver allows California to reduce benefits for recipients who can be expected to work (e.g., those without disabilities). In addition, to stimulate work the waiver increases the earnings disregard and removes the limitation on two-parent families working more than 100 hours per month. A waiver has also been received that ties benefit levels to schooling performance. Under the Cal-LEARN program, pregnant and parenting teens receive rewards for completing high school and receiving good grades.

Impending Changes in the Welfare System

As of this writing, the nature of the federal legislation that will finally emerge remains uncertain. However, the House and Senate Conference Committee has agreed to a reform proposal that was passed by the House on December 21 and the Senate on December 22, 1995. Although President Clinton vetoed the bill on January 9, 1996, he has agreed to many of the fundamental aspects of the proposed overhaul. Thus, the system that eventually emerges will probably have many of the characteristics of the Conference Committee agreement. [4] The following are the most important elements.

Cash Welfare Assistance

Structure of Grants and Eligibility Requirements

  • Cash welfare assistance that has historically been distributed through the AFDC program as an entitlement will be turned into a block grant allocation to states, called the Temporary Assistance for Needy Families Block Grant.
  • Welfare eligibility, except for a minority of cases (up to 15 percent of caseload) experiencing unusual hardship, will be restricted to a cumulative lifetime total of no more than two years of cash assistance without working. Individuals can receive no more than an additional three years of benefits if they work in either a subsidized private sector job or a public sector job. It is currently estimated that 70 percent of AFDC participants use welfare for more than two years over their lifetimes, and 45 percent are enrolled for more than five years (Ellwood, 1986). Therefore, a large share of recipients are likely to be affected by the time limits.
  • Over time, an increasing percentage of welfare recipients in each state will be required to engage in work-related activities. For example, among all families participating in the program, adults in 15 percent of families must be working in 1996, increasing by about 5 percentage points in each year until the year 2002, when 50 percent of single parents receiving cash subsidies must be working.[5] Among two-parent families receiving cash assistance, 90 percent of families must have at least one working family member by 1999. Those states whose caseloads do not satisfy the work requirements will have grants reduced by 5 percent.
  • States must deny benefits to single mothers under age 18 unless the mother lives with her parents or in a supervised setting and is enrolled in school or a training program.
  • Cash benefits cannot be increased if a mother has an additional child while receiving welfare, unless the birth is due to rape or incest. However, there is a state "opt-out," so any state can pass a law overturning this prohibition.
  • Applicants who have resided in the state for less than 12 months can be subject to more-stringent eligibility criteria and benefit levels. For example, benefits are likely to be limited to those available in the previous place of residence. Thus, individuals would have less incentive to migrate in search of higher welfare benefits.
  • Parents must assist in determining paternity in order to maintain their cash assistance, unless contacting the father may lead to harmful consequences for the mother or child.

Level of Funding to States, and State Requirements and Opportunities

  • The amount of federal assistance that a state receives will be the higher of the funding levels from 1995, 1994, or the average of 1992-1994.
  • Over the first five years of implementation, federal funds for welfare, including AFDC, SSI, and food stamps, can be expected to decline in real terms for many states. This is because there will be no adjustment for inflation or for increases in populations who would have been eligible under previous rules (e.g., noncitizens).
  • Higher funding will be awarded to those states that decrease their abortion and illegitimacy rates. If illegitimacy declines by at least 1 percent (2 percent) and the abortion rate declines, then the state will receive an additional 5 percent (10 percent) in welfare block grants.
  • Funding allocations will not be very responsive to future changes in local economies and/or populations. In other words, increases in "need" will not be met with equivalent increases in funding. However, if California has population growth higher than the average across all states, or its population grew by 10 percent or more between 1990 and 1994 (which California's did not--it grew by 7.2 percent), or its welfare spending per poor person in the state is less than the national average (which it is not), then it will receive an additional 2.5 percent in federal grants.
  • States will be able to borrow from a federal contingency fund (for an amount up to 20 percent of total state welfare expenditures) in order to provide welfare services if the state is experiencing financial difficulties; the loan must be repaid with interest. States are also permitted to save an unlimited amount of their block grant funds for use in later years.
  • States must establish an electronic data tracking system that will allow them to determine the total lifetime benefits received by all applicants and recipients. This must include information on receipt of benefits in other states and the applicant's immigration status, including the date at which an immigrant became a citizen.
  • In general, some program requirements for design and administrative reporting may be eliminated, giving states opportunities to consolidate, avoid redundant administrative costs, and design innovative programs that are more cost-effective and provide target populations with better incentives to reduce their dependence on government subsidies in the short and long run.

Changes in Other Social Services

The changes in assistance to the poor are not restricted to modifications of the cash assistance program. For example, eligibility requirements for food stamps are currently the same across all states, but the legislation permits non-uniform standards of eligibility. However, federal costs of any new state food stamp plan must not exceed previous federal costs (after adjusting for changes in the cost of the Thrifty Food Plan). States will also have the option to receive food stamp funds through a block grant, with the requirement that they establish an electronic benefit transfer system. Able-bodied beneficiaries 18 to 50 years old who do not have dependents would have to work to receive assistance.

Various nutrition programs (e.g., the National School Lunch and Breakfast programs and Women Infants and Children program) are currently provided by the federal government as entitlements to low-income children and adults. States determine the income eligibility guidelines (within federally mandated limits). Much of the assistance is in the form of actual food items or nutritional screening services. The Conference agreement makes no cuts in school lunch or school breakfast programs and maintains the Women, Infants, and Children program. However, it does allow as many as seven states to receive their allocations in the form of nutrition block grants.

Supplemental Security Income (SSI) will remain a federal program (with optional state supplementation), but it will undergo change. Individuals who are currently eligible for SSI assistance because of drug or alcohol addiction will no longer receive benefits. The eligibility requirements for children are likely to change, perhaps reducing the number of eligible children. In addition, parents' resources will not be counted against children in determining their eligibility. Currently, unless a child is institutionalized, the parents' income is counted against the child, which provides an incentive for parents to move their children into institutionalized care.

Child support enforcement will continue as a matching grant program, with the federal government currently reimbursing each state 66 percent of the costs of administering its Child Support Enforcement program. (Ninety percent of the states' costs for developing and improving management information systems is reimbursed by the federal government.) The Conference agreement will permit states to withhold, suspend, or restrict the use of driver's licenses, professional or occupational licenses, and recreational licenses of people not abiding by child support laws. Passports will be refused to those owing $5000 or more. The agreement expands the Federal Parent Locator Service to better track down deadbeat parents. In addition, states must establish an Automated Case Registry for each case in which services are provided, including the names, Social Security numbers, and dates of birth of both parents. States would also be required to work with financial institutions to retrieve financial data on delinquent child support payers. A lien may be placed on the assets held by the institution. Those who owe child support will not be eligible for food stamps.

Under child protection services, foster care and adoption maintenance payments remain open-ended entitlement programs receiving matching funding from the federal government. A Child Protection Block Grant will focus on prevention of abuse and services to abused children. A Child and Family Services Block Grant will be established to replace the Child Abuse Prevention and Treatment Act, the Abandoned Infants Assistance Act, adoption opportunities under the Child Abuse Prevention and Treatment and Adoption Reform Act, family support centers under the McKinney Homeless Assistance Act, and the temporary Child Care and Crisis Nurseries Act. Federal funding for the block grants will increase from $2.047 billion in 1987 to $2.766 billion in 2002. Each state's share of these resources will be determined by the share they have received historically. Clearly, this will hurt states that experience an increase in eligible populations.

Finally, for child care, a Child Care Block Grant will be established as part of the Child Care and Development Block Grant (CCDBG). As in other block grants, each state will receive a grant equal in size to the amount it currently receives from the total of AFDC Child Care, transitional Child Care, and At-Risk Child Care programs.

Benefits and eligibility for immigrants will be reduced in most social service programs. Legal resident aliens and those arriving after the enactment of the legislation may not receive SSI or food stamps unless they have worked in the United States for 10 years or until they receive citizenship. States will have the option of providing benefits to legal aliens under the Temporary Assistance for Needy Families Block Grant, Medicaid, and Title XX programs. Legal noncitizens arriving in the future will be denied benefits under all federal means-tested programs (except emergency medical services under Medicaid, short-term emergency disaster relief, school lunch, foster care and adoption, and immunization) for five years after their entry into the United States. (Refugees, asylees, veterans, and those working in the United States for at least 10 years are exempted.) Those who remain legal noncitizens after being in the United States five years will continue to be denied SSI and food stamps; states will have the option of providing these immigrants with cash welfare and Medicaid and Title XX services.

As indicated earlier, Medicaid is viewed as an entitlement, providing basic health care coverage to poor children, including all AFDC recipients. However, recent proposals seek to limit Medicaid costs, which have grown over 400 percent since 1985. By giving states lump-sum block grants, future funding is expected to be limited even though demand for services is likely to grow along with eligible populations of destitute children. Also, as insurance coverage via private sector employment is reduced, many more children of employed parents will require Medicaid. State options to save money include reducing benefits, reducing eligible populations, and mandating (or encouraging) employer-based coverage. In the face of budget pressures, states are more likely to maintain coverage for the very poor, nonworking families. In contrast, the benefits received by poor working families are more vulnerable to reduction or eventual elimination. This could enhance the attractiveness of AFDC relative to employment, thereby working against the reform goals of reducing long-term dependence by encouraging welfare recipients to work.

Issues Facing the State and Localities

Federal Reform Includes Reduced Funding

These likely elements of change will create several challenges for California. One of the most salient issues is that block grants will go hand-in-hand with significantly lower funding levels in the long run. Although future allocations to states will be set at the highest level of recent years, inflation will quickly erode the purchasing power of these payments and the funding level will not increase, even if the demand for services or the cost of providing them continue to increase in the future. If California agencies maintain "business as usual," at the very least they will have to reduce the amount of services provided, increase revenues, or reallocate funds from other components of their budgets. California's Legislative Analyst's Office estimates that, relative to projections under the status quo, the congressional welfare reform legislation would cost the state $8 billion in federal aid for AFDC, SSI, and food stamps over the next five years. This represents about 20 percent of the total budget for these programs.

California and all other states will not have the incentive to supplement cash welfare funding at the same level as they have in the past. In the past, the federal government matched state funding levels. As indicated earlier, the match was between 50 to 80 percent, depending on the income of the state. This match disappears under block granting, which implicitly increases the state's price of an additional dollar of welfare spending and might be expected to lead to a decrease in the amount of assistance the state wishes to allocate to welfare.[6] In the past, Governor Wilson has sought decreases in benefits through waivers, and the reduction in federal requirements may provide some opportunity to make these changes. On the other hand, there is currently a requirement that states must spend at least 75 percent of their 1994 level of spending; states that are most successful in moving families off of welfare will be allowed to reduce spending below 75 percent.

Will the State Plan for "Rainy Days"?

The amount of future federal funding to states will be based primarily on current funding levels. Except to a limited degree, changes in federal funding streams will not adjust in the face of demographic or economic trends experienced at the state level. As a consequence, states will have to shoulder the burden of planning for contingencies. This could be problematic if, for example, caseloads increase substantially during a recession and if the states do not have the resources or budget flexibility to serve many of those in need. Most of the federal legislative proposals do allow for some adjustment for states experiencing larger-than-average population increases or increases in unemployment. There are also mechanisms (similar to a "loan bank") for adjusting the funding steam in response to cyclical fluctuations in need. However, under the proposed formulas for adjustment, federal funding will not be as elastic as the potential change in need under plausible scenarios likely to affect many states, and these "loans" must be repaid.

Will Programs be Integrated?

It is clear that the new regime will require more comprehensive approaches to both policymaking and service delivery. At the very least, the replacement of categorical programs with more-flexible (albeit lower-level) block grant funding will require states to consider broader cost-benefit tradeoffs and will provide unique opportunities to consider systemic, coordinated approaches to program design. In addition, states will have new incentives to develop overarching strategic plans. Moreover, many of the mandated changes in the structure of particular programs will have important implications that transcend narrow domains or individual agencies. For example, the more optimistic goals regarding time-limitation and work requirements for welfare will be unattainable without coordinated efforts to facilitate the transition to work. To meet federal mandates, effective JOBS (Job Opportunities and Basic Skills) programs, alternative health insurance coverage, daycare options, and transportation to work will have to be available. In addition, many of the proposed program attributes, including work requirements, time limitations, and restrictions on eligibility for immigrants, could leave many populations, especially children, without a safety net. States may wish to design alternative means of providing benefits in such circumstances.

Currently, state agencies are structured and organized to parallel the narrow and restricted funding streams provided by the federal government. State and local agencies are, for the most part, charged with implementing predetermined policies and adhering to a cumbersome set of administrative regulations. State and local officials are mostly concerned with following the rules, avoiding risk, counting the number of families served, and keeping the federal funding flowing. There is little room for innovation, coordinated service delivery, new program design and evaluation, or comprehensive strategic planning across traditional functional domains. However, if states are to meet the new challenges and opportunities posed by block grants, fundamental reform must be made in the way social services are provided. This will require enhanced flows of information, new decision rules, and broader criteria for evaluating policy options. At the very least, California will have to establish new mechanisms for facilitating interactions across programs and make complex decisions based on multiple, often conflicting objectives. It is likely that organizational barriers, bureaucratic inertia, and deeply imbedded cultural resistance will make the requisite changes difficult to accomplish without a complete restructuring of social service agencies and the way in which they conduct their business. Unfortunately, there are no obvious examples of comprehensive reform of similar organizations that could provide a useful blueprint for implementing such dramatic change.

What Decisions Will be Passed on to Localities?

States will have to determine how much of the resources and control will be passed on to local communities and counties. In the past, the design of federal programs left relatively little discretion to the state and local governments. However, increased flexibility over program design, eligibility, and funding allocations will require another level of decisionmaking that could well lead to new conflicts. In particular, state and local communities will need to decide how control over social services will be distributed between Sacramento and local communities. What is the likely regional distribution of welfare money under block grants? Counties and local communities that have political leverage are likely to receive greater resources. If urban perspectives are better represented at the national level, then one might expect that there will be a shifting of resources from cities to rural or suburban areas.

One of the attractions of the legislation is that it moves the decisionmaking closer to the problem, which will hopefully lead to more innovative and effective policies. However, it is not clear that the state government will be much better in addressing California's problems. California is a large, diverse state whose needy population has differing problems. Perhaps the decisionmaking needs to be shifted down to a lower level, but it is unlikely that the state legislature will willingly give up control. As a result, tension between California and the federal government may simply be replaced by the tension between Sacramento and the localities.

California is one of a handful of states in which local communities are currently involved in the welfare delivery system, which can be an asset for the state. However, if local agencies currently implement federal policies merely as part of a bureaucracy--a way of doing business that will need to be overhauled, their involvement may be viewed as an additional barrier to effective reform. Whether the state will build on this existing infrastructure to meet its goals remains an open question. Alternatively, perhaps other institutions, such as private organizations, would be better suited to contract with the state to deliver welfare services.

Database Infrastructure Must be Built

In order to meet new federal requirements and support systemic approaches to policy design and implementation, California will have to invest heavily in new database infrastructure. At the very least, it will become necessary to track individuals over their lifetimes as they enter, leave, and reenter episodes of receiving public assistance. These tracking systems will need to be developed in collaboration with other states since proposed regulations limit the cumulative time on welfare, regardless of location. This is an ambitious and expensive project. However, with some foresight, the data source that is used to track individuals over time may be quite valuable for examining the effectiveness of programs. This will be especially true if the information system is capable of assessing the movement of individuals and family members across multiple social service programs.

Work Programs Must be Expanded

The legislation requires that eventually, at least 50 percent of the AFDC-FG caseload and 90 percent of the AFDC-U caseload be engaged in work-related activities. Nationally, the General Accounting Office estimates that only about 11 percent of the AFDC caseload is participating in the JOBS program, and in California the rate is under 9 percent. Is California ready to expand the employment-related dimension of assistance? What job-related activities will be provided for participants? How will work be monitored? How much will workers be paid? How will California work with unions to ensure that these government-subsidized workers do not displace other workers? Will there be a training component of the program? Does there exist an infrastructure for managing this dimension of the program? Is California ready and willing to expand the GAIN program to meet these requirements? Or should the state pursue alternative strategies? The fundamental goal of most welfare reform policies is self-sufficiency, which is derived largely through employment.[7] As a result, the work and job training component of welfare reform is fundamental.

Although jobs and training programs may reduce welfare dependency and save money in the long run, some states may be reluctant to pursue these options. Clearly, such programs require an initial investment of tax money, with benefits unlikely to be forthcoming until several years later. The up-front investment is likely to be even greater if the effort is supported by other programs designed to enhance the attractiveness of employment and/or job preparation activities. These include transportation and child care subsidies. Rather than make such an investment, California might prefer the 5-percent penalty on federal funds in the event that it does not achieve federal targets.

Some AFDC recipients cannot work because of physical or mental health problems. What will happen to these individuals when their two-year time limit expires? To what extent are job agencies, with no experience in working with needy populations, equipped to serve these people? Failure of these programs will likely cause enrollment in other programs, such as SSI or homeless assistance, to expand. What will happen to the children of mothers who become ineligible for benefits? There may be a rise in the demand for child care services for the low-income population. If the mother cannot find work, a growing number of children may become wards of the state, entering foster care or institutionalized care, which is often expensive. These cross-program effects are likely to be quite important.

Be Aware of Consequences of Actions for Related Federal Funds

State-determined eligibility and benefit requirements for AFDC can indirectly affect other sources of federal funding. If states decide to decrease cash welfare assistance, then the amount of food stamp benefits available to an applicant would increase because cash assistance is counted against food stamps (at least under current law). As another example, if the federal government continues noncapped assistance for foster care and adoption services, then local and state government would have a financial incentive to move children into foster care and adoption services and away from other types of services that are funded solely out of the state's available funds. Of course, the extent of movement will depend on the cost of out-placement to foster care and adoption services, as well as those services' capacities to expand.

Incentives to Work May Actually Attract Nonparticipants to AFDC

To encourage self-sufficiency, welfare reform will include proposals that will make work while on welfare more attractive. For example, child care benefits may increase, AFDC-related job training may expand, and the amount of earnings disregarded may increase. An unintended side effect of such changes is that participating in the program will become more attractive, and, at least in the short run, participation in AFDC may actually increase.

Realize that Changes in Welfare Benefits May Induce Migration

The increased state and, potentially, local flexibility in program design could, in theory, create more inter- and intrastate variation in benefit levels or standards of eligibility. If adjoining states or localities within a state offer different programs, they must realize that some individuals may migrate to the areas with more generous benefits.[8] To prevent this, states will be reluctant to increase benefit levels and, instead, may begin a competitive "spiral to the bottom." The current legislation allows states to restrict benefits for new entrants, at least in the short run, thereby reducing migration in search of assistance. Alternatively, states (and jurisdictions within states that are empowered to design programs independently) should collaborate with contiguous government jurisdictions to see that these migration effects and responses geared to reduce migration are not realized.

Be Prepared to Demonstrate the Usefulness of Federal Funds

Along with devolution and increased flexibility in designing welfare programs comes an increased responsibility to compare policy alternatives and evaluate decisions after the fact. Because the federal government is allocating block grants with (in theory) fewer strings attached, the state should expect to become more accountable for the use of these resources. Specifically, two to three years from now states should be prepared to demonstrate that the money is being spent wisely and that the programs they have established are working. States will need to do more than simply calculate caseloads and comply with regulations. The design of effective policies will require answers to substantive questions such as: Are the programs helping people achieve long-term self-sufficiency? Are they reducing out-of-wedlock and teenage pregnancy? Have they decreased the abortion rate? What elements of program design are critical in achieving observed levels of success? States, as a group, should strive to plan, conduct, and disseminate comparable, high-quality evaluations of their programs. Otherwise there will be no pool of common knowledge upon which to improve program effectiveness or to justify continuing levels of support for the welfare block grants.

Looking Toward the Future, but Learning from the Past

Under block grants, California will be given greater control over its welfare policy. A variety of policy options will likely be debated, and in fact Governor Wilson's proposal for reform, which anticipates federal action, has been discussed widely. To inform these decisions, it is important to integrate the lessons we have learned from welfare policies of the past.

Some 75 percent of all initial welfare applicants apply following the birth of a child or the break up of a marriage. However, there is little evidence that such events are, in fact, induced by the prospects of receiving welfare. For example, women are eligible for AFDC only if they have a child (or are pregnant), and the amount of the benefit increases with the number of children, providing, in theory, an incentive for women to have children. A number of empirical studies testing this claim with alternative approaches have found no consistent evidence that women have children because of the generosity of welfare benefits. (See reviews of the evidence by Moffitt, 1992, and Robbins and Fronstein, 1993, and new evidence from Jackson and Klerman, 1995). There are several reasons why welfare benefits may not induce women to have children. As summarized by Jackson and Klerman (1995), women (and men), especially teens, may not make childbearing decisions using rational, forward-looking calculations based on the costs and benefits of having a child. Second, the amount of assistance received under welfare may be so small that it does not provide enough of an incentive to have a child. Third, the observed differences in welfare payments across states and over time within states are not large enough to induce changes in childbearing. However, if benefits were altered by a much larger amount (e.g., if they were completely eliminated), then we may observe changes in childbearing.

The AFDC eligibility criteria provide an incentive to remain unmarried, or once married, to divorce. Some of the empirical evidence does find an effect of AFDC on female household headship, but the magnitude of the effect is small. (See Moffitt, 1992, for a review.)

As discussed, policymakers have expressed concerns that differences in generosity might induce women to migrate to states with higher benefits. However, the empirical evidence does not support this claim either (Roan, 1996; Walker, 1994). As is the case with fertility, it may be that there is no effect because the differences in benefits across states are not large enough to offset the costs of moving, which may include losing ties to family members who are important resources. However, welfare-motivated migration may occur in major cities that straddle state lines, where the cost of migrating is lower. California does not have any major metropolitan areas along its borders, which minimizes this problem. However, if the design of programs is eventually left to the discretion of individual counties, this could emerge as a future issue.

In theory, increases in AFDC benefits should reduce labor force participation as well as the hours worked by employed women who receive welfare[9] because earned income merely reduces the level of assistance on a dollar-for-dollar basis. On this topic, the evidence in fact suggests that AFDC reduces the number of hours of work. Using the mid-point of the range of estimates reviewed by Danziger et al. (1981), Moffitt (1992) calculates that if AFDC were completely eliminated, the number of hours worked would increase by about 5.4 hours per week.[10] Moreover, the studies find that very few women who would not be eligible for AFDC in the absence of the program change their hours of work enough to become eligible in the presence of the program. Or, as stated by Moffitt (1992, p. 17): "95 percent of those on the AFDC rolls would, if off the program, retain eligibility for benefits." At the same time, Moffitt (1996) notes that if participation in job training programs is basically voluntary and the programs are valuable, then offering training may increase the welfare caseload. On the other hand, if job training is mandatory, which is the direction in which current proposals are moving, and some potential AFDC participants would rather not have to enroll in job training, then welfare caseloads may decrease when job training is expanded.

California's job training program, GAIN, currently enrolls about 78,000 participants monthly. The Manpower Demonstration Research Corporation completed an evaluation of the program in 1994, with some encouraging findings (MDRC, 1994). In some counties, GAIN increased participants' earnings by 49 percent and decreased their welfare payments by 15 percent. And the return on the investment was about $3 for every $1 spent by the government. Although the positive effects in other counties were not as large, they were still substantial in many settings.

With an estimated one-half of all AFDC recipients having had a teen birth, addressing the special problems faced by teenagers provides perhaps the most significant leverage point for increasing self-sufficiency.[11] Summarizing the work on teenagers, Maynard (1994) finds that employment is the primary avenue out of poverty, and that support services such as child care and transportation are effective in promoting education and employment. At the same time, financial penalties (such as reduced benefits), in conjunction with an outreach effort by case managers to those who have violated the rules, also play an important role in achieving positive results.


Despite the current impasse on welfare reform, it is clear that the "new federalism" will pose significant challenges for the delivery of social services in California. Regardless of the exact form of the final legislation, several emerging trends are inevitable. First, it is important to recognize that there is a broad consensus that traditional approaches to welfare and related public assistance have not been effective. The categorical programs will be replaced by more-flexible block grants that, at least in theory, will provide a unique opportunity to redesign social programs. However, available funding will be severely constrained, with fiscal pressures surely leading to reduced budgets overall. Perhaps more dramatically, the probable end (or at least reduction) in entitlements will place increasing responsibility on states to address the changing needs of its citizens. Initial projections suggest that prospective federal funding for California under the proposed block grant legislation will eventually be reduced by an amount exceeding 20 percent of the total cost of the programs. Finally, devolution of authority will, in reality, come with many new strings attached. Federal restrictions on eligibility, the duration of benefits, and work participation requirements will have significant effects on overall funding levels, administrative costs, and the design of alternative policy solutions. Unfortunately, there is not much evidence that the new restrictions will make these programs any more effective.

Despite the daunting challenge, the wave of welfare reform can be viewed as a unique opportunity to reinvent the manner in which social services are provided. Short of a complete organizational overhaul (an ultimate outcome that may indeed be desirable), reformefforts will, at the very least, require the use of information technologies, common data sets, and mechanisms for enhanced communication across agency boundaries and different levels of influence, including government, private industry, and the nonprofitsector. Effective reform will also require a dramatic change in the values, objectives, and culture of policymakers in state and local government. In the past, the provision of social services has focused narrowly on expanding budgets, meeting administrative requirements, avoiding controversy, and counting the number of program recipients. In the future, it may be possible to link previously independent efforts, plan strategic approaches that span traditional departments, and design more innovative programs that respond and listen to the populations they are intended to serve.

Clearly, such a complete overhaul will be difficult to accomplish and there exists no convincing evidence that it will succeed. Thus, as an intermediate solution between the status quo and instantaneous systemic change, expansion of the current waiver system should be considered. Over two-thirds of the states have been granted waivers that allow them to experiment with alternative program designs. Recent evaluations of these changes have begun to inform decisionmakers of the efficacy of a variety of policies. Perhaps an expansion of the waiver system, including expediting states' requests for waivers and lengthening the time for experimentation, would be a judicious direction for policy. As the federal and state governments learn from these changes, they can implement on a broader scale the policies that have been proven to be successful.


Bureau of Labor Statistics, Employment and Earnings, February 1996.

California Department of Social Services (1996). Proposed Redesign of the Welfare System--Key Facts Booklet, January.

------ (1995a). Public Welfare in California, Information Services Bureau, Health and Welfare Agency, Statistical Series PA3-431.

------ (1995b). Aid to Families with Dependent Children Characteristics Survey. Information Services Bureau, Health and Welfare Agency, Program Information Series Report 1995-04.

Danziger, Sheldon, Robert Haveman, and Robert Plotnick (1981). "How Income Transfer Programs Affect Work, Savings, and the Income Distribution: A Critical Review," Journal of Economic Literature, 19:975-1028.

Ellwood, David T. (1986). Targeting Would-Be Long-Term Recipients of AFDC, Mathematica Policy Research, Inc., Reference No. 7617-953.

Hill, Elizabeth G. (1996). Analysis of the 1996-97 Budget Bill, Legislative Analyst's Office, State of California, Sacramento, California.

Hoynes, Hilary Williamson, and Thomas MaCurdy (1993). "Welfare Spells Over the Last Two Decades: Do Changes in Benefits Explain the Trends?" Mimeo, University of California, Berkeley.

Jackson, Catherine A., and Jacob Alex Klerman (1995). "Welfare and American Fertility," paper presented at National Bureau of Economic Research Summer Institute, Cambridge, Mass., July 1995.

Klerman, Jacob Alex, and Arleen Leibowitz (1993). Employment Continuity Among Young Mothers, RAND Labor and Population Program Working Paper Series, DRU-504-NICHD.

Manpower Demonstration Research Corporation (1994). Final Report: Benefits, Costs, and Three-Year Impacts. Manpower Demonstration Research Corporation, New York.

Maynard, Rebecca (1994). "Teenage Childbearing and Welfare Reform: Lessons from a Decade of Demonstration and Evaluation Research," mimeo, University of Pennsylvania, July 29, 1994.

Moffitt, Robert (1992). "Incentive Effects of the U.S. Welfare System," Journal of Economic Literature, 15(1):1-61.

Moffitt, Robert A. (1996). "The Effect of Employment and Training Programs on Entry and Exit from the Welfare Caseload," Journal of Policy Analysis and Management, 15(1): 32-50.

Roan, Carole E. (1995). "The Role of AFDC Benefits in Location Choice," Mimeo, Brown University.

Robbins, Philip K., and Paul Fronstein (1993). "Welfare Benefits and Family-Size Decisions of Never-Married Women," Institute for Research on Poverty Disccusion Paper 1022-93, University of Wisconsin, September.

U.S. House of Representatives, Committee on Ways and Means, (1995). Overview of Entitlement Programs, Washington, DC: Government Printing Office.

Walker, James R. (1994). "Migration Among Low-Income Households: Helping the Witch Doctors Reach Consensus," Institute for Research on Poverty Discussion Papers Dp 1031-94.


[1] Unfortunately there is no accurate measure of state-level cost-of-living differences that would allow us to make reliable adjustments across states.

[2] Administrative expenses represent about 8 percent of all AFDC-related costs in California. This compares with the U.S. average of just over 11 percent, primarily reflecting California's higher level of benefits per capita.

[3] Most versions of welfare reform allow for some percentage of the population (e.g., those with disabilities) to be exempt from the limitations on duration. However, the allowable percentage is likely to be significantly less than 19 percent.

[4] It is worth noting that the political clamor for welfare reform is pervasive at all levels of government. As our earlier discussion suggests, several states have received federal waivers and are initiating the process of change in advance of the federal legislation. Thus, these issues have current relevance even if the welfare reform bills remain bogged down.

[5] The work requirements will be reduced (on a percentage-for-percentage basis) for states that successfully get families off public assistance. If this were not the case, states would be penalized if the most employable individuals left the welfare roles, thereby making it more difficult to meet federal targets for work participation.

[6] In theory, the total effect of reform on the level of state-provided funding is uncertain. In the past, states have generally not altered funding levels in response to federal budget cuts. Of course, if the state is truly given more flexibility to design more innovative programs, the resulting gains in effectiveness could lead to increases in supplemental funding.

[7] Becoming married is another major pathway off of welfare.

[8] As we note below, there is little solid evidence to support this claim.

[9] Of course, there may be work-related benefits and costs that are independent of income. The former may include untaxed fringe benefits or on-the-job human capital development. Costs include transportation costs, disutility of time working or absence from the home, or daycare expenses.

[10] Although the complete elimination of AFDC is an unlikely policy scenario, it is clear that certain populations will be entirely cut off after reaching the time limitation on the duration of assistance. The question of what happens to these people (and to their children) is partially addressed by knowing how many of them are likely to be motivated and successful at finding work.

[11] We are not arguing that teen births are necessarily the root cause of subsequent welfare dependency. Indeed, the correlation between such births and the subsequent need for assistance may result from other factors.